“3.0 (quilt)” is the most widely used Debian source package format

My goal with the “3.0 (quilt)” source format has always been to standardize the patch management in Debian source packages. This message seems to have been well understood. dbs and dpatch have been deprecated by their respective maintainers.

I made numerous efforts to make this source format useful in as many use cases as possible (but some improvements are still possible) and I have added hints to encourage maintainers to switch. Thanks to this, the adoption rate of this new source format has been very good and it’s now the most widely used source package format in Debian—only two years after its introduction in Debian unstable.

With 9829 source package using “3.0 (quilt)”, it surpassed the number of source package still using “1.0″ (7368). (Those numbers have been taken from http://upsilon.cc/~zack/stuff/dpkg-v3/ on december 13th 2011.) The number of source packages using “3.0 (quilt)” doubled this year.

(Click on the picture to see it full size)

Of the 7368 packages using the old format, 6816 packages trigger the missing-debian-source-format lintian tag. This means that only 552 source packages have explicitly opted to keep using the old format and that the bulk of the remaining packages are rarely updated packages that have not been switched yet.

Debian 6.0 is out, Wheezy kicks off

As you probably already know, Debian released Squeeze aka Debian 6.0 this week-end. It was a really great week-end.

I saw quite a few release already, but none with so much online social activity. Alexander and Meike Schmehl live commented the release on the Debian identi.ca account and Joey Hess held his Debian Party Line.

On top of this, the team in charge of the website rolled out a new design on quite a few online services during the week-end including www.debian.org, wiki.debian.org, planet.debian.org and more.

Congratulations to all the people who made this happen (and the release team in particular). It’s great to see Debian achieve all of this.

Do you know that it’s the third time in a row that we manage to release in the 18-24 months timeframe? 3.1: June 2005 → 4.0: April 2007 → 5.0: February 2009 → 6.0: February 2011.

Yes, despite our size and the fact that we are all volunteers, we have managed to stick to a reasonable schedule for a stable distribution that is deployed on a large scale.

Wheezy kicks off

The best part of the release for us — the developers — is that wheezy is now open for development and we can work on new features for the next release. ;-)

And it started quickly: according to UDD, wheezy already features 488 new source packages that are not in squeeze, 1713 updated source packages and among those 1246 are new upstream versions.

I really look forward to the upcoming projects and related discussions.

Click here to subscribe to my free newsletter and get my monthly analysis on what’s going on in Debian and Ubuntu. Or just follow along via the RSS feed, Identi.ca, Twitter or Facebook.


For the curious, here are the UDD queries I used:

# Updated packages in wheezy
select count(source) from sources_uniq as su where (select version from sources_uniq where release='squeeze' and distribution='debian' and source = su.source) < su.version and release='wheezy' and distribution='debian';
# Updated with a new upstream version
select count(source) from sources_uniq as su where debversion(regexp_replace((select version from sources_uniq where release='squeeze' and distribution='debian' and source = su.source), '-.*', '')) < debversion(regexp_replace(su.version, '-.*', '')) and release='wheezy' and distribution='debian';
# New package in wheezy
select count(source) from sources_uniq as su where source not in (select source from sources_uniq where release='squeeze' and distribution='debian' and source = su.source) and release='wheezy' and distribution='debian';

Secret figures of a Debian/Ubuntu blogger: what you liked most on raphaelhertzog.com

Chart goes up on screenI launched raphaelhertzog.com this summer (taking over the English content of my former multi-lingual blog), when I decided that I would be more serious about blogging on Debian/Ubuntu related topics. On September, I decided to write 2 articles per week and up to now I managed to keep the schedule.

Two of my articles were published by Linux Weekly News, those are much more researched than the average blog article (they are tagged with [LWN] in the list below).

The most popular articles

Most people read my blog through the RSS feed which happens to be syndicated on Planet Debian and Planet Ubuntu. According to the feedburner’s statistics, the top-5 articles are:

  1. 5 reasons why I still contribute to Debian after 12 years (32700 views)
  2. [LWN] Understanding Membership Structures in Debian and Ubuntu (31700 views)
  3. Social Micropayment Can Foster Free Software, Discover Flattr (30100 views)
  4. Everything you need to know about conffiles: configuration files managed by dpkg (29900 views)
  5. How to make 110.28 EUR in one month with free software and Flattr (29400 views)

But I also have occasional readers visiting my blog because my articles are announced on Identi.ca, Twitter and Facebook (and they circulate on social networks, thanks to those who are sharing them!). The top-5 articles according to the statistics of my website are:

  1. 5 reasons why I still contribute to Debian after 12 years (6000 views)
  2. [LWN] Can Debian offer a Constantly Usable Testing distribution? (5000 views)
  3. Understanding Debian’s release process (1500 views)
  4. Flattr FOSS (1400 views, not an article but I regularly blog about this project)
  5. Can Debian achieve world domination without being on Facebook? (1100 views)

The most flattered

Since I am using Flattr on my blog, it can be interesting to see the articles which generated lots of flattr micro-donations. The top-3 articles are my articles about Flattr (1, 2, 3). Excluding articles related to Flattr, the top-5 is:

  1. 5 reasons why I still contribute to Debian after 12 years (12 flattr)
  2. The secret plan behind the “3.0 (quilt)” Debian source package format (10 flattr)
  3. How to use multiple upstream tarballs in Debian source packages? (5 flattr)
  4. [LWN] Understanding Membership Structures in Debian and Ubuntu (4 flattr)
  5. Do You Want a Free Debian Book? Read on. (4 flattr)

Most articles get 2 to 3 flattr clicks.

The most commented

I usually get 4-5 comments on most articles but some generate much more feedback:

  1. [LWN] Can Debian offer a Constantly Usable Testing distribution? (40 comments)
  2. 5 reasons why I still contribute to Debian after 12 years (22 comments)
  3. Can Debian achieve world domination without being on Facebook? (15 comments)
  4. How to generate different dependencies on Debian and Ubuntu with a common source package (14 comments)
  5. [LWN] Understanding Membership Structures in Debian and Ubuntu (12 comments)

Factoids

Here are my conclusions based on the above figures:

  • Writing about your Debian/Ubuntu work and your long term involvement makes for highly popular content that spreads well.
  • In-depth and well researched articles (like those written for LWN) do not generate more flattr revenues than the average article even if they take 4 to 8 times as long to write.
  • People are more likely to flattr you for your free software contribution than for the value they get out of your article.
  • People care a lot about the Debian release process, and like to discuss the topic.

If you also appreciate the above-linked articles, you should click here to subscribe to my email newsletter.