[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Public WebGL] Changing extension spec boilerplate



Hi David,

On 22/06/2012 03:33, David Sheets wrote:
None of the official extensions appear to use this section and I know
of no unofficial extensions that use this tag. Also, I don't think
WebGL extensions should be adding any normative text to OpenGL ES 2.0
chapters. The present section title and associated meaning are
entirely due to the direct transcription of the OpenGL ES 2.0
extension template.

Unless there are other objections, I believe <additions> should change
to WebGL spec additions. Here is the bug and patch:
<https://www.khronos.org/bugzilla/show_bug.cgi?id=662>
I agree. Somehow the patch was not attached to the bug. Once it is, I'll apply it in my work area and submit it.
warning: failed to load external entity "/extension.xml"
...

A copy of the stdout/stderr would be helpful. Something is failing in
the map over the extension list, EXTS. What operating environment are
you executing "make" under?

I have submitted a minor patch at
<https://www.khronos.org/bugzilla/show_bug.cgi?id=661> that may fix
your problem if you have extensively edited the EXTS list in Makefile.
Like bug #662, there is no patch attached. However I figured out what you are talking about and make the change myself. The effect is to change the ",," between each file in the extensions list submitted to xsltproc with a single ",". Somehow this cured the above problem, but I don't understand why.

For what it's worth I'm running the make under cygwin.
...
XMLMind. It understands DTD's W3D XML Schema and RELAX NG Schema.
I've got another issue. XMLMind is reformatting the text within <idl></idl> basically converting all occurences of white space to a single space and creating a single long line. This means the generated html is incorrect. I tried fixing it by using <br/> and &nbsp; (after defining that entity). XMLMind is retaining the &nbsp; but not the <br/> so it's still broken.

I think it probably related to the "Guess ignorable whitespace" preference in a box labelled "When there is no DTD or schema." Is it possibly to provide a minimal DTD that tells it not to ignore whitespace inside <idl>? I don't know XML well enough to know how to do that. I'd rather not change the preference because it is global.

Alternatively any chance of improving the <idl> parsing so it's not totally dependent on the text layout?
Great! One of the most useful applications of these specification
documents is the automated extraction of GLSL symbols and types for
shading language processors. If your extension proposal extends the
GLSL global namespace with new macros, types, variables, or functions,
I would like to work with you to declare this in extension.xml in a
structured way.
I hope to post the first draft early next week. I'd be happy to work with you improve extension.xml.

Regards

    -Mark