“smart ass” answer? Oh so when I give a logical and reasoned counter to how I didn’t actually call you a pillock, you simply dismiss it without any logical reasoning of your own and call my response a smart ass smokescreen yet remaining with your misinterpretation. You’re trying to excuse yourself out of the debate in a very disingenuous manner. Nice try! What you’ve done is make it obvious that you’ve feigned a hissy fit to try and cover up what are obviously ill-thought through arguments, thereby confirming my suspicion that you’re sincere. You possibly started out sincerely but I think you’re far too invested into your nickname, or your belief of it, to be able to let go. ------------- I have not changed any premise, the “solo” ‘no server’ issue is what I was talking about for all along (since 22 Nov); I was very clear about it, how on earth you misinterpreted that is entirely your own doing. So how did you interpret “no server and no connection”? Do you even understand these terms? Despite my subsequent repeated attempts to get you to walk through what happens when no distant server is involved (and I gave good reasoning to accept that it remains uncountered), you still keep wanting to continue on the premise that that one has to be involved; you can’t even see which of us brought this nonsense into the argument. Moreover * You still haven’t acknowledged the difference between controlling and major/prominent. * You don’t know the difference between implied and inferred * You didn’t acknowledge the different between “don’t” and “not always” * And you still don’t know the difference between pillock and inept (I guess someone who is both won’t understand that difference anyway). -------------- The simple fact is that your two statements I quoted are contradictory and not reconcilable “will not on average be equal simply to the sum …//... [on average they *will* simply be additional] this is nothing more that I have been saying all along” – you’ve been tripped up. You also don’t know what goes on technically because “[you] was of the belief …” of something; belief is not understanding, it’s merely faith. You clearly don’t have any idea how connectivity works (not that it was relevant anyway). You don’t even understand where the data from an outgoing connection is queued (it’s not queued at the output port, I would have thought my example would have made that really, really obvious – again, nevermind). And you have the nerve to call yourself technogiant! – Demonstrating another contradiction by quoting two of your other statements: “the latencies actually don't add together” “I can agree with your comment that latencies obviously do add together,to slow the system”. Perhaps we should have stopped there, but your silly pillock comment got me to bite again. ------------ Your last post was nothing but hot air – where was the substance? The real difference between you and I is that I have substantiated myself with all the appropriate quotes for comparison and explained the (dis)connecting logic, whereas you merely made claims. You had turned this argument into a battle of wits for which you were gravely prepared; you are indeed as per your “choice of word”. You’ll probably want to think twice before repeating such silly arguments in future ;c)

So just like I stated...you again have denied the meaning that you implied....I really have to congratulate you on your extraordinary ability to twist the truth beyond all recognition, and never answer the question as posed...you should have been a politician. I have interpreted you comments/insults in the same way as most people would...so trying to deny that you intended them that way with a "smart ass" answer does not alter anything..........
As regards the argument, instead of answering the last question I posed you have now changed the entire premise on which the discussion was based....we were discussing our gamers pc being used in an online game but were restricting our discussion to the gamers pc side of things......you are now trying to change the argument and talk about a pc being used for a "solo" game so as too avoid answering my question about the outgoing port.....Yet another contrived misrepresentation of the situation to avoid answering the question and protect your argument from the ultimate conclusion that you are incorrect.
In fact that really describes your behaviour throughout this discussion...all this to be honest contorted nonsense about me misrepresenting you has been nothing but a smoke screen to avoid answering the questions I've asked and the obvious conclusion of the examples I've posed..that conclusion being that my original comment was in fact correct.
I'm going to conclude with this reply now for several reasons.....firstly one of us has to be adult and big enough to allow the other the last word otherwise this will go on forever....so you can go ahead and say what you wish, I'm confident that most readers will see the truth even through your contorted and contrived web of nonsense.....secondly I see little point in discussing anything with someone such as yourself when you continually deny the obvious meaning of what you have said, try to change and deny the entire basis of the discussion just to facilitate your argument and uses a contrived smoke screen whereby you deliberately misinterpret the situation in order to accuse me of misrepresenting you.....Finally I think that the way you have conducted yourself has shown you to be a conceited and condescending individual who is much more worthy of the insults you have used than I...at least my insults are direct,not implied, at least I mean and stand by what I say.

My replies are very consistent; you keep misrepresenting them to make them appear inconsistent. Being “immature at both IT and critical thinking” does not make one a pillock, it actually makes one immature at both IT and critical thinking, nothing more (and not generally inept). My parents are immature at both IT and critical thinking yet I know they are not pillocks or inept people. They don’t know a strawman argument when they hear one – obviously I do (and did, taa daaaa!) However, if I were to explain this to them and they didn’t want to accept it without countering ...... There is a great difference between not being proficient at something, and generally being foolish/simpleton/annoying, hence being inept at something does not make one a pillock (unless you want to disagree with the dictionary definitions). “so you are again calling me inept or a pillock” you may have well said “so you are again calling me inept or a pillock or a terrorist, or a murderer” and it would have been just as correct (the point being the cunning use of the OR statement). I don’t know why you keep remaining with the term pillock; if you continue to do so then again it will be difficult to conclude you aren’t. ----------

“I was of the belief that outgoing data was queued on the gamers pc” to repeat myself: what “outgoing port”? Don’t forget that we’re restricting the argument purely to our gamers PC (no connection, no server). How/where/why would such data be queued if there are no connections (nor server) to consider? ------------

Repeating something else: "but that increase will not on average be equal simply to the sum of the speed increases" Excuse me, are you disagreeing with my statement that the average increase of system response time (slow connection or not) is not equal to an additional display lag? I thought you had already agreed that it is? ""Statistically, the latencies *will* add; on average they *will* simply be additional. Also the latencies *will* tend towards adding even for individual packets." this is nothing more that I have been saying all along". So what are you saying? You could not be any less clear!

Well at least we do agree about something...even if it is only the new web site "comment" format is not so good......
Other than that I'm afraid I find your replies quite inconsistent....you start by making a comment that quite clearly implies to anyone that reads it, and is not simply a matter of me misinterpreting it, that I am "inept" or to use my choice of word a "pillock".....you then deny that you implied this but then proceed to imply the same thing and I quote..."IMO, this has made you look very immature at both IT and critical thinking (and perhaps even in the English language); my error is that I indulged you"......so you are again calling me inept or a pillock, just insert whatever derogatory term you choose.....I'm sure you will again deny this was you intention....please if you intend to insult me at least have the courage of your convictions and don't subsequently deny it.
Back to the main argument and I wanted to pick up on something you said...."How can we consider an outgoing queued data stack on the PC when we’re bounding the argument to purely the gamers PC? Even if there was a slow connection involved, there wouldn’t be a data stack on the PC anyway"....I am quite willing to admit if I have misconstrued what is happening here. I was of the belief that outgoing data was queued on the gamers pc...does it not happen that way? What happens if the pc generates more outgoing data than the outgoing port can deal with? It must be stored or queued somewhere on the users pc?

"but that increase will not on average be equal simply to the sum of the speed increases" Excuse me, are you disagreeing with my statement that the average increase of system response time (slow connection or not) is not equal to an additional display lag? I thought you had already agreed? "Statistically, the latencies *will* add; on average they *will* simply be additional. Also the latencies *will* tend towards adding even for individual packets." this is nothing more that I have been saying all along". What are you saying here? You could not be any less clear!

I don’t know about you but I hate the new formatting of the comments section of the revamped website. I’ve added some “----------“s to split sections to hopefully ease reading. ----------- I’m not backtracking, its not like I called you a pillock or anything like that. It is you who has had to be guided into interpreting my supposed ‘insult’ in the correct manner. I’ve given my explanation; if you can’t counter it then don’t waste further time with what is essentially a diversion. If this really is so significant to you then I advise you to refrain from visiting forums in general - the ‘insult’ here is extremely mild by comparison with what normally goes. -------------- “"removed" implies a deliberate act” No it doesn’t; it could well have been accidental or an oversight resulting from the new website structure (of course, neglect could be considered deliberate if one dug deep enough). Yes everything happens for a reason, but it doesn’t necessitate intent. My wording in this sole case could well be misinterpreted, the difference between your input and my own is that the repeated occurrences of yours is clearly misrepresentation i.e. I supposedly called you a pillock – that’s a direct implication. I can’t prove intent or lack of, nor is it necessary. This sole one of mine can only misinterpreted - if you insist in reading too much into it i.e. your misinterpretation is only inferred. There is a big difference between implied and inferred. IMO, this has made you look very immature at both IT and critical thinking (and perhaps even in the English language); my error is that I indulged you.------------“and considering the speed increases that were being discussed were so small, then they were almost completely irelevant" Yes this was one of your points and I had agreed with that one from my very first post, but this wasn’t the one I was questioning. ---------“please don't split hairs” you call highlighting assumptions critical to the argument splitting hairs? “If you wish you can restrict my argument purely to our gamers PC and its outgoing queued data stack on that pc” wait a minute, that’s illogical. How can we consider an outgoing queued data stack on the PC when we’re bounding the argument to purely the gamers PC? Even if there was a slow connection involved, there wouldn’t be a data stack on the PC anyway – the slow connection forms the delay (holds the transit data in cyberspace), not the connection going to it. Are you seriously telling the world that long pings result with outgoing data queuing at the PC? An example: if I’m sending data to the sun, will it still remain resident at my PC for 8.5 minutes, then somehow mysteriously disappear immediately after when the data in transit has reached the sun? If so then I think I now know why you have misunderstood how all this works. Regardless, if we’re restrict the argument purely to our gamers PC then there is no network connection to consider (and hence no external server). So where is that stack and where is that outgoing data going? How large (and temporally random) can that queue be? As you see, resolution of your assumption does everything to forward this argument.-----------------“I fully accept you have superior knowledge in your chosen field” as I do yours (of course), but my chosen field (the one at degree level), as well as my additional cross subject expertise, is just a little bit more relevant to this debate than the “qualification” (not necessarily a degree) in chemical kinetics; so which field has the advantage in this case? Answers to be queued in one’s mind, sorry – I mean lost in cyberspace…

As regards the "Pillock" comment...well of course as your comment was implied rather than specific then it's meaning is subjective and open to interpretation of the reader...while you seem now to be back tracking on the strength of your comment which started almost immediately with the comment "W.T.F." I believe that most would have come to the same interpretation as I have.
As regards server,network connection, gamers pc and who is discussing what...please don't split hairs, it does nothing to forward the argument.
If you wish you can restrict my argument purely to our gamers PC and its outgoing queued data stack on that pc....this does not affect my argument in any way.
As regards your comment.."I notice your last response was removed"..well it was not accurate, more than my last comment were lost, several of your and mine were lost. Also your use of the word "removed" implies a deliberate act for some specific reason...only a minor consideration of course but as you seem to be concerned with examining my wording in such detail perhaps you should apply the same critical eye to your own comments to save misinterpretation of your meaning.(seems to be a common thread here)
Regardless of all this tit for tat nonsense which to be quite honest is starting to make us both look like children in the playground and not becoming of people of our age and academic achievement I think we are finally agreed that any speed increases will add together in the first instance, but that due to the way the gamers pc outgoing queued data stack is dealt with then on average this will lead to an increased speed of response but that increase will not on average be equal simply to the sum of the speed increases.
I accept that my initial comment needed further clarification and have been trying to explain my meaning ever since.It was simply an attempt to convey in a couple of sentences that the speed increases will not add together in a simple 1+1 =2 fashion and considering the speed increases that were being discussed were so small, then they were almost completely irelevant...and I certainly have not been disingenuous and have not tried deliberately to misrepresent you.
One final thing, your comment..."Chemical kinetics trumps communication electronics every time within an IT technical discussion doesn’t it. "....well of course I fully accept you have superior knowledge in your chosen field....but as is also widely accepted an outside view and cross subject expertise can often prove to be advantageous.

I did not call or imply you are a pillock; only you inferred it. My clear implication is that you are not as your nick suggests: ‘technogiant’ i.e. not as good at IT as the nick suggests, as opposed to being a fool or outright stupid (in general). Not being a giant at something doesn’t automatically make one inept; other categories exist inbetween (good, average, mediocre, poor). Pillock is a general term; my implication was obviously one bound purely within IT. However, if you still don’t get these differences then it will be difficult to avoid the perception that you are indeed as you erroneously claimed my latter description of you.

“I have restricted my discussion to our gamers pc as this is what we are discussing.”
You first said said ” connection lag”, then “network connection is the bottleneck” in a next comment, before anyone else mentioned any related relevance to these at all. Why should an “outgoing data queue” factor into this when there isn’t a means for a queue?

“So the outcome of the model that I am suggesting is correct” you are now at risk of contradicting yourself. If you insist on standing by your statement of “I have restricted my discussion to our gamers pc as this is what we are discussing.” then we can completely disregard all you stated about network connections and connection latencies and you will have to accept that without this large, unpredictable, external lag, the latencies do add.

How on earth you call “I compiled this just before the site shut down for its revamp. I may as well post it. I notice your last response was removed.” as “misrepresentation”? What exactly was misrepresented? I made no implication as to the reason of the removal of the posts, merely an observation for anyone else who is suffering our ping-ping as to why my last response appears as being out of the blue (I would have thought this was really obvious – nevermind). You can accuse me of misrepresentation all you want, but it won’t mean fudge if you can’t substantiate your claim.

“you imply that other contributors are in agreement with you” I didn’t imply it, I outright claimed it - well it IS pretty obvious (especially if we’re limiting ourselves only to what happens only at the gamers PC).

“So the outcome of the model that I am suggesting is correct is that the speed increases will not lead to a slowing of the system and will most of the time speed it up but not by as much as if you simply added together the individual speed increases”
On average it will; that’s what people care about, as opposed to worrying about individual strays (or the standard deviation if you will) within a very large group – assuming there is a large, unrelated, random additional factor, but we can’t assume that if we’re restricting this discussion to our gamers pc. How does: “Statistically, the latencies *will* add; on average they *will* simply be additional. Also the latencies *will* tend towards adding even for individual packets. …. [if there is] no server, the latencies will indeed add " equate to “the latencies actually don't add together”?

“I could of chosen the wording more carefully” Yes you should have, would you not agree a subsequent clarification would have saved you an awful lot of effort? It seems quibbling over the wording makes all the difference. You might want to reconsider a lot of your wording! Chemical kinetics trumps communication electronics every time within an IT technical discussion doesn’t it.

I was not quoting you and while you didn't use the word "pillock" that was your implied meaning...so you are incorrect to say this is a misrepresentation and as you are so keen to try and accuse me of misrepresenting you, I now accuse you of misrepresenting the situation...I have not misrepresented you at all at any point...this is just another example of you deliberately misinterpreting the situation in order to accuse me of misrepresenting you.
You have further misrepresented the situation in respect of the "server" discussion...it is you not me that has been discussing server involvement, remember your comments in relation to quantization effect?...I have restricted my discussion to our gamers pc as this is what we are discussing. So this again is another example of you deliberately misinterpreting the situation and then accusing me of misrepresenting you.
Another misrepresentation of the situation was your comment "I notice your last response was removed."..well some may well interpret that to mean there was something wrong with it and it was deliberately removed...which is not the situation, as you seem to pay much attention the the words used rather than to their meaning you would more correctly have have said that my last reply together with several others (both yours and mine) were lost (rather than removed as that implies a deliberate act) when the site was revamped.
Your scrutiny of my use of the word "don't" when I said that latencies don't add together at the beginning of this discussion ....well perhaps I could of chosen the wording more carefully, if I thought it was going to be so closely scrutinized...but as it was at that time I was merely trying to say something in a quick sentence or two rather than going into the in's and out's that we subsequently have.
I think you will find that since I have started to elaborate I have said and repeatedly said the same thing.
That is that the speed advantages on the faster system, be that due to faster screen refresh rate, faster gamers response etc do add together in the first instance but that due to the way in which the outgoing data queue is dealt with then that speed advantage will not be preserved so simply adding together all the speed advantages as you have suggested is correct does not give an accurate representation of the final outcome. So the outcome of the model that I am suggesting is correct is that the speed increases will not lead to a slowing of the system and will most of the time speed it up but not by as much as if you simply added together the individual speed increases.(hence the reason for me saying they don't add together as they do not in a simple mathematical fashion).You seem to now be agreeing with this and I quote...". Statistically, the latencies *will* add; on average they *will* simply be additional. Also the latencies *will* tend towards adding even for individual packets." this is nothing more that I have been saying all along and I quote from a comment of mine dated 22/11/08...."In effect reducing the number and size of the latencies prior to the connection will obviously not make the process any slower but may or may not make it any quicker depending on the queued data traffic at the time".
Simply remove where I say "may or may not" with what you say "tend towards adding" and you have the same meaning.
So why could you not agree with this point earlier and post it now as if some new and sudden enlightement? That is all that I have been trying to say.
Just a few further points...you imply that other contributors are in agreement with you...again not a strictly true representation as since we have started discussing this in any depth it has only been you and me involved.
Also your comment about my "qualification"...well mine like yours is an hons degree but I'm not into certificate waving...and I still think if you were more familiar with chemical kinetics you would have more quickly been able to see the argument I was making, and that what I was saying was correct as you have finally agreed bar your quibbling over wording.

I did not 'call you a pillock', or anything else for that matter; such a claim that I did is another clear occurrence of misrepresentation of my statements and behaviour. Also, "not always" is not the same as "don’t". I can’t help but feel that this could be the lesson we take from this!

I've only briefly read your reply Steve...it all seems rather pointless now a good proportion of our replies have been lost.
I think I can agree with your comment that latencies obviously do add together,to slow the system, I think we are also in agreement that this will not always be in a strictly 1+1=2 fashion due to the outgoing connection and the way it operates.This was the reason for my initial comment.
I really don't see the point in laboring on with this...perhaps we can both take a lesson from this.
Firstly perhaps I should have explained the reasoning I have for making my comment in the first place.
Secondly perhaps you should have questioned my reason for my comment before calling me a pillock...
Anyway as I said it's all a bit of a storm in a tea cup.

I compiled this just before the site shut down for its revamp. I may as well post it. I notice your last response was removed.

I am well known to bite and bite hard if I suspect someone is behaving disingenuously; misrepresentation of statements are usually a good givaway, something your done a couple of times. It may not be the best attribute but it’s one I’m comfortable with. Anyway, to respond to the facts:

“You are discussing what happens on the server side” you brought up what happens at the server side, not me; I only showed how arguing what happens there is irrelevant. This leads to a critical point: you made the assumption that there is a server involved; the original comment you responded to didn’t imply use of a server (a ‘cyber battlefield’ needn’t make use of internet connectivity); your argument is redundant if there is no server – the latencies will indeed add therefore nulling your original comment – you r argument is lost straight away. Even if there is a separate server involved: if the game machine is near the server (assuming the server isn’t overloaded) then your argument still tends towards being redundant (the connection need not be the dominant factor). Furthermore, players won’t want to play on a server where the connection is the dominant delay factor (they will be killed by those closer to the server), especially those worried about a 32ms delay.

To close the loop by continuing on with your assumption of a distant server:

“..."because the connection is the bottleneck then the way it operates becomes the controlling factor.".....” No, it becomes the major/prominent factor; do you still not understand the difference between this and ‘controlling’? As I’ve already said: the latencies are independent - one latency does not control or influence the other, larger or not – unless you want to explain how one influences the other? I don’t know why you persist with something that is independent and hence unrelated. Shoe on other foot: perhaps if you had an honours degree (let alone a ‘qualification’) in communication electronics (let alone ‘chemical kinetics’) it would have helped you.

“the advantage of the faster system may well be lost a certain proportion of the time”
It is regained by an equal amount for an equal portion of time. On average the advantage of the faster system is preserved (contradicting your claim that I supposedly said “summative advantages on the faster system are not preserved” – please don’t misrepresent my statements [I said packets, not system]).

With that I think I know where we differ; it is not on the facts but on how latency is viewed. Let me offer a reasonable and logical end to this for both of us; I will repeat what I’ve stated within a prior post. Individually (in some circumstances), the latencies *may not* always perfectly add for each and every packet due to the random nature of other independent factors (I don’t believe this what you were originally heading given your earlier posts regarding ‘bottlenecks’ – but you seem to be now so I’ll go with it). Statistically, the latencies *will* add; on average they *will* simply be additional. Also the latencies *will* tend towards adding even for individual packets. Something is very wrong if we don’t agree on these.

To get to the nub of this: I don’t believe your view of additional latency is the one generally used. If you want to use physical ‘kinetics’, I can explain various physical examples to prove this point, such as with air pressure, temperature and light intensity - all have individually random ‘packets’ but are widely accepted as being simply additional. This is supported by the other respondents also disagreeing with you. As far as practically all people will view this, the latencies will *seemingly* add and will be considered as additional for all intents and purposes, regardless of whether you view my responses as ‘going off on a tangent’.

You can disagree with the generally accepted meaning of latency if you wish (you don’t get fairer than that). Regardless, your original assertion of “the latencies actually don't add together.....” is debunked because they clearly can. Also the advantage of a faster system will generally be preserved, by an amount equal to the reduction of such a delay – which was the point of the original comment you responded to.

This issue is one of logic, not faith, hence we shouldn’t have any reason to disagree; you’ll also be disagreeing with all the other posters in this thread too. You’ll probably want to think twice before repeating such claims in future ;c)

Oh well...I think we are just going to have to agree to disagree....no hard feelings....all the best...Ken (technogiant)

Sorry but I still cannot accept what you say, at least within the context of your original comment.

Critically, the two lags are independent; one does not compensate for, or absorb, the other. Even if you wanted to argue the random nature of the connection lag, I can still say that on average, the average difference in processed lag is equal to the difference in display lag. This doesn’t matter anyway. Yes the connection delay likely won’t be equal (preserved) for all packets, but that is irrelevant: the display lag will still be in addition to the connection lag regardless of whatever the connection lag is for each packet because they are independent. The latencies add, both statistically and individually.

“So as I have repeatedly said decreasing the gamers response time by what ever means obviously will not slow down the overall response”
I should hope so. However, increasing the gamers response time by whatever means (display lag, input control lag, distraction, alcohol) obviously WILL slow down the overall response (and vice versa). My ‘reductio ad absurdum’ argument still stands.

We are almost in agreement.....but not quite.....imagine two packet of data arriving at the same connection which is busy....the one packet of data the equivalent of that in the faster situation arrives 100ms before the second parcel of data equivalent to that on our slower system.
All I am saying is that 100ms advantage may not be preserved....as the connection is busy with the other tasks both the packets have arrived at the connection and are waiting to be sent....once the connection has dealt with those prior tasks it will then go on to deal with our two packets....assuming that it will deal with those in the order in which they arrived then the faster one will be sent first followed by the second one...however the time difference between them will not now be 100ms but will be the time difference that the connection took between sending the first and second packet.
If the connection takes less then 100ms after sending the first packet to send the second packet then the 100ms advantage will not have been preserved the second slower packet of data may now be following 30ms behind the first.
This is why I've said that the latencies don't add up as the second packet of data is not now simply following 100ms behind the first as you would expect if you were to simply add the latencies as you have suggested.
So as I have repeatedly said decreasing the gamers response time by what ever means obviously will not slow down the overall response but it may or may not speed it up by as much as you may think by simply adding together all the speed increases.
I do hope we have finally reached a point at which we can agree....(ps don't know where I got boogle from lol)

Why am I boogle ;c)

I see where you were going, but it still doesn’t add up. I hope we can agree on the following:
1) “Incoming data from the server arrives containing fresh data... at our gamers pc....is processed and displayed on the screen”
- the displayed image would now be later than would have been due to the additional display lag (all else equal, the additional delay being equal to the display lag)
2) “our gamer responds.”
- the gamer would now respond later than would have been due to earlier additional display lag
3) “It is that data that is being returned to the server that is all important rather than what is being displayed locally on the gamers screen as it is the server that is controlling the game mechanics.”
- the returned data to the server would now be later than would have been due to earlier additional display lag and gamer response lag.

From here on is where we seem to differ.

“Any minor changes in response speed will effectively be negated by the connection because it is the slowest point …”
This is a non-sequitur. Regardless of how comparatively great the connection lag is, the returned packet from a gamer using a delayed display will still arrive at the server with an additional lag equal to that of the display; this delay won’t be negated. Even if you assume a quantisation effect where the server is processing returned data of discrete time slots (the earlier packets are queued [delayed] such that some delayed packets are processed within the same frame) statistically (average - with lots of such events) it won’t matter because at occurences the delayed returned packet (no matter how small the additional delay) will be bunched with the next slot, so suffering a greater delay (possibly greater than what could be from a tiny display lag). In fact, the statistical additional delay will be equal to the difference of image delivery times between the refresh delays (trust me, I have an A-level in pure stats).

Completing the post example (with the quantisation effect): the recipient of the package from mr fast refresh sends a 1 hour courier to tell the sender to say he got the package and that all is good. A few minutes later, the recipient of the package from mr slow refresh sends his 1 hour courier to tell the sender his (same) message, but the sender’s office has just closed for lunch (or worse yet for the rest of the day), so all that courier can do is wait until the office is open again. Given enough occurrences, the average additional delay of getting the message through will be equal to the difference of delivery times between mr slow refresh and mr fast refresh.

Yes the bottleneck will be elsewhere, but the lag is still additional.

.

Case in point: you might as well say that slower reacting gamers statistically won’t be at a disadvantage – reductio ad absurdum.

No sorry...you are incorrect...the sequence I have described is not in the wrong order...
"...within the context of graphics display, the described order of the events is wrong: the display (refresh) event can only occur AFTER the network processing...."
The above is only true of incoming data and not the outgoing data that we are discussing.
Let me explain further.....Incoming data from the server arrives containing fresh data...perhaps the sudden appearance of an opponent... at our gamers pc....is processed and displayed on the screen....our gamer responds...his screen is updated with his response straight away (locally without involvement of the server or connection traffic) and the data relating to his response is sent as I described in the previous scenario with the sorting office to the server.
It is that data that is being returned to the server that is all important rather than what is being displayed locally on the gamers screen as it is the server that is controlling the game mechanics. As I described previously the arrival of this data at the server is subject to the bottleneck and delay of the connection. Any minor changes in response speed will effectively be negated by the connection because it is the slowest point and works in an out of order fashion.
So like I said previously a faster response by the gamer be that because of a faster screen refresh or some other reason or in fact the cumulative effect of several things will not necessarily result in the information relating to his response being received by the server any sooner....also as I've said previously it wouldn't slow it down but it wouldn't necessarily get there any sooner it would depend on the tasks waiting on the connection.
So returning to our initial argument this is why I say that all the latencies don't add together...while simplistically they obviously do but in terms of the over all response as received by the server (which is after all the most important thing as this decides the game out come) they do not.

Your example is invalid; within the context of graphics display, the described order of the events is wrong: the display (refresh) event can only occur AFTER the network processing. To correct your example: the parcels may well be held up in the same queue within the sorting office (network connection) for days (yes, this might well be the bottleneck), but after that the packages still be delivered to their respective destinations. All else equal, the package given to Mr Fast Refresh (the sharp guy in his sports car) to deliver will arrive at its intended destination in a shorter time than the one given to Mr Slow Refresh (the laid back guy in his clapped out car). Therefore, in this case, the latencies (subsequent to the processing lag) clearly add, even if the bottleneck is elsewhere.

Sorry in advance if this seems simplistic or patronizing...but I'm going to use a simple example to explain what I am saying....
Two brothers...shall we call them Mr Fast Refresh and Mr Slow Refresh...they both want to send a parcel (containing the gamers response to an on screen event)...to outer Mongolia.
So Mr Fast Refresh being a very sharp and motivated type of guy is up at 6am jumps in his sports car and speeds to the postal sorting office where he deposits his parcel at about 6:15am.
Now Mr Slow Refresh is much more laid back...he jumps in his slow and clapped out car at 9am...along the way he gets stuck in the rush hour traffic and encounters all manor of sequential delays (latencies)....he doesn't get his parcel to the sorting office until 11am.
Sadly all is not well at the sorting office and due to staffing shortages they are always the bottleneck and slowest part of the postal service (ie network connection) they have a backlog of parcels many of them marked high priority and destined for the Houses of Parliament London. 
They simply have to be dealt with first and the parcels received this morning wont be dealt with until 3pm that day.
So Mr Fast Refresh wont get his parcel to outer Mongolia any faster than his brother Mr Slow Refresh regardless of all the sequential delays that added together and made Mr Slow Refresh so late.
Do you see now? if not then forget it..I can't put it any more simply

"Once the gamer has responded to the new information on the screen (albeit 8ms faster) then graphics rendering and display do not come into it."
That’s a fallaciously nested argument. You effectively said ‘once the delay has happened, it subsequently doesn’t play a part’ – which is fairly obvious; the key thing here is that the delay does happen.

"Your argument is that all the small lags and latencies along this process are sequential and add together."
That’s misrepresentation of my statements. Some of the latencies will be parallel; others will be serial. The process of displaying 3D images (all else equal) gives an additional, serial lag. The network packets, being the bottleneck or otherwise, will be largely unchanged.

"My argument is that as the network connection is the slowest point in that path then the manner in which it operates actually dictates the overall process."
Likely correct, but in this case it is utterly irrelevant; only you had referred to ‘network connections’, everyone else was talking about display lag.

One thing I do agree on: our little ping-pong is as significant as the additional lag from displaying 3D images :c)

Sorry...without meaning to be rude...but either I've completely missed the meaning of your last comment or you've gone off on a tangent.

We were not discussing graphics rendering and display...or at least only in as much as how a faster screen refresh rate would (or more correctly would not) improve online gaming performance....
Once the gamer has responded to the new information on the screen (albeit 8ms faster) then graphics rendering and display do not come into it.
All that is of concern is how fast the gamers response is transmitted to the server.
Your argument is that all the small lags and latencies along this process are sequential and add together.
My argument is that as the network connection is the slowest point in that path then the manner in which it operates actually dictates the overall process. As I have pointed out the connection/data transmission does not operate in the inorder sequential way you are describing and so it is not correct to say that all the latencies simply add together.
In effect reducing the number and size of the latencies prior to the connection will obviously not make the process any slower but may or may not make it any quicker depending on the queued data traffic at the time.
This discussion seems to be getting pointless as I am now just repeating what I've said previously in different ways...either we just don't agree and will have to agree to differ or I am not explaining myself adequately enough for you to understand my argument.
Either way it's pointless bashing on with it as at the very most it is an extremely minor consideration.

"queue of data waiting ", "data packet are given priority"

What you say is true but I see how these apply to graphics rendering for display devices - in this case there's no such queue (unless the unit knows the future), only the occasional dropped packet.

In terms of graphics rendering and display, how do the latencies not add up.

I do also understand what you are saying and you may argue that even if there is a queue of data waiting to be sent to the server then the sooner your game data got into the queue the better...and if that queue were dealt with strictly in the order of arrival of the data then again your argument would be correct....but the queue is not dealt with in an in order fashion as I understand it...certain types of data packet are given priority...lost packets are prioritized for resending and your particular data may sit there for a longer or shorter period of time depending on what data is queued at the time.
This removes the emphasis from the time of arrival of your data packet (obviously within certain limits) and places it on the queue, its content of data packets and the way it is dealt with.
That is why I said they don't add up, because a data packet received by the queue 8ms sooner than it would normally have been may or may not be dealt with 8ms sooner it depends on the data queue....so you cannot simply add all the small latencies together...because the connection is the bottleneck then the way it operates becomes the controlling factor.

I do agree with you regarding the significance of the issue (I've more or less already said what you stated at the end of your reply), but I’m not sure the reply progressed what I asked:

- How can periods of events on a critical path (the next event not starting until the finish of another) not add together? 

Don't forget, this is in relation to your comment: "the latencies actually don't add together"

The latency either adds or it doesn’t, in this case how can they not add together?
Yes "bottlenecks" do exist elsewhere, but we’re talking about the overall response time, not where the largest delay is.


I suppose either argument could be correct....if the network connection is not bottle necking the data then your argument would be correct...if the connection is bottle necking the game data then any minimal increase in speed this may give would depend on how that queued data is dealt with by the connection...if the data is just waiting while the connection deals with higher priority data then any increase would be lost.(that's why killer nic cards increase performance as they prioritize game data traffic in the queue)
Either way the difference in speed going from a 60Hz to 120Hz refresh rate screen or as i work it out from 16ms to 8 ms response only represents an 8 ms response improvement is irrelevant even if in your scenario was completely correct...an don't forget of course this is only the screen refresh rate... if your graphics card is only pumping out 60 fps then you will not see any difference as your screen will simply redisplay the same image twice in succession with a 120Hz screen.
.
So bottom line is this is irrelevant in real terms and to suggest that gamers or even twitch gamers will be left for dead in droves by not having such a screen is rubbish

I noticed how you didn't actually answer the critical question:

- How can periods of events on a critical path (the next event not starting until the finish of another) not add together?

I already understand what a bottleneck is. In this case, the response time of the gamer is going to be significantly greater than an additional 32ms delay, hence the latter obviously cannot be the bottleneck (I’m not saying the former is it either) - but this is utterly irrelevant. You clearly stated "the latencies actually don't add together" which unrelated to bottlenecking and in this case is illogical.

lol...yes it was an unfortunate choice of nickname as I'm not from an IT background but is one I use routinely.....there's no need to be insulting though.

I take it you're familiar with the term "bottleneck" where the overall speed of the process it determined by the slowest part of the system?
ie perhaps a program where the speed of execution is being determined by the slow speed which the hard drive is delivering data to the cpu...no point buying a faster cpu when its the hard drive is the problem.
Likewise makes no difference if your driving a ferrarri or a ford escort...when your in a traffic jam you'll be going the same speed.
Do you get my drift...the network connetion is the bottleneck so the screen response is irrelevant.

"the latencies actually don't add together"

W . T . F ?

How can periods of events on a critical path (the next event not starting until the finish of another) not add together?
<boggle>

Yet you have the nerve to nickname yourself 'technogiant'?

the latencies actually don't add together.....
In effect you will be bottle necked by the slowest part of the system which will probably be the connection lag....so other lesser delays like 32m/s on the screen will make no difference whatsoever

Yes, but assuming you have the same speed reflexes as your competitor, they're still killing you 32ms before you've reacted!

seen this technology three times now. Once at a private tech event, at Nvision, and at a private showing. I've seen it demonstrated on Mitsubishi DLP's and on 120Hz LCD's (Viewsonic and a pre-production Asus model at the time) with the old clunky glasses and the newer type shown at Nvision. The tech is pretty darn good, and it is the first shutter tech that doesn't give me a headache.

>sorry Richard but twitch gamer or living cyborg you cannot react in 32 m/s
>posted by : technogiant, 08 November 2008 

He made no such claim.

A person whose monitor sports a 32ms lag will be able to react 32ms later than the optimal.

Having played online games, a 30ms lag is a lot.

you might not think that 32ms comes into play, but compound a wide range of latencies across hardware and software and it adds up fast.

"... you cannot react in 32 m/s"

Although that's true, it's not the point. His overall response time will be 32ms faster, regardless for his reaction time. It's possible that 32ms (on top of the ~200ms reaction time) might make some difference, but such occurrences would be quite infrequent.



sorry Richard but twitch gamer or living cyborg you cannot react in 32 m/s

I have been using nvidia graphics cards and drivers with my edimensional shutter glasses and 21" 100 Hz CRT for years (they use to work brilliantly together). 

Its only since nvidia mysteriously stopped shutter glass support in their drivers for the last couple of years of cards (and then only for iZ3D) that this stopped. 

We don't need new shutter glasses we just need nvidia to switch support back on for shutter glasses (I hope this story means nvidia is doing that now). 

Till then I can't fly any flight simulators (I could never go back to 2D simulators having experienced 3D) and I'm not about to fork out a couple thousand for a new monitor.

You can't trust nvidia for continued support though, nor for releasing hardware without severe bugs, at least initially (and no I'm not just talking about the substrate issues right now).
Note, I said nividia because this is about nvidia, I might add that ATI also had it share of dropped projects and hardware failures, although sometimes because the world/developers simply didn't want to pick up on it, no matter how smart it was.

..is a 64-bit driver. I used to have a pair of glasses back in the day and I'd love to have another.

Is that a TFT, or LCD panel ? Or even Plasma ?
Or is it going to be a final iteration of the venerable old CRT that still kicks them all out for longevity, color rendition and image quality at all resolutions ?
Not to mention contrast ratio and refresh rates, of course.

are totally what I want 3d or no 3d

twitch gamers die in droves for not using higher refresh CRTs and the evidence is a lot of LCDs have an image processing delay up to 32ms which is life or death on the cyber battlefield.

It used to be the case that if you bought a monitor you wouldn't have to upgrade it for several years....I would love to have 3d gaming like this but having recently shelled out £350 on my 24 inch monitor which I hoped would last me at least 5 years I really couldn't stomach replacing it so soon.....is this something that will work with any game with the correct drivers and hardware or does the game programming have to support it as well...if the latter is correct then it probably won't be widely available for some years yet.