The pragmatist pledge

I had a revelation a few days ago.

Years ago, I had a tendency to hit my computer screen when the computer would not do what I wanted it to do. I stopped that at some point; like I said, it was years ago. That's not the big revelation, though.

The big revelation was the fact that the day of me stopping to hit the screen happened to coincide with the day of me reformatting my Windows partition to free up space for /home.

I guess that, more than anything, is the main reason why I've never looked back. Though I subscribe more to the Free Software philosophy than I do to the Open Source philosophy (that is, "software should be free" rather than "open source improves software quality"), I do not think Free Software is the end solution to all problems relating to software. In that regard, I am a pragmatist: if there is a free alternative, I'll prefer to use that, but I do not shirk away from using something which is non-free—or not entirely free—when it will solve a problem.

In that context, I have recently been somewhat depressed and demotivated because of some things that have been happening of late in Debian. Without wanting to comment on the case at hand, I must say that after having been a Debian Developer for almost eight years now, I am starting to see a pattern:

  1. Debian releases
  2. Developers are happy, and start working on all sorts of new code to put in Debian
  3. We work a lot, and interesting things happen all over our code base.
  4. Somebody (presumably someone from the release team) calls a freeze.
  5. Because new features are a no-go now, suddenly loads and loads of people have too much time on their hands.
  6. Loads and loads of people spend time looking at other people's bugs, and see things they do not like.
  7. Suddenly, it becomes immensely important to resolve that issue now. Never mind that we're about to release, that can wait. Never mind that some people actually depend and/or wait on the next Debian release. That doesn't matter. All that matters is that this issue must be resolved now, because it's not free!!!1!!.

As a pragmatist, I'm extremely unhappy about that. Yes, I agree that getting Debian rid of non-free software is a worthy goal, but not at every cost. Yes, I agree that releasing Debian with non-free software is a bad thing, but did people really have to wait until we were about to release to bring that issue up?

The problems were out there for a very long time. Ignorance is not an excuse; if you cared about these things so much, you could have looked for them earlier, too.

I'm not about to say what people should and should not do during the freeze. But as a pragmatist, I do not think that bringing up such issues when we're trying to release tings is a good idea. So I'm putting forward this pledge, which I call "the pragmatic DD's pledge":

I will not, deliberately or otherwise, delay the release of the next Debian stable, by proposing or seconding a GR that will have an impact on the release schedule when the freeze is in progress. I will instead make sure I detect such issues before the freeze has started, or will deliberately wait until after the release has happened before bringing these issues up.

As for the current Lenny vote, here's my ballot:

[ 4 ] Choice 1: Reaffirm the Social Contract
[ 2 ] Choice 2: Allow Lenny to release [...]
[ 2 ] Choice 3: Allow Lenny to release [...]
[ 2 ] Choice 4: Empower the relase team to decide about allowing DFSG violations
[ 2 ] Choice 5: Assume blobs comply with GPL unless proven otherwise
[ 1 ] Choice 6: Exclude source requirements for firmware (defined)
[ 3 ] Choice 7: Further discussion

My only wish is that I would now also be able to find some more time to fix some RC bugs. Sigh...

(as an exception to my policy, comments on this item will be silently dropped. UYOB)

update: Only after I posted this blog entry did I see that many other people had issues with this insanity; and I see some problems with some of the voting strategies that people propose. Let me make something clear: if you do not want choice 1 to win, then you must not vote any of the other options below "Further Discussion". The reason is that since those need a 3:1 supermajority, and since supermajority is decided by whether an option on the ballot beats Further Discussion in our voting system, this means that if more than 25% of our voters vote everything below "Further Discussion", Choice 1 will win by virtue of being the only option reaching its required majority.

If you think, like I do, that "releasing Debian" is more important than "being able to tell people that vrms isn't lying to you", then you should vote Choice 1 below "Further Discussion", and vote all the other options above it, even those which you think are not the best option; because if your preference does not win, then at least you allow the other options that you can live with to stand a chance of reaching the required supermajority, and will allow them to win.

Also note that if you've already voted, you can still change your vote!