[Plib-devel] Release 1.8.5....and a glimpse of the future.

[Steve B. <st...@sj...>]

I guess I have no excuses left then!

I'll get to it at the weekend.

Actually, I've been doing work on "Son of PLIB" in odd idle moments - I 
have a version that'll work with OpenGL 3.0 (or whatever they are going 
to call it) that uses 100% shaders - no standard pipeline at all.  As 
with OpenGL, my main effort is to slim-down, simplify and provide more 
flexibility for data formats and plugins for algorithmic flexibility.

But I'm not making any effort on backwards, forwards or sideways 
compatibility...anything I need to change, or merely feel like changing 
- I'll change.

Multitextures and decent multipass, bucket sorting and a 'posteffects' 
mechanism are all there - plus Java scripting and XML-based 
configuration files.  Bulk vertex data can have user-specified data 
types and you can add your own per-vertex attributes to feed data to 
your shaders - so stuff like normal-mapping is now easy.

But I'm developing it for a rather specific purpose and I don't plan to 
do the work collaboratively because I have a clear vision of what I need 
and I don't want to cloud it with other people's needs until I'm ready - 
I will, however OpenSource it under LGPL when it's done.

The myriad of SSG node types has been slimmed down to just two - "Leaf" 
and "Branch" - the Leaf will hide the data transfer mechanisms to the 
GPU and the Branch will use plugins that operate at cull-time to 
simulate the old ssgTransform, ssgRangeSelector, etc.  ssgTexture gets 
some love and care so it can do 1D, 2D, 3D and Cube maps - ssgState 
becomes a container for ssgShaderPair and a bunch of ssgUniform 
variables that can (of course) be bool, int, half-float or float...or 
ssgTexture handles - all of the messy lazy state switching code 'goes 
away' (thank God for that!).   The SG math library supports every data 
type from char to short to int to halffloat to float to double with 
almost all of the usual functions - and to keep it all up to date, it's 
mostly auto-generated code.

What mostly needs to be done is to have the PUI library "go away" and be 
implemented as a layer on top of SSG - and the SL audio code "go away" 
and be replaced with an OpenAL wrapper...I'll probably release what I 
have before I attack those things.  Sadly, none of the file loaders will 
survive intact and it's too much work to rewrite them all.  My son is 
Blender fanatic (<sigh>) and I have a blender plugin that writes files 
in a ".plib" format (which is XML).

Anyway - this is off in the future somewhere - I'm not making promises 
or commitments.


Jan Reucker wrote:
> Hello,
>
> here's a patch to take away the last excuses not to release PLIB-1.8.5.
> It sets the version information in configure.in and ul.h to 1.8.5 and
> contains all necessary changes to the web site.
>
> Needless to say that PLIB SVN passes "make distcheck" without errors.
> And there hasn't been a bug report or submitted patch for a while, so
> a new release is overdue.
>
> What's left to do (from my point of view):
>
> - Someone with SVN access should apply the patch.
> - Someone has to run "make dist" to create the 1.8.5 tarball
> - Someone with SSH and SVN access must update the web site
>   (emulate the transferCvs2WebSite script)
> - Someone with file release privileges must upload the tarball
>   to SF.net
> - "make distcheck" fails for the examples, at least on my PC,
>   but I don't know how to fix it.
>
> Did I forget something?
>
> Kind regards,
> Jan R.
>
>   
> ------------------------------------------------------------------------
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2008.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> ------------------------------------------------------------------------
>
> _______________________________________________
> plib-devel mailing list
> pli...@li...
> https://lists.sourceforge.net/lists/listinfo/plib-devel