>Mitchell L Model wrote: > >>kdeaddons is not in fink, and when I tried to compile the downloaded source I >>got errors I couldn't get around -- even compiling just the one specific >>addon I wanted (the kpython plugin for kate). >> >>Is anyone working on kdeaddons? Is there some known general problem that >>affects kdeaddons in addition to other packages? >> >>It has occurred to me before and now seems particularly important that it >>would be very useful for fink to include a list of packages that are NOT >>included, what their status is, who is working on them (if anyone), why they >>haven't been included, what technical problems have blocked their inclusion, >>etc. That way someone who wanted a particular package could learn what >>others already had discovered, offer to help, take responsibility for the >>package, etc. Is there something like this somewhere? >> >Are you volunteering? What you're suggesting is very hard to do--the >"available packages" list (on the website) is generated from those packages >which are currently in the distribution. What method would you propose to >access _all_ of the various open-source packages that aren't included? If >it's not automatable then it's not going to happen--this is a lot of effort.
I really don't know enough about how the Fink packages come into being to clearly ask the questions I want to ask. Answers like yours help me understand what is possible and what I'm really looking for, as well as give more focused examples of what I want to see happen. You're right, of course: it wouldn't be possible to track all open source packages -- where would we even find them all? What I had in mind was packages that had come to the attention of either users who want them or developers who are already providing packages, especially packages relating to the ones requested. Part of what I'm concerned about is avoiding duplication of effort. I'm also concerned that if several people start to work on a package they either share their experience as they work or one of the people who could do it the fastest and easiest takes over. > >Some of what you've proposed on the package submission tracker--people post >trial packages for evaluation and comment. There's also a package request >tracker, to request packages not in the distribution. Both of these are >linked from http://fink.sourceforge.net. I didn't know about either of these. I realized after I sent the message that my comments about finding out about tracking package status -- which were a reflection on my real request, which was specifically about kdeaddons -- were premature and that I should look at the FAQ and the site first. I apologize. I am an extremely experienced developer and consultant, and I thank the community for Fink every day, but I haven't been involved at all (yet?) -- from the outside getting involved in Fink package development, or even finding out the status of packages, is a pretty intimidating prospect. I'm sure it's no big deal once I learn the basics, but I didn't really even know where to start. Now I do, so thanks. From your description, though, there is still a significant hole: there's no tracking of efforts to port a package that didn't reach the point where they could be released as trials. For instance, I spend some time trying to compile kdeaddons, but the compilation ran aground with some pretty serious C++ errors for both the whole package and for just the kpython addon I really wanted. So I gave up and posted my question. For all we know that's what happened with 10 other people. In fact, the thing I am most worried about is not the waste of duplicated effort but the waste of struggling to get something to work when someone has already tried and figured out why it won't or what major task would have to be accomplished before it could. I'm certainly willing to poke around the code a little and try to hack around the errors, but if someone already knows there's no way to get it to work or that fixing or bypassing one error leads to another, for a long sequence of one-at-a-time fixes, I don't want to pursue this any further. That's the sort of information I couldn't figure out where to find. I'll certainly post my request to the package request page you mentioned. Maybe there should just be a Wiki for discussions of efforts to port specific packages, with a Wiki page for each package and a few pages about how to work with packages, release packages, compilation hints, problems common to many packages with their common solution, etc. I could go there and look for the kdeaddons page; if it wasn't there, I could start one and say what problems I ran into; if it was there, I could see if other people had hit the same problem and whether they had gotten past it, etc. Or even just a text file somewhere with entries such as "kdeaddons: [EMAIL PROTECTED], 2005-05-12; tried to compile and ran into too many C++ errors; gave up." > >The core developers frequently just commit test packages to their experimental >areas on the fink main tree, so those packages are harder to track. Yes, that's part of the problem. For all anyone knows someone is working on kdeaddons right now and almost ready for the first release. Or maybe no-one has looked at it at all. That's one of the problems I'm identifying. --- Mitchell ------------------------------------------------------- This SF.Net email is sponsored by Oracle Space Sweepstakes Want to be the first software developer in space? Enter now for the Oracle Space Sweepstakes! http://ads.osdn.com/?ad_ids93&alloc_id281&op=click _______________________________________________ Fink-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/fink-users
