Switch Theme:

GW Official FAQ - Drop Pod Doors  [RSS] Share on facebook Share on Twitter Submit to Reddit
»
Author Message
Advert


Forum adverts like this one are shown to any user who is not logged in. Join us by filling out a tiny 3 field form and you will get your own, free, dakka user account which gives a good range of benefits to you:
  • No adverts like this in the forums anymore.
  • Times and dates in your local timezone.
  • Full tracking of what you have read so you can skip to your first unread post, easily see what has changed since you last logged in, and easily see what is new at a glance.
  • Email notifications for threads you want to watch closely.
  • Being a part of the oldest wargaming community on the net.
If you are already a member then feel free to login now.




Made in us
Longtime Dakkanaut




 Charistoph wrote:
col_impact wrote:
Incorrect.

Spoiler:
Line of sight literally represents your warriors’ view of the enemy – they must be able to see their foes through, under or over the battlefield terrain and other models (whether friendly or enemy).

For one model to have line of sight to another, you must be able to trace a straight, unblocked line from its body (the head, torso, arms or legs) to any part of the target’s body.

Sometimes, all that will be visible of a model is a weapon, banner or other ornament he is carrying.


Visibility is discussed in the 3rd paragraph. It refers to "a model" and so can refer to any model pertinent to the situation. A model in the way is a model that is visibly in the way. Similarly, proximity or direction alone does not allow a model to block line of sight, a model must visibly block line of sight.

If it was "any model", it would say, "any model". If it was every model, it would say, "every model". It is only "a model", singular, though, so this interpretation that we are to consider any and all models at the same time puts a lie to the singular use.


If it were "only one model" the BRB would say "only one model". If it were "not more than one model" it would say not more than one model. If it were "a target model" the BRB would say "a target model."

I have already proven that the the BRB is referring to "a model" in a nonspecific general way. When the indefinite articles is used in that way, it can apply to any instance of "a model" in the situation.

For example

You have 3 weapons on your person.

A computer terminal asks you "do you have a weapon on your person? select yes or no".
The answer is yes. There are 3 instances of "a model" on your person.
The answer is not no since the computer terminal did not refer to "only one weapon" just "a weapon"

An indefinite article indicates that its noun is not a particular one (or ones) identifiable to the listener. It may be something that the speaker is mentioning for the first time, or its precise identity may be irrelevant or hypothetical, or the speaker may be making a general statement about any such thing. ( https://en.wikipedia.org/wiki/Article_(grammar)#Definite_article )

 Charistoph wrote:

Other models in this situation are classed in the same group as battlefield terrain. We are not told to consider those other models' visibility at the time of the 3rd Paragraph. Other models need to be considered, but only the same way as battlefield terrain, as indicated by the "and" preceding "other models". Again, you focus on two words when there is so much more to be considered in all of this.


The BRB chooses to say "all that will be visible of a model" in a situation where more than one model could be at play. Obviously "a model" would include a model in the way just as easily as "a target model"

 Charistoph wrote:

"A model" does not always mean "any model", context must still be considered. "A model" definitely does not mean "all models" or "every model". We are told to consider one model at the time. So, where does it state to consider other models in this? It does not.


The indefinite article is being used in a non-specific general way in a situation where there are any number of things that might mean "a model". Any instance of "a model" is to be considered

 Charistoph wrote:

We are told to consider one model at the time.


Do you have rules quote on this? You are literally making this up. The rule uses "sometimes" and "these cases" so we are actually talking about any number of times.

 Charistoph wrote:

col_impact wrote:
 Charistoph wrote:

Where does it state to not consider models invisible by this standard for other models?

The extent of the view is the view of the enemy of model. The rule is only concerned sorting out the visibility of models that are along that view. And in fact, any models not along that view might as well be considered invisible for the line of sight check since they are simply not relevant.

"Enemy model" is not stated. One is not always looking for a view of an "enemy model". As stated before, we could be looking at a possessor of a Special Rule which affects all models who can draw Line of Sight to it. In these cases, a significant portion of them are actually friendly models.

Incorrect.

Spoiler:
Line of sight literally represents your warriors’ view of the enemy


Infiltrate also refers only to enemy

Spoiler:
Infiltrators can be set up anywhere on the table that is more than 12" from any enemy unit, as long as no deployed enemy unit can draw line of sight to them.


 Charistoph wrote:


If the rule was "only concerned (with) sorting out the visibility of models that are along that view", then, as stated before, words like "any" and "all" would have been used rather than "a". The third paragraph is only considering one model at a time. The inclusion of other models in this is beyond its constructed paradigm.


Still waiting for the rules quote on "at a time". No where in the entire rule on Line of Sight is "at a time" mentioned. If the BRB meant "at a time" like it does for Wound Allocation then it would say "at a time". Your argument is invalid so long as you insist on adding words that are simply not there.

 Charistoph wrote:

col_impact wrote:

Incorrect.

There is no mention of "at a time". The rule actually says "sometimes" and "these cases" so there can be any number of times and any number of cases.

Only if one chooses to ignore the context presented earlier in the section. We are only told to be considering "a model" at this point in the writing of the rules, not "all models", nor "any models".

Sometimes is including the differences between "Hey, you're out in the open" and "Hey, all I can see is your gun!". Remember, context and and relevance. I know these are advanced concepts for some, and it is just easier to focus on one or two words to make a judgement on it, but it actually flows better when all such things are brought together instead of piecemeal or just picking and choosing what you will pay attention to.


No where in the whole rule is "at a time" mentioned so you are the one ignoring context. If you make up something that is not in the context such as adding "at a time" to the context then your argument is invalid. If you want to take a look at a rule that uses "at a time" look at Wound Allocation.

"Sometimes" can occur any number of times. If you are going to ignore the straightforward meaning of words then your argument is invalid.

This message was edited 1 time. Last update was at 2016/10/04 00:54:19


 
   
Made in us
Not as Good as a Minion





Astonished of Heck

col_impact wrote:
If it were "only one model" the BRB would say "only one model". If it were "not more than one model" it would say not more than one model. If it were "a target model" the BRB would say "a target model."

I have already proven that the the BRB is referring to "a model" in a nonspecific general way. When the indefinite articles is used in that way, it can apply to any instance of "a model" in the situation.

For example

You have 3 weapons on your person.

A computer terminal asks you "do you have a weapon on your person? select yes or no".
The answer is yes. There are 3 instances of "a model" on your person.
The answer is not no since the computer terminal did not refer to "only one weapon" just "a weapon"

An indefinite article indicates that its noun is not a particular one (or ones) identifiable to the listener. It may be something that the speaker is mentioning for the first time, or its precise identity may be irrelevant or hypothetical, or the speaker may be making a general statement about any such thing. ( https://en.wikipedia.org/wiki/Article_(grammar)#Definite_article )

Nevertheless, and you are reminded of this farther down your grammar page, it is still, just one. Not many, not all, not any, just one. You are instructed on doing this for one model, partially by the use of "a model", but later by the use of "he" and "the model".

You have brought this up before and it was not accepted as relevant on its own. Why do you keep insisting on using it?

col_impact wrote:
 Charistoph wrote:

Other models in this situation are classed in the same group as battlefield terrain. We are not told to consider those other models' visibility at the time of the 3rd Paragraph. Other models need to be considered, but only the same way as battlefield terrain, as indicated by the "and" preceding "other models". Again, you focus on two words when there is so much more to be considered in all of this.

The BRB chooses to say "all that will be visible of a model" in a situation where more than one model could be at play. Obviously "a model" would include a model in the way just as easily as "a target model"

And then it says, "he" and "the model" later on, which are still specifically singular in their use.

Now, where does it tell us in which order we are to be considering the visibility of other models before considering the one we WANT to see? This is, what, the fifth time I've asked and you haven't properly answered?

col_impact wrote:
 Charistoph wrote:

"A model" does not always mean "any model", context must still be considered. "A model" definitely does not mean "all models" or "every model". We are told to consider one model at the time. So, where does it state to consider other models in this? It does not.

The indefinite article is being used in a non-specific general way in a situation where there are any number of things that might mean "a model". Any instance of "a model" is to be considered

An assumption based on consideration of only two words. No consideration made for other words being used or the context in place. The use of an indefinite doesn't restrict us to one specific model, but it still does restrict it to one model for the context of the sentence in question. Other words are needed to expand this "a model" to be beyond the one being considered by this sentence.

Context and relevance, they are things in the rulebook not to be casually dismissed. Quit picking and choosing what you will adhere to.

col_impact wrote:
 Charistoph wrote:

We are told to consider one model at the time.

Do you have rules quote on this? You are literally making this up. The rule uses "sometimes" and "these cases" so we are actually talking about any number of times.

I've explained this. By the use of the singular terms, "a model", "he", and "the model". Nothing else allows me to consider these instructions beyond an individual model's basis. Therefore, one model at a time.

col_impact wrote:
 Charistoph wrote:

col_impact wrote:
 Charistoph wrote:

Where does it state to not consider models invisible by this standard for other models?

The extent of the view is the view of the enemy of model. The rule is only concerned sorting out the visibility of models that are along that view. And in fact, any models not along that view might as well be considered invisible for the line of sight check since they are simply not relevant.

"Enemy model" is not stated. One is not always looking for a view of an "enemy model". As stated before, we could be looking at a possessor of a Special Rule which affects all models who can draw Line of Sight to it. In these cases, a significant portion of them are actually friendly models.

Incorrect.

Spoiler:
Line of sight literally represents your warriors’ view of the enemy


Infiltrate also refers only to enemy

Spoiler:
Infiltrators can be set up anywhere on the table that is more than 12" from any enemy unit, as long as no deployed enemy unit can draw line of sight to them.

Upgrade your reading comprehension and remember things like context. You were referencing the Line of Sight rules when you brought up "enemy models", I highlighted it in red to help you find it. I was referencing the Line of Sight rules when stating "'Enemy model' is not stated".

Remember the point of an indefinite article. It is not just limited to being used on an enemy model. "Enemy model" is not used in the second and third paragraphs of Line of Sight. If there is a special rule that allows friendly models to benefit that are within 6" of it and can draw Line of Sight to it, we would still be considering if the only thing on that model that could be seen was his weapon, banner, etc. And yes, there are several of those rules out there. And it is to those rules which I was referencing, "In these cases, a significant portion of them are actually friendly models."

col_impact wrote:
 Charistoph wrote:
If the rule was "only concerned (with) sorting out the visibility of models that are along that view", then, as stated before, words like "any" and "all" would have been used rather than "a". The third paragraph is only considering one model at a time. The inclusion of other models in this is beyond its constructed paradigm.

Still waiting for the rules quote on "at a time". No where in the entire rule on Line of Sight is "at a time" mentioned. If the BRB meant "at a time" like it does for Wound Allocation then it would say "at a time". Your argument is invalid so long as you insist on adding words that are simply not there.

It is in the context of the instructions up to this point. It is in the use of the singular uses of nouns and the indefinite article. I've mentioned this before, actually, and not just in this post.

col_impact wrote:
 Charistoph wrote:

col_impact wrote:

Incorrect.

There is no mention of "at a time". The rule actually says "sometimes" and "these cases" so there can be any number of times and any number of cases.

Only if one chooses to ignore the context presented earlier in the section. We are only told to be considering "a model" at this point in the writing of the rules, not "all models", nor "any models".

Sometimes is including the differences between "Hey, you're out in the open" and "Hey, all I can see is your gun!". Remember, context and and relevance. I know these are advanced concepts for some, and it is just easier to focus on one or two words to make a judgement on it, but it actually flows better when all such things are brought together instead of piecemeal or just picking and choosing what you will pay attention to.

No where in the whole rule is "at a time" mentioned so you are the one ignoring context. If you make up something that is not in the context such as adding "at a time" to the context then your argument is invalid. If you want to take a look at a rule that uses "at a time" look at Wound Allocation.

"Sometimes" can occur any number of times. If you are going to ignore the straightforward meaning of words then your argument is invalid.

Yes, it can occur any number of times. But it does not mean we allow this to be repeated all at once, either, which is what you are proposing. "Sometimes" is being used to join the context of this sentence with the one before it.

Remember the context being provided as an instruction set.

We are told that to establish line of sight from one model to another, we draw a straight, unblocked line between the two bodies. Sometimes, certain parts of a model are all that can be seen. The model is not considered seen by only seeing these parts. This is so a model with certain impressive bits are not unduly penalized.

Tell me, where in that paragraph above are we to be considering all models under this judgement of visibility, even though we only are seeking one?

Are you a Wolf, a Sheep, or a Hound?
Megavolt wrote:They called me crazy…they called me insane…THEY CALLED ME LOONEY!! and boy, were they right.
 
   
Made in us
Longtime Dakkanaut




 Charistoph wrote:
col_impact wrote:
If it were "only one model" the BRB would say "only one model". If it were "not more than one model" it would say not more than one model. If it were "a target model" the BRB would say "a target model."

I have already proven that the the BRB is referring to "a model" in a nonspecific general way. When the indefinite articles is used in that way, it can apply to any instance of "a model" in the situation.

For example

You have 3 weapons on your person.

A computer terminal asks you "do you have a weapon on your person? select yes or no".
The answer is yes. There are 3 instances of "a model" on your person.
The answer is not no since the computer terminal did not refer to "only one weapon" just "a weapon"

An indefinite article indicates that its noun is not a particular one (or ones) identifiable to the listener. It may be something that the speaker is mentioning for the first time, or its precise identity may be irrelevant or hypothetical, or the speaker may be making a general statement about any such thing. ( https://en.wikipedia.org/wiki/Article_(grammar)#Definite_article )

Nevertheless, and you are reminded of this farther down your grammar page, it is still, just one. Not many, not all, not any, just one. You are instructed on doing this for one model, partially by the use of "a model", but later by the use of "he" and "the model".

You have brought this up before and it was not accepted as relevant on its own. Why do you keep insisting on using it?


We are not limited on the number of instances on "a model".

 Charistoph wrote:

col_impact wrote:
 Charistoph wrote:

Other models in this situation are classed in the same group as battlefield terrain. We are not told to consider those other models' visibility at the time of the 3rd Paragraph. Other models need to be considered, but only the same way as battlefield terrain, as indicated by the "and" preceding "other models". Again, you focus on two words when there is so much more to be considered in all of this.

The BRB chooses to say "all that will be visible of a model" in a situation where more than one model could be at play. Obviously "a model" would include a model in the way just as easily as "a target model"

And then it says, "he" and "the model" later on, which are still specifically singular in their use.

Now, where does it tell us in which order we are to be considering the visibility of other models before considering the one we WANT to see? This is, what, the fifth time I've asked and you haven't properly answered?


By using "a model" the BRB intentionally makes no distinction between a model in the way and a target and there is similarly no order specified.

If you think there is some order specified in the rule please show it.

This is the fifth time you have ignored my completely supported answer to the question. How many times do I have to tell you the answer to 3+4 is 7? Repeating your question does not make my correct answer incorrect.

 Charistoph wrote:

col_impact wrote:
 Charistoph wrote:

"A model" does not always mean "any model", context must still be considered. "A model" definitely does not mean "all models" or "every model". We are told to consider one model at the time. So, where does it state to consider other models in this? It does not.

The indefinite article is being used in a non-specific general way in a situation where there are any number of things that might mean "a model". Any instance of "a model" is to be considered

An assumption based on consideration of only two words. No consideration made for other words being used or the context in place. The use of an indefinite doesn't restrict us to one specific model, but it still does restrict it to one model for the context of the sentence in question. Other words are needed to expand this "a model" to be beyond the one being considered by this sentence.

Context and relevance, they are things in the rulebook not to be casually dismissed. Quit picking and choosing what you will adhere to.


You keep trying to sneak in an unwarranted restriction, a sense of "only one", into the sentence. You are not allowed to unless you can prove that it must be read that way.

"During a soccer match, a player is allowed to score a goal." does not mean the same thing as "During only one soccer match, only one player is allowed to score only one goal."

In fact, "During a soccer match, a player is allowed to score a goal." means the same thing as "During soccer matches, players are allowed to score goals."


As long as you keep reading "a" as "only one" then your argument is invalid.

 Charistoph wrote:

col_impact wrote:
 Charistoph wrote:

We are told to consider one model at the time.

Do you have rules quote on this? You are literally making this up. The rule uses "sometimes" and "these cases" so we are actually talking about any number of times.

I've explained this. By the use of the singular terms, "a model", "he", and "the model". Nothing else allows me to consider these instructions beyond an individual model's basis. Therefore, one model at a time.


The use of "a" does not carry with it "at a time". You are literally making that up. Feel free to research up some grammar documentation to prove otherwise. Until you stick with the words in the BRB your argument is invalid.

 Charistoph wrote:

Upgrade your reading comprehension and remember things like context. You were referencing the Line of Sight rules when you brought up "enemy models", I highlighted it in red to help you find it. I was referencing the Line of Sight rules when stating "'Enemy model' is not stated".

Remember the point of an indefinite article. It is not just limited to being used on an enemy model. "Enemy model" is not used in the second and third paragraphs of Line of Sight. If there is a special rule that allows friendly models to benefit that are within 6" of it and can draw Line of Sight to it, we would still be considering if the only thing on that model that could be seen was his weapon, banner, etc. And yes, there are several of those rules out there. And it is to those rules which I was referencing, "In these cases, a significant portion of them are actually friendly models."


Incorrect.

"Line of sight literally represents your warriors’ view of the enemy"

If there are several rules out there which represent your warriors' view the friendly, please point them out. To my knowledge, there are none.

This would appear to be another instance where you just make stuff up when the words on the page don't suit your needs.

 Charistoph wrote:

col_impact wrote:

No where in the whole rule is "at a time" mentioned so you are the one ignoring context. If you make up something that is not in the context such as adding "at a time" to the context then your argument is invalid. If you want to take a look at a rule that uses "at a time" look at Wound Allocation.

"Sometimes" can occur any number of times. If you are going to ignore the straightforward meaning of words then your argument is invalid.

Yes, it can occur any number of times. But it does not mean we allow this to be repeated all at once, either, which is what you are proposing. "Sometimes" is being used to join the context of this sentence with the one before it.

Remember the context being provided as an instruction set.

We are told that to establish line of sight from one model to another, we draw a straight, unblocked line between the two bodies. Sometimes, certain parts of a model are all that can be seen. The model is not considered seen by only seeing these parts. This is so a model with certain impressive bits are not unduly penalized.

Tell me, where in that paragraph above are we to be considering all models under this judgement of visibility, even though we only are seeking one?


I am proposing that you deal with each and all instances of "a model".

The BRB explicitly tells us to generally deal with the visibility of anything that is "a model". So if you think the BRB is specifying beyond that then you will need to prove it. The BRB tells us to care about the visibility of anything counting as "a model" in this line of sight check since it generally refers to "a model".

For a model to be in the way it must be visibly in the way and not in the way by some other measure.

For a model to be blocking it must be visibly blocking and not by some other manner of blocking.

If a model in the way is deemed partially or wholly invisible by the BRB then that affects whether it is in the way for the purposes of line of sight.
   
Made in us
Not as Good as a Minion





Astonished of Heck

col_impact wrote:
We are not limited on the number of instances on "a model".

Nor are we instructed to look beyond "a model" before another determination of Line of Sight is engaged, hence "at a time".

col_impact wrote:
By using "a model" the BRB intentionally makes no distinction between a model in the way and a target and there is similarly no order specified.

If you think there is some order specified in the rule please show it.

This is the fifth time you have ignored my completely supported answer to the question. How many times do I have to tell you the answer to 3+4 is 7? Repeating your question does not make my correct answer incorrect.

I have been asking you for an order of operations because that would support your statement and assertion. I explained the reasoning behind the question quite clearly and in context. You have provided nothing but your assertions that this is so while ignoring everything else connected to what you bring up. You have focused on only one thing while ignoring all around it.

You are to consider one model at a time, this is supported by the use of singular terms throughout the paragraph. Why should I or must I consider any other models before the one I am trying to draw Line of Sight to it? The use of an indefinite article does not impose any requirements of such.

col_impact wrote:
You keep trying to sneak in an unwarranted restriction, a sense of "only one", into the sentence. You are not allowed to unless you can prove that it must be read that way.

"During a soccer match, a player is allowed to score a goal." does not mean the same thing as "During only one soccer match, only one player is allowed to score only one goal."

In fact, "During a soccer match, a player is allowed to score a goal." means the same thing as "During soccer matches, players are allowed to score goals."

As long as you keep reading "a" as "only one" then your argument is invalid.

Your lies become you. My insistence is "one at a time" because we are not instructed to make this consideration for more than "a model". I have never been reading it as "only one". That is purely a fabrication of yours. The only time I sated "only one" is not in a rules reference, but a reference in practical application. Note the difference.

With your insistence that "a" means "any", "all", "more than one", your argument falls flat on its face.

col_impact wrote:
 Charistoph wrote:
I've explained this. By the use of the singular terms, "a model", "he", and "the model". Nothing else allows me to consider these instructions beyond an individual model's basis. Therefore, one model at a time.

The use of "a" does not carry with it "at a time". You are literally making that up. Feel free to research up some grammar documentation to prove otherwise. Until you stick with the words in the BRB your argument is invalid.

And your lack of desire to apply reading comprehension causes you to fail again. You didn't bother to note I didn't keep it to just "a model". And I used words from the paragraph in question. Would you like me to highlight them for you?
Sometimes, all that will be visible of a model is a weapon, banner or other ornament he is carrying. In these cases, the model is not visible. Similarly, we ignore wings, tails and antennae even though they are technically part of a model’s body.

See, all words from the paragraph in question. All singular when being discussed.

col_impact wrote:
Incorrect.

"Line of sight literally represents your warriors’ view of the enemy"

If there are several rules out there which represent your warriors' view the friendly, please point them out. To my knowledge, there are none.

This would appear to be another instance where you just make stuff up when the words on the page don't suit your needs.

Not making it up, just remembering old rules. They weren't very common. A difference.

Not to mention, weren't you the one talking about it not being about "target models"?

Not to mention, the portion that states "enemy models" is not in the instruction set we are looking at in the third paragraph. You really have an issue with applying things in context.

Now, a reminder, will you address it, or go politician again and answer the question with your own unrelated answer?

Remember the context being provided as an instruction set.

We are told that to establish line of sight from one model to another, we draw a straight, unblocked line between the two bodies. Sometimes, certain parts of a model are all that can be seen. The model is not considered seen by only seeing these parts. This is so a model with certain impressive bits are not unduly penalized.

Tell me, where in that paragraph above are we to be considering all models under this judgement of visibility, even though we only are seeking one?

col_impact wrote:
I am proposing that you deal with each and all instances of "a model".

The BRB explicitly tells us to generally deal with the visibility of anything that is "a model". So if you think the BRB is specifying beyond that then you will need to prove it. The BRB tells us to care about the visibility of anything counting as "a model" in this line of sight check since it generally refers to "a model".

No, it does not. "Anything" is not a term used in the third paragraph. This is just more of your only looking at one thing and ignoring all else. We are only told to consider the visibility of "a model" right after we are told to be drawing a line from one model to another. Not "all models", not "any models", just "a model".

col_impact wrote:
For a model to be in the way it must be visibly in the way and not in the way by some other measure.

Where are we told to consider the visibility of other models under this standard?

col_impact wrote:
For a model to be blocking it must be visibly blocking and not by some other manner of blocking.

Quote, please.

col_impact wrote:
If a model in the way is deemed partially or wholly invisible by the BRB then that affects whether it is in the way for the purposes of line of sight.

True, by where are we told to make that differentiation here in the third paragraph of Line of Sight?

Are you a Wolf, a Sheep, or a Hound?
Megavolt wrote:They called me crazy…they called me insane…THEY CALLED ME LOONEY!! and boy, were they right.
 
   
Made in us
Longtime Dakkanaut




 Charistoph wrote:
col_impact wrote:
We are not limited on the number of instances on "a model".

Nor are we instructed to look beyond "a model" before another determination of Line of Sight is engaged, hence "at a time".


Incorrect. There is no "at a time" anywhere in the rule. Insisting upon the presence of words that simply aren't there shows that your argument is invalid. Stick to the rules as given.

 Charistoph wrote:
col_impact wrote:
By using "a model" the BRB intentionally makes no distinction between a model in the way and a target and there is similarly no order specified.

If you think there is some order specified in the rule please show it.

This is the fifth time you have ignored my completely supported answer to the question. How many times do I have to tell you the answer to 3+4 is 7? Repeating your question does not make my correct answer incorrect.

I have been asking you for an order of operations because that would support your statement and assertion. I explained the reasoning behind the question quite clearly and in context. You have provided nothing but your assertions that this is so while ignoring everything else connected to what you bring up. You have focused on only one thing while ignoring all around it.

You are to consider one model at a time, this is supported by the use of singular terms throughout the paragraph. Why should I or must I consider any other models before the one I am trying to draw Line of Sight to it? The use of an indefinite article does not impose any requirements of such.


The rules specify no order of operations nor do they mention "at a time". Your insistence on processes which simply aren't present in the rules makes your argument.

The 3rd paragraph make no specification about what "a model" is. Therefore you are not allowed to. Therefore you resolve the "sometimes" of a rule for each instance of "a model". This might involve a target model or a model in the way or a model in the way and a target model.

 Charistoph wrote:


col_impact wrote:
You keep trying to sneak in an unwarranted restriction, a sense of "only one", into the sentence. You are not allowed to unless you can prove that it must be read that way.

"During a soccer match, a player is allowed to score a goal." does not mean the same thing as "During only one soccer match, only one player is allowed to score only one goal."

In fact, "During a soccer match, a player is allowed to score a goal." means the same thing as "During soccer matches, players are allowed to score goals."

As long as you keep reading "a" as "only one" then your argument is invalid.

Your lies become you. My insistence is "one at a time" because we are not instructed to make this consideration for more than "a model".


There is no "one at a time" at all in the rule. If you feel otherwise, prove it with a rules quote or grammatical reference.

 Charistoph wrote:

I have never been reading it as "only one". That is purely a fabrication of yours. The only time I sated "only one" is not in a rules reference, but a reference in practical application. Note the difference.

With your insistence that "a" means "any", "all", "more than one", your argument falls flat on its face.


Cool. Then "a" does not mean "only one" and there is no quantitative restriction. The 3rd paragraph can refer to each instance of "a model" in the scenario.

 Charistoph wrote:

col_impact wrote:
 Charistoph wrote:
I've explained this. By the use of the singular terms, "a model", "he", and "the model". Nothing else allows me to consider these instructions beyond an individual model's basis. Therefore, one model at a time.

The use of "a" does not carry with it "at a time". You are literally making that up. Feel free to research up some grammar documentation to prove otherwise. Until you stick with the words in the BRB your argument is invalid.

And your lack of desire to apply reading comprehension causes you to fail again. You didn't bother to note I didn't keep it to just "a model". And I used words from the paragraph in question. Would you like me to highlight them for you?
Sometimes, all that will be visible of a model is a weapon, banner or other ornament he is carrying. In these cases, the model is not visible. Similarly, we ignore wings, tails and antennae even though they are technically part of a model’s body.

See, all words from the paragraph in question. All singular when being discussed.


Cool. You have just pointed out the grammatical property known as Agreement. The 3rd paragraph starts talking about "a model" that could be a model in the way or a target model. Later references cannot repeat the indefinite article and must agree in case. You have discovered Agreement!

You really need to polish up on your grammar because you are seeing castles where there are windmills.

 Charistoph wrote:


col_impact wrote:
Incorrect.

"Line of sight literally represents your warriors’ view of the enemy"

If there are several rules out there which represent your warriors' view the friendly, please point them out. To my knowledge, there are none.

This would appear to be another instance where you just make stuff up when the words on the page don't suit your needs.

Not making it up, just remembering old rules. They weren't very common. A difference.


Can you point to one? Your foggy memory isn't an official rules resource.

 Charistoph wrote:

Not to mention, weren't you the one talking about it not being about "target models"?

Not to mention, the portion that states "enemy models" is not in the instruction set we are looking at in the third paragraph. You really have an issue with applying things in context.

Now, a reminder, will you address it, or go politician again and answer the question with your own unrelated answer?

Remember the context being provided as an instruction set.

We are told that to establish line of sight from one model to another, we draw a straight, unblocked line between the two bodies. Sometimes, certain parts of a model are all that can be seen. The model is not considered seen by only seeing these parts. This is so a model with certain impressive bits are not unduly penalized.

Tell me, where in that paragraph above are we to be considering all models under this judgement of visibility, even though we only are seeking one?


Line of sight represents the warriors view of the enemy. That is the extent of the view and the viewing model is not looking outside that extent.

When the viewing model traces line of sight according to the extent of that view, it will see models. Whether its a model in the way or a target model, the 3rd paragraph might affect what counts as visible or ignored on the model.

 Charistoph wrote:


col_impact wrote:
I am proposing that you deal with each and all instances of "a model".

The BRB explicitly tells us to generally deal with the visibility of anything that is "a model". So if you think the BRB is specifying beyond that then you will need to prove it. The BRB tells us to care about the visibility of anything counting as "a model" in this line of sight check since it generally refers to "a model".

No, it does not. "Anything" is not a term used in the third paragraph. This is just more of your only looking at one thing and ignoring all else. We are only told to consider the visibility of "a model" right after we are told to be drawing a line from one model to another. Not "all models", not "any models", just "a model".


Incorrect. The 3rd paragraph uses "sometimes" and "these cases" to indicate that we can deal with each instance of "a model" that the rule instructs us to deal with.

 Charistoph wrote:

col_impact wrote:
For a model to be in the way it must be visibly in the way and not in the way by some other measure.

Where are we told to consider the visibility of other models under this standard?

col_impact wrote:
For a model to be blocking it must be visibly blocking and not by some other manner of blocking.

Quote, please.


The BRB uses line of sight and therefore relies on the visibility of models in the way to determine line of sight. If the models in the way are not visibly blocking line of sight then they are not blocking line of sight. Models that are considered invisible do not block line of sight. These are all simple truisms.

If you don't think it's a truism then feel free to treat all models in the way as invisible and see what happens.

 Charistoph wrote:

col_impact wrote:
If a model in the way is deemed partially or wholly invisible by the BRB then that affects whether it is in the way for the purposes of line of sight.

True, by where are we told to make that differentiation here in the third paragraph of Line of Sight?


The 3rd paragraph instructs about cases that affect the visibility of a model, whether its a target model or a model in the way.

This message was edited 1 time. Last update was at 2016/10/04 05:36:41


 
   
Made in us
Not as Good as a Minion





Astonished of Heck

col_impact wrote:
Incorrect. There is no "at a time" anywhere in the rule. Insisting upon the presence of words that simply aren't there shows that your argument is invalid. Stick to the rules as given.

Failure at reading comprehension does not help this statement. I never stated it actually said, "at a time". I am saying that due to the context of the information at hand, and the lack of direction to consider more than one model in these instructions, we are to do it one at a time.

col_impact wrote:
The rules specify no order of operations nor do they mention "at a time". Your insistence on processes which simply aren't present in the rules makes your argument.

The 3rd paragraph make no specification about what "a model" is. Therefore you are not allowed to. Therefore you resolve the "sometimes" of a rule for each instance of "a model". This might involve a target model or a model in the way or a model in the way and a target model.

Failure at consistent reading comprehension does not help this statement. The instructions are to consider "a model", as in "one model". Not "any models", not "all models", just one.

In order for us to consider more than "a model", there needs to be an order of operations, or the direction has to be for "all models" or "any models. You cannot present them, as they do not exist. Therefore, with a lack of direction to do so, I am only required to consider one model's visibility per Line of Sight check, and that will be whatever I'm trying to establish Line of Sight to.

Remember, a lack of instructions does not equate to a permission to do so.

col_impact wrote:
There is no "one at a time" at all in the rule. If you feel otherwise, prove it with a rules quote or grammatical reference.

Failure at reading comprehension is no excuse. I have given it before. Look it up.

col_impact wrote:
Cool. Then "a" does not mean "only one" and there is no quantitative restriction. The 3rd paragraph can refer to each instance of "a model" in the scenario.

Failure at reading comprehension cannot help your case. "A model" cannot mean more than one at a time considering the context in which the section is written.

col_impact wrote:
Cool. You have just pointed out the grammatical property known as Agreement. The 3rd paragraph starts talking about "a model" that could be a model in the way or a target model. Later references cannot repeat the indefinite article and must agree in case. You have discovered Agreement!

You really need to polish up on your grammar because you are seeing castles where there are windmills.

Incorrect. "A" does not equal "all" or "any", as you continue to profess. It means even less when considered in context from the previous sentence where it is being considered going from one model to another.

Quit pursuing your sandcastles in the sand.

col_impact wrote:
Can you point to one? Your foggy memory isn't an official rules resource.

Out of a hundred codices worth of material, most of which I do not have any more in any format? You jest.

col_impact wrote:
 Charistoph wrote:

Not to mention, weren't you the one talking about it not being about "target models"?

Not to mention, the portion that states "enemy models" is not in the instruction set we are looking at in the third paragraph. You really have an issue with applying things in context.

Now, a reminder, will you address it, or go politician again and answer the question with your own unrelated answer?

Remember the context being provided as an instruction set.

We are told that to establish line of sight from one model to another, we draw a straight, unblocked line between the two bodies. Sometimes, certain parts of a model are all that can be seen. The model is not considered seen by only seeing these parts. This is so a model with certain impressive bits are not unduly penalized.

Tell me, where in that paragraph above are we to be considering all models under this judgement of visibility, even though we only are seeking one?

Line of sight represents the warriors view of the enemy. That is the extent of the view and the viewing model is not looking outside that extent.

When the viewing model traces line of sight according to the extent of that view, it will see models. Whether its a model in the way or a target model, the 3rd paragraph might affect what counts as visible or ignored on the model.

So, the Politician's Answer it is. The question wasn't about the 3rd paragraph, it was about the restate mentioned above. This one here:
Spoiler:
We are told that to establish line of sight from one model to another, we draw a straight, unblocked line between the two bodies. Sometimes, certain parts of a model are all that can be seen. The model is not considered seen by only seeing these parts. This is so a model with certain impressive bits are not unduly penalized.

Tell me, where in that paragraph in that spoiler are we to be considering all models under this judgement of visibility, even though we only are seeking one?

col_impact wrote:
Incorrect. The 3rd paragraph uses "sometimes" and "these cases" to indicate that we can deal with each instance of "a model" that the rule instructs us to deal with.

That does not perfectly jell when combined with the preceding sentence. "Sometimes" is referring to a portion of the times the preceding sentence engender, due to its conjunctive nature. It is not carte blanche to ignore the singularity of the nouns being used or to treat all these times at the same time.

col_impact wrote:
 Charistoph wrote:

col_impact wrote:
For a model to be in the way it must be visibly in the way and not in the way by some other measure.

Where are we told to consider the visibility of other models under this standard?

col_impact wrote:
For a model to be blocking it must be visibly blocking and not by some other manner of blocking.

Quote, please.

The BRB uses line of sight and therefore relies on the visibility of models in the way to determine line of sight. If the models in the way are not visibly blocking line of sight then they are not blocking line of sight. Models that are considered invisible do not block line of sight. These are all simple truisms.

If you don't think it's a truism then feel free to treat all models in the way as invisible and see what happens.

So, no quote, just an assertion. The 3rd paragraph does not consider the visibility of other models when determining the visibility of a model. It is only considering the visibility of one model. It does not literally state to consider other models who are only showing non-body bits in this assertion. Other rules can insert themselves here, of course, as in "Own Unit", but this 3rd paragraph is not trying to make a determination of any models a focus is not upon. Only a model you do have a focus on.

col_impact wrote:
 Charistoph wrote:

col_impact wrote:
If a model in the way is deemed partially or wholly invisible by the BRB then that affects whether it is in the way for the purposes of line of sight.

True, by where are we told to make that differentiation here in the third paragraph of Line of Sight?[

The 3rd paragraph instructs about cases that affect the visibility of a model, whether its a target model or a model in the way.

But it does not actually state this for a model in the way. It is stating this consideration after making a point of establishing visibility from one body to another. This rule is to not penalize a model for having grand bits. Why would an intervening model be penalized for having epic bits if this rule was not in place?

Why put it in such a general way, when there are other ways far more clear for "any model" or "all models" in view? Why just use statements which indicate one model is being addressed?

This is because the context of the previous sentence is still in play. The "Sometimes" is in place because it is distinguishing between "hey, I'm in the open" and "hey, I'm behind a wall". "A model" is in place because Line of Sight is not always used against a targeted model. "A model" is used because not every case of trying to draw Line of Sight will result in this situation. Could it have been more specific, yes.

Was it intended to be this way? I cannot say. I do honestly think that when they write these rules they rarely think of the weird situations, and their writing shows as only being directed at the here and now with no concept of fitting things together. They allow future writings to intrude on the present, with no care as to how that may coordinate with other future writings.

Are you a Wolf, a Sheep, or a Hound?
Megavolt wrote:They called me crazy…they called me insane…THEY CALLED ME LOONEY!! and boy, were they right.
 
   
Made in us
Longtime Dakkanaut





col_impact wrote:

If it was "any model", it would say, "any model". If it was every model, it would say, "every model". It is only "a model", singular, though, so this interpretation that we are to consider any and all models at the same time puts a lie to the singular use.

If it were "only one model" the BRB would say "only one model". If it were "not more than one model" it would say not more than one model. If it were "a target model" the BRB would say "a target model."

I have already proven that the the BRB is referring to "a model" in a nonspecific general way. When the indefinite articles is used in that way, it can apply to any instance of "a model" in the situation


The very next sentence after using "a model", referring back to that model it says "the model". Model singular, not models plural. Are you going to say "the model" is nospecific also? This disproves that the BRB is referring to "a model" in a nonspecific way. I brought this up before and Charistoph has brought it up several times. "The model" is referring to the same model that was mentioned in "a model". You choose to ignore this so that you can say that you have proven that "a model" is nonspecific. At this point it starts to feel like intellectual dishonesty on your part to just keep going back to "a model" and ignore what is said about "the model".

This message was edited 3 times. Last update was at 2016/10/04 14:03:31


 
   
Made in us
Longtime Dakkanaut




 Charistoph wrote:
col_impact wrote:
Incorrect. There is no "at a time" anywhere in the rule. Insisting upon the presence of words that simply aren't there shows that your argument is invalid. Stick to the rules as given.

Failure at reading comprehension does not help this statement. I never stated it actually said, "at a time". I am saying that due to the context of the information at hand, and the lack of direction to consider more than one model in these instructions, we are to do it one at a time.


You can say whatever you want but you are not a rules resource. You can choose to play it "one at a time" but that is your house rule. The BRB makes no mention of "at a time" therefore no such procedure is being required.

The 3rd paragraph uses "sometimes" and "these cases" and "a model" refers to a non-specific, general model. The rule applies to each instance of "a model" bases on the words the BRB actually uses.

 Charistoph wrote:
col_impact wrote:
The rules specify no order of operations nor do they mention "at a time". Your insistence on processes which simply aren't present in the rules makes your argument.

The 3rd paragraph make no specification about what "a model" is. Therefore you are not allowed to. Therefore you resolve the "sometimes" of a rule for each instance of "a model". This might involve a target model or a model in the way or a model in the way and a target model.

Failure at consistent reading comprehension does not help this statement. The instructions are to consider "a model", as in "one model". Not "any models", not "all models", just one.

In order for us to consider more than "a model", there needs to be an order of operations, or the direction has to be for "all models" or "any models. You cannot present them, as they do not exist. Therefore, with a lack of direction to do so, I am only required to consider one model's visibility per Line of Sight check, and that will be whatever I'm trying to establish Line of Sight to.

Remember, a lack of instructions does not equate to a permission to do so.


The 3rd paragraph does not specify "only one model" so we are not restricted to one.

There is no order of operations so we are free to consider the visibility of "a model" in whatever order or all at once to our hearts desire.

The 3rd paragraph states "sometimes" and "these cases" which allows for any number of times and specifies that multiple cases can result from application of the rule.

The words in the BRB simply do not support you.

 Charistoph wrote:


col_impact wrote:
There is no "one at a time" at all in the rule. If you feel otherwise, prove it with a rules quote or grammatical reference.

Failure at reading comprehension is no excuse. I have given it before. Look it up.


So no rules quote and no grammatical reference. Got it.

If you are unable to support what you say with rules quotes and references, your argument is invalid.

 Charistoph wrote:


col_impact wrote:
Cool. Then "a" does not mean "only one" and there is no quantitative restriction. The 3rd paragraph can refer to each instance of "a model" in the scenario.

Failure at reading comprehension cannot help your case. "A model" cannot mean more than one at a time considering the context in which the section is written.


The Line of Sight rules make no mention of "one at a time". Therefore "one at a time" is not the context in which the section is written.

You are not permitted to add your context to the rules without labeling it a house rule. I will stick with the rules as they are actually written.

 Charistoph wrote:

col_impact wrote:
Cool. You have just pointed out the grammatical property known as Agreement. The 3rd paragraph starts talking about "a model" that could be a model in the way or a target model. Later references cannot repeat the indefinite article and must agree in case. You have discovered Agreement!

You really need to polish up on your grammar because you are seeing castles where there are windmills.

Incorrect. "A" does not equal "all" or "any", as you continue to profess. It means even less when considered in context from the previous sentence where it is being considered going from one model to another.

Quit pursuing your sandcastles in the sand.


If the indefinite article is being used generally to non-specifically refer to something, then it does mean"any"

Spoiler:
A and AN are called indefinite articles. "Indefinite" means "not specific". Use A(AN) when you are talking about a thing in general, NOT a specific thing.

Examples:

I need a phone. Not a specific phone, any phone
Mark wants a bicycle. Not a particular bicycle, a bicycle in general
Do you have a driver's license? In general


 Charistoph wrote:

 Charistoph wrote:


col_impact wrote:
Incorrect.

"Line of sight literally represents your warriors’ view of the enemy"

If there are several rules out there which represent your warriors' view the friendly, please point them out. To my knowledge, there are none.

This would appear to be another instance where you just make stuff up when the words on the page don't suit your needs.

Not making it up, just remembering old rules. They weren't very common. A difference.

col_impact wrote:
Can you point to one? Your foggy memory isn't an official rules resource.

Out of a hundred codices worth of material, most of which I do not have any more in any format? You jest.


Then you have failed to substantiate your argument with evidence. Simple as that.

If you cannot substantiate your argument with evidence, then your argument is invalid.

 Charistoph wrote:

So, the Politician's Answer it is. The question wasn't about the 3rd paragraph, it was about the restate mentioned above. This one here:
Spoiler:
We are told that to establish line of sight from one model to another, we draw a straight, unblocked line between the two bodies. Sometimes, certain parts of a model are all that can be seen. The model is not considered seen by only seeing these parts. This is so a model with certain impressive bits are not unduly penalized.

Tell me, where in that paragraph in that spoiler are we to be considering all models under this judgement of visibility, even though we only are seeking one?


Your spoiler isn't in the BRB. However this bit of text is in the BRB.

Spoiler:
Line of sight literally represents your warriors’ view of the enemy – they must be able to see their foes through, under or over the battlefield terrain and other models (whether friendly or enemy).
For one model to have line of sight to another, you must be able to trace a straight, unblocked line from its body (the head, torso, arms or legs) to any part of the target’s body.
Sometimes, all that will be visible of a model is a weapon, banner or other ornament he is carrying. In these cases, the model is not visible. Similarly, we ignore wings, tails and antennae even though they are technically part of a model’s body. These rules are intended to ensure that models don’t get penalised for having impressive banners, weaponry, and so on.


 Charistoph wrote:

col_impact wrote:
Incorrect. The 3rd paragraph uses "sometimes" and "these cases" to indicate that we can deal with each instance of "a model" that the rule instructs us to deal with.

That does not perfectly jell when combined with the preceding sentence. "Sometimes" is referring to a portion of the times the preceding sentence engender, due to its conjunctive nature. It is not carte blanche to ignore the singularity of the nouns being used or to treat all these times at the same time.


"Sometimes" can occur any number of times and "these cases" indicates the rule can produce multiple cases. My argument is proven, the rule allows for any instances of "a model".

 Charistoph wrote:


col_impact wrote:
 Charistoph wrote:

col_impact wrote:
For a model to be in the way it must be visibly in the way and not in the way by some other measure.

Where are we told to consider the visibility of other models under this standard?

col_impact wrote:
For a model to be blocking it must be visibly blocking and not by some other manner of blocking.

Quote, please.

The BRB uses line of sight and therefore relies on the visibility of models in the way to determine line of sight. If the models in the way are not visibly blocking line of sight then they are not blocking line of sight. Models that are considered invisible do not block line of sight. These are all simple truisms.

If you don't think it's a truism then feel free to treat all models in the way as invisible and see what happens.

So, no quote, just an assertion. The 3rd paragraph does not consider the visibility of other models when determining the visibility of a model. It is only considering the visibility of one model. It does not literally state to consider other models who are only showing non-body bits in this assertion. Other rules can insert themselves here, of course, as in "Own Unit", but this 3rd paragraph is not trying to make a determination of any models a focus is not upon. Only a model you do have a focus on.


A truism means that it is true by use of the terms involved and in this case the terms involved are those of a Line of sight system. A model in the way cannot be blocking unless it is visibly blocking. If you feel that it is not a truism then point to a case in a Line of Sight system where a model can be considered blocking without it being visibly blocking.

 Charistoph wrote:

col_impact wrote:
 Charistoph wrote:

col_impact wrote:
If a model in the way is deemed partially or wholly invisible by the BRB then that affects whether it is in the way for the purposes of line of sight.

True, by where are we told to make that differentiation here in the third paragraph of Line of Sight?[

The 3rd paragraph instructs about cases that affect the visibility of a model, whether its a target model or a model in the way.

But it does not actually state this for a model in the way. It is stating this consideration after making a point of establishing visibility from one body to another.


Incorrect. There is no mention of establishing visibility in the prior sentence.

Spoiler:
For one model to have line of sight to another, you must be able to trace a straight, unblocked line from its body (the head, torso, arms or legs) to any part of the target’s body.


What counts as "visible of a model" is entirely the concern of the 3rd paragraph.

 Charistoph wrote:

This rule is to not penalize a model for having grand bits. Why would an intervening model be penalized for having epic bits if this rule was not in place?


The 3rd paragraph provides a reason. The BRB does not need to provide an exhaustive set of reasons that covers even the corner cases. The BRB is simply providing info here and no actual rule and the info provided does a good job at covering the non-corner cases.

Further, a model with a gargantuan banner would get in trouble for Modeling for Advantage shenanigans. Even though Modeling for Advantage is not a rule in the BRB, a player that presents a model with a gargantuan banner gets in trouble with his opponent and now they must sort out a way to resolve the issue. Having the portion of the banner that exceeds the size of a normal banner provides a way to incorporate a model with a giant banner that would otherwise have to be excluded.

 Charistoph wrote:

Why put it in such a general way, when there are other ways far more clear for "any model" or "all models" in view? Why just use statements which indicate one model is being addressed?


"A" does not mean "only one".

"Sometimes" means there can be any number of times. "These cases" allows for multiple cases.

The 3rd paragraph allows for multiple instances of "a model".

 Charistoph wrote:

This is because the context of the previous sentence is still in play. The "Sometimes" is in place because it is distinguishing between "hey, I'm in the open" and "hey, I'm behind a wall".

That's not what "Sometimes" means. "Sometimes" means "at times" and can occur any number of times which debunks your interpretation.

If you don't adhere to what words actually mean then your argument is invalid.

 Charistoph wrote:

"A model" is in place because Line of Sight is not always used against a targeted model.


Incorrect. The process always involves a target model.

So long as you continue to make stuff up, your argument remains invalid.

 Charistoph wrote:

"A model" is used because not every case of trying to draw Line of Sight will result in this situation. Could it have been more specific, yes.

Was it intended to be this way? I cannot say.


Here is where you arguing intent.

I will accept what the BRB says. The BRB chose to use "a model" in a non-specific, general way.

That is the rules as they are written. If you go against the rules as they are written you are house ruling.

 Charistoph wrote:

I do honestly think that when they write these rules they rarely think of the weird situations, and their writing shows as only being directed at the here and now with no concept of fitting things together. They allow future writings to intrude on the present, with no care as to how that may coordinate with other future writings.


I think you just need to accept the rules as they are written. My line of reasoning adheres exactly to the rules as they are written and it provides a bullet-proof way for dealing with Line of Sight where many longstanding problems simply vanish.

Perhaps those longstanding issues come from people reading into the rule and not reading what the rules actually say.

Maybe you should try not reading into the rules.

This message was edited 1 time. Last update was at 2016/10/04 20:05:15


 
   
Made in us
Longtime Dakkanaut





col_impact wrote:

I think you just need to accept the rules as they are written. My line of reasoning adheres exactly to the rules as they are written and it provides a bullet-proof way for dealing with Line of Sight where many longstanding problems simply vanish.

Perhaps those longstanding issues come from people reading into the rule and not reading what the rules actually say.

Maybe you should try not reading into the rules.


No, your line of reasoning does not adhere exactly to the rules, not as long as you ignore that the model referred to as "a model" is also referred to as "the model" in the next sentence, indicating that they are referring to one model. You haven't presented a bullet proof way for dealing with Line of Sight, merely the bullet-ridden corpse of the actual rules.

Third time I've asked you to address it. Also ignored and not adequately addressed by you when Charistoph has brought it up numerous times. Think you can get around to addressing the point, or are you just going to post the same points that this refutes?

This message was edited 1 time. Last update was at 2016/10/04 20:32:25


 
   
Made in us
Longtime Dakkanaut




 doctortom wrote:
col_impact wrote:

I think you just need to accept the rules as they are written. My line of reasoning adheres exactly to the rules as they are written and it provides a bullet-proof way for dealing with Line of Sight where many longstanding problems simply vanish.

Perhaps those longstanding issues come from people reading into the rule and not reading what the rules actually say.

Maybe you should try not reading into the rules.


No, your line of reasoning does not adhere exactly to the rules, not as long as you ignore that the model referred to as "a model" is also referred to as "the model" in the next sentence, indicating that they are referring to one model. You haven't presented a bullet proof way for dealing with Line of Sight, merely the bullet-ridden corpse of the actual rules.

Third time I've asked you to address it. Also ignored and not adequately addressed by you when Charistoph has brought it up numerous times. Think you can get around to addressing the point, or are you just going to post the same points that this refutes?


You haven't been reading my posts.

From above . . .

The 3rd paragraph does not specify "only one model" so we are not restricted to one.

There is no order of operations so we are free to consider the visibility of "a model" in whatever order or all at once to our hearts desire.

The 3rd paragraph states "sometimes" and "these cases" which allows for any number of times and specifies that multiple cases can result from application of the rule.

The words in the BRB simply do not support you.
   
Made in us
Longtime Dakkanaut





col_impact wrote:
 doctortom wrote:
col_impact wrote:

I think you just need to accept the rules as they are written. My line of reasoning adheres exactly to the rules as they are written and it provides a bullet-proof way for dealing with Line of Sight where many longstanding problems simply vanish.

Perhaps those longstanding issues come from people reading into the rule and not reading what the rules actually say.

Maybe you should try not reading into the rules.


No, your line of reasoning does not adhere exactly to the rules, not as long as you ignore that the model referred to as "a model" is also referred to as "the model" in the next sentence, indicating that they are referring to one model. You haven't presented a bullet proof way for dealing with Line of Sight, merely the bullet-ridden corpse of the actual rules.

Third time I've asked you to address it. Also ignored and not adequately addressed by you when Charistoph has brought it up numerous times. Think you can get around to addressing the point, or are you just going to post the same points that this refutes?


You haven't been reading my posts.

From above . . .

The 3rd paragraph does not specify "only one model" so we are not restricted to one.

There is no order of operations so we are free to consider the visibility of "a model" in whatever order or all at once to our hearts desire.

The 3rd paragraph states "sometimes" and "these cases" which allows for any number of times and specifies that multiple cases can result from application of the rule.

The words in the BRB simply do not support you.


Okay, let's help you out - the first two sentences of the third paragraph:

"Sometimes, all that will be visible of a model is a weapon, banner or other ornament he is carrying. In these cases, the model is not visible."

From this, "these cases" means "the model" in that sentence is referring to "a model" in the sentence before. It does not extend beyond that. It does allow for any number of times and mean "the model" becomes "the models" plural at that time, or they would have said "the model(s)" Cases plural does not mean plural all at once when dealing with the singular version there. Obviiously it's referring to at different times in the game, not different models at the same time.

The English language does not support you.

This message was edited 1 time. Last update was at 2016/10/04 20:53:44


 
   
Made in us
Longtime Dakkanaut




 doctortom wrote:


Okay, let's help you out - the first two sentences of the third paragraph:

"Sometimes, all that will be visible of a model is a weapon, banner or other ornament he is carrying. In these cases, the model is not visible."

From this, "these cases" means "the model" in that sentence is referring to "a model" in the sentence before. It does not extend beyond that. It does allow for any number of times and mean "the model" becomes "the models" plural at that time, or they would have said "the model(s)" Cases plural does not mean plural all at once when dealing with the singular version there. Obviiously it's referring to at different times in the game, not different models at the same time.

The English language does not support you.


Incorrect.

There is no difference in meaning between . . .

"During a soccer match a player is allowed to score a goal" and "During soccer matches players are allowed to score goals".

Each of those sentences speaks generally (e.g., of "a player" or of "players") and so they mean the same thing. They differ only stylistically.

As has been proven numerous times in this thread "a model" does not mean "only one model". Why do you keep on insisting that it does?

Since the 3rd paragraph is referring to "a model" non-specifically and generally and there is absolutely no quantitative restriction in the sentence and "sometimes" and "these cases" allow for multiple instances of "a model", the English language fully supports me and disproves your argument.

This message was edited 2 times. Last update was at 2016/10/04 21:08:27


 
   
Made in us
Not as Good as a Minion





Astonished of Heck

col_impact wrote:
You can say whatever you want but you are not a rules resource. You can choose to play it "one at a time" but that is your house rule. The BRB makes no mention of "at a time" therefore no such procedure is being required.

The 3rd paragraph uses "sometimes" and "these cases" and "a model" refers to a non-specific, general model. The rule applies to each instance of "a model" bases on the words the BRB actually uses.

You are presenting a double standard here. You expect us to believe "a model" actually means "all models" when everything else indicates otherwise.

I never actually tried quoting the rule as "at any time" or suggested it actually stated it, but the nature of the rules combined with the context of the situation are what makes it "at a time". I have said this several times now. Try to pay attention to what other people say. At this point, I think all you do is use me as an excuse to spam the same thing over and over again hoping that people will agree with you or give up because you won't shut up about it.

col_impact wrote:
The 3rd paragraph does not specify "only one model" so we are not restricted to one.

Then demonstrate how a singular term can mean many. A generic non-specific does not mean "all at one time", it just means that it doesn't have to be a "target model" or "enemy model" or a model of any specific type.

col_impact wrote:
There is no order of operations so we are free to consider the visibility of "a model" in whatever order or all at once to our hearts desire.

Incorrect. As soon as you consider more than one model, you are considering "models" and not "a model". This is a very basic English concept here. Therefore, an order of operations would be required in order to make this "a model" consideration extend beyond the singular term.

col_impact wrote:
The 3rd paragraph states "sometimes" and "these cases" which allows for any number of times and specifies that multiple cases can result from application of the rule.

The words in the BRB simply do not support you.

Because it is not only happening once in a game. It can happen any number of times you try to determine Line of Sight. "Sometimes" does not convert "a model" to "any model" or "many models". It does not translate to doing it many times during a consideration. The consideration itself happens sometimes.

col_impact wrote:
 Charistoph wrote:

col_impact wrote:
There is no "one at a time" at all in the rule. If you feel otherwise, prove it with a rules quote or grammatical reference.

Failure at reading comprehension is no excuse. I have given it before. Look it up.

So no rules quote and no grammatical reference. Got it.

If you are unable to support what you say with rules quotes and references, your argument is invalid.

I have no desire to spam like you. Look it up.

col_impact wrote:
The Line of Sight rules make no mention of "one at a time". Therefore "one at a time" is not the context in which the section is written.

You are not permitted to add your context to the rules without labeling it a house rule. I will stick with the rules as they are actually written.

Says the one who ignores what others say and considers "a model" to mean "any model" and "all models".

col_impact wrote:
If the indefinite article is being used generally to non-specifically refer to something, then it does mean"any"

Spoiler:
A and AN are called indefinite articles. "Indefinite" means "not specific". Use A(AN) when you are talking about a thing in general, NOT a specific thing.

Examples:

I need a phone. Not a specific phone, any phone
Mark wants a bicycle. Not a particular bicycle, a bicycle in general
Do you have a driver's license? In general

Not exactly. The use of "any" I am referring to you using is "used to refer to one or some of a thing or number of things, no matter how much or man". The actual use of the indefinite "any" in this sentence would be properly "whichever of a specified class might be chosen". Bascially, you are translating "a model" to be "any model I can see" which can be just as easily be translated as "every model". This concept separates the sentence's use of singular nouns, though, so we cannot be looking at more than one for the assessment of this paragraph when it is called.

col_impact wrote:
 Charistoph wrote:

So, the Politician's Answer it is. The question wasn't about the 3rd paragraph, it was about the restate mentioned above. This one here:
Spoiler:
We are told that to establish line of sight from one model to another, we draw a straight, unblocked line between the two bodies. Sometimes, certain parts of a model are all that can be seen. The model is not considered seen by only seeing these parts. This is so a model with certain impressive bits are not unduly penalized.

Tell me, where in that paragraph in that spoiler are we to be considering all models under this judgement of visibility, even though we only are seeking one?

Your spoiler isn't in the BRB. However this bit of text is in the BRB.

Still a Politician's Answer. I didn't say what was in the spoiler was in the BRB. This was an exercise in understanding context.

I will be ignoring the rest. You tend to spam when it gets long.

col_impact wrote:
I think you just need to accept the rules as they are written. My line of reasoning adheres exactly to the rules as they are written and it provides a bullet-proof way for dealing with Line of Sight where many longstanding problems simply vanish.

Perhaps those longstanding issues come from people reading into the rule and not reading what the rules actually say.

Maybe you should try not reading into the rules.

That is a laugh considering everything you've written. I think you need to accept that your interpretation is not gospel to everyone. Very few people seem to see grammar the way you do. Very few people seem to accept your hyper-focus of one thing and ignoring the same context almost everyone else associated with things.

Are you a Wolf, a Sheep, or a Hound?
Megavolt wrote:They called me crazy…they called me insane…THEY CALLED ME LOONEY!! and boy, were they right.
 
   
Made in us
Longtime Dakkanaut





col_impact wrote:

Incorrect.

There is no difference in meaning between . . .

"During a soccer match a player is allowed to score a goal" and "During soccer matches players are allowed to score goals".

Each of those sentences speaks generally (e.g., of "a player" or of "players") and so they mean the same thing. They differ only stylistically.

As has been proven numerous times in this thread "a model" does not mean "only one model". Why do you keep on insisting that it does?

Since the 3rd paragraph is referring to "a model" non-specifically and generally and there is absolutely no quantitative restriction in the sentence and "sometimes" and "these cases" allow for multiple instances of "a model", the English language fully supports me and disproves your argument.


Ah, the soccer match example. The problem is though that it doesn't match up with what it stated in the rules. It's not "during a soccer match a player is allowed to score the goal".

You respond, yet once again "THE" model doesn't matter at all to you, "A" model has not been proven at all to not mean "only one model" in this thread when the next sentence refers to that as "THE model". The 3rd paragraph is referring to "THE model as well as "A model". Quit ignoring "THE model, as you did once again. Fourth time I bring this up. Are you going to deflect back to "a model" without adequately explaining "the model", which has been shown in this thread to mean a singular model?
   
Made in us
Longtime Dakkanaut




 Charistoph wrote:
col_impact wrote:
You can say whatever you want but you are not a rules resource. You can choose to play it "one at a time" but that is your house rule. The BRB makes no mention of "at a time" therefore no such procedure is being required.

The 3rd paragraph uses "sometimes" and "these cases" and "a model" refers to a non-specific, general model. The rule applies to each instance of "a model" bases on the words the BRB actually uses.

You are presenting a double standard here. You expect us to believe "a model" actually means "all models" when everything else indicates otherwise.


"A model" refers to each instance of "a model" that can occur "sometimes" (which can occur any number of times) to produce "these cases" (which allows for multiple instances.

 Charistoph wrote:

I never actually tried quoting the rule as "at any time" or suggested it actually stated it, but the nature of the rules combined with the context of the situation are what makes it "at a time". I have said this several times now. Try to pay attention to what other people say. At this point, I think all you do is use me as an excuse to spam the same thing over and over again hoping that people will agree with you or give up because you won't shut up about it.



There is nothing in the rules that say or suggest "at a time". No matter how many times you say this will not change that fact. Until you can prove that the rule says "at a time" I suggest you stay quiet on the issue. Reading into the rule makes your argument invalid.

 Charistoph wrote:


col_impact wrote:
The 3rd paragraph does not specify "only one model" so we are not restricted to one.

Then demonstrate how a singular term can mean many. A generic non-specific does not mean "all at one time", it just means that it doesn't have to be a "target model" or "enemy model" or a model of any specific type.



When the indefinite article is used non-specifically and generally, a singular term can mean each instance of the singular term in the situation.

You have 3 weapons on your person.
A computer terminal asks you "Do you have a weapon on your person? select yes or no"
The answer is "yes" since you have multiple instances of "a weapon" on your person.

"During a soccer match, a player is allowed to score a goal" means the same thing as "During soccer matches players are allowed to score goals"
"During a soccer match, a player is allowed to score a goal" does not mean "During only one soccer match, only one player is allowed to score only one goal"

Quit trying to sneak in a quantitative restriction where there is none. "Sometimes", "these cases", and the indefinite article used non-specifically allow for multiple instances.

 Charistoph wrote:

col_impact wrote:
There is no order of operations so we are free to consider the visibility of "a model" in whatever order or all at once to our hearts desire.

Incorrect. As soon as you consider more than one model, you are considering "models" and not "a model". This is a very basic English concept here. Therefore, an order of operations would be required in order to make this "a model" consideration extend beyond the singular term.


"Sometimes", "these cases", and the indefinite article used non-specifically allow for multiple instances. No order of operations required.

 Charistoph wrote:

col_impact wrote:
The 3rd paragraph states "sometimes" and "these cases" which allows for any number of times and specifies that multiple cases can result from application of the rule.

The words in the BRB simply do not support you.

Because it is not only happening once in a game. It can happen any number of times you try to determine Line of Sight. "Sometimes" does not convert "a model" to "any model" or "many models". It does not translate to doing it many times during a consideration. The consideration itself happens sometimes.


The context according to your standards would be a single line of sight determination between "one model to . . . another" or are you suddenly not going to adhere to your own standards?

"For one model to have line of sight to another, you must be able to trace a straight, unblocked line from its body (the head, torso, arms or legs) to any part of the target’s body."

Without a doubt the context is determining line of sight from "one model to . . . another" which is a singular occurence.

Sounds like you are being hypocritical. If you are hypocritical and inconsistent with your standards then your argument is invalid.

Of course if you actually understand grammar and its uses then an indefinite article can be used non-specifically and generally to refer to each instance.

 Charistoph wrote:


col_impact wrote:
 Charistoph wrote:

col_impact wrote:
There is no "one at a time" at all in the rule. If you feel otherwise, prove it with a rules quote or grammatical reference.

Failure at reading comprehension is no excuse. I have given it before. Look it up.

So no rules quote and no grammatical reference. Got it.

If you are unable to support what you say with rules quotes and references, your argument is invalid.

I have no desire to spam like you. Look it up.

It's not spamming when I request you to back up what you say.
What you are doing is called 'dodging' which shows there is nothing but smokes and mirrors to your argument.
So back up what you say with a rules quote or a grammatical reference.

If you are unable to support what you say with rules quotes and references, your argument is invalid.

quote=Charistoph 703020 8943483 f45b9b45c8512f4742541d2a5a08a68d.jpg]

col_impact wrote:
The Line of Sight rules make no mention of "one at a time". Therefore "one at a time" is not the context in which the section is written.

You are not permitted to add your context to the rules without labeling it a house rule. I will stick with the rules as they are actually written.

Says the one who ignores what others say and considers "a model" to mean "any model" and "all models".

col_impact wrote:
If the indefinite article is being used generally to non-specifically refer to something, then it does mean"any"

Spoiler:
A and AN are called indefinite articles. "Indefinite" means "not specific". Use A(AN) when you are talking about a thing in general, NOT a specific thing.

Examples:

I need a phone. Not a specific phone, any phone
Mark wants a bicycle. Not a particular bicycle, a bicycle in general
Do you have a driver's license? In general

Not exactly. The use of "any" I am referring to you using is "used to refer to one or some of a thing or number of things, no matter how much or man". The actual use of the indefinite "any" in this sentence would be properly "whichever of a specified class might be chosen". Bascially, you are translating "a model" to be "any model I can see" which can be just as easily be translated as "every model". This concept separates the sentence's use of singular nouns, though, so we cannot be looking at more than one for the assessment of this paragraph when it is called.


Incorrect. I am translating "a model" to be 'any instance of a model' which is appropriate to the use of the indefinite article to refer to something in a non-specific, general way.

You have 3 weapons on your person.
Do you have a weapon on your person, yes or no?
The answer is yes.
There are multiple instances of "a weapon" on your person. Until you no longer have an instance of a weapon on your person the answer will be yes.

 Charistoph wrote:

That is a laugh considering everything you've written. I think you need to accept that your interpretation is not gospel to everyone. Very few people seem to see grammar the way you do. Very few people seem to accept your hyper-focus of one thing and ignoring the same context almost everyone else associated with things.


It seems like only you and doctortom take issue.

Anyone is welcome to read through this thread and form their own opinion after witnessing how much my argument shreds yours to pieces.


Automatically Appended Next Post:
 doctortom wrote:
col_impact wrote:

Incorrect.

There is no difference in meaning between . . .

"During a soccer match a player is allowed to score a goal" and "During soccer matches players are allowed to score goals".

Each of those sentences speaks generally (e.g., of "a player" or of "players") and so they mean the same thing. They differ only stylistically.

As has been proven numerous times in this thread "a model" does not mean "only one model". Why do you keep on insisting that it does?

Since the 3rd paragraph is referring to "a model" non-specifically and generally and there is absolutely no quantitative restriction in the sentence and "sometimes" and "these cases" allow for multiple instances of "a model", the English language fully supports me and disproves your argument.


Ah, the soccer match example. The problem is though that it doesn't match up with what it stated in the rules. It's not "during a soccer match a player is allowed to score the goal".

You respond, yet once again "THE" model doesn't matter at all to you, "A" model has not been proven at all to not mean "only one model" in this thread when the next sentence refers to that as "THE model". The 3rd paragraph is referring to "THE model as well as "A model". Quit ignoring "THE model, as you did once again. Fourth time I bring this up. Are you going to deflect back to "a model" without adequately explaining "the model", which has been shown in this thread to mean a singular model?


As you might recall from basic grammar class, "he" and "the" are simply instances of Agreement with "a model". You have to use "he" or "the model" since you are still referring to "a model". "The model" refers to "a model" (the instance of "a model" we are talking about).

Shall I point you to articles on grammar and agreement?

This message was edited 1 time. Last update was at 2016/10/04 22:44:48


 
   
Made in us
Not as Good as a Minion





Astonished of Heck

col_impact wrote:
There is nothing in the rules that say or suggest "at a time". No matter how many times you say this will not change that fact. Until you can prove that the rule says "at a time" I suggest you stay quiet on the issue. Reading into the rule makes your argument invalid.

That just indicates you know nothing of the grammar of instructions or how to derive context. More hilariously, you point out why later on.

col_impact wrote:
When the indefinite article is used non-specifically and generally, a singular term can mean each instance of the singular term in the situation.

You have 3 weapons on your person.
A computer terminal asks you "Do you have a weapon on your person? select yes or no"
The answer is "yes" since you have multiple instances of "a weapon" on your person.

"During a soccer match, a player is allowed to score a goal" means the same thing as "During soccer matches players are allowed to score goals"
"During a soccer match, a player is allowed to score a goal" does not mean "During only one soccer match, only one player is allowed to score only one goal"

Quit trying to sneak in a quantitative restriction where there is none. "Sometimes", "these cases", and the indefinite article used non-specifically allow for multiple instances.

And none of your examples allows one to be considering more than one of the "a (noun)" to be happening at the same time. More importantly, if it is at the same time, you are ignoring the singular context of the statement and removing the "a" in favor of "every" or "all", and making it plural. This instruction is only about "a model", as in one. The case of "non-specific" does not get to override the quantitative nature of the words being used. It is not about many models, or any other models around when making this consideration. As soon as you start considering anything but one model, you have violated the statements of the paragraph.

But I've said that before, and even your link has said that, and you still ignore it in favor of your paradigm that is focused on the one tiny portion of a sentence. From your link:
REMEMBER: You cannot use A(AN) with plural nouns because A(AN) means "one" or "a single".

Examples:

I saw a bears in Yellowstone National Park. Not Correct
I saw bears in Yellowstone National Park. Correct

The statement indicates one model when you are making this consideration, not any around not everyone in sight, and most definitely not all. This consideration happens occasionally, so "sometimes".

But I've said it before and you've rejected it without warrant.

col_impact wrote:
"Sometimes", "these cases", and the indefinite article used non-specifically allow for multiple instances. No order of operations required.

But not at the same time. Non-specific does not translate to "any number", just "any type of, not detailed or exact". You can do it again when it comes time to consider another model for Line of Sight.

col_impact wrote:
 Charistoph wrote:

col_impact wrote:
The 3rd paragraph states "sometimes" and "these cases" which allows for any number of times and specifies that multiple cases can result from application of the rule.

The words in the BRB simply do not support you.

Because it is not only happening once in a game. It can happen any number of times you try to determine Line of Sight. "Sometimes" does not convert "a model" to "any model" or "many models". It does not translate to doing it many times during a consideration. The consideration itself happens sometimes.

The context according to your standards would be a single line of sight determination between "one model to . . . another" or are you suddenly not going to adhere to your own standards?

"For one model to have line of sight to another, you must be able to trace a straight, unblocked line from its body (the head, torso, arms or legs) to any part of the target’s body."

Without a doubt the context is determining line of sight from "one model to . . . another" which is a singular occurence.

Sounds like you are being hypocritical. If you are hypocritical and inconsistent with your standards then your argument is invalid.

Of course if you actually understand grammar and its uses then an indefinite article can be used non-specifically and generally to refer to each instance.

Your lack of reading comprehension or bizarre sense of grammar is why you think this is hypocritical. Care to actually explain what I am being hypocritical about?

If I understand what you are implying, I am saying Line of Sight checks happen numerous times during a game. During some of these checks, but not all, the only part of a model that can be seen is a bit on the list. During these times, the model cannot be considered as seeable. It is only during these Line of Sight checks that we can make this consideration, and only for one model.

col_impact wrote:

It's not spamming when I request you to back up what you say.
What you are doing is called 'dodging' which shows there is nothing but smokes and mirrors to your argument.
So back up what you say with a rules quote or a grammatical reference.

If you are unable to support what you say with rules quotes and references, your argument is invalid.

As I said, I have. You want it so badly, go look it up again. You seem to think such searches are easy and anyone can do it, so do it yourself. I am not your secretary.

col_impact wrote:
 Charistoph wrote:
Not exactly. The use of "any" I am referring to you using is "used to refer to one or some of a thing or number of things, no matter how much or man". The actual use of the indefinite "any" in this sentence would be properly "whichever of a specified class might be chosen". Bascially, you are translating "a model" to be "any model I can see" which can be just as easily be translated as "every model". This concept separates the sentence's use of singular nouns, though, so we cannot be looking at more than one for the assessment of this paragraph when it is called.

Incorrect. I am translating "a model" to be 'any instance of a model' which is appropriate to the use of the indefinite article to refer to something in a non-specific, general way.

You have 3 weapons on your person.
Do you have a weapon on your person, yes or no?
The answer is yes.
There are multiple instances of "a weapon" on your person. Until you no longer have an instance of a weapon on your person the answer will be yes.

Actually, that IS what I said you were saying. When you put "a model" in to action, you are taking the "any" as "used to refer to one or some of a thing or number of things, no matter how much or many". This puts a lie to the reminder at the bottom of your page, though, regarding quantities. The "any" in this case is "whichever of a specified class might be chosen", which is the only way to interpret it while maintaining the singular tense of the words used.

And should I bother to mention that if you have three of something, you still have one of it, but if you have one of something, you don't have three?

col_impact wrote:
It seems like only you and doctortom take issue.

Anyone is welcome to read through this thread and form their own opinion after witnessing how much my argument shreds yours to pieces.

Actually, I've talked over some of these grammar discussions with my wife whose Masters degree is about getting people to speak properly. And she does not agree with your assessments. But you have only my word on that. And you probably couldn't understand much of what I say, anyway, because my grammatical usage does not match yours. You certainly have misrepresented my statements enough to demonstrate this.

More importantly, the only person who ever actually seems to support your bizarre concepts of grammar is you.

col_impact wrote:
As you might recall from basic grammar class, "he" and "the" are simply instances of Agreement with "a model". You have to use "he" or "the model" since you are still referring to "a model". "The model" refers to "a model" (the instance of "a model" we are talking about).

Shall I point you to articles on grammar and agreement?

Articles of agreement do not allow a change in quantity measured, though. Some nouns can be used both singularly and pluralistically, such as "you" and "they", depending on the situation, and the conjugation of the verb usually indicates which is being used.

And DrT was spot on that your soccer example does not mesh with the words being used in the paragraph in question. Your soccer example is a general statement used to introduce a concept. The paragraph question is getting in to particulars and specific cases, such as the consideration of a goal attempt by a player who is out of bounds.

This message was edited 2 times. Last update was at 2016/10/05 04:41:53


Are you a Wolf, a Sheep, or a Hound?
Megavolt wrote:They called me crazy…they called me insane…THEY CALLED ME LOONEY!! and boy, were they right.
 
   
Made in us
Longtime Dakkanaut




 Charistoph wrote:
col_impact wrote:
There is nothing in the rules that say or suggest "at a time". No matter how many times you say this will not change that fact. Until you can prove that the rule says "at a time" I suggest you stay quiet on the issue. Reading into the rule makes your argument invalid.

That just indicates you know nothing of the grammar of instructions or how to derive context. More hilariously, you point out why later on.

By declining to comment on the fact that there is nothing in the rules that say or suggest "at a time" then you concede the point.

 Charistoph wrote:

col_impact wrote:
When the indefinite article is used non-specifically and generally, a singular term can mean each instance of the singular term in the situation.

You have 3 weapons on your person.
A computer terminal asks you "Do you have a weapon on your person? select yes or no"
The answer is "yes" since you have multiple instances of "a weapon" on your person.

"During a soccer match, a player is allowed to score a goal" means the same thing as "During soccer matches players are allowed to score goals"
"During a soccer match, a player is allowed to score a goal" does not mean "During only one soccer match, only one player is allowed to score only one goal"

Quit trying to sneak in a quantitative restriction where there is none. "Sometimes", "these cases", and the indefinite article used non-specifically allow for multiple instances.

And none of your examples allows one to be considering more than one of the "a (noun)" to be happening at the same time. More importantly, if it is at the same time, you are ignoring the singular context of the statement and removing the "a" in favor of "every" or "all", and making it plural. This instruction is only about "a model", as in one. The case of "non-specific" does not get to override the quantitative nature of the words being used. It is not about many models, or any other models around when making this consideration. As soon as you start considering anything but one model, you have violated the statements of the paragraph.


Incorrect. The case of the non-specific use of the indefinite article does actually override the quantitative nature of the words being used.

Spoiler:
You have 3 weapons on your person.
A computer terminal asks you "Do you have a weapon on your person? select yes or no"
The answer is "yes" since you have multiple instances of "a weapon" on your person


You have 3 instances of "a weapon" on your person all at the same time.

Spoiler:
"During a soccer match, a player is allowed to score a goal"


The above statement affects each instance of a soccer match and each instance of a player on the team all at the same time.

Whenever there is no explicit quantitative restriction and the indefinite article is used non-specifically and generally then "a" can refer to a single instance or multiple occurrences of a single instance.

 Charistoph wrote:

But I've said that before, and even your link has said that, and you still ignore it in favor of your paradigm that is focused on the one tiny portion of a sentence. From your link:
REMEMBER: You cannot use A(AN) with plural nouns because A(AN) means "one" or "a single".

Examples:

I saw a bears in Yellowstone National Park. Not Correct
I saw bears in Yellowstone National Park. Correct

The statement indicates one model when you are making this consideration, not any around not everyone in sight, and most definitely not all. This consideration happens occasionally, so "sometimes".

But I've said it before and you've rejected it without warrant.


You are confused. "A bear" is not being used in a non-specific, general way in your example above. No wonder you are struggling with the grammar since you can't keep the different uses of the indefinite article straight. The indefinite article is being used because the bear is being mentioned for the first time to the listener but it is a specific one (one that the speaker saw).

Now if the sentence read "a bear is a mammal" does that sentence only speak about one bear or bears generally? "A bear is a mammal" means the exact same thing as "Bears are mammals".

Are you beginning to see the difference? I feel that I should start charging you for the grammar lessons.

 Charistoph wrote:

col_impact wrote:
"Sometimes", "these cases", and the indefinite article used non-specifically allow for multiple instances. No order of operations required.

But not at the same time. Non-specific does not translate to "any number", just "any type of, not detailed or exact". You can do it again when it comes time to consider another model for Line of Sight.

Incorrect. The non-specific use of the indefinite allows for multiple instances. "A bear is a mammal" does not mean that there is only one bear that is a mammal.
And since "sometimes" and "these cases" allow for multiple times and these cases, it looks like the writing on the page simply does not support your argument.

 Charistoph wrote:

col_impact wrote:
 Charistoph wrote:

col_impact wrote:
The 3rd paragraph states "sometimes" and "these cases" which allows for any number of times and specifies that multiple cases can result from application of the rule.

The words in the BRB simply do not support you.

Because it is not only happening once in a game. It can happen any number of times you try to determine Line of Sight. "Sometimes" does not convert "a model" to "any model" or "many models". It does not translate to doing it many times during a consideration. The consideration itself happens sometimes.

The context according to your standards would be a single line of sight determination between "one model to . . . another" or are you suddenly not going to adhere to your own standards?

"For one model to have line of sight to another, you must be able to trace a straight, unblocked line from its body (the head, torso, arms or legs) to any part of the target’s body."

Without a doubt the context is determining line of sight from "one model to . . . another" which is a singular occurence.

Sounds like you are being hypocritical. If you are hypocritical and inconsistent with your standards then your argument is invalid.

Of course if you actually understand grammar and its uses then an indefinite article can be used non-specifically and generally to refer to each instance.

Your lack of reading comprehension or bizarre sense of grammar is why you think this is hypocritical. Care to actually explain what I am being hypocritical about?

If I understand what you are implying, I am saying Line of Sight checks happen numerous times during a game. During some of these checks, but not all, the only part of a model that can be seen is a bit on the list. During these times, the model cannot be considered as seeable. It is only during these Line of Sight checks that we can make this consideration, and only for one model.


How many straight unblocked lines are being traced in this sentence? Are we dealing with a singular line or plural lines?

"For one model to have line of sight to another, you must be able to trace a straight, unblocked line from its body (the head, torso, arms or legs) to any part of the target’s body."

Since you have decided to treat "trace a straight, unblocked line" as "trace straight, unblocked lines", you are being hypocritical and cannot even keep to your own standards.

 Charistoph wrote:

col_impact wrote:

It's not spamming when I request you to back up what you say.
What you are doing is called 'dodging' which shows there is nothing but smokes and mirrors to your argument.
So back up what you say with a rules quote or a grammatical reference.

If you are unable to support what you say with rules quotes and references, your argument is invalid.

As I said, I have. You want it so badly, go look it up again. You seem to think such searches are easy and anyone can do it, so do it yourself. I am not your secretary.


I called you out and you cannot deliver. Nothing but smoke and mirrors to your argument. I always support what I say with rules quotes and references. That's what makes my argument valid.

 Charistoph wrote:

col_impact wrote:
 Charistoph wrote:
Not exactly. The use of "any" I am referring to you using is "used to refer to one or some of a thing or number of things, no matter how much or man". The actual use of the indefinite "any" in this sentence would be properly "whichever of a specified class might be chosen". Bascially, you are translating "a model" to be "any model I can see" which can be just as easily be translated as "every model". This concept separates the sentence's use of singular nouns, though, so we cannot be looking at more than one for the assessment of this paragraph when it is called.

Incorrect. I am translating "a model" to be 'any instance of a model' which is appropriate to the use of the indefinite article to refer to something in a non-specific, general way.

You have 3 weapons on your person.
Do you have a weapon on your person, yes or no?
The answer is yes.
There are multiple instances of "a weapon" on your person. Until you no longer have an instance of a weapon on your person the answer will be yes.

Actually, that IS what I said you were saying. When you put "a model" in to action, you are taking the "any" as "used to refer to one or some of a thing or number of things, no matter how much or many". This puts a lie to the reminder at the bottom of your page, though, regarding quantities. The "any" in this case is "whichever of a specified class might be chosen", which is the only way to interpret it while maintaining the singular tense of the words used.

And should I bother to mention that if you have three of something, you still have one of it, but if you have one of something, you don't have three?


And if you have three models that are only showing wings, then you still have multiple instances of "a model" that is only showing wings. The rule allows you handle "these cases" and not just "this case" any number of times ("sometimes").

 Charistoph wrote:

col_impact wrote:
It seems like only you and doctortom take issue.

Anyone is welcome to read through this thread and form their own opinion after witnessing how much my argument shreds yours to pieces.

Actually, I've talked over some of these grammar discussions with my wife whose Masters degree is about getting people to speak properly. And she does not agree with your assessments. But you have only my word on that. And you probably couldn't understand much of what I say, anyway, because my grammatical usage does not match yours. You certainly have misrepresented my statements enough to demonstrate this.

More importantly, the only person who ever actually seems to support your bizarre concepts of grammar is you.


If your wife has something to say on the matter then by all means have her open up an account on DakkaDakka so she can school me on grammar. (Although her Masters degree in "getting people to speak properly" probably refers to speech pathology so I am not sure how her expertise on stuttering, autism, etc. will help the discussion).

Otherwise, any statements about her or hearsay attributed to her do nothing for your argument. She could be a complete fabrication on your part, and given your track record with the truth, she likely is a complete fabrication.

But hey, have her join in on the thread. After all, you have already discussed it with her, so she should be ready to chime in. She can speak for herself. Just make sure she is ready. My education on these matters exceeds hers.

 Charistoph wrote:

col_impact wrote:
As you might recall from basic grammar class, "he" and "the" are simply instances of Agreement with "a model". You have to use "he" or "the model" since you are still referring to "a model". "The model" refers to "a model" (the instance of "a model" we are talking about).

Shall I point you to articles on grammar and agreement?

Articles of agreement do not allow a change in quantity measured, though. Some nouns can be used both singularly and pluralistically, such as "you" and "they", depending on the situation, and the conjugation of the verb usually indicates which is being used.

And DrT was spot on that your soccer example does not mesh with the words being used in the paragraph in question. Your soccer example is a general statement used to introduce a concept. The paragraph question is getting in to particulars and specific cases, such as the consideration of a goal attempt by a player who is out of bounds.


You have a completely confused sense of basic grammar. Let's sort out a simple example.

"A bear is a mammal."

So, following from that statement, how many bears are mammals? Just one bear on the planet? Each instance of a bear on the planet? Any and all bears on the planet?

This message was edited 1 time. Last update was at 2016/10/05 07:26:39


 
   
Made in gr
Furious Fire Dragon





Athens Greece

What bears have to do with drop pods...?

Got milk?

All I can say about painting is that VMC tastes much better than VMA... especially black...

PM me if you are interested in Commission work.
 
   
Made in us
Lone Wolf Sentinel Pilot






 Capamaru wrote:
What bears have to do with drop pods...?


An extremely prolonged flame war where weak analogies and countless points and counterpoints have been called by both sides... it's YMDC, after all...

Revel in the glory of the site's greatest thread or be edetid and baned!
 BobtheInquisitor wrote:
Every trip to the FLGS is a rollercoaster of lust and shame.

DQ:90S++G+M+B++I+Pw40k13#+D+A++/sWD331R++T(S)DM+ 
   
Made in us
Longtime Dakkanaut





 Capamaru wrote:
What bears have to do with drop pods...?


It came about because someone made the point that door block line of sight to things behind them even if they are decorative (part of an earlier argument about whether the doors count as part of the hull or are included in the "ornamental" clause - something that neither side would be able to prove). col_impact then insisted that the doors wouldn't block line of sight to something behind them if they were ornamental because of his interpretaion of line of sight rules. And his condescension about anybody disagreeing with him not knowing basic grammar. (The most amusing thing being he's assuming that GW rules writers are using his rules of grammar, which....to be charitable, don't seem to match up with normal rules of grammar.) The regular drop pod thing has been ignored for quite a while, since it comes down to "you have to decide for your own games how to handle them"
   
Made in us
Not as Good as a Minion





Astonished of Heck

col_impact wrote:

By declining to comment on the fact that there is nothing in the rules that say or suggest "at a time" then you concede the point.

I have commented. Look it up.

col_impact wrote:
Incorrect. The case of the non-specific use of the indefinite article does actually override the quantitative nature of the words being used.

Spoiler:
You have 3 weapons on your person.
A computer terminal asks you "Do you have a weapon on your person? select yes or no"
The answer is "yes" since you have multiple instances of "a weapon" on your person


You have 3 instances of "a weapon" on your person all at the same time.

Spoiler:
"During a soccer match, a player is allowed to score a goal"


The above statement affects each instance of a soccer match and each instance of a player on the team all at the same time.

Whenever there is no explicit quantitative restriction and the indefinite article is used non-specifically and generally then "a" can refer to a single instance or multiple occurrences of a single instance.

And you fail at math and use of numbers used in grammar. If you have three, do you not have one? Your interpretation fails here because you are taking my statement of "one" to be "only one".

In addition, your example does not completely follow the exact same use as the paragraph in question.

col_impact wrote:
You are confused. "A bear" is not being used in a non-specific, general way in your example above. No wonder you are struggling with the grammar since you can't keep the different uses of the indefinite article straight. The indefinite article is being used because the bear is being mentioned for the first time to the listener but it is a specific one (one that the speaker saw).

Now if the sentence read "a bear is a mammal" does that sentence only speak about one bear or bears generally? "A bear is a mammal" means the exact same thing as "Bears are mammals".

Are you beginning to see the difference? I feel that I should start charging you for the grammar lessons.

And your reading comprehension fails. You fail to acknowledge the reminder which comes from the same source as you are using for your assertion because the examples are not specific. As bad as these examples are, they are as relevant as the ones you insist on using.

Your bad grammar lessons which ignore HUGE things about writing and English wouldn't be worth my daughter's allowance. In order to be a teacher, your lessons must be accepted.

col_impact wrote:
Incorrect. The non-specific use of the indefinite allows for multiple instances. "A bear is a mammal" does not mean that there is only one bear that is a mammal.
And since "sometimes" and "these cases" allow for multiple times and these cases, it looks like the writing on the page simply does not support your argument.

But not at the same time. As soon as you consider more than one model, it is no longer "a model", but "models". The multiple instances come later. You have multiple opportunities to perform Line of Sight Checks. For each of these checks, this consideration is for one model, and only one model.

col_impact wrote:
How many straight unblocked lines are being traced in this sentence? Are we dealing with a singular line or plural lines?

"For one model to have line of sight to another, you must be able to trace a straight, unblocked line from its body (the head, torso, arms or legs) to any part of the target’s body."

Since you have decided to treat "trace a straight, unblocked line" as "trace straight, unblocked lines", you are being hypocritical and cannot even keep to your own standards.

Ah, I see now. You cannot read, that's the problem.

You took the, "It can happen any number of times you try to determine Line of Sight", to mean that I was trying to draw multiple lines during the same check. This is not true. You are confusing an attempt at the process with the key step of the process. In programming terms, I am speaking of the number of times you use the Function, not the number of times a line is showing in a Function. Do you understand the difference?

And the third paragraph is an instruction line in that Function, and it is only looking at one model when it does its check.

col_impact wrote:
I called you out and you cannot deliver. Nothing but smoke and mirrors to your argument. I always support what I say with rules quotes and references. That's what makes my argument valid.

Do not confuse "cannot" with "will not". And considering that half of what you do is spam the same thing over and over again, even when it is not accepted, is what I am trying to avoid.

col_impact wrote:
And if you have three models that are only showing wings, then you still have multiple instances of "a model" that is only showing wings. The rule allows you handle "these cases" and not just "this case" any number of times ("sometimes").

At which point you fail at grammatical math without supporting instructions in game relevance. You are using a real world instance to support something that is not mentioned in game. But I've said that before, haven't I?

col_impact wrote:
 Charistoph wrote:
And DrT was spot on that your soccer example does not mesh with the words being used in the paragraph in question. Your soccer example is a general statement used to introduce a concept. The paragraph question is getting in to particulars and specific cases, such as the consideration of a goal attempt by a player who is out of bounds.

You have a completely confused sense of basic grammar. Let's sort out a simple example.

"A bear is a mammal."

So, following from that statement, how many bears are mammals? Just one bear on the planet? Each instance of a bear on the planet? Any and all bears on the planet?

And you ignored what was said to repeat the same thing again. You have stated the examples regarding quantitative syntax do not apply, but neither does your soccer or bear example above.

Are you a Wolf, a Sheep, or a Hound?
Megavolt wrote:They called me crazy…they called me insane…THEY CALLED ME LOONEY!! and boy, were they right.
 
   
Made in us
Longtime Dakkanaut




Charistophe,

You ignored a simple question.

"A bear is a mammal."

So, following from that statement, how many bears are mammals? Just one bear on the planet? Each instance of a bear on the planet? Any and all bears on the planet?


Oh and doctortom feel free to let us know your thoughts as well on the matter.

This message was edited 2 times. Last update was at 2016/10/05 19:42:12


 
   
Made in us
Longtime Dakkanaut





col_impact wrote:
Charistophe,

You ignored a simple question.

"A bear is a mammal."

So, following from that statement, how many bears are mammals? Just one bear on the planet? Each instance of a bear on the planet? Any and all bears on the planet?


Oh and doctortom feel free to let us know your thoughts as well on the matter.


Is the statement "A bear is a mammal" followed by the statement "The bear may try to steal my picnic basket"? Because if not, this is entirely irrelevant as you are ignoring something to indicate that it is singular. Quit being ridiculous with useless analogies, please.

This message was edited 2 times. Last update was at 2016/10/05 20:42:39


 
   
Made in us
Longtime Dakkanaut





 Capamaru wrote:
What bears have to do with drop pods...?


It's an insult to Merica. Obviously this ties back to the right to bear arms, and how Americans want drop pod doors to blow off and not be part of the hull as if they were an armament because Americans ha e the right to blow things up. However GW is British and their use of blow off in the fluff which is RAI and the new CEO direction for the RAW in 40k is more of another type of blow reference in regards to quality of find a spot to DS onto the table with a huge footprint from the deployed doors counting as hull which is really just the brits calling out the yanks for not knowing the queens English.

So yes, bears.

This message was edited 1 time. Last update was at 2016/10/05 20:51:30


 
   
Made in us
Longtime Dakkanaut





blaktoof wrote:
 Capamaru wrote:
What bears have to do with drop pods...?


It's an insult to Merica. Obviously this ties back to the right to bear arms, and how Americans want drop pod doors to blow off and not be part of the hull as if they were an armament because Americans ha e the right to blow things up. However GW is British and their use of blow off in the fluff which is RAI and the new CEO direction for the RAW in 40k is more of another type of blow reference in regards to quality of find a spot to DS onto the table with a huge footprint from the deployed doors counting as hull which is really just the brits calling out the yanks for not knowing the queens English.

So yes, bears.


Actually, with col_impact's ananalogies, I don't think it's so much the right to bear arms as it is the right to arm bears. Mabye ursoids are going to be a new army released for 40k (possibly even cyber-bears) and he's trying to subtly warn us about it in the rules forum.
   
Made in us
Longtime Dakkanaut




 doctortom wrote:
col_impact wrote:
Charistophe,

You ignored a simple question.

"A bear is a mammal."

So, following from that statement, how many bears are mammals? Just one bear on the planet? Each instance of a bear on the planet? Any and all bears on the planet?


Oh and doctortom feel free to let us know your thoughts as well on the matter.


Is the statement "A bear is a mammal" followed by the statement "The bear may try to steal my picnic basket"? Because if not, this is entirely irrelevant as you are ignoring something to indicate that it is singular. Quit being ridiculous with useless analogies, please.


So you refuse to answer a simple question. Your argument is invalid then.

"A bear is a mammal" applies generally to each instance of a bear on the planet since "a bear" is being used in a non-specific, general way (and that is true in any context). Similarly, as already proven, the 3rd paragraph uses the indefinite article in "a model" in a non-specific general way. When something applies generally, it applies to each case. That is literally what 'generally' means.

This message was edited 1 time. Last update was at 2016/10/05 23:36:22


 
   
Made in us
Not as Good as a Minion





Astonished of Heck

col_impact wrote:
Charistophe,

You ignored a simple question.

"A bear is a mammal."

So, following from that statement, how many bears are mammals? Just one bear on the planet? Each instance of a bear on the planet? Any and all bears on the planet?

You ignored the point of what you quoted to repeat the same error while doing so? Awesome. You should run for public office. Nobody understands half of what they write anyway.

"One bear is one mammal". Adds up to me.

Your confusion lies in you conflating the "one" use of "a" with "only one". You keep adding "only" to it in your counter-assertion, even if unstated, when none of us have been actually applying it anywhere.

In addition, this is not used in the same context as the paragraph in question. For example, "I have shot a gun". Have I shot all guns available? Nope, that would be taking the comment out of context. Any number of guns? Possibly, but no way to tell without additional information. All that can be confirmed is that one gun is was shot. The type of gun is not specified, but that doesn't mean I shot all the same ordinance used by both invaders and defenders on Omaha Beach just because I was at Omaha Beach once.

Moving along with the same example, from the Shooting Phase:
First, select a weapon that one or more models in your unit are equipped with. The selected weapon cannot be one that the unit has shot with during this phase. All models in the unit that are equipped with the selected weapon can now shoot at the target unit with that weapon.

By using your assertion regarding "a model" in Line of Sight's third paragraph, I can use any number of weapons the model is equipped with in this instruction because I am to select non-specified weapon.

This message was edited 1 time. Last update was at 2016/10/06 02:06:23


Are you a Wolf, a Sheep, or a Hound?
Megavolt wrote:They called me crazy…they called me insane…THEY CALLED ME LOONEY!! and boy, were they right.
 
   
Made in us
Longtime Dakkanaut




 Charistoph wrote:
col_impact wrote:
Charistophe,

You ignored a simple question.

"A bear is a mammal."

So, following from that statement, how many bears are mammals? Just one bear on the planet? Each instance of a bear on the planet? Any and all bears on the planet?

You ignored the point of what you quoted to repeat the same error while doing so? Awesome. You should run for public office. Nobody understands half of what they write anyway.

"One bear is one mammal". Adds up to me.


"A bear is a mammal" does not mean the same thing as "one bear is one mammal". Saying "one bear is one mammal" would require you to be counting the number of mammals for some odd reason (maybe for an audit for a mammal exhibit?) You are required to produce sentences exactly the same in meaning.
In fact, "a bear is a mammal" means exactly the same as "Bears are mammals". Both sentences speak generally about bears.

 Charistoph wrote:

Your confusion lies in you conflating the "one" use of "a" with "only one". You keep adding "only" to it in your counter-assertion, even if unstated, when none of us have been actually applying it anywhere.

Nope. You keep adding a restriction to the use of the indefinite article where there is none. The indefinite article in the 3rd paragraph is being used to refer to something in a non-specific general way. Generally means it applies to each case of "a model" in the situation.

 Charistoph wrote:

In addition, this is not used in the same context as the paragraph in question. For example, "I have shot a gun". Have I shot all guns available? Nope, that would be taking the comment out of context. Any number of guns? Possibly, but no way to tell without additional information. All that can be confirmed is that one gun is was shot. The type of gun is not specified, but that doesn't mean I shot all the same ordinance used by both invaders and defenders on Omaha Beach just because I was at Omaha Beach once.


Your example underscores that the case of more than one gun still satisfies "a gun". If I have shot 3 guns I have shot a gun. If I have 3 guns on my person I have a gun on my person.

However, shooting a gun is not really the general case of the indefinite article.

But lets bring it closer to the situation in the 3rd paragraph which is the general case of the indefinite article.

Consider:
'Sometimes, all that will be visible of a gun on your person is its holster. If this is the case it cannot fire'

If you have 3 guns on your person that match that criteria all 3 will not be able to fire. "A gun" will generally refer to any that match the criteria since it refers to each case of a gun on your person.

By your read, if you have 3 guns on your person that match that criteria only one gun will not be able to fire and the other 2 will be able to fire even though only their holsters are visible.

In order to actually say that the rule would have to read . . .
"Sometimes, all that will be visible of a gun on your person is its holster. If this is the case, choose one gun on your person that only has its holster visible. That gun cannot fire."

 Charistoph wrote:

Moving along with the same example, from the Shooting Phase:
First, select a weapon that one or more models in your unit are equipped with. The selected weapon cannot be one that the unit has shot with during this phase. All models in the unit that are equipped with the selected weapon can now shoot at the target unit with that weapon.

By using your assertion regarding "a model" in Line of Sight's third paragraph, I can use any number of weapons the model is equipped with in this instruction because I am to select non-specified weapon.


Incorrect. Select means to choose out of a group so that is a quantitative case of the indefinite article (e.g. pick a tool, choose a color, buy a loaf, kick a toad). We are not dealing with the general use of the indefinite article but simply a different use ("to refer to one of a group").


This message was edited 5 times. Last update was at 2016/10/06 03:54:34


 
   
Made in us
Not as Good as a Minion





Astonished of Heck

col_impact wrote:
"A bear is a mammal" does not mean the same thing as "one bear is one mammal". You are required to produce sentences exactly the same in meaning.
In fact, "a bear is a mammal" means exactly the same as "Bears are mammals". Both sentences speak generally about bears.

Oh, I thought we were supposed to be answering the questions we wanted to answer here. After all, that's what you've been doing for some time now.

col_impact wrote:
Nope. You keep adding a restriction to the use of the indefinite article where there is none. The indefinite article in the 3rd paragraph is being used to refer to something in a non-specific general way. Generally means it applies to each case of "a model" in the situation.

Incorrect. I am using all of the aspects as they apply appropriately.

col_impact wrote:
Your example underscores that the case of more than one gun still satisfies "a gun". If I have shot 3 guns I have shot a gun. If I have 3 guns on my person I have a gun on my person.

But without further context to make it more than one, such as mentioning having 3 guns available, I have no reasonable expectation to extend it past one. And that is the point you keep missing.

col_impact wrote:
 Charistoph wrote:

Moving along with the same example, from the Shooting Phase:
First, select a weapon that one or more models in your unit are equipped with. The selected weapon cannot be one that the unit has shot with during this phase. All models in the unit that are equipped with the selected weapon can now shoot at the target unit with that weapon.

By using your assertion regarding "a model" in Line of Sight's third paragraph, I can use any number of weapons the model is equipped with in this instruction because I am to select non-specified weapon.

Incorrect. Select means to choose out of a group so that is a quantitative case of the indefinite article (e.g. pick a tool, choose a color, buy a loaf, kick a toad). We are not dealing with the general use of the indefinite article but simply a different use ("to refer to one of a group").

It is still a generic, non-specific weapon we are addressing. Your assertion is that "generic non-specic" use allows for any and all to do this during the same event.

Are you a Wolf, a Sheep, or a Hound?
Megavolt wrote:They called me crazy…they called me insane…THEY CALLED ME LOONEY!! and boy, were they right.
 
   
Made in us
Longtime Dakkanaut




 Charistoph wrote:

col_impact wrote:
Your example underscores that the case of more than one gun still satisfies "a gun". If I have shot 3 guns I have shot a gun. If I have 3 guns on my person I have a gun on my person.

But without further context to make it more than one, such as mentioning having 3 guns available, I have no reasonable expectation to extend it past one. And that is the point you keep missing.


The situation of the 3rd paragraph will often involve more than one model as I have shown. In fact, without models in the way you have nothing that could be visibly blocking line of sight and therewith no need to actually check line of sight. What would block line of sight in the case of only a target model? The table? At any rate, it's pretty clear that the situation will often involve more than a target model and that is all that matters to our discussion.

 Charistoph wrote:

col_impact wrote:
 Charistoph wrote:

Moving along with the same example, from the Shooting Phase:
First, select a weapon that one or more models in your unit are equipped with. The selected weapon cannot be one that the unit has shot with during this phase. All models in the unit that are equipped with the selected weapon can now shoot at the target unit with that weapon.

By using your assertion regarding "a model" in Line of Sight's third paragraph, I can use any number of weapons the model is equipped with in this instruction because I am to select non-specified weapon.

Incorrect. Select means to choose out of a group so that is a quantitative case of the indefinite article (e.g. pick a tool, choose a color, buy a loaf, kick a toad). We are not dealing with the general use of the indefinite article but simply a different use ("to refer to one of a group").

It is still a generic, non-specific weapon we are addressing. Your assertion is that "generic non-specic" use allows for any and all to do this during the same event.


Ah, you are confusing "generic" with "general". No wonder your understanding is all out of whack.

"Generic" means non-specific so it's redundant to say along with non-specific.

"General" means "applying to each case" or "applying to all cases".

Ah, so that's it! Your understanding of the meaning of the words involved is a little off. Cool, so once you adjust your understanding to "general" and not "generic" you should be right on track.


Also, you didn't really comment on this.

col_impact wrote:
But lets bring it closer to the situation in the 3rd paragraph which is the general case of the indefinite article.

Consider:
'Sometimes, all that will be visible of a gun on your person is its holster. If this is the case it cannot fire'

If you have 3 guns on your person that match that criteria all 3 will not be able to fire. "A gun" will generally refer to any that match the criteria since it refers to each case of a gun on your person.

By your read, if you have 3 guns on your person that match that criteria only one gun will not be able to fire and the other 2 will be able to fire even though only their holsters are visible.

In order to actually say that the rule would have to read . . .
"Sometimes, all that will be visible of a gun on your person is its holster. If this is the case, choose one gun on your person that only has its holster visible. That gun cannot fire."


I would like your thoughts on the above.

This message was edited 2 times. Last update was at 2016/10/06 04:17:04


 
   
 
Forum Index » 40K You Make Da Call
Go to: