[Not a bug] Poison clouds make the engine lag

[Lunever]

When I played RTC on my desktop (1GHz Athlon AMD, 768 MB RAM, GForce2, chaching on of course) I noticed that casting 3 to 5 Mon-poison-clouds on the same tile will slow down the engine down to a virtually unusable state (which at high levels usually results in the parties death, because the characters don't react anymore the way you want them to, ar at least not when you want them to react).
Since I also play RTC often on a much older laptop (pII 297 MHz, 256 MB RAM) and it generally performs fine even on that machine, I suppose the engine might possibly handle clouds in an inefficient way; does it draw each new cloud over previous clouds? If so, it might be better to only have a single cloud object on each tile that has just the propper added up power-level of all the clouds cast unto that tile.

[George Gilbert]

Unfortunately the suggestion you made isn't possible because custom clouds might have alpha channels so that multiple clouds on the same tile appear "thicker" than a single one.

I have a think to see what can be done, but I suspect that at ~300 MHz your computer is only 40x more powerful than the 8 MHz ST / Amiga, but it's trying to write 32x as much data to screen (because of the increased colour depth and resolution) and so its performance will be roughly the same as the original (which also slowed down greatly when more than a couple of clouds were on screen at once).

[Lunever]

Well, since I have the problem not only on the 300 MHz but also on the 1GHz, and on both machines I do not have much other processes running (WinXP, Sygate personal firewall, AntiVir from Avira, and a broadband connection; on the Athlon I leave them on, on the pII I turn off even these), I suppose it'd be good if you'd find a way to optimize that a bit.

If I understand you correctly, it could be optimized as long as no alpha-channels are used, which custom clouds might do, but standard poison (or monster-smoke) clouds don't. Is it maybe possible then to optimize it just for the non-alpha-channel clouds?

[Sophia]

I suspect that at ~300 MHz your computer is only 40x more powerful than the 8 MHz ST / Amiga

Not even that much.

The ST and Amiga did amazing things with their 8 MHz because of all of the custom chips and specialization... try doing that stuff on even a 12Mhz (150% faster!) 286.

[copperman]

Ooo, sounds like a challange

[George Gilbert]

Well, since I have the problem not only on the 300 MHz but also on the 1GHz, and on both machines I do not have much other processes running (WinXP...

Ah - that's your problem. WinXP hogs the CPU like nobody's business. Even Microsoft don't recomend you run it on that spec machine (and everyone else recomends at least double, if not triple that!).

There are several threads about this (specifically relating to ADGE and Adamo trying to run it on WinXP on too lower spec box - for example http://www.dungeon-master.com/forum/vie ... hp?t=24552 ) on these forums, and millions of threads about the general case if you type "WinXP" and "performance" into Google!

[Paul Stevens]

I did type that into Google. Found the Microsoft
page. I quote:

In general, adding memory is the easiest and most effective way to improve a computer's performance. Although it is recommended, Windows XP does not require 128 MB of RAM. The operating system can run with 64 MB of RAM.

While Windows XP does not require a state-of-the-art processor for a good user experience—the minimum is a 300-megahertz (MHz) Pentium II-class processor—

I have a laptop with XP. (Only because it came that way
and I'm to lazy to change it). Relatively slow processor
but I have never seen any performance problems.
I would suspect drivers/video/background tasks rather
than Windows XP itself.

Also look at some things like:
http://www.microsoft.com/windowsxp/usin ... eperf.mspx
http://kadaitcha.cx/performance.html

[Lunever]

It may take a couple of weeks until I have time to do so, but to verify this I will create a new and otherwise clean Win98SE installation on the Athlon and seem whether that performance problem persists.

[beowuuf]

Remember that win98 has it's own perfromance problems when running new software with it - ie the hard resource limit coded inside it!

[mikko]

I suspect that at ~300 MHz your computer is only 40x more powerful than the 8 MHz ST / Amiga

Not even that much.

The ST and Amiga did amazing things with their 8 MHz because of all of the custom chips and specialization... try doing that stuff on even a 12Mhz (150% faster!) 286.

Well, that's not exactly true. You cannot compare processors with different architecture that way. Especially when they have something like 20 years in between! Too much has changed. Even with the latest processors the ones from AMD have slower clock speed but are faster than Intel's. And what comes to the custom chips, I doubt that they compare even closely to modern(ish) sound and video hardware..

With a quick search I found this page for reference http://www.geocities.com/Hollywood/1930/sgi_speed.html. It happends to have a 300 MHz Pentium II processor listed and, while there's no Amiga 500 or Atari ST, there is an Amiga 4000 (25MHz). And the speed difference with that test is some 58x. So with an Amiga 500 it could be something like 150x..

Of course that is not a definitive comparison but only a test with one application. But I'd say the figures look reasonable (maybe not for the Amiga 1000 there, dunno)..

[Lunever]

I hope it can be optimized a bit. I honestly don't care how pretty the clouds look a s long as I can use them to set a deadly trap with a power-cloud-tile. I'm currently playing for the first time since quite a while the RTC dungeon, also at high level, and without firballs and stat-potions you have to become even more inventive in regard to means of combat.

[Sophia]

You cannot compare processors with different
architecture that way.

Then why did you post a link to a web page trying to do exactly that?

Even with the latest processors the ones from AMD have slower clock speed but are faster than Intel's. And what comes to the custom chips, I doubt that they compare even closely to modern(ish) sound and video hardware..

I was not comparing the custom chips to modern hardware. I compared an 8Mhz Amiga or ST to a 12 MHz PC with a 286. They were contemporaries. An 8Mhz Amiga, with its custom chips doing a share of the processing, could do things even a 12Mhz PC, with its comparatively "dumb" hardware, couldn't. In modern terms, we got a similar breakthrough in graphics when the burden of rendering the 3D started being offloaded to the graphics card itself.

I used the word "better," not "faster," because, the point I was trying to make is similar to the point you were trying to make about AMD vs. Intel: clock speed isn't everything.

[beowuuf]

So lunever, basically you have all the monsters going ultra fast at archmaster level and all the stats draining and other and game engine mechanics working overtime etc ultra fast...I think that could be a big factor in why you are experiencing these problems, though GG would have to confirm that.

Basically can you play a normal game of CSB with powerful enough characters to see if this still occurs? Because really, I don't liek the idea of say cloud mechanics getting reduced and stopping inventive graphical things (you realise the cloud mechanics are currently the only way to get overlays in RTC, for example) and only have it benefit someone playing a game at a stupidly high level of difficulty

[Lunever]

Right Beo, I will test it with a new low-level-party soon (although I got to play that party to be a bit first so they can cast a couple of clouds to reproduce the effect). Since there was only 1 monster about at the last test and George already said in another thread the the actual handling of monsters and dungeon levels by the engine does not take much ressources I suspect that the result will be no different though.

Besides, whether to play at high level is to be rated as "stupidly" is a matter of perspektive.

[George Gilbert]

George already said in another thread the the actual handling of monsters and dungeon levels by the engine does not take much ressources I suspect that the result will be no different though.

Yep - I'd be very surprised if the difficulty level makes any difference. Nearly all the time (90%+) is spent drawing the view; even if the time taken to update the dungeon doubles (which is probably a reasonable estimate), then the total time will only change by < 10%.

[beowuuf]

Ah well, then ignore me : )

[mikko]

Then why did you post a link to a web page trying to do exactly that?

Hmm.. I meant that you cannot compare just the clock speeds. The page I posted did not compare clock speeds but actual performance of the processors. Sorry, if I was a bit unclear on that.

I used the word "better," not "faster," because, the point I was trying to make is similar to the point you were trying to make about AMD vs. Intel: clock speed isn't everything.

Ok. Actually I can't find the word "better" in your previous post. But yeah, I was mainly commenting the original statement that a Pentium II would be "just" 40 times faster than Atari ST based on the clock speeds of the processors. I do know that 286 was a piece of crap and how superior Amiga was at the time compared to PCs and IMHO also to ST due to its specialized chips.

[Sophia]

Ok. Actually I can't find the word "better" in your previous post.

Oops, that's because I didn't use it. George did, though. Does that count?

I do know that 286 was a piece of crap and how superior Amiga was at the time compared to PCs and IMHO also to ST due to its specialized chips.

I agree... I just didn't want to start an ST/Amiga fight on the side.

[George Gilbert]

I agree... I just didn't want to start an ST/Amiga fight on the side.

Indeed, there'd be no point doing that - everyone knows what the result would be...

[mikko]

Oops, that's because I didn't use it. George did, though. Does that count?

Nope. He said "powerful". And I even did a search for "better". The results are not very convincing..

[Sophia]

Power = better

[Sophia]

Indeed, there'd be no point doing that - everyone knows what the result would be...

Oh! oh! oh! Those sound like... fighting words!

... Maybe. I actually don't know.

[beowuuf]

*cough amiga was betterer cough*

[Trantor]

I never had an ST, so of course the Amiga was better!

[Sophia]

I never had one either.

Who needs 'em

[Lunever]

I first used my father's ST but later decided for an Amiga for myself.
In that age I preferred to trade in my C64 for an Amiga because it offered the better oportunities for playing games.
Yet I liked ST-OS better than Amiga-DOS.

[Lunever]

BTW: The new config option did solve this problem on my machines.