> -----Original Message-----
> From: [email protected] [mailto:[EMAIL PROTECTED] On
> Behalf Of Endre Bakka
> Sent: Thursday, June 19, 2008 2:37 PM
> To: [email protected]
> Subject: [Trac-dev] Re: Is Redmine a better Trac? What's gone wrong
> with Trac?

[snip]

> Example: Just two weeks ago, Jeff Hammel asked about including
> AutoQueryPlugin into the core, and he listed excellent reasons why it
> should be included (I agree with all of them). Nobody answered his e-
> mail.

Lately we have been slightly more focused on final testing before getting
0.11 out the door. I do agree that feature should be added, in fact I
thought it already was when I read the description.

> Example: I wrote a patch for subcomponents since our company need this.
> It
> was ignored, both the ticket and an e-mail. I started asking around on
> IRC
> and was told it was clean, but way too large to include in the core so
> I
> should make a plugin.
> 
> It took me a couple of evenings to implement, and 90% of the time was
> spent trying to understand the Trac codebase so I was suprised when
> told
> it was too big (btw, "very high quality code" should be easy to read
> and
> understand - I think Trac still has some potential in that respect).
> Well,
> I wrote the plugin, but it was not a pleasant experience. It was much
> harder to write, uglier, required a plethora of hacks due to
> shortcomings
> in the plug-in API, but I managed. I figured I would wait until 0.11
> was
> released in case last minute changes affected my plugin. I've been
> waiting
> since December.

I would rather see the API fixes in Trac core than the original patch.
Subcomponents really are a feature I would rather see in a plugin, because
the vast majority of people don't need or want them.

> If I had gotten at least some feedback (even "your code sucks, clean it
> up"), you would have seen many more patches from me.
> 
> Suggestions:
> - Live up to your mission statement. Do it the Bazaar way (see book for
> details)
> - Spend a bit less time on development and a bit more time at looking
> at
> other peoples patches that could be included. If you do, Trac will move
> forward at a much higher pace. Users will be encouraged and more people
> will join and write patches to include.
> 
> Add functionality in baby steps. Some crappy support is better than
> zip,
> zilch, nothing. Users can give feedback (even in the form of patches)
> as
> you go along steering you towards a better product. Nobody (maybe
> except a
> handful in the world) are able to design and write the perfect system
> without a couple of iterations. Everything can be rewritten (with a
> proper
> testsystem, without fear), and database tables can be altered anyway
> you
> like during an upgrade.

Apparently you didn't read what I said. We specifically avoid this model. It
leads to poor design and even worse implementations.

> There is a reason why so many "other FOSS projects" work this way. It
> is
> more efficient. It is the Bazaar way.

It also leads to giant, messy piles of garbage. It shows when something has
been built with a single, coherent vision, and I think this is a large part
of why people like Trac in the first place.

--Noah


--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "Trac 
Development" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at 
http://groups.google.com/group/trac-dev?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to