PCOWS a memory hog ?!?

[black_cat]

Sorry to bother you with a (small) complaint, but I found the game to be a bit greedy on my computer resources :

(excuse my French [;)])

This particular memory usage was measured during command input (no AI planning or phase execution) in the second mission of the Russian campaign, around turn 2. But the amount seems pretty stable.
The settings seemed quite reasonable to me : no shadows, 3 men per squad, medium quality, 1024x768 16 bits.

I also noticed the game swapped A LOT when reloading (I mean, first load after game start is sort of quick, but the subsequent ones are a real pain : 3 minutes plus horrible sound stutters as the game swaps like mad).

I have a 2800+ Athlon with 512 MB RAM, A GEForce 6800 GPU with 256 MB DDR and indeed I have some other applications running (which contradicts the readme, ok).
But frankly I find it annoying to have to shut down my antivirus, mailer and data miner software just to play a game - however fun it might be.

Of course I could change my motherboard and buy more RAM, but frankly I suspect a bit of .NET tweaking (from the programmer's side [:)]) could reduce the memory consumption dramatically.


Since I am on it, another minor annoyance is the way the save game function works : type the name, click save, click exit, click exit.
Hitting ESC will only lead you out of the game (if you insist [:)]).
How about implementing the usual <click or type name> then <click save or hit return> usual interface ?
Or at least change the ESC or ENTER key binding so that a few key presses could get the player back to the game (istead of returning to desktop !).

[Laryngoscope]

Yes it does seem rather thirsty for what it is.

Your graphic settings probably won't make much of a difference to the memory footprint.

On the whole I think given memory prices today increasing to 1024MB would help not just here but also everywhere.

Of course I could change my motherboard and buy more RAM, but frankly I suspect a bit of .NET tweaking (from the programmer's side ) could reduce the memory consumption dramatically.

Could the code be better optimized? I am sure it could. All code can. The problem with code bloat is an industry issue.

There is no $$$ in improving code efficiency for the most part. Once it works well enough on a target machine, money will be better spent on adding more features or improving graphics rather than reducing save file reload times by X seconds. It is just the nature of the beast.

[black_cat]

Yeah. That is one of the reasons why I quit computer engineering after 15 years or so of trying to conciliate my love of for programming with my bosses time to market reduction efforts [:)].
Now seriously folks, I think the current implementation is really bloated. I only hope the reasons why this relatively low consumption design (we're not speaking of a flight sim 100x100 km landscape here) gobbles megabytes of RAM for breakfast is well known and kept in check by the developpers.
Looks more like usign unbound high level data structures like XML records all the way instead of switching to efficient internal representations. Just an educated guess, of course. I may be completely wrong, but still... An exemple of this is the horrible memory consumption for a reload : looks like previous data are not completely released before reloading, more like a garbage collection outisde programmer's direct control in charge of cleaning up the orphaned megabytes.
Considering the planned improvements and extensions, I would say the basic engine may benefit from a major version number increment.

[Prince of Eckmühl]

Out of curiosity, I fired up MTW2.

I ran the Agincourt scenario, a middle-sized battle.

WinXP Task Manager indicated that the memory usage for the EXE was 347mb.

Memory prices, especially those for obsolescent DDR, are really low right now.

I'd strongly recommend you getting some more, if not for PzC, then for the rest of your games.

PoE