Archive for May, 2008

pommed v1.18: small things

Sunday, May 18th, 2008

I’ve just released pommed v1.18, which is mainly a maintenance release.

I’ve added some more USB IDs for Apple external keyboards and relaxed the event devices identification for internal keyboards.

pommed will now happily start on a machine fitted with a keyboard+trackpad assembly that normally isn’t found on this model. This can happen when the topcase of the machine is replaced with the wrong part. I expect this situation will happen more and more often as people buy parts from EBay or similar to repair their laptops.

Obligatory loldebian post

Thursday, May 15th, 2008

Because a lolcat is worth a thousand jokes, here are 3 of them.

Thanks to rominet for coming up with those :-)

Of course it’s far worse. Did I tell otherwise?

Wednesday, May 14th, 2008

Dear Daniel,

It looks like you are referring to my post, though you got my name wrong so that wasn’t immediately obvious.

Of course this is far worse than the 2003 compromise in terms of the direct, known and quantifiable impact it has on our users. I don’t think I stated otherwise, so I hardly see why your post starts with “I disagree”.

Making code valgrind-clean …

Tuesday, May 13th, 2008

NOT.

Also see #363516.

Genius.

Now regenerating every single SSH key, SSL certificate and whatever else I can identify that’s been produced by one of the Valgrind-clean openssl. Also expiring and changing every single password I’ve ever typed in a vulnerable SSH session (be it at login or in the session).

Updating the packages on the machines was fun already.

Worst Debian day ever since the 2003 compromise. And that was a BAD one.

I guess we need a new openssl maintainer, we obviously cannot trust the current one(s).

SANE in Lenny

Wednesday, May 7th, 2008

The SANE project is working on improving SANE, extending the API and ABI in a backward-compatible manner and bumping the version from 1.0.x to 1.1.x to celebrate that.

The timetable has been posted, and calls for a release of SANE 1.1.0 on July, 30th.

This will be too late for the Lenny freeze by a few weeks, which means Lenny is set to be released with SANE 1.0.19.

SANE 1.0.19 is a good, solid release, which is good news. I’m not sure 1.1.0 will be as solid as 1.0.19 is, so I won’t try to rush 1.1.0 into Lenny at the last minute.

Until the Lenny freeze, I’m going to augment the current SANE 1.0.19 with code from the SANE CVS, concentrating mostly on bugfixes and self-contained new hardware support and features.

Hence, if there is something in the SANE CVS that you would like to see in Lenny: test it, then tell me about it. You have until the end of June to do so.

Currently on my TODO list:

  • pixma backend update

Currently in experimental, sane-backends 1.0.19-7:

  • saned & net backend with mDNS/DNS-SD support
  • debconf support for enabling saned

Comments and feedback welcome.

Etch 1/2 considered harmful for kittens

Tuesday, May 6th, 2008

Trying to upgrade a craptastic server that’s proving problematic under 2.6.18 to the Etch 1/2 kernel, not only is the machine extremely sloooooow to boot, but it turns out that it’s partly due to bnx2 now requesting a firmware file, whereas the firmware is built-in in the 2.6.18 Etch kernel.

Of course, the machine has no working network access due to this, and, to make things even worse, the firmware file is nowhere to be found. No firmware-nonfree in etch-proposed-updates and it’s not in the firmware-nonfree package in unstable.

You’d better hide your kittens, for Etch 1/2 makes me want to kill a large number of kittens.