Saturday, January 3, 2009

No more brainstorming

As some of you have already noticed, the Brainstorm site was taken down not long after the launch (for those of you subscribed to the Pidgin feed, apologies for the redundancy).

This is not a post for debating why this happened. If you want to learn more, read the development mailing list and maybe even send an email.

In lieu of Brainstorm, you can now vote on individual issues on the development site.

I'm interested in knowing what you think you think of the alternative (since comments are disabled in the other posts). If it's better, great! If not, why?

If you did not see Brainstorm, here's an identical example.
Posted by Casey Ho at 5:24 PM
Labels:

17 comments:

  1. Thanks for trying.

    I'm not coming anywhere near Pidgin's trac again - way too make gtfo-like comments or "give us code then talk" (and then they ignore the code).
    ReplyDelete
  2. >I'm not coming anywhere near Pidgin's trac again - way too make gtfo-like comments or "give us code then talk" (and then they ignore the code).

    And that's exactly what this exercise showed - the majority of user feature requests were already known to pidgin developers:

    Whiteboard - pidgin/gtkwhiteboard.c , Already included but not enabled by default?
    Jabber Priority setting - #1345, submitted 20 months ago.
    Jabber Transport Registration - #556, submitted 20 months ago, *has a patch included*.
    Voice and Video - #34, submitted 2 years ago. I noticed that this milestone is set for Feburary 2009 - that's exciting to hear!
    Gnome Keyring Master password - #673, submitted 20 months ago, has a GSoC implementation!
    ICQ xStatus - #4508, submitted 12 months ago, will be in 2.6.0!
    Saving/Using Animated Smileys - #2067, submitted 12 months ago.
    Message Styles - #730, submitted 20 months ago, there is an improved version attached!
    Email notification improvements - #780, submitted 20 months ago, full reviewed patch attached. I personally tried to push this one in #5763 only to be told it had been forgotten and will be in 2.5.0.

    In many of these cases the code is even *right there*, waiting for review and inclusion. I'm not oblivious to the fact that fixing some of these things would require considerable time and experience but there should be a concerted effort to include patches where someone has provided them.
    ReplyDelete
  3. I think that while the Brainstorm thing was an interesting idea, the concerns about Drupal and duplication were probably valid. I think the best thing to do would be to do a better implementation of the voting plugin in Trac. Currently, it's worse than Brainstorm for the users but I suppose it can be improved over time. In the end, I think it's better to do something that the developers will actually look at and will actually have an effect on Pidgin development.

    I think there are a few major problems with Trac though, particularly massive duplication of tickets (such as the recent one about the Windows hang on exit). I don't know whether there's a way to check for duplicates in Trac but it would be a nice feature to have or try implementing. It would also be nice to have more lists of commonly requested tickets in a more visible place to cut down on duplicate tickets. I might download the Trac source and mess around with it to see what I can do on that front, although my Python proficiency is pretty low currently. I also feel like the Trac wiki is a bit disorganized but I'm not sure how it might be better organized.
    ReplyDelete
  4. The Trac voting feature sucks. It's slow and there's no visibility of the most voted features.

    The worst is that some brainstorm ideas don't qualify as track issues.
    ReplyDelete
  5. All this voting business is very nice, but I think you should concentrate on releasing a flawless windows version of pidgin. I'm getting complaints from many people over the last days who don't understand why pidgin is crashing everytime they want to close it and are losing faith in the program. It's very hard to be an evangelist for alternative IM-clients if water is not turned into wine but into mud.
    (Nevertheless Pidgin is a really great tool and I'm very thankful for you all to be developing it - but releasing a flawed version really scares people away)
    ReplyDelete
  6. How about a brainstorm idea: reduce the number of crashes in windows.

    That can't be translated into a trac ticket, can it?
    ReplyDelete
  7. "reduce the number of crashes in windows" is also a useless idea. We obviously don't want Pidgin to crash, but the individual crashes have to be reported to us. If everyone votes for fewer crashes (hell, I would too) that isn't going to help anyone at all, ever, period. The actual issues have to be reported in trac or directly to a developer to get something done, or the user has to fix the problem and put a patch in Trac. See it all comes back to Trac.
    ReplyDelete
  8. Kevin: "reduce the number of crashes in windows" is just an idea. Maybe nobody cares about that and nobody votes for it, but maybe it's important to a lot of people.

    Let's suppose that a lot of people vote for that idea. It will become more and more visible and it would gather a lot of comments.

    The discussion would probably generate some other ideas, perhaps some people would try to narrow down the issues. Possibly to some protocol(s) or Windows versions.

    I bet then people will create trac tickets for the issues and link them to the brainstorm idea, just like in Ubuntu brainstorm and as the trac tickets get solved eventually the brainstorm idea is marked as done and everybody is happy.

    Unrelated trac tickets don't get that much visibility.
    ReplyDelete
  9. Drag and drop names into chatrooms or creating chatrooms by just dragging and dropping names into a window just like Skype. That way I don't have to go through some strange way to set it up. No one else I know knows how to even set one up, and I've had it crash Pidgin multiple times trying. I just try to find someone using a real client to do it for me and add me in.
    ReplyDelete
  10. Asking for a Ubuntu package isn't exactly a 'bug' either, and given the attitude on trac, it most likely will be closed as "WONTFIX GTFO MY TRAC" right away. Yet people wanted it, as Brainstorm showed.
    ReplyDelete
  11. felipec,

    "Reduce the number of crashes in Pidgin" is obviously a priority. I don't think any skilled dev in the world would let a segfault slip in their code.

    Obviously it's a problem and needs to be fixed, but that's a whole another tangent.
    ReplyDelete
  12. Maxious,

    When it comes to patches, I think John and some others have been doing a decent job of yelling at people to get them integrated. Obviously not everything is in, but progress was made the past few months.
    ReplyDelete
  13. Vadi,

    Point taken. Don't worry, more improvements should come in the future.
    ReplyDelete
  14. Voting on Trac is certainly not better -- it's not made to be a voting system. There is no easy way to come along and see which issues are the most voted on, and there is no easy way of voting; you have to view the entire issue before you can vote.

    I really liked having the Brainstorm just for the fact that it was easy to see what kind of ideas people had out there. They had better visibility. With Trac, not so much.

    Though, I guess it doesn't really matter anyway since they've practically said they'll be working on what they want and not what gets voted on.
    ReplyDelete
  15. From Maxious's comment, I found one patch that was mis-categorized as an enhancement, this has been fixed, and it should show up on John's, Etan's, and other developer's radar.

    The Webkit one is something Sean worked on for a while, but in a manner unfortunately typical of him, never put into monotone, even in a separate branch. I also don't see a release of it there on the launchpad site, though apparently a number of people are using it with at least some success. A patch will have to be submitted at some point to us.
    ReplyDelete
  16. Since developers don't want to support a brainstorm, I've initiated an "unofficial" brainstorm.

    Nicolas Deschildre was kind enough to provide the infraestructure required, and the initial setup. Thanks!

    So vote up for your ideas!

    http://pidgin.ideatorrent.org/
    ReplyDelete
  17. "I don't think any skilled dev in the world would let a segfault slip in their code."

    But they do, and a lot. And they usually don't care, as long as it works for them. Pidgin devs have a clear tendency to scratch their own itch exclusively.

    Proof of that is the increasing number of crash reports for msnp15, but nobody cares.

    A proposed solution would be to have priority and severity just as any decent bug tracking system does. That way it would be easier to track down critical issues.

    But a blog is not the place to discuss that; a brainstorm is.
    ReplyDelete

Subscribe to: Post Comments (Atom)