[MPlayer-dev-eng] MplayerXP WILL NOT HELP YOU (Official answer)

Nick Kurshev nickols_k at mail.ru
Sun Mar 31 20:17:24 CEST 2002


Hello, Arpi!

On Sun, 31 Mar 2002 20:12:37 +0200 you wrote:

> Hi,
> 
> > > i really want my system to spend negative times with i/o stuff ;)
> > > more io - less cpu usage - sounds gerat! i want threads Now!!! :)
> > Negative values are your stuff - I didn't touch it yet!
> 
> my code NEVER printed negative values... it cannot print as it's impossible
But always rounded all results on 100% ;)
> to happen if te other 3 values are correct...
> it means that something is really bad with your other numbers...
> 
REAL RESULTS are independed from this stuff.
They are real things which were computed by mplayer's core (old and new)
during movie playback.
> anyway i wouldn't call your paste as benchmark - no environment, no -v logs,
> no public accessible file give etc... in short: it isn't reproducable.
> you probably never learnt anything about iso 9001 and such standards, but
> there are basic rules of benchmarking, just cut'n'pasting few nice numbers
> has no meanings at all.
There is published fourcc, WxH, bitrate and all other info!
Anyone can reproduce benchmarks with any other file which he like.
(except using of excplictly bad formats like: ASF, MOV_QT, FILM, VIVO)
It doesn't matter which file is used - just relative comparision of working
of two cores.
> 
> > I really don't know why you compute systime ;)
> > Who intersested to know time of other processes?
> 
> not only other processes - it's idle time and system i/o time
> 
which is rounded to 100%.
> > > > REAL RESULTS: dropped=938 too slow=222 too fast=0
> > > 
> > > it shouldn't drop a single frame on such system with vidix and DR.
> > > 
> > Yes with vidix there is 0 dropped frames.
> > But it's not VIDIX test but XP mode only; when movie is decoded and scaled
> > on CPU only (without using of HW accel). Same as Intel is presenting quake3
> > which is run with help of UP P4 only.
> 
> but who needs XP with software stuff and dropped frames when he can get full
> rate playback with vidix or direct rendering ?
> 
Who writes new drivers for vidix? nvidia users? you (matrox user)?

> anyway the frame dropping code in mplayer is shit, it's fact.
My core uses not yours frame dropping code!
> it was hacked together in 10 minutes and was never improved, because of lack
> of such slow hardware. maybe gabucino could help here, using his 386 :)
>
But sorry! I started my branch simply because mplayer drops frames
on my Duron-700 with some kind of movies.
> i think with some clever fps control and dynamic dropping control it could
> be much better, even better than your nice xp...
> 
Also you could say: Nick, implement HW MPEG-1,2,4 decoding extension for VIDIX
to use something like: fread(&video_memory...
and all problems with non monotomous decoding of stream will disappeared.
But nobody send me ATI's documentation without prohibition of using its
context under GPL.
and there are many other users which have top strange software (like ET6000)
without any hopes to get vidix driver at least!

But motivation of this thread is other: MplayerXP HELPS when mplayer doesn't help.
And I want to watch this fact on mplayer's news page or remove that news at all.
It seems that I prove that, at least!
Else I'll prone to decice it as redistributing of FUD.
That's all.
> 
> A'rpi / Astral & ESP-team
> 
> --
> Developer of MPlayer, the Movie Player for Linux - http://www.MPlayerHQ.hu
> _______________________________________________
> MPlayer-dev-eng mailing list
> MPlayer-dev-eng at mplayerhq.hu
> http://mplayerhq.hu/mailman/listinfo/mplayer-dev-eng
> 


Best regards! Nick
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
URL: <http://lists.mplayerhq.hu/pipermail/mplayer-dev-eng/attachments/20020331/d7756091/attachment.pgp>


More information about the MPlayer-dev-eng mailing list