I did a CVS checkout of the head, jakarta-turbine-2, which is where all the commits seem to be going.
I then did a grep over the sources looking for the various points at which classes in the org.apache.velocity.* hierarchy are referenced.
I found 21 (!) different source files which reference Velocity classes. Now, there was again a reaction of irritation with this. Henning and a couple of other people had proposed (with a certain aggressive subtext) that there was some onus on *me* to submit a patch if I "wanted" the FM support back.
Well, as things stand, I have to question whether this proposal was really made in good faith, since obviously, with that many points of coupling, it would be very hard for me to submit a patch.
For example, I just looked at two different frameworks that support Vel, WM, and FM. Niggle is one I wrote myself. There is also JPublish, by Anthony Eden. I did a similar grep and found 3 files that reference Velocity classes in Niggle and 4 such files in JPublish.
In the above cases, it is quite easy to submit a patch to support another template engine, since with that level of coupling, you can quickly identify the points at which these things are used, and which hooks need to be implemented and how.
So, okay, where does that put us? People have made noises about my being helpful or not being helpful.
I previously said that I wasn't going to submit any patches for FM support in Turbine, because my position was that if you guys pulled it out for whatever reasons, it was your task to put it back in. I still feel that way, but I have decided to recant on that statement and put the ball in your court.
If you guys can get the codebase to the point where the coupling is reasonable -- i.e. only a few points in the entire framework reference org.apache.velocity.* (as is the case with Niggle or Jpublish) I will commit to doing the rest. Ideally, you would also localize the references to Velocity in the docs, so that I could simply patch in something that tells how to use FM, and thus Velocity and FM could be supported for the first time on a somewhat equal footing.
IOW, I'll put in the FM support, but you have to clean house first.
I think that's a reasonable proposal and I think it puts the ball in your court in terms of there being some seriousness about doing something on this.
Regards,
Jonathan Revusky -- lead developer, FreeMarker project, http://freemarker.org/
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
