An interesting idea

So, I've been thinking quite a bit about my impact on the environment lately (largely thanks to our getting airconditioning installed, and using it pretty regularly due to the hot weather we've been having). I'd really like to put some solar panels up on the roof so that I'd feel less evil for using the airconditioning, but we're renting this place so it's not possible. I've done a few things to reduce our power consumption, but there's only so much you can do . . .

So, I've been thinking about that sort of thing. One of the things I've been thinking about is the impact of driving - the trip out to Angler's reach over new year's, and the fact that it made me hanker for a 4wd, highlighted it for me a bit. Since our Echo is petrol powered there aren't many real options (the various petrol replacements aren't currently very satisfactory); if it was diesel powered we could switch to biodiesel and it'd make a real difference. For that reason I'd like to make our next car diesel powered (hopefully by the time we want to replace the Echo it'll be much easier to get diesel equivalents).

So, biodiesel. It's not too hard to make (at least, one version), from waste cooking oil. All it needs is some reasonably simple processing, done in (basically) a giant stockpot. The setup cost of that wouldn't be too bad, but still enough to make it more of a challenge than most people would be up for; likewise the processing would be more than most people could realistically do. It's also enough trouble that most people wouldn't bother.

The idea I had in the shower this morning was this: a 'biodiesel club'. Basically, a bunch of people would get together and pool the money needed for setup and the skill/experience needed for the processing. Then, each person would supply the waste oil needed for their consumption (plus or minus a bit, of course - I imagine there'd be some averaging between people), and it'd be processed by the club's experts. Club members would pay a yearly membership fee, which would cover maintenance of the processing equipment and any expansion needed to keep up with consumption changes.

What all this would provide is a cheap and easy way for people who are interested to switch over to biodiesel - all that subsequent members need to do is find suppliers for waste cooking oil, and collect enough to provide their needs. The core people (the creators of the club, and the processing experts) would get a share of the overall supply - they'd effectively be paid for their work in fuel. And everyone would be both saving money and reducing their environmental impact.

Sandra suggested there might be some legal issues for the club - someone using the biodiesel has a mechanical problem that they then blame on the fuel and sue the club. Personally I think this is a bit pessimistict, since everyone would have volunteered to use it, in full knowledge of the fact that it doesn't have any quality guarantees, but it'd certainly be a good idea to make new members sign an agreement disallowing that sort of thing.

I think this is a pretty good idea for making biodiesel more easily available. Of course, it's based on the assumption that people aren't already commercially collecting all the waste oil there is - if they are, there's probably not going to be enough available. I haven't looked into it yet, but I will . . .

By the way, although this may seem like a relatively trivial thing, the numbers I've seen suggest that driving is about a quarter of our carbon footprint - converting to a carbon neutral fuel is definitely significant, and if biodiesel clubs help a fair number of people do that, then it has the potential to be significant.

  • Current Mood

Driving graagh!

So, I've been driving to work earlier than usual recently - I try to leave no later than 8am (if I can't get organised before then I'll leave around 9, rather than trying to deal with peak morning traffic). Leaving this early generally means I can get to work in about the same time it takes during the off hours, give or take a bit. However, it's /just/ borderline early enough to avoid the annoying traffic - if I leave at five to eight it'll generally be smooth, but if I leave /at/ eight it's a bit of a crapshoot.

This morning was a case in point: mostly smooth traffic, with a longer queue at the lights than normal, and average speed through most of the drive closer to 70 than 80. However, there's one spot where two relatively large streams of traffic /merge/ (Gungahlin drive and William Slim ave merging together and joining Tuggeranong parkway, for those who know the area). Normally it's reasonably smooth, though it slows down significantly in anything more than light traffic; this morning there was a line of walking-pace cars from the turn off William Slim all the way to Tuggeranong parkway. No accident or anything that might have caused it - just cars driven by Canberra drivers.

There are many things that annoy the hell out of me about Canberra drivers, but one of the biggest issues I have is their complete inability to merge. I don't mean they're slow and inefficient about it - they're /incapable/ of merging. If there's naturally enough space to allow them to change lanes without having a car in the way they can manage, but as soon as they have someone blocking that they have /no/ clue what to do. Most of the time what they'll do is drive side by side with minimal following distance until the road starts to narrow, and then they'll /stop/, crawl forward, still with no space between the cars, until a car in one lane or the other stops and lets the car next to it get ahead, and then drives in behind them. Thus you end up with a long line of cars crawling along, jostling for that precious position a car length or two ahead of some other randomly picked car.

It's /insanely/ stupid and infuriating, because if only they didn't insist on tailgating so much, there'd be enough space between cars for them to merge without all that jostling and crap - they could just zip the two lines of traffic together smoothly, without even having to slow down /at all/. Hell, you wouldn't even need to be using a two second gap between cars - even just three car lengths would make it ten times smoother. But no, they just /have/ to be following the person in front as closely as they can possibly manage, leading to this kind of crap.

They even do it when they're merging at speed, believe it or not. Say I'm ahead and to the side of them: rather than driving to get /behind/ me, they'll drive /next/ to me - either driving right up beside me as if they're trying to cut me off, or with their nose about in line with my rear wheels, meaning that when the road narrows they have to tap the brakes in order to get in behind. Which, of course, means that the people behind and next to /them/ have to tap their brakes to avoid rear ending them, with the braking cascading back through the line. People do this even when they're merging at /80/, let alone at 50 or 60.

If only these morons would learn what a proper following distance was the carrying capacity of the roads would probably triple, and the aggregate fuel consumption of the ACT would halve.

  • Current Mood

Wot I did on my holidays

Well, not exactly a holiday . . . More like the opposite, actually . . .

I figured I should post a bit of an update, since it's been a while and I've been doing stuff in the meantime (though not much that pertains to my various projects - work and money has been getting in the way there).

So, what have I been up to? I've spent a large part of the last six months working hard on project stuff for CSIRO - mainly the DHCP system I was working on last year, too. It's finally finished and done now, thankfully, and I'm back to working on slightly less exotic stuff.

Sandra's been here for six months now, and that's a significant part of the money issue - she's been getting a whole lot of dental work done, catching up on pretty much a lifetime of neglected teeth. Unfortunately, that ends up being pretty pricey . . . It's largely done now, but there's still a few things that need to be finished off.

Wot I /actually/ did . . .Collapse )

And what about the boats? I need cash to make real progress with them. Cash to do things like buy materials and tools to start experimenting with construction techniques, and to buy parts for the electronics, and so forth. Sadly, cash has not been available lately . . .

Ta-ta for now,

  • Current Mood


Well, we finally heard back from Immigration about Sandra's visa application. It's been accepted, and she has a deadline of March 2, 2008 for getting into the country.

We're sorting out tickets and such like, but due to a number of dental problems Sandra's had we're scraping a bit to get the money together . . . We should be able to make it, but only with a bit of difficulty.

That said, it's still really great news. The wait is finally over!

  • Current Mood
    happy, but tired


So, I haven't posted anything recently (for values of 'recently' along the lines of 'two months'). This isn't because I died or something, I've just been /stupidly/ busy with work. The whole of July was taken up with a bunch of stuff leading up to a four day trip to Sydney, then August was swallowed up by an insane amount of work on a project (185 hours 'officially' for August, plus 15-20 hours each weekend doing stuff like coding that I couldn't get done in the cubicle farm at work).

What project? This will probably sound a bit silly given the amount of work that's gone into it, but it's basically a standardised and consolidated DHCP system for the whole of CSIRO - replacing every single DHCP server in the organisation with a considerably smaller number of standardised servers, and a centralised management system for their configuration.

It doesn't sound like something that should be consuming my entire life for the three months that it has (and a considerable amount of work before that), but it's a bigger job than it sounds like at first. This affects /every/ machine in CSIRO aside from a relatively small number of statically configured machines - at last count, I believe it works out at something like 8,000 computers, on fifty or sixty networks, most of which have been using whatever DHCP system was hacked up five or more years ago, with no standardisation, not even vaguely sane policies in many cases. All of that legacy configuration information needs to be migrated into the new system, without losing any of it and without the users noticing the change.

On top of that, all the software it uses (aside from the actual dhcpd and RDBMS) was written from scratch - tools to migrate legacy configurations, automatically generate new configs, and provide an interface for managing the configurations. It's not an enormous amount of code (around 4600 lines of Perl for the backend, and a couple of thousand lines of PHP for the interface), but it's ended up being a lot more difficult (and fun, sometimes) to write than I expected. The backend code has a fair bit in comon with a compiler - it parses a set of configuration templates into an intermediate representation, walks the contents of the IR doing stuff to it, and then spits out a version of the processed IR (it's generally transforming a template into a dhcpd.conf fragment, or a legacy configuration into a dhcpd.conf fragment or a bunch of SQL commands to load the config into the database).

I've written (in Perl) a mostly complete parser for the ISC dhcpd.conf format - it's a big, fairly complicated, somewhat hairy, and (I think) reasonably sane recursive descent parser, for a format that not only isn't documented anywhere, it's really annoyingly inconsistent. I'd /like/ to say that it lets you do stuff like use keywords as object names (i.e. you can declare something like 'host unknown {...}', where 'unknown' is a word that's used elsewhere as part of the fixed syntax), but the reality is that I don't think the real parser /has/ any concept of keywords or stuff like that. It actually seems to have the lexer and parser so intimately interwined that the lexer knows that in that context it should treat the word 'unknown' as a plain word rather than a keyword. There's also all sorts of conceptual inconsistencies, like the way that it handles parameters vs options (for those who know the format, options are the 'option foo' type things, whereas parameters are the ones that are just a bare word). In /general/ options control what gets sent to the clients in a response, and parameters control the server's behaviour; however, there are a /lot/ of cases where parameters get sent to the client, and where options drive the server. All told it's really a pain to deal with.

In any case, the whole of CSIRO in the ACT is now using it, and in Tasmania - next week we'll be adding WA and SA, then Queensland, NSW and Victoria (with those last two being the biggest and hairiest of them all). By the end of the first week of October, the world will be mine! . . . um . . . yeah.

So, that's why I've been silent here, and why I haven't been socialising or anything.

  • Current Mood
    What's been going on in my life?

Structural design issues . . .

(the mood thing is from the t-shirt I'm wearing at the moment - yay Firefly!)

This was originally going to be a big long technical post about the fine details of how I'm planning to build these models, and why I made those decisions. However, it was turning out kind of annoying to write, so I've tossed it and I'm now going to just write something shorter (yes, this is shorter than the previous version was), simpler and hopefully saner . . .

All your structural design issues are belong to this post! (and I am /so/ going to hell . . .)Collapse )
  • Current Mood
    Curse your sudden but inevitable betrayal!

(no subject)

Okay, so Sandra and I are actually still waiting on the FBI to complete their background check on her and get back to her, which could take another four to six weeks, so in order to keep us sane we've decided to bring her over here for a visit - a week over Easter, then back to the US to wait . . .

It's so nice having the option of throwing money around occasionally . . .

She gets here on the 4th of April, leaves on the 12th, and will probably be around town a bit in between.


Welding classes

So, I went to enroll in a welding for beginners class at Lake Ginninderra college's night school thing, after hearing about it from iki_maska. I enroll online, at around 8:30pm on friday. I then get a call at 11am satruday morning from the guy who organises the classes saying (paraphrased) "Yeah, I'm sorry about that class you enrolled in being closed - we didn't get enough people enrolled before the deadline. But we've had some more interest since then so we're thinking about openning it - I'll get back to you early next week if it works out that way" Rather confusing when you haven't had any coffee . . .

The upshot of it is that any Canberrans interested in learning basic welding and having $185 plus the cost of materials, and free wednsday evenings from 6:30-9:30, can sacrifice themselves on the altar of getting this class up and running this term should they so choose . . .

Antti, did you or Adam actually enroll?