[WM] Any WebMake Activity?

[Bryce Harrington]

On Sat, 15 Mar 2003, Justin Mason wrote:
> Bryce Harrington said:
> > We've been using EtText along with template-toolkit at the Open Source
> > Development Labs for the website and some general documentation.
> > However, we're probably going to move away from ettext in favor of
> > Text::WikiFormat, due to issues with our JP lab processing Japanese
> > language pages with it.  So we'd love to hear if there are any plans for
> > further development of EtText.
>
> Definitely.  I think EtText is the best bit, myself ;)
> I'd be interested in trying to fix those issues BTW...

That'd be cool, we'd love to see them addressed!  I've cc'd the
relevant OSDL folks (jun is the original requester, dave is a heavy
ettext user here, and has several thoughts for improving it I hope he'll
pass along.)  I'll also cc the list of issues your way.

> > Btw, I created an additional tool called 'tgen' to make it a little
> > easier to use template-toolkit for processing ettext-formatted pages;
> > it's somewhat analogous in operation to WebMake but bases processing
> > decisions on file extensions rather than Makefiles.  I'm likely to be
> > putting out a new release in the near future, so would love to hear
> > feature requests.
>
> BTW, I think you're on the right track -- I've been contemplating redoing
> WebMake using the ttk for templating, since the WM templating system is
> quite simplistic, and has some issues as a result.

I've been quite impressed with TT2.  I've also noticed that others have
worked on some integration of this and CGI::Wiki, and I've been
considering that tying in with that could result in a pretty powerful
toolset (allowing for online-editing, for instance).  Have you looked at
CGI::Wiki?  Any thoughts on it?

> What I'd do would be ttk+ettext+webmake's metadata model+webmake's
> dependencies tracking, with possibly some new tags, one to make it more
> intuitive to convert a tree of files into output files, and a tag to draw
> in media from a location that doesn't already exist in the output tree.
> That's the basic TODO list right there ;)

What I've done with tgen is built a plugin system for handling
pre-processors; currently implemented are handlers for ettext, man, and
pod, but it's straightforward to add new processors.  tgen is smart
enough to only build changed files, but doesn't do dependency checking
(I don't believe tt2 does either).  It's unlikely I'll need it though.
It doesn't do page-level metadata management either, though that is
something I'm going to need in the near future.

> But it's a big task, and with all the work on SpamAssassin, I haven't had
> much time to do this... :(

Wow, I didn't realize you did SpamAssassin too.  Coolness.  In fact
we're looking into setting that up here for the various mailing lists we
host for the linux kernel development community.

Thanks,
Bryce