Skip to content
Forum Archive

Cockos Happy AVS

39 posts

justin#
Oops, sorry I forgot to update CHAVS. Here are new builds (v2.91):



and the code:



(the code now includes the required bits from WDL, too, so no need to git clone anything).

This should fix the rand() bug, and also updates to even newer EEL-compilation, which has some fancy stuff such as basic code optimization.
DrO#
oh the irony of two v2.91 on the same day.

as i've been looking into fixing things from your 2.9 and just put up this 20mins before you (as trying to get something stable in the distribution again).

were there any code changes on your side in the last 2 weeks? as that's within the time scale of when i pulled down the code from git.

-daz
justin#
Originally Posted by DrO View Post
oh the irony of two v2.91 on the same day.

as i've been looking into fixing things from your 2.9 and just put up this 20mins before you (as trying to get something stable in the distribution again).

were there any code changes on your side in the last 2 weeks? as that's within the time scale of when i pulled down the code from git.

-daz
Hah, bizarre!

I actually intended to do this two days ago, but that ended up getting derailed...

As far as recent changes to code goes, EEL2 hasn't changed recently, so you should be fine there, but avs_eelif.cpp has been updated:


 void AVS_EEL_IF_resetvars(NSEEL_VMCTX ctx)

 {

   NSEEL_VM_freeRAM(ctx);

-  NSEEL_VM_resetvars(ctx);

+  NSEEL_VM_remove_all_nonreg_vars(ctx);

 }
(NSEEL_VM_resetvars() is no longer defined so old chavs.zip shouldn't compile against the newer EEL versions). There are also changes to NSEEL_addfunctionex() lines in that file, but they aren't critical (just some casts removed).


---

Also, I forgot to mention, also new in the latest CHAVS is support for functions, see:

DrO#
i know that feeling well, heh.

rightio, will sort out a diff from my 2.91 against things and roll out a test 2.92 in a day or so in order to get us back on an even footing with all of the changes (as 2 x v2.91 isn't going to help with the fragmentation of things as is).

-daz
Yathosho#
Originally Posted by TomyLobo View Post
you should really modify aperesize instead, so it resizes dynamically 🙂
that way you also don't have to patch again for each new binary release.

I can explain you the code if you want, too
actually, that sounds good. talk on #avs?
TomyLobo#
that offer is 2 years old, i'm not even sure i have the source anymore or that i understand it any better than you would
drivetheory#
Originally Posted by justin View Post
Oops, sorry I forgot to update CHAVS. Here are new builds (v2.91):



and the code:



(the code now includes the required bits from WDL, too, so no need to git clone anything).

This should fix the rand() bug, and also updates to even newer EEL-compilation, which has some fancy stuff such as basic code optimization.


vis_chavs.dll
MD5: b7adf0e8f6c63c8c5160f8190255860e

shows as "Advanced Visualzation Studio v.2.81b [vis_chavs.dll]" inside winamp and is giving me a CTD when ever i try to run it in Winamp 5.63...

vis_chavsmod-2.9_avsmod-0.5.5.dll still works perfectly fine for me but the new one listed above doesn't

anyone else having this problem?
drivetheory#edited
The problem i was having with vis_chavs.dll it seems was caused by the box for "SMP support" (AVS Editor -> Settings -> Display) being checkmarked
Gourmand#
WOW! I'm waiting for new AVS through years! Dreaming see my VU-meters (vapourMachine, BatmanRadio and others) presets om my Android tablet and iBasso DX50 player in Rockbox. AVS FOR HANDHELDS! But I do not think that browser JScript code is the best way. I'm sure Qt-based renderer would be much better and faster. But I will wait for any implementation.