Older blog entries for gobry (starting at number 19)

Pybliographer

a new release on the stable branch, no big surprises. Plans are quite clear for an early preview of the development code, now I just need to find some time to actually put some code in it :-)

proski

regarding arch being more centralized than subversion, it's rather the opposite :-) First, unless Gna! forces you to do so, more than one person can commit to an arch repository. The development pattern you refer to is the one where there is an "official" tree, which collects patches from other trees. Note that anyone can start a mirror of your tree, or a branch with local fixes which is kept in sync with the official tree. The difference with subversion is that these people can also see their changes be merged in the official version with no problem (ie they won't see their patch be applied a second time to their own branch when they update from the official tree for instance).

So you can imagine multiple topologies depending on your needs. For instance if you have many small contributors and a few more regular developers, you might imagine having a first level of integrators that take the contributions, and you only merge from those.

Pybliographer: the mailing list is now available via NNTP thanks to Gmane. I didn't know this service until someone asked me if it could be used for pyblio. Nifty.

Work: struggling with cygwin and mingw to bring our company build tools (intended for embedded development) on an XP box... The tools themselves are in Python, which gives me at least some hope, but it's messy anyway.

PHP: even if I dislike the language, I managed to merge Wordpress and Gallery in the same web site, with a common visual appearence. Hopefully I won't have too many changes to do, as I feel like walking on eggs (what's the idiomatic english equivalent of this?)

Plugged TMDA in the exim server that handles mail for my domains. Now I need to list where I registered with my previous email and put some tagged address instead, in order to do some SPAM source tracking :-)

Ahh, the pleasure of overengineering solutions to non existent problems... I really like that when I get bored :-) Made a distributed Bluetooth scanner with BlueZ, Python, XML-RPC, lots of unnecessary threads,... just to be able to check when people are about to enter or leave the office.

web Released a first draft version of baikal-bangkok.org, (so far, only in french) which integrates WordPress in the site. Stripped down its features to a minimum, as it will not be intended for the casual geek... Worked the whole week-end with Caroline, who will do the actual journey, to define the site's content and layout. Making a pleasing design is definitely something that is not related to programming skills (and something I confess I don't know), but can be made even harder if one tries not to completely go against usability concerns... Anyway, I will probably try to replace all my table hacks with CSS in a second iteration. I started to work with Bluefish, which seems to handle php and html mixes quite well, but went back to emacs when I had to bulk-modify lots of files...

trs80 Thanks for the info about iTunes DRM. Regarding deekayen's remark about burning the AAC on a CD, I think this is where Apple has it right: block what would be a sharing highway (ie, download from iTMS and make your iTunes directory directly shareable from some p2p tool), but do not be completely in the way people would like to use their legitimate files.

I installed WordPress yesterday. I'm now in the process of customizing it for a friend of mine who will travel two years, from Lake Baikal to Bangkok. She's interested in the relationship between water and people's culture and way of living. Her trip will lead her to several places where this relationship is especially important. She would like to have some way of sharing her experiences as they happen, and I proposed to use a blog for that. Let's hope I won't have too much support requests, or the travel expenses will be high :-)

Work: slowly managed to convince the team at work to move over to Arch instead of CVS. The last issue is related to the fact that our Linux-only business has seen some Windows projects creep in... And Arch on windows is not yet ready for prime time: some hopes seem to come from using the Unix comptibility layer from Microsoft. An attempt by a friend of mine to install the Unix Services left his XP box crashing immediately after boot. I think we'll rather access the code via samba and perform the versioning on a Linux box...

Pybliographer: things were a bit slowed down by a lack of free time, or rather a decision to spend it differently. However, I still think of releasing the development code soon. I find it already useful in its current shape, maybe I won't be the only one :-)

Spent a few hours trying to configure GSSAPI for my OpenLDAP server, only to discover that GSSAPI does not work through NAT... Grmbl

It's weird to feel a loss for people you don't actually know. I guess this is what "community" means.

Pybliographer: Released a new stable version. Now, I've moved the arch repositories on my own server, with a repository browser. This was also the opportunity to apply the pgp-signature feature of the tla-1.2pre development series.

check before you code... I wanted to start a small project to implement a shell for ldap browsing, and imaginatively called it ldapsh... just to discover the sourceforge project called by that name, which happens to be what I wanted to do (be it in Perl instead of Python :-))

I'm giving a try to Bytemark's Debian virtual servers: so far, it works well, even if it's strange to feel that the CPU is sometimes used for something else than looks like your own box. I'll use the machine to host my Arch repositories, but this is also the opportunity to put common data (bookmarks, vcards,...) while my grand replication framework is still vaporware...