I just found Byron. He was around the side of the house in some moderately thick undergrowth, dead, and as stiff as a board.

Needless to say, Sarah and I are quite upset. You couldn't ask for a better cat. He was so affectionate and personable. The last time Sarah saw him yesterday he was chasing something, we can only assume it was a snake or something that didn't agree with him. It's just such a shame. He will be missed.

I know how Mike must have felt now. Byron, our cat, has done something very out of character for himself, and not turned up at dinner time. We've been for a doorknock around the neighbourhood, and none of our immediate neighbours have seen him. He never really leaves the yard, so we're a bit worried. Hopefully he'll turn up in the morning. Here's to a crap night's sleep...

Today I made QA uploads of yydecode and gpppkill, setting the maintainer to the QA Group.

Whilst doing this, I noticed that libtool1.4 is orphaned, and if Scott had his way, would be removed from the archive forthwith. As there's 7 packages that still have build dependencies on this, I thought I'd seek advice on whether I should file bugs against the six packages that don't already have such bugs filed.

Today I had a bit of spare time (the chapter of the textbook I'm trying to read is an exceptionally long and dry read), so I thought I'd take a look at some of the release critical bugs.

I made 3-day delayed uploads of cacti and laptop-netconf, to fix #273261 and #261939, respectively. I made them 3-day delayed because the first bug had been open for 31 days, and the maintainer of laptop-netconf hadn't been seen by Echelon since the end of September. Neither upload is terribly intrusive, and so I don't think I'll get anyone's nose out of joint.

Tonight Sarah and I, along with Beth, Kylie and Andrew and Alice, went around to Denise's for dinner.

Denise made a trifle for dessert, along with all the trimmings. What caught my attention was what was printed on the can of (light) whipped cream:

Dairy Whip Lite contains 48% less fat than cream containing 35% fat

Must be the latest marketing ploy. Bewilder the consumer into assuming it's not that bad for you...

Today I had one of those rare wins with d-i, where I identified something it didn't do, that I wanted it to do, and could actually do something about it.

As I thought yesterday, it was relatively easy to add the ability to create a RAID1 array consisting of one 1 active device. I just had to remove the logic that said "thou must specify two devices if you've said there are two active devices" and duplicated the existing logic used for hot spare devices where if you specified less hot spare devices than what you initially indicated, it assumed the rest were missing.

The slowest part has been testing the patch. The upside of testing the install inside qemu is I could have mdcfg open in gvim in one window, hack it with nice syntax highlighting, then pop over to the qemu window, and duplicate the changes to the live mdcfg with nano, then fiddle with my md devices. Rinse and repeat, until it was all working.

Thanks also to Joey's terrific instructions, building a full d-i image to test my modified udeb was a breeze. Pity I couldn't say the same for performing the install in qemu. It's running exceptinally slow, compared to yesterday. It only took about three hours to do a complete install with LVM on RAID1, but today, with the monolithic mini.iso I built from a current checkout of d-i, it's 51% through installing the base system on a single degraded RAID1 device, some 5-6 hours after I started it off.

I'm on a bit of a mission to get the RAID support in d-i to allow the creation of a degraded array. So this afternoon, I thought I'd refresh myself with how mdcfg currently did it's thing by installing a RAID1 system in qemu.

Last time I played with qemu and d-i, it was all good up until it came to the network, which ended up being more of a notwork. Today, I tried the approach I used about 18 months ago with User Mode Linux. Using bridge-utils, I rejigged the host to be using a bridge device, br0, and added eth0 and the tun0 device to the bridge. (I had to load the tun kernel module and because I was running qemu as myself, I also made /dev/net/tun mode 666 for good measure). This immediately gave network access to d-i running in qemu, and DHCP worked flawlessly.

I created two 1GB empty files by dd'ing /dev/zero, one for hda and one for hdb and kicked off an install. Something like 3 hours later, the install completed. It was sooo sloooow. I was reminded of a driver reviver TV commercial, where a family of koalas are driving somewhere, and the kid koala in the back is going "Are we there yet?" every five minutes, because I felt like I was constantly doing that, jumping between virtual consoles to check the installation's progress.

But it wasn't in vain. I spent time poking around the innards of mdcfg, and it should be relatively easy to hack it so that, for example, if you go to make a RAID1 device, and only select 1 partition, it'll create a degraded mirror. It seems to do something of this ilk already. If you specify that there are a number of hot spare devices, and then fail to select that many partitions, it assumes the remainder are missing.

All in all, I'm very impressed with how d-i has progressed in terms of block device flexibility. We've come from boot floppies, supporting only vanilla partitions, to partman, that until recently supported LVM and RAID, to the situation today, where it'll finally do LVM on RAID. That's the icing on the cake for me. If I can help get to the point where you can create a degraded array, so you can add disks post-install, I think we'll be at the pinnacle of partitioning flexibility.

When I first discovered the Internet, I was a windoze user, and so when I discovered IRC, mIRC was it. So I grew up on mIRC, and never quite managed to adapt to command-line IRC clients like ircII. Well, I could single-task in them fine, but that wasn't how I was used to using IRC.

Whenever I was on a machine with a GUI, I'd generally use XChat, and didn't feel too much like a fish out of water. Problem is, I rarely use a GUI. I usually SSH into the box that is my gateway, reattach screen, in which I have a screen for Mutt, a screen for micq, and a screen for bashing away at a shell prompt.

So the net result of all this is I don't IRC much any more. People who know me know that I tend to flit in and out, depending on if I'm after someone, or some information. I wouldn't mind changing that, and having an idle IRC session in another screen, but till now, I haven't found the IRC client to make it worth my while.

Today, I discovered irssi, and it might just be that IRC client. At least for a start, it didn't bugger up my terminal, the way ircII invariably seems to. The main issue I have is I feel totally uncoordinated in it at the moment. I can split the screen into windows, which is half the battle, I'd like to be able follow a couple of channels at once. It's totally bizarre trying to get my head around what happens when you split the screen, join a channel and then part. It seems to close the "window" resulting in your splits going to hell. The other issue is colours. I haven't mastered how to customise that yet, and the doco is, well, light on in places. So if you run into me on IRC in the next little while, please be patient, while I figure out how the heck to drive my IRC client.

Well I've just made by very first NMU. Here's hoping I haven't made a royal mess. Hopefully I've done everything right. I pinged the maintainer (who according to the database, hasn't been seen for a few months) and sent a patch to the BTS. I made a 7 day delayed upload, and have subscribed to the package in the PTS for a while.

It's all Steve's fault. He asked me to look at #274955 last night at the LCA meeting, and I figured it wouldn't be too hard to fix, because the patches were already in the BTS.

Ha! Famous last words. All I can say is thank $DEITY Ryan Murray was lurking on (presumably) debian-bugs-dist, and offered a few tips to get things building properly. Of course, I learned a thing or two about autoconf in the process, so it was a good educational exercise. I'd also like to say that dpatch continues to rock my world.

I decide it's high time to replace my running shoes, so I get a spanky new pair to go with my new fitness regime, and it's been raining ever since.

On the other hand, I could investigate a new career as a droughtbreaker...

With lectures winding down, and the other side of final exams less than a month away, I thought I'd take some time out (read: procrastinate some) and look at the job situation.

My current contract probably only has about 4 weeks left in it (more if I rejig my hours so I'm working less around exams), and I'm not relying on it being extended or renewed.

I had a browse of what was going on Seek, and also discovered this job mentioned on the CLUG mailing list. It sounded interesting (but it would have to be very interesting for that sort of money) so I jumped through their hoops and put my resume in.

Once I've finished Uni exams, I plan on having a crack at doing the CCNA exam cold-turkey, just to boost the resume a bit. Everyone I've worked with keeps telling me I'll pass it no problems. I often wonder if people have an inflated perception of my abilities, or I just underrate myself.

In the last 24 hours (probably less), I've received 7 unique spam emails pertaining to Rolex watches. That's 7 emails that got through all my spam defenses (and a lot more seem to be getting through lately, dammit).

I just find it odd that all of a sudden, Rolex spam is flavour of the moment.

Well this is a new concept. A programming assignment that I just can't do. It's due tomorrow afternoon, I can't get it to produce what the assignment specification says is the correct result, and I can't see what's wrong with my algorithm, so I don't see the point in losing sleep over it...

Today I attended a talk given by Richard Stallman on Free Software and whatnot. It was a pretty stock standard GNU spiel in hindsight, a lot of it is pretty much word for word from parts of the GNU website.

The thing that really took me back was Richard's behaviour before the talk started.

I arrived about 20 minutes early and grabbed a seat in the lecture theatre. Tom Worthington was the MC, and he was running around trying to get things organised. Shortly afterwards, Richard turned up and squatted down at the front table of the theatre, bashing away at his laptop.

Not long after that, Tom asked him if he was happy. Richard's response was: "I'd be happy if I wasn't interrupted. If I wanted anything I would ask."

The guy I was sitting next to and I just went "whoa" and looked at each other. First impressions count, and my first impression of RMS is that of a rude prick.

Rude prickness aside, he gave a fairly entertaining talk, but I didn't learn much new. Unfortunately I didn't get an opportunity to ask about the whereabouts of a new GFDL draft.

Recursion is obviously one of those things you either get, or you don't get. The lecturer said it's one of the hardest things to get if you've got programming experience and haven't been using it.

I went around to Rick's to get some help with the current COMP1110 assignment, and we ended up scribbling on a whiteboard from about 7pm till after 4am. I think if either of us got recursion, it would have been trivial, but because neither of us seemed to have that enlightenment, we had to resort to what Rick calls "programming by accident" - having a stab until it works, which for a recursive algorithm is all the harder. Having a whiteboard helped, but I still think we managed to make a meal of the problem.

So now I can annotate my tree, with an implementation that is questionable in my opinion, compared to what Google had to about the algorithm.

Next stop, pruning the tree. I'm hoping (and thinking) that that is going to be a rather trivial exercise in wandering around the tree lopping off branches.

I'm absolutely appalled that Howard has been returned to office.

I can't believe that the Iraq war wasn't more of an election issue. The fact that the Australian people have not only chosen to reelect, but swing towards, a government that supported the US' unilateral invasion of Iraq sends the wrong message to the rest of the world. It says that the Australian people are okay with this kind of thing. It wasn't that long ago that thousands of people marched in protest against the military action in Iraq. Did all these people just forget when they got to the ballot box?

I voted fairly early in the day today, but got stuck in the early morning rush. It's interesting to see the demographic of my suburb. A lot of elderly people, a few young families and a few couples. All the elderly were getting priority in the queue, which meant it took me about half an hour to get to the front of the line.

I couldn't vote for Howard, even though I think his government has done a good job of economic management, I'm just too pissed off about Iraq. Being an environmentalist, I voted Green, which is just an indirect vote for Labor. Damn preferential voting system.

I was just thinking on Friday, how it would be rather interesting to see the impact on the Canberra micro-economy if there was a change of Federal government in this election, followed by a change of ACT government in next weekend's election. Apparently last time there was a change of Federal government, the Canberra property market took a real dive, because there was a lof of upheaval in the public service.

Gave pre-rc2 a spin on my desktop PC today, and it went fine

It's good that one can have the entire system on LVM now, but I kind of miss GRUB. I only switched from LILO to GRUB around the time I started playing with d-i, and realised how much more flexible GRUB is compared to LILO, but GRUB's not so good at dealing with RAID and LVM...

I'll have to try it on my Sony VAIO laptop to see if my WEP issues are indeed resolved now, and I should also do another install on my Ultra 5. I miss the talking alarm clock under the bed.

I had a bit of a foray into greylisting with MIMEDefang today, as it's been something I wanted to implement for a while.

The main example linked from the greylisting website is a monster filter written by Jonas Eckerman. I wanted to stick as closely to the filter that ships with the Debian package (because that's what I'm currently running), so I pulled out all the greylisting code and added it to the standard filter, coming up with this cross-breed.

I've got it running on my backup MX at the moment. I'm not sure if I'm game to put it on my main mailserver yet, just in case I've screwed something up. I can also see a few optimisations that could be made. Rather than opening a text file for each message to examine whitelists, I'd like to just tie to Sendmail's access.db hash, and add whitelist entries to that, if I can. I suspect there might be some file permission issues, as the access file has fairly restrictive read access, and I don't think the filter runs as root.

Now all I need to do is stop point-blank rejecting mail that originates from IPs in various DNSbls (scoring them up in SpamAssassin instead) and I should be in business.

I've wanted to make this all singing, all dancing gym membership system for my gym for a number of years. I wanted to use biometrics instead of membership cards.

I thought I'd found the perfect fingerprint reader, but of course it doesn't work on Linux, and the vendor did not give a terribly useful response to my product enquiry email.

After reading an article in Linux Journal, I wonder how easy it would be reverse engineer the device? I don't know a lot about how biometric devices work. I don't know whether they store the biometric data in the device itself, or if they spit out some sort of hash based on a fingerprint, and it's up to the application to match that with previous data.

If you know anything about biometrics, particularly with Linux, I'd love to get an education.

I was quite blown away tonight to receive a phone call from my Dad (I was expecting to have to talk him through the latest Microsoft Windows patch installation) with news that my Mum had had a heart attack last night.

She's spending a few days in hospital while they run some tests. Apparently it wasn't a "normal" heart attack, as she's reportedly got good arteries. At this stage, it could be one of:

• stress
• a blot clot
• a hole in the heart

I'm putting my money on stress. Mum's always been a worry wort, and generally highly strung at the best of times. Apparently prior to having heart issues last night, she'd had a heated telephone argument with a family friend.

It's just a bit of a shock to the system, and totally unexpected (I guess this is always the case with such things though).

So I've finally decided, after 4 years, not to bother renewing my gym membership. It's a bit of a shame. I think making the decision to join that gym, and doing it, has had a big effect on my overall fitness and appearance. I just haven't been motivated enough lately to go, so I'm not getting my money's worth. I did the maths on the way home from the gym tonight, and if I train four times a week, for the whole 10 week membership, it's equivalent to paying $8 a visit. So if I start missing appointments, it just starts getting ridiculously expensive. I'd be better off paying $20 once a week for a proper one-on-one personal training session elsewhere, and getting thrashed.

I've decided, instead of going to the gym, I'll join Sarah's triathlon club and just train with them, but not compete in any of the triathlons (there's no way I'm swimming in the lake). I'll probably go back to the gym next year, but while I'm unmotivated like this, it's just a waste of my time and money.