This is my monthly summary of my free software related activities. If you’re among the people who made a donation to support my work (78.31 €, thanks everybody!), then you can learn how I spent your money. Otherwise it’s just an interesting status update on my various projects.
Debian packaging
I wanted to update publican to the latest upstream release but I stopped after a few hours of work during which I filed two bugs that a modicum of testing should have caught before release. So I decided to wait for the next minor release.
I uploaded python-django 1.4.4 and 1.4.5, new upstream maintenance and security releases which thus went into wheezy. I also prepared a stable update of Django (1.2.3-3+squeeze5) which required me to backport the last 2 sets of security patches.
I uploaded a new revision of wordpress to fix a problem with TinyMCE (#700289) and to update/add many translation files (#697208).
Bug reporting and misc fixes
Live-build issue. I experienced some intermittent failures when building HDD live images with live-build on armel. Daniel Baumann directed me to the problematic piece of code (the “oversizing” of the image size was not enough) so I committed a small fix by increasing the oversizing factor to 6%.
Live-config issue. I also reported another issue that I diagnosed in live-config (#701788), namely that the script which setups sudo was failing when the default user is root.
git-buildpackage issue. I filed #700411 after noticing that git-import-orig imported the debian directory provided by upstream. Those directories are not used with “3.0 (quilt)” source package and their presence in the upstream branch is thus harmful: any change to the upstream debian directory will result in conflicts when you merge a new upstream release in your packaging branch.
rubygems integration. Later I had to package a bunch of ruby applications that were using Bundler and I wanted to reuse as many packaged ruby modules that I could. But for this, those modules had to provide the required rubygems meta-information. I filed #700419 to request those on rake-compiler and with the help of Cédric Boutillier (and others on #debian-ruby), we identified a bunch of ruby modules which could get those with a simple recompilation. I filed bin-nmu requests in #700605.
Misc bugs. simple-cdd offers to select profiles to install but I noticed that the associated debconf template was not translated (#700915). The startup scripts (provided by initscripts) in charge of activating the swap are supposed to handle a “noswap” kernel command line option to disable swap. In #701301, I reported that the option was not working correctly if “quiet” was present first in the command line due to spurious “break” statements.
Debian France
Administrative work. We were late for some legal procedures so I wrote the report of the last general assembly and sent it to the “Tribunal d’instance of Sarreguemines” to record the changes in the administrative board. I also completed the “special register” of the association, it’s a notebook that is legally required and that must document any important change in the governance structure of the association (new members of the board, headquarters change, new bylaws, etc.).
Galette developments. Debian France is funding a few enhancements to the Galette free software that we’re using to manage the association. I am in touch with the Galette developer to answer his questions and ensure that his work will meet our needs.
Librement
I have been looking for talented developers who have a genuine interest in my Librement project. I want to fund the initial development of the project but I don’t have the means to fund it entirely. So I really wanted to find developers who would find an interest beside the money that I would pay.
I got in touch with the team of developers from Scopyleft and they look like very good candidates. But they’re heavy users of the Scrum development method and asked me to play the role of “product owner”. So I started to describe the project with “user stories” (i.e. “create the backlog” in the Scrum jargon), you can have a look at them here on trello.com. If you’re interested by the topic of free software funding, feel free to review and to send me your comments.
My goal is clearly to have a “minimal viable product” with the first iteration(s) that I fund and then use the platform itself to fund further developments of the project.
Thanks
See you next month for a new summary of my activities.
The link for the first bug in the section “Debian packaging” seems to be wrong, it links to a bug report for the apache package from 2003.
Thanks, I was missing a trailing 6. Fixed!
I love how you regularly put up your activities for the month. Makes for a great portfolio, as well as log for yourself.
Thanks for telling! It’s good to know that people appreciate it because it’s a serious commitment to have to take notes of what you’re doing and then taking 2-3 hours to review everything and jot it down in a nice article.
Regarding Librement: if there’s one thing I’ve learnt over the years of being in a startup environment, it’s that even the MVP way of thinking can be overengineering. One way of thinking about it is where the big risks in the product are – in the technical side (unlikely for many of us) or in actually finding a product + marketing combination that gives enough value to people that enough will bother using it/paying for it?
Sometimes it’s possible to think up a way to do a lot of marketing/selling work up front (i.e. the sort of thing that makes the venture a success or not) without even having any product yet, but doing things manually instead.
Anyway, just a thought. 🙂
That’s what I have been doing on https://raphaelhertzog.com/support-my-work/ I update the amounts and the graphs manually but it’s a bit tiresome. Now I want to automate this and to offer it to other free software developers.
For selling documentation, it’s a bit the same, I did it for my own on https://debian-handbook.info/ but now I want to go to the next level.
Yeah, you’ve field-tested this with great success. What I’m thinking is that you for instance could try to line up a bunch of projects/people if you haven’t already and get commitment from them. That could take out some of the risk and also make it easier to figure out what to target first. It could be that what is most important to you doesn’t matter to the projects that are most interested in Librement.
For instance, PyPy is taking donations: http://pypy.org/
I don’t know how they do it, but it looks to me like some manual work is involved.