On Sat, Oct 12, 2002 at 12:10:50AM -0700, Tom Lord wrote:
>
> > On this point, I very much agree. In fact, I agree so much, that I even went
> > and coded this some time ago:
>
> http://subwiki.tigris.org/
> http://svn.webdav.org/repos/projects/subwiki/trunk/
> http://test.webdav.org/wiki/Welcome
>
> > I wouldn't even call it alpha code right now. You can't add a page except by
> > checking out the pages, doing 'svn add', and committing it back. After I get
> > unburied from some other work, I'll be doing that. And now that Mike got
> > ViewSVN support going, I'll be hooking that in, too. And then... :-)
>...
> Take Twiki (verify that it's free software). It's in perl and I
> remember someone (you?) having objection to perl.
I don't "object" to Perl. I just think it is a poor choice for most uses,
and I personally choose to use Python for as much as possible.
> There's also
> a wiki of good repute or two in python (i forget what they're called).
Piki, PikiePikie, MoinMoin, and many others.
> Twiki, at least, already has RCS hooks. Replace those with svn hooks.
> This should be trivial.
I'm not sure why you suggest to "take" any of these others. Or to replace
their RCS hooks, or anything.
As I wrote above, I've *already* implemented a Wiki that uses SVN for its
storage. A person can modify pages through the Wiki or through SVN itself.
You can browse the Wiki-fied content, the bare pages via a web browser (or
other DAV client), or through checking out the pages. etc etc.
> Start writing wiki extension modules that take advantage of svn's
> transactional nature. For example, a user filling out a form can
> result in changing, atomicly, a bunch of "correlated" files in the
> wiki.
Oh yah... there are definitely a lot of interesting avenues. I just haven't
had time to complete some of the basic SubWiki features. But the basic
architecture and features are there.
> This needs no feature of svn that isn't already there. Application of
> the resulting very flexible and very scalable framework can drive
> subsequent optimization efforts and clean-ups.
Sure.
> I think svn 1.0 is pretty much done.
I beg to differ. Look at our issue tracker. There are quite a few things
that (IMO) really need to be done for a 1.0 release.
We didn't even call our code "alpha" until we'd self-hosted for over 10
months. We aren't planning to call it "beta" until we drive out almost *all*
of the known bugs. And 1.0? After it "percolates" for a while in users
hands.
Personally, I think a version control system's #1 feature is "stability and
bug-free." I want to tell people, "hell ya, it's solid. just try and break
it. that thing is gonna be central to your development organization, and it
WILL NOT let you down."
Cheers,
-g
--
Greg Stein, http://www.lyra.org/
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Sun Oct 13 07:49:42 2002