Merry Christmas everyone! Best wishes to all of you.

Some stuff has moved over here. First of all: me. I'm now on the Canary Islands. Nice and warm for Christmas. Second of all, I've managed to find the problem with Probe Request with ACX100 hardware.

A quick patch later, all is good and we discover the same APs as the Windows driver does. We may now actually be more functional than the Windows driver, which really says a lot about the Open Source/Free Software way of doing things. It also says a lot about the people who reverse-engineered the driver. Kudos to them.

And talking about that, some progress has been made with the reverse-engineering of the IPN2220 driver. I think IDA Pro might have mis-disassembled some stuff. Either that or the compiler is doing really weird stuff. A lot of negative numbers are being used as offsets. Also, IDA Pro is marking subroutines wrong. There are references from one subroutine to inside another one.

There has to be few things that are more frustrating that looking for a verb during a Philosophy exam and have the same verb repeating over and over in German, finally managing to come up with a suitable verb in Spanish, and have the idea the rest of the exam that the German verb was actually far better suited for the occasion. I still think that "Gefüle hervorrufen" is far better than "apelar a los sentimientos", but I don't think my teacher would have agreed.

And whilst we're on language mixing, there's a jumper going around (well, I'm guessing they made several and sold them) which has the following text on it: "Διφερεντε νο εσ δεσιγυαλ". I have no problem with the text being in Spanish using greek characters. I also write stuff in different scripts just for the sake of it. The problem I have is that it's not even written properly. The "εσ" should be "ες". If you're going to sell something which can actually be read (unlike chinese-like symbols on clothing which don't have any actual meaning) the least you should do it at least check it's written properly. There's not much point to this semi-rant, I just though I'd share.

In completely unrelated news, I now have a new desk, which means I have shitloads of space to waste. I can have both my CRT screens (which I'll have to get working as dual-head under X at some point, though that's another story), my laptop on my desk and still have some writing space left.

From the ACX driver front, I've confirmed that the ACX100 can send Probe Request packets, but then something weird happens with RX packets. PCI says it can't find any free. I haven't looked at the USB, but it does timeout whilst unlinking during rmmod. This only happens if something's really wrong with the device or I did something stupid.

I had too little sleep last night. Not even four hours, and when I got up if was really cold. This was followed by a two-hour Chemistry exam (which I think went relatively well), and then the rest of the 'working day' was just waiting for the weekend. There's nothing interesting in school anymore...

In my attempts to reverse-engineer the Inprocomm IPN2220 driver I stumbled upon the Linux 2.4 driver. This is distributed by D-Link, and is of course a MIPS binary blob, which gets linked with some wrapper. I'm thinking about bringing this to the attention of the gpl-violations.org people. The problem is mainly that Inprocomm doesn't exist anymore. I'm just hoping that D-Link knows something about the source. GPL-Violations.org has lawyers and people with experience (a few people anyway) which I'm also hoping would know how to get information about a deceased company.

Qemu runs quite well, and compared to Bochs, it just flies. One problem I still have is that if I want to run qemu/bochs via SSH, there's always some processor time left free, so it runs slow. Another problem is that WinXP says it couldn't validate the license or somesuch, which didn't happen under bochs. Strange.

I'm hoping to be able to run something on qemu with its PPC emulation, but so far, I've managed to get the LFS livecd to display Tux and halt.

Today, I received an e-mail from someone from VMware, who was asking for some peer-review. Sadly, I do not know of anybody who would be able to work for them. I'm assuming they saw my e-mail with the ChangeLog of the ACX driver and assumed I was 'in the biz', but don't tell anyone I used that expression.

My router is now in the Linux Counter (http://counter.li.org) although it doesn't have very good uptimes. At most, it could last 14 hours up. After that it needs to be shut down so we can sleep.

What I'd really like is a Mini-ATX box which I could have as allways-on machine for real. This would imply lowish power consumption and no to very low noise. I'd really like to have it hooked up to the sattelite aerial so it can record the shows I never remember to watch (most notably Global Edition) and with which I could try to have a high uptime. The uptime would be limited by the quality of the power supply here, which means the box will get rebooted at least twice a year.