Commercial use? No. Open source modification? No. Redistribution? Nope.

This is really annoying, but thankfully there’s another option. For your bookmarking pleasure, enter https://github.com/joahua/AusPostcode. It’s a CC-BY licensed version of suburb/postcode data in both CSV and JSON formats (feel free to fork and submit a pull request if you’ve got others) based on Australian census data that’s much more permissively licensed.

Happy free/open-source postcoding!

Turns out it doesn’t use Eclipse’s HTTP Proxy settings, but needs setting elsewhere.

On Windows XP, this will be in your Application Data path under Subversion. Mine is as follows:

C:\Documents and Settings\joshs\Application Data\Subversion

I haven’t got a Vista machine to test on, but it will still be the Application Data\Subversion folder within the user’s path. (I will confirm this next time I’m on a Vista box.)

Linux users, look in ~/.subversion/

Open the file “servers” (no extension) and scroll to the bottom section, [Global].

Un-comment and edit the http-proxy-host and http-proxy-port settings (and user/password if required, it wasn’t for me) as appropriate and everything will start working. You don’t even need to reload Eclipse.

Productivity just soared!

And people wonder why I refuse to buy music online.

On the plus side, I got all system configuration stuff (esp. Samba, which can be a lot more difficult than perhaps it should be at times) worked out last week (i.e. the system was nearly perfect, but for being utterly unable to install even SRPM packages of a more recent Python version), and Michael went through installing everything with me at work… we had to battle Windows a little there, but even it relented. So close. Then I’ll spend heaps of time cutting layouts to markup and seeing them working, and non-Youthworks time taking Satchmo for a spin (which will hopefully lend itself to a certain application very nicely). The lovely thing about all this is I need Django to work for CYIADA, so I’m supported in getting it up and running, but then have enough ‘spare’ hours in the week that I can engage in freelance projects that ultimately mean I know what’s going on with CYIADA and am mildly more competent to make minor modifications as required accordingly.

Some of those projects might even feed back into the project, which would be a bonus — but even if they come to nothing, it’s worthwhile for skills development alone.

Michael’s pulling the plug on the server that this website has run on since 2003.

The ‘server’ has changed dramatically in constitution since it all began way back when, but… wow. An astonishingly large part of my teenage years. For the longest time, it seemed as though the Internet had altogether ceased to exist everytime Dale’s connection went out. In the early days, we were all running servers on port 1200 to circumvent ISP restrictions on port 80. phpBB was the order of the day, running Apache — on a pirated copy of Windows 2000 (those were the days in which “legitmate software” constituted an oxymoron). Operating on an early ADSL link with 64kbps upload, forum emoticons were hosted on free web space provided by iiNet in order to conserve bandwidth. You laugh now, but the speed boost was incredible. Every time iiNet dropped out (to future readers: that’s what happens when the internet goes out for a couple of hours, none of this occasional connection time-out rubbish), an irate explanatory post from mwdmeyer would emerge and life would continue as normal. Until parents discovered the server running and turned it off again, which would spark an effort to conceal yet another computer in a room crowded full of equipment. About halfway through 2004, they gave up searching.

These were the days (for me) of NE2000 clones powering Smoothwall/m0n0wall routers, recycling hardware, a subscription to Atomic before all the other kids (I bought more geeky magazines than anyone I know–I think it was that strange meeting place of compters, creativity, and cant that I later became comfortable with), when GeForce 2’s and Pentium 4’s (the first ones with RDRAM that everyone despised) and DDR-supporting Athlons were still zippy. When frame-based redirects passed for domain names — .tk, anyone?

Mostly, it was about the forums… but as for personal publishing, this was no small resource. My first dynamic website was a blog hosted on that server — I don’t think it yet had a name — we all rolled our own web software in those days (it’s not that long ago). Some of us still do. The first domain name acquired was Dale’s, in March 2004, co-inciding (more or less) with the forums’ first birthday. Twelve US dollars later (Joker.com’s prices still haven’t changed), we were all still using frame-based redirects — static IPs were the stuff of pipe-dreams, and Dynamic DNS, though around, was outside of the experience of most of us. Steve ran a notoriously-flaky IIS server with real domains and Exchange, but paid about $150 a month for the privilege: static IPs being available only on business grade internet connections.

These are mere details. The forums themselves constitute an amazing chronicle of the lives of mwdmeyer, ucosty, Sammy, i_am_a_n00bie, Smile:), smKz, n|cktangents, angelicdeity, baibai, Sphinx^, ludvikas, and a handful of others over a fairly tumultuous time. There is so much not recorded explicitly that surrounds the nearly 16,000 messages from these eleven users alone. Some has been suppressed, other parts forgotten, but all of it inextricably linked together in the momentum of time. There are some things about that time which will never be shared with those who weren’t around.

The forums didn’t survive post-school. This shouldn’t be surprising, given the amount of research that says this will be the case for any given relationships faced with that manner of transition, but it was still bizarre witnessing what would have been several months of time spent on a single website evaporate into (not much). The server moved from Balmain to Marian Street, eventually finding its way into a rack there. This is where things get hazy for me. I think the last time I saw Michael might’ve been New Years’ Eve 2005/2006… I feel some sense of guilt about that, but recognise mutual busy-ness had a role such that neither of us should be blamed alone. I don’t believe that a blameless “but things changed” is ever sufficient when talking about close relationships. I’m fairly certain my closest friend for about two years at school is someone that I no longer have anything to do with, but can’t explain why. And I know that I can’t in any way blame him, because I’m so guilty of failing to keep working on relationships myself.

I suppose the point of all this is that the computer formally known as ‘Metro’, now ‘Loki’ (I don’t know how it got that name — Loki to me is an amazing contributor to Linux-based gaming, 2000–2002 RIP, but it could just as easily have been named after the Norse trickster and Odin’s wily accomplice!) isn’t just the latest in a series of bits of electronic gear that some markup and pixels have been piped off for a couple of years. This is just one step closer to a complete closure of a very large chapter of my life… and, yeah, that’s incredibly sad.

Please don’t for a minute consider this to be my arguing that Loki should stay switched on — it’s about something far greater and more personal than a startlingly reliable FreeBSD web server that just happened to host a website for free for a long time.

There aren’t too many people you can make sit in the back of a car on their 18th birthday, much less who will laugh along with as it happens.

This isn’t an obituary, just a poor expression of remorse at the (human) disconnection and ‘drifted’ relationships of that era. Michael, once all this stupid uni crap gets out of the way (maybe after you move again?), I owe you a fairly large drink.

Thankyou.

I haven’t had a great day with Firefox. Well… I spent 3–4 hours in meetings today, so I didn’t even have that much time with Firefox! Still managed to let me down twice, though.

Damn its indisposable development tools *sobs uncontrollably*

I think I’ll switch back to Opera for all non-development Internet-related activity for a while… unless anyone has any other browser recommendations? I’ve seriously thought about IE7, but its rendering is still just a little too patchy for me to be able to live with myself as an Internet user.

Bleh. Let it be observed: even high-profile open source does not always lead to a good product. Its memory management is nothing short of repulsive. It will regularly use more memory than Photoshop and Illustrator combined — admittedly, I use Photoshop mostly for web production and not high resolution print stuff (though that does happen a few times a week, and it won’t often go far beyond the 350MB that Firefox seems to manage fairly regularly)

I’m still using CS2, so there aren’t any magical CS3 memory management advances that make such a claim possible… Firefox just sucks :P

I’d blame Windows being in need of a reinstall (it’s been running since October… more than six months without death :P Plus I started out not being happy with it because it’d been installed from the guy I bought the computer off, I just hacked it to use my CD key instead of the one he’d used to test things… so it’s never been perfect), but really, it’s not that bad for any other application. I normally do a reboot once a week and things are fine… heavy duty graphics editing, occasional video editing, constant mail and occasional wordprocessing… and of all those things it is a web browser that can’t get it right. Perhaps I shouldn’t be so derisive about it seeing as I make a living off developing in this relatively simple world… but I am.

The flip side to all of that, of course, is that I’ve been trying to live (more) like a normal user the past few years. Essentially, recognising that it’s simpler to buy software than write it (WordPress, Flickr), using hackably-open technologies instead of truly open ones (WMA Lossless sans DRM), and a general abandonment of open source principles in favour of vastly improved productivity (Photoshop, Premiere, Office 2007, royalty-free stock).

It’s certainly paid off in terms of professional development and enhanced creative potential… but there’s something lost in not being able to hack visualisations hooked up to a webcam together on a command-line anymore. Admittedly, that sort of thing only comes around half a dozen times a year! But no matter, it’s all good fun. Given more friends who were into that sort of thing and some good music, I’d so live in the party house. I’ve not figured out how to do the same command-line video tricks using Windows just yet, so next time I’ll probably use Windows for visualisations (woo particle emitters!) and a separate Linux-powered laptop (maybe?) for webcam trickery. Then I’ll take webcam stuff straight out into Windows capture and skip my vis mixer altogether for once… I gotta learn to travel lighter anyway!

I stumbled across this post announcing support for OpenID on the Pixelsoup blog of JadedPixel, makers of Shopify. I’m not entirely convinced this even makes sense here, but whatever — it’s fascinating.

On the plus side, Shopify appeared to have disappeared off the face of the planet (in terms of stagnation) for a couple of months there, but their blog at least is alive and kicking. Certainly not for consumption by tomorrow’s audience, but potentially good for some e-commerce applications.

There is no WordPress of online shop software, which is slightly irritating. A great opportunity for a very high profile open source project: take OsCommerce and make it user-friendly and generally not-crappily-interfaced right out of the box! WordPress has done a few innovative feature things, but I reckon most of its success has been built on the grounds that it’s free software (heck, it wouldn’t exist if it couldn’t have forked B2) and the work they’ve poured into the interface.

I’m posting too much geeky stuff. Something non-geeky to come soon, I hope.

It’s almost frustrating to find such good and mature CMS products on the market and not have any need to use them for CYIADA… I spent the last week mostly trying to shape fairly simple data models for different aspects of the website and it’s rapidly becoming clear just how structurally complex multi-tiered community/community generated content sites are. At least in terms of relational DB complexity, yes, this is bigger than MySpace.

Anyway, SilverStripe looks worth a look for simpler endevours.

The complexity is mostly introduced where users become authors, which defies traditional CMS workflow altogether. It’s also far more structured than Wiki systems are, and far more private. Privacy is being balanced against flexibility which is being balanced against community and all of these are being met with time/cost concerns.

But I like to keep telling myself I don’t really know what I’m talking about and getting a programmer will solve all these dilemmas … yeah, right.

I’ve got sick of sitting on my hands and just want to be a web practitioner again. I know the product inside out, it’s been planned to the hilt, stakeholders are universally intrigued/waiting for it, and I’m being impatient and feeling generally like charging forwards. Which is, in all probability, not the best way to be approaching things.

These two consultants came in a week ago and we explained the project to them and (what I heard was) they said “we want flowcharts and scope documents”. I’ve killed a few trees in my time, but the next person to help me in that isn’t going to be a consultant telling me to rehash (yet again — I’ve written documents in so many forms, website copy so many times, etc.) what I’ve got without any further input. I’m in this weird place now where waiting for a programmer is nearly required for further planning action, and everyone but me appears to want more planning before action… meanwhile, I’m writing models for Django and scaring myself with the complexity and learning Adobe products better and generally landing squarely back in front-end territory, which is where I’ve comfortably been for the last two or so years. Well, with the exception of Adobe products, which I only finally caved to last year… whatever :P

Actually, that’s a lie, but I’m feeling a little vitriolic (oxymoron?). Ubuntu didn’t work at all, and of a sudden CentOS did. It’s not quite as polished but I could grow to love it (maybe). I just need to look past this whole RPM thing, which really is ugly compared to the breadth and depth of apt options available. IMO, of course. And the whole ports gig just scares me so I’m gonna stay away from FreeBSD here for a while longer (until this one breaks in another two years?)

I’d forgotten how much work I’d put into making Samba shares behave as well as they had been for the last forever, too. And was convinced there was nothing of value left on the computer (I didn’t delete the home directories, just in case… that was really easy because they’re even on a physically separate volume, it was basically less effort to just leave them there) and consequently (yet again) deleted a MySQL database with StuffOfValue™ in it. In this case a CYIADA survey database I’d built because there were no other options available and (you probably guessed it) I hadn’t sorted out hosting by IT at work yet.

So the aim now is to setup a similarly solid server that’ll last me another two years, barring hardware upgrades (this thing desparately needs more RAM even though it’s got half a gig–I have no idea where it all goes). This time around it’ll be more web-production-esque in its role, which basically means it’ll have more than just being a quiet Samba PDC and file server and web dumping ground on its plate, at least until everything I’ve got planned for it today reaches maturity, or my situation changes to the point that paying for a VPS or real dedicated server somewhere else is a viable option. Loki does, indeed, work quite well, but I can’t screw with it quite as much as some things make me want to (not that I’d want to do that to Loki… in between catastrophic hardware failure it’s amazingly stable and the lack of general screwing-around-ness is probably a big part of that! Probably… :P)

No aspirations surrounding the idea of a media server this time around. Though there’s a possibility I’ll look at maybe building a terabyte RAID server later this year, which would mean rethinking whisper’s role somewhat. It’d probably be relocated to downstairs (it’s cooler there) and replaced by a case with better ventilation and requiring better ventilation. The EPIA board I’ve got isn’t passively cooled, but I reckon it can deal with getting toasty that much more because it’s got a fan stuck to it. It’s a borderline fan requirement, anyway — the hard drives get hotter than the processor (highest I’ve seen the drives is about 62° C, the processor would only hit 55, tops) on forty-something degree Sydney days. If the storage upgrade is called for I’d probably look at getting something with a bit more grunt though, just because if the space requires better ventilation then that lets me stop constraining the system power according to temperature!

Anyway. Now I’m a CentOS kid. Which makes me feel kinda dirty inside because of the whole Prominent North American Upstream Provider All In Title Case issue, but I think I can live with myself for the time being.

OpenOffice doesn’t use that much memory. The GIMP’s been a running and used process nearly as long and it’s still using less than 200MB of memory.

Where, o Firefox, is your clone tool for designing from various ‘inspirations’ online, hey?

Small consolation is that it crashed and didn’t close properly (hence my thinking to check how much memory the monster was using… 2GB is generally pretty roomy for me, so I’m not too stressed about resources), so obviously everything is not okay and they’ve got some work to do for the next release still.

Or I’ll go back to using Opera, I tells ya!

(Slightly tongue in cheek — I can’t go back to Opera, on account of having discovered the wondrous tool that is Firebug. Sigh. Damn extensions. They don’t even use much memory.)

So all I want is an enum field. Or a multiple choice box, easy to get in Excel.

Neither of these are available at time of writing. The term “enum” has only been mentioned on any OO.o mailing lists pertaining to Base nine times, ever. And it supposedly connects to a MySQL server. Yeah, right.

I guess it’s back to rapid prototyping of a web interface to deal with data entry, or using Excel/Access… sigh. This was meant to be the quick and easy (and open source) solution.