An Object Lesson in How to Respond to Criticism

I can honestly say - I was happy at the reaction to Eric's talk at DjangoCon, and I loved being part of the discussions after the fact.

After seeing the way the Django core team has since "reacted" - maybe not just to Eric's talk, but to the general criticisms levied against them as a whole - I can honestly say I'm proud to have them all be a part of the Python community.

Eric is perfectly right: That's how you respond to things.

Speaking from the other "side" in this:

There were, if you were paying attention, actually two keynotes which bashed on Django, thus equaling the combined number of Django-bashing keynotes from the first two DjangoCons :)

Anyway. The committer policy was obviously becoming more of a hindrance than a help, and the focus in the last couple releases on landing big stuff (and, thus, letting less-flashy but still valid and useful feature requests languish) was probably good at the time but just doesn't work long-term, as evidenced by the backlog of feature-request tickets.

I think there was a general understanding -- among the core team and among the broader community -- that these in particular were the two biggest issues, and having them aired out publicly and discussed in person instead of on a mailing list or a forum was important. DjangoCon's traditionally been good at that sort of thing, but this time around it was even more important than normal, and I think that's one of the great things about the conference. It's much harder to flame someone who's standing a few feet away than it is to go after somebody on a mailing list, and it's much more likely that calm will prevail and productive discussion will ensue.

And speaking of productive discussion, that's the other great thing about the conference: it's about the only time all year where we can get most of the committers into the same actual room to talk about stuff. There were a couple meetings -- one at the hotel, one over dinner -- which were fairly productive; the new committer policy came out of that, was proposed on the committers' list shortly after the conference, and approved pretty quickly since most of the issues had already been talked over. Doing the same thing electronically likely would have taken months; thanks to DjangoCon it basically only took a couple hours, followed by the formal proposal and vote, and then the drafting of the announcement and the process of offering the new commit bits.

Making software is an inventive process and invention is just as much about tearing down ideas as it is about building them up.

Entire phases of development, tracking systems, and personnel are dedicated to fixing defects. In other fields of engineering, it would be considered a failure for such defects to exist in the first place.

Mistakes and criticism are fundamental to software development. Accepting this and not turning it into an emotional roller coaster is one of the biggest hurdles for new developers and for managers.

I don't think that someone would react different, all things considered. The critiques were firm and had ground. Django core is used to criticism (think of Cal Henderson's Keynote), and the community slowly became aggravated with the things we hate in django. It was coming.

I really liked the keynote, and i salute the man that had the courage to do this. Lets hope this keynote makes deep impression on django core and sticks for some time so we can see the community and the project become better and better.

Kudos Eric.