Jump to content
Sign in to follow this  
normannobrot

info on "out of memory error"

Recommended Posts

Mr JRama.... you are exactly what Mr Culliton predicted...

A piece of info from Wiki ( i suppose Mr Culliton was not the writer):

"A memory leak, in computer science (or leakage, in this context), occurs when a computer program consumes memory but is unable to release it back to the operating system.

This is on wikipedia? I'm surprised, it's garbage.

If the application is actually unable to release the memory back then the application simply requires that much memory. That is not a leak. Just because the computer doesn't have enough RAM+swap doesn't mean it is a leak. A redesign to do the same thing with less memory might still fix it.

A memory leak in programming terms is being able to give pieces back but not doing it. The application does not need them anymore, but the programmer doesn't realize it and fails to do so. Or somebody holds on to things by accident in a garbage collected language.

Share this post


Link to post
Share on other sites
A piece of info from Wiki ( i suppose Mr Culliton was not the writer):

"A memory leak...however, many people refer to any unwanted increase in memory usage as a memory leak, though this is not strictly accurate."

And now my dear Lt Bull, do you see what Mr Culliton was refering to?

Yes. Forgive me for applying the "looks like sh*t, smells like sh*t, feels like sh*t, must be sh*t" approach to trying to give a suitable enough name to what I observed. Didn't realise there was any reason to be so strict.

Phil, Phil, Phil - no one is doubting your abilities as a coder, but is being on the receiving end of customers identifying and reporting (in as much detail as they can) problems in software of which you are a coder of really your thing?

Lt Bull - I see now that you're just looking for an argument. When "CM's problems can only be caused by bad programming" became "what? I didn't mean *you* guys are bad programmers" it should have been a clue for me.

Great. You now resort to blatant fake quoting. Classy. You have some nerve to say I'm looking for an argument when it's your posts that get all personal. FWIW, if this problem is not the result of "bad programming" (by whoever I don't care), what are they a result of? Good programming?

Now you conveniently backpedal again - you go in one post from disagreeing with me with no technical or factual basis (and yes, your personal lack of knowledge does have bearing here), to the next, wondering where I got the idea that you're disagreeing with me

(hint: search for the word "disagree" in your post, if you haven't edited it again).

It gets worse folks.

"Backpedaling" and implying I would edit my posts as if I have something to hide. Again, classy. Anyway I did find the word "disagree" in my post where I supposedly have "no technical or factual basis (and yes, your personal lack of knowledge does have bearing here)". I said:

But if you are saying that steady gradual memory increases/"creep" (like in my example) over the course of a game is normal, then I would disagree.

So I have "no technical or factual" basis to say that? No technical basis? I precisely recorded the CMBN memory usage in Mb/Gb tracked over the course of about 40 WEGO turns using a handy simple well known Microsoft diagnostic utility used by thousand of people worldwide and presented my data in chart form. The results are repeatable. But your saying that isn't technical? And no factual basis? So I really didn't do this, I am making it up? The trends I recorded...unfactual. The "out of memory" crash occurring when the CMBN memory usage as reported by Task Manager exceeds a certain limit...unfactual.

So I guess you want us to believe that the the steady gradual memory increases/"creep" in CMBN WEGO (like in my example) is:

a) unfactual

B) is actually NORMAL (in the good sense of the word, as in desirable "meant to be")?

So BFC actually expect the game to crash if you play WEGO for too long? Interesting. Are we wasting our time reporting on an "normal" issue BFC was probably already aware of? Just update the manual and make users aware of this fact to stop them finding out the hard way.

Engaging you simply seems to worsen your behavior

What behaviour? The one where I just question assertions that just make no sense? Yeah how inconsiderate of me. Or the one where I share my findings with other players and encourage and instruct them to do the same so we can all get some correlation? Yes, just terrible behaviour right there.

and as Dan Dare indicates is very likely not the best use of my time. So, I will *stop* engaging with you.

I couldn't agree more with him. As I said, I am surprised you took such a pre-emptive personal angle in your responses to me. When one starts getting all personal in a technical help thread it's time to move on.

Your tickets will be answered, your information will be gratefully used, but further discussion regarding your advice or opinions about this problem is finished. I would ask again, very politely, that you confine your posts to actual facts.

LOL, that's what I have been doing, reporting "actual facts" I have observed on my PC (and can probably be confirmed and observed on other peoples machines if they care to check themsleves). The irony is that this is exactly what you are refuting and objecting to.

Please no more. There was absolutely no need to single me out like this. This is getting laughable.

Share this post


Link to post
Share on other sites
Guest

Yes, I've had reports along those lines. Whether they have the same root cause I don't know.

Share this post


Link to post
Share on other sites
Combat Mission is a 32-bit process and can't use more than 2GB of RAM.

that is realy interesting fact there...

so you mean to say that no matter how much ram you have its only gona use 2gig?

mmmmmm.....??? kinda shows something is going wrong somewhere..

Share this post


Link to post
Share on other sites
Does anyone suffering this out of memory crash also have the same problem when/if they played CMSF?

YES !! have had this error on both games...

tried making a massive scenario, on CMSF...stupid size....FULL EDITOR SIZE..

was gona try and do a black hawk down scenario, just with a tank instead...well... editor couldnt handle it....

got the main large part of town handled with vast amounts of damage around, but then everytime i went to preview, it crashed with the error..

Share this post


Link to post
Share on other sites
that is realy interesting fact there...

so you mean to say that no matter how much ram you have its only gona use 2gig?

mmmmmm.....??? kinda shows something is going wrong somewhere..

not wrong. its just the way its designed. its not something they can fix without recoding the whole game (probably).

Share this post


Link to post
Share on other sites

and for the record, i too use WIN XP... as i cant stand the un-freindlyness of newer systems on older games,

hell....i dont realy consider the CMX1 series that old yet, and their already running half supported on win vista / 7...

its kinda sad realy as our systems die out over time, we will eventualy not be able to play the CM X 1 series, as its not fully supported. LOL.... how epic of me....

**SOB**SOB** LOL..

Share this post


Link to post
Share on other sites
not wrong. its just the way its designed. its not something they can fix without recoding the whole game (probably).

BAH ! chin up son and keep the faith....LOL

it'll probably be a while, and im sure their seriously looking into it, as phil states, plus with alott of people having this issue they need to..

i can emagine it will be a very large patch tho...lol :(

hell, they fix this, and the game WOULD realy be worth 10/10 LOL :D

Share this post


Link to post
Share on other sites
BAH ! chin up son and keep the faith....LOL

it'll probably be a while, and im sure their seriously looking into it, as phil states, plus with alott of people having this issue they need to..

i can emagine it will be a very large patch tho...lol :(

hell, they fix this, and the game WOULD realy be worth 10/10 LOL :D

Unless you are using Windows Xp 64 bit, you cannot use more than 2Gigs (give or take) of RAM. The machine can only until 2^32, and that gives you a maximum limit on the addresses for memory.

Share this post


Link to post
Share on other sites

My observations in this case:

- not only on huge/unit-heavy maps. I'm still in the tutorial campaign and it happens often

- no slowdowns before the crash. The game works perfectly smooth and then suddenly: bang!

- the game sometimes produces corrupted saves. It often crashes in the phase after saving a corrupted file (not yet verified, if this is ALWAYS the case).

Share this post


Link to post
Share on other sites
Unless you are using Windows Xp 64 bit, you cannot use more than 2Gigs (give or take) of RAM. The machine can only until 2^32, and that gives you a maximum limit on the addresses for memory.

I asked this earlier but didnt get an answer. Why is the game coded/programmed in a way where it doe not make full use of up to date and more tech if it is available?

Share this post


Link to post
Share on other sites

i was just playing busting the bobcage again. i quit at turn 14 (furthest ive gone without crashing) and went back to the menu. i then loaded it again (to play as the other side) and got the 'out of memory crash' during the loading screen at 25%. this, and LT Bulls graph, do sugest that it is holding onto memory it dosnt need. i.e. remembering a scenario ive quit while at the main menu.

Share this post


Link to post
Share on other sites
I asked this earlier but didnt get an answer. Why is the game coded/programmed in a way where it doe not make full use of up to date and more tech if it is available?

How would I know? I'm guessing here, but if you code the game in 64bit, most people using still using Windows XP or people using Vista/Win7 32bit, will not be able to run the game. I believe the original code dates from before 2007 (when CMSF was released). And that would have been the sensible choice at the time.

As far as I understand it is not easy/possible to make a program that works on 32bit machine, but can use 64bit's memory on the right computer. Big software programs such as Matlab come in two versions, 32bit and 64bit, and they have more people programming than BFC, so I suspect it's no trivial task.

I have a 64bit computer myself, but I understand there are other people besides using different configurations. I have only come across out of memory error sporadically (CMSF mainly), but only in large scenarios or scenarios that run for more than an 2 hours or so.

Share this post


Link to post
Share on other sites
i was just playing busting the bobcage again. i quit at turn 14 (furthest ive gone without crashing) and went back to the menu. i then loaded it again (to play as the other side) and got the 'out of memory crash' during the loading screen at 25%. this, and LT Bulls graph, do sugest that it is holding onto memory it dosnt need. i.e. remembering a scenario ive quit while at the main menu.

I have noticed in CMSF, that if I load PBEM files in a certain order, the sound would stutter, whereas if I loaded it in a different order there would be no problem.

Share this post


Link to post
Share on other sites
How would I know? I'm guessing here, but if you code the game in 64bit, most people using still using Windows XP or people using Vista/Win7 32bit, will not be able to run the game. I believe the original code dates from before 2007 (when CMSF was released). And that would have been the sensible choice at the time.

As far as I understand it is not easy/possible to make a program that works on 32bit machine, but can use 64bit's memory on the right computer. Big software programs such as Matlab come in two versions, 32bit and 64bit, and they have more people programming than BFC, so I suspect it's no trivial task.

I have a 64bit computer myself, but I understand there are other people besides using different configurations. I have only come across out of memory error sporadically (CMSF mainly), but only in large scenarios or scenarios that run for more than an 2 hours or so.

i expect its a limitation imposed by the engine and the fact they are using openGL not direct x. if support for openGL has been dropped i doubt Microsoft developed it to accommodate for 64bit (im guessing too btw :P). when was the cmx2 engine developed? before 2007 anyway. back then 2gb was plenty for any game.

also i dont think its realy a lack of ram problem because like people have said it dosnt slow down or stutter, it just dies and other people are also running it fine on 2gb also in the graph it crashed at 1.6gb, it had another 0.4 to go.

Share this post


Link to post
Share on other sites
i expect its a limitation imposed by the engine and the fact they are using openGL not direct x. if support for openGL has been dropped i doubt Microsoft developed it to accommodate for 64bit (im guessing too btw :P). when was the cmx2 engine developed? before 2007 anyway. back then 2gb was plenty for any game.

also i dont think its realy a lack of ram problem because like people have said it dosnt slow down or stutter, it just dies and other people are also running it fine on 2gb also in the graph it crashed at 1.6gb, it had another 0.4 to go.

No, it's to do with your operating system (and the hardware it runs on). Windows Xp simply cannot count further than 2^32. Doesn't matter if you use Directx or OpenGL.

Share this post


Link to post
Share on other sites

Just had this happen with a quick battle, medium size all round. I have 3GB of VRAM on my GTX580 and 8GB of DDR3 system RAM, on W7-64. I'm on a 120GB SSD, pagefile (12GB) is on a regular 7200 HD.

Is there any official word on this (haven't read the entire thread)? If this is going to be an unfixable flaw online is out the window, me and my bro had enough of this with CMBB and CMAK.

Share this post


Link to post
Share on other sites

I've read a bit more of the thread. I wasn't playing WEGO, my game was turn-based. The game timed into the red about 26 minutes before it crashed (think it was a 50 minute game to start with). Forces on my side were not particularly large, I had about eight PaKs and assorted inf (computer picked them), no idea what the Yanks had (computer picked them too).

I'm incredibly disappointed to discover this game is apparently hobbled by being 32-bit only, non multicore and using a crippled API (Open-GL). None of which would matter if it ran as it does WITHOUT then crashing. With a problem I seem to remember being endemic with the old series (CMBB etc). The crashes were fundamentally why me and my brother stopped playing those games and if this is more of the same this game will be going on the shelf almost before I've begun playing it properly. A game that crashes randomly and without warning is simply not viable. I should mention I have no such problems playing RoF, CoD, RO, DH, GRID and a few other games I've run on my system as it stands now.

If, as one of the programmers in this thread states, the API was, and is, a known issue, this should have been clearly flagged at the point of sale. I, for one, would not then have purchased this game. The absolute rage we used to experience when an online game repeatedly crashed, to the point where even constantly saving it no longer helped, is not something I expected to have repeated with the new engine.

Share this post


Link to post
Share on other sites
How would I know? I'm guessing here, but if you code the game in 64bit, most people using still using Windows XP or people using Vista/Win7 32bit, will not be able to run the game. I believe the original code dates from before 2007 (when CMSF was released). And that would have been the sensible choice at the time.

As far as I understand it is not easy/possible to make a program that works on 32bit machine, but can use 64bit's memory on the right computer. Big software programs such as Matlab come in two versions, 32bit and 64bit, and they have more people programming than BFC, so I suspect it's no trivial task.

I have a 64bit computer myself, but I understand there are other people besides using different configurations. I have only come across out of memory error sporadically (CMSF mainly), but only in large scenarios or scenarios that run for more than an 2 hours or so.

cheers not criticising just trying to learn and understand why things are done a certain way. I love the game but have been getting out of memory errors on maps that are not that large or unit heavy.

Share this post


Link to post
Share on other sites

I'm incredibly disappointed to discover this game is apparently hobbled by being 32-bit only, non multicore and using a crippled API (Open-GL).

The other games you listed are all 32-bit as well. The problem lies elsewhere probably and will hopefully be fixed. Even if it was 64-bit, and if the problem is a memory leak it will just delay the time before the game crashes (and I'm not saying that it crashes). I have yet to experience a crash in CMBN and like I said only sporadically ran into trouble in CMSF.

Share this post


Link to post
Share on other sites
cheers not criticising just trying to learn and understand why things are done a certain way. I love the game but have been getting out of memory errors on maps that are not that large or unit heavy.

If wikipedia is to be trusted, it seems my memory is not that shoddy and making thing run on both 32bit and 64bit is not that easy :).

http://en.wikipedia.org/wiki/64_bit#32-bit_vs_64-bit

Share this post


Link to post
Share on other sites
Guest
This topic is now closed to further replies.
Sign in to follow this  

×
×
  • Create New...