9 replies to this topic

[ETsneak]

I'm just now deciding to run a silent server as the 6th server im hosting lately, but I have played several hours on silent servers..

1. weapalt/attack2
For nine years it's been "bind key +weapalt", changing it now to "bind key +attack2" seems wrong to me. I know it makes more logical sense, and it probably should have been done by the makers of ET.. but this game is 9 years old and we are used to "+weapalt". Or do you think we should change the QWERTY keyboard because the most commonly used keys are separated in order to slow us down? Should everyone get used to a "more efficiently designed" key layout?

2. FPS
Even when I ran an AMD Athlon64 6000+ 3Ghz x2, DDR2 4-8GB, and 9800GTX+ or 9600GT I never had issues with FPS in any mod but NoQuarter (because of it's flawed design causing performance decreases). I now run an FX-4100 (hotfixes applied), GTX285, and DDR3 1333mhz 8GB and I keep a solid 125FPS in all mods (on occasion in etpub in poorly designed maps like karsiah_te2, the FPS will drop. It seems to also be somehow related to specific servers, my FPS drops alot on tjw(etpub), yet on my own servers I keep a steady 125FPS.

It seems to me like silent also has some issues with certain maps, it would be nice if this could be something at the top of priorities to work on, because I feel performance is much more important than visual effects and fancy things. I want a solid 125FPS wherever I go, and 76 for high recoil weapons.

3. FixedPhysicsFPS cvar
I saw that you list this cvar as "experimental". This is odd since it's been used by etpro and etpub for several years now without an issue. Care to explain the reason behind saying something like this about that cvar?

That's all for now :>

Edited by ETsneak, 17 July 2012 - 01:46 PM.

[gaoesa]

1. Changin this was necessary because the knife throwing behaviour was changed. It is not possible to hold the knife armed using weapalt.The attack2 is used in Jaymod, NoQuarter and in Nitmod also.

2. We always try to look for ways to improve performance.

3. I have no idea why ETPub team sees it as experimental. Nevertheless, that decision was made by them and we kept it like that. Furthermore, we haven't even looked at it's implementation which is all the more reason to maintain the ETPub teams recommendations.

[ETsneak]

3. I have no idea why ETPub team sees it as experimental. Nevertheless, that decision was made by them and we kept it like that. Furthermore, we haven't even looked at it's implementation which is all the more reason to maintain the ETPub teams recommendations.

Strange, never seen anyone from etpub say this, or any docs that say it.. including the 0.9.1 and 1.0 docs.

[gaoesa]

They may have changed the opinion of it being experimental and it is not regarded as such any more. The comment was made based on the information available in the source code and in documentation, but it may have have been bad interpretation.

EDIT:
This all I could find about it from the changelog:

* forty: #367 - added preliminary fixed physics. This is just Haste's
pmove_accurate code from q3.
only options 0 and 1 implemented for now.

That is for the 0.8.1 version.

[ETsneak]

They may have changed the opinion of it being experimental and it is not regarded as such any more. The comment was made based on the information available in the source code and in documentation, but it may have have been bad interpretation.

EDIT:
This all I could find about it from the changelog:


That is for the 0.8.1 version.

oh okay, and what version was out when you began working on silent? 0.9.1? (just asking)

[gaoesa]

0.9.1

[Dragonji]

Unfortunately I have to agree with Sneak about FPS... Just played a bit on an ETPub server and the game looked smoother and my FPS were way higher. One thing which is fastly noticeable: FPS goes down so bad when you use Flamethower on silEnT when ETPub does not have any problems handling Flamethower's flame animation (FPS stable all the time).

Everything tested on the same config on the same machine (Intel C2D 2x3,2GHz, GF 9600GSO, 2GB DDR2 RAM).

PS. Just wondering, can this be caused by this feature (just a thought)? (yes, I do realize it was my suggestion, shame on me )

Edited by Dragon, 02 August 2012 - 11:16 PM.

[gaoesa]

The ETPub flamethrower and silEnT flamethrower is the same thing.

EDIT:
I did quick testing on my local computer with nightly ETPub build in places I know there are FPS drops, and couldn't notice ETPub having better performance. This is of course not valid benchmarking, but still makes me think that the way better is a result from other things.

Also, as a response for the weapon animation fix. I doubt it has significant impact if any. I don't remember it exactly though.

I remember the flamethrower causes FPS drops in every mod I have tested, when you step into the flames.

[TheSilencerPL]

Also, as a response for the weapon animation fix. I doubt it has significant impact if any. I don't remember it exactly though.

But I remember it, it does not have any impact on the FPS.

[Dragonji]

Weird, I did some tests on my local PC too this morning and didn't notice any FPS drops on silEnT.

Edited by Dragon, 03 August 2012 - 09:10 AM.