Older blog entries for epx (starting at number 6)

A lot of small changes in initscripts. It became a real good program IMHO, with the best algorithms to solve dependency trees etc. This weekend I will make a "proofread" in it to try to squash the last conceptual bugs, if any.

The "sound server" chapter of my life is finished ;) KDE/aRts team accepted a patch from me that implements a "suspend" method, for softwares like KRecord that need to own /dev/dsp. It will be available in KDE 2.1. For KDE 2.01, it introduces binary incompatibility, so I cannot use the patch right now, because as a distribution we have to enforce binary compatibility with KDE-2-compliant binary packages out there.

Talked about CUPS, sound servers, KDE and initscriptsNG in irc://irc.uninet.edu/#linux last Saturday. Got a bit nervous, so my English output was not the best possible ;) Next time it will roll more smoothly. People seemed to like.

Modified initscriptsNG, so /sbin/iniscripts now supports user calls like that:
initservices start smb sshd bla bla bla
It solves dependencies, starts services that are needed for those explicitely asked etc. In "initservices stop", it will ignore services that user has asked to stop but are needed by other ones that would be still running after initservices stop.

pam_mount 0.3.1 has been released a few days ago, with another bugfix patch from Norway.

In the day-to-day work, lots of small things here and there that take a lot of time. I expect to finish KDE 2.01 packaging this week once and for all.

Another interesting thing I've been working on is the use of sound servers, to allow (among a lot of other things) the mix of n sound sources without hardware support for multiple DSP streams. I have tested KDE's aRts a lot. With a LD_PRELOAD trick, most programs will be fooled to play into aRts than to /dev/dsp. The only thing aRts is really lacking is some program to support RECORDING. aRts needs full-duplex in soundboard to support recording, and most of soundboards out there (e.g. notebooks) does not support that.

Today I proposed my initscriptsNG (initscripts with dependencies between services) to LSB. I exchanged some ideas with Richard Gooch but it seems that we have very different ideas of what a boot system should be (he wants something that is easy/clean for people that maintain the system using vi; while I aim for compatibility and transparency for newbies). So I sent the proposal to LSB, since some persons there seem to want the same things as me.

Anyway, what I want is to provoke a decision. Whatever boot system "wins" (Gooch, Clausen, mine), I will be happy to work to improve it.

Ok, pam_mount version 0.3 has been released, in the same place as before ;)

The day began pretty weird. It is warm AND raining. I could not sleep@night again, so I went to work at 6:30am. The supermarket is closed for balance (again!) till 9:00am and I am hungry :P

Since pam_mount is making almost everything I imagined for it, it's time to choose another near-term target. KDE 2.01 will ship way before I imagined; its pt_BR translations are OK. Even KDE 2.1 is said to ship in the end of January 2001, with a preview for a 2.1.1 some weeks later! <

Resolving boring, "physical-world" things (car is leaking oil, and has a thick pipe - which function I haven't figured out - hanging free near the ground).

Dad has bought a cheapo flatbed scanner. My chemical-film camera got an extended useful life ;)

Wow, pam_mount has received a patch from Kjetil Torgrim Homme <kjetilho@ifi.uio.no> less than 24 hours after pam_mount 0.2 announce. It implements the only TODO feature in 0.2: template volumes (that are the same for all users except by the user name).

The version 0.3 will ship anytime soon, I just need to test the template thing.

At last, the 0.2 version of pam_mount is shipping! This PAM module has been useful since the very first version (0.01) but this is the first "decent" version, with an RPM etc.
This module can mount remote volumes at login time, reusing the password the user has just typed. It can be pretty useful in environments where there are Windows NT and Netware private volumes that any user needs to have access during a Unix session.
Tarball and RPMs can be found here.

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!