20 Apr 2004 thomasvs   » (Master)

Weekend

Had quite a productive weekend. First of all, I got off my ass and tried out the Intel Pro Wireless 2100 drivers on my laptop. I had bought driverloader, which worked quite well, but I'm getting tired of having to download stuff each time I upgrade my kernel, and of course I actually forget to download the RPM before upgrading in the first place.

Of course this gave me a good change to test my kernel module packaging strategy again. Fifteen minutes of work gave me a loadable ipw2100 module (without WEP, at first), just by running ./configure and make. Five more minutes gave me a set of packages for it.

Then, I enabled WEP and rebuilt the hostap packages from the QA submission queue at www.fedora.us. I had a problem with the function call being used from hostap not being versioned. After some thinking, I figured out that this was because the hostap package didn't include a hostap.ver file that actually does the symbol redefining. So I changed the hostap package to include that, rebuilt it for four archs and four kernels, and then rebuilt the ipw2100 packages for the same. And lo and behold, the packages worked. So if you have Fedora Core 1 and an ipw2100 card, *please* test these packages together with these..

Don't forget to download the firmware as per the instructions on the ipw2100 project site, and install it in /etc/firmware.

Next step is to update my kernel module stuff to 2.6, but I'm not looking forward to that. AFAICT from discussing with people it seems there is no decent way of building kernel modules against a read-only kernel-source tree. Moreover, Arjan seems to say that to build kernel modules for a kernel/arch combo, you need that exact kernel package installed as well. That will probably make it harder to do mass builds of kernel modules as well. Sigh :)

DocBook

I've been wanting to write a usable DocBook template tarball for quite some time. There are a couple of "guides" I'd like to do and every time I work on projects that use DocBook there is always something tripping me up. Between xmlto breaking in TeX processing for PS and PDF, or the docbook2 tools insisting on downloading SYSTEM identifier stuff from the net, there just is no foolproof way of building this stuff.

So, after a day of trying to write somewhat clean make rules and .m4's, I have a template tarball project that builds documentation, passes make distcheck, and easy to use in other projects. Yay me.

Fluendo

So, it's official: Fluendo is launched ! In a nutshell, we're going to write a free software streaming media server, on top of GStreamer, making it possible to do completely free software-based video streaming, using royalty-free codecs.

We decided some time ago to fund Xiph.org, since to reach our goal we actually *need* a decent royalty-free video codec. Theora is very close to what we need at this point; as soon as the bitstream specification is fixed, all videos created with that version of Theora will be playable by future versions of the library. This will hopefully have the same effect as the Vorbis Beta 1 release had for audio.

And even if it doesn't, it still enables us to provide this server working completely, for free, and hopefully, allow distributors of Linux to pick it up, distribute it, as well as the GStreamer stack with playback and recording applications. I can understand it doesn't make sense to do so if you can't distribute video codecs as well, so once Theora is ready to be distributed, I hope this changes the field a little.

Anyway, I'm pretty psyched we can announce all of this. There's always a fear of turning over to the dark side without realizing, but being able to start a company with these goals is exciting, and I hope we do well. We're moving in our new office this week, and our new collague is arriving next week.

The nice thing was being able to use my docbook-xml-template to generate the press release easily :) So the next one we do will just be a matter of filling in the content and running make.

Names

Ross is not the first one to link Fluendo to influenza. I really like the name, it took us long enough to come up with something that we liked, still had a domain name available, and not too much GoogleJuice. I'm wondering though if non-hypochondriac people make the same link between Fluendo and influenza ?

Also, Ross ran away with a "might happen" newsbit and posted it, probably to put some pressure on us to deliver :) All I can say is that we'd like to, but aren't sure yet if all the pieces will be in place.

Muine

Read Jorn's latest entry. Good to see Muine progressing. Only, Jorn is switching backends (again). I have a lot of respect for Jorn, and he's a talented coder, but I really have to wonder *how hard it can be to do some bug reporting*. The number of times jorn was in our channel, multiplied by the number of bug reports by him, multiplied by the number of mails to our mailing list, are easily countable on the fingers of my two hands.

People seem to think stuff should just fall out of the sky. Sometimes it does, but when it doesn't, it doesn't hurt to poke the clouds a little so they drop some more stuff you want :) How's about some simple feedback about the framework you're coding against ?

Latest blog entries     Older blog entries

New Advogato Features

New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.

Keep up with the latest Advogato features by reading the Advogato status blog.

If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!