I’m going to leave a copy of the archives here because they’re well-indexed by search engines, but if you want to keep up on the latest, go here.

I used Moin on my main site back in the early 1.5-ish days and it worked rather well. I can’t remember why I changed, though it could well have been one of the many service provider moves I made. At any rate, Moin gives me the flexibility that I want, and I’m not terribly scared of Python if I want to get into some of the code that runs it.

I’ll eventually get my blog moved back there. While WordPress does blogs well I would rather keep everything together on one site. For some reason it gives me warm fuzzies.

A couple of the obvious features I like are that it’s written in Perl and has a good security policy. Of course, TWiki is written in Perl and has a security alerts page, so that doesn’t guarantee much. WebGUI is shipped ready-to-go in binary form, and is ostensibly pre-configured, which I figure should save me from making too many foolish mistakes. Time will tell.

I’m going to give it a go on my main site soon. Photos don’t put themselves on-line!

I’m not much past the first chapter in either, but both seem pretty good so far. I’ll post my take on each when I have one.

twiki install centos 5.1: Don’t do it unless you know what you’re getting into. I had a bad security experience likely caused by Twiki, and I’ve seen a lot of people reporting similar experiences that they are convinced was Twiki’s fault. At the very least, google “twiki security” before you do it. The codebase is much too large for a single person to audit it in a reasonable amount of time. If you need a good Wiki in general, give Moin a try. Consider using a Wiki hosting service, where someone else has to worry about the security of the underlying machine. Or just don’t use a Wiki at all — unless you truly want what a Wiki in specific has to offer.

iphone unofficial toolchain easiest way: There is no easy way. There is drudge’s way that I point to in another entry, but aside from that I didn’t find any way that was at all easy. binutils can be convinced to generate (possibly-working) tools for arm-apple-darwin fairly easily, but I think GCC is a bit trickier. Don’t try to do it on your own unless you want to spend a lot of time making the tools work rather than using them.

do peanuts have gluten?: Nope, peanuts don’t have gluten in them. They may have gluten on them, though — it depends on how they’ve been processed or flavoured. When in doubt, read the label and assume that “spices” includes something gluten-based. I’ve seen some with and some without gluten, so have a look around.

Maybe I’m just grumpy, but I’m thinking that FOAF needs an evil twin, maybe called DIAF, which would allow one to express their contempt for such a scheme.

I’m also running into walls left, right, and centre in my plan to get GCC running on the iPod Touch. I had also forgotten about Perl‘s perverse configuration process, which doesn’t lend itself easily to cross-compiling.

At this point I think I’m going to punt on building Perl for the iPod until the official SDK is released and I can get on the developer program. The fact that this will almost certainly take several months (until Apple is ready to accept non-Americans) is annoying to me, but there’s naught that can be done about it — I simply don’t have the knowledge to port GCC to the iPod, and that seems to be the most sane way to get these things going.

Is there anyone out there who knows a bit about autoconf or how bison (or similar projects) fit together who can give me some clue as to what’s going on? I’m a total autotools newbie.

I’ll see what other deps I can pull together in the meantime.

So far I’ve got flex 2.5.35, libiconv 1.12, libtool 2.2, and m4 1.4.10. They’ve all been pretty easy, with the only hiccup coming when the flex build #defined malloc to rpl_malloc, apparently because the autoconf macros AC_FUNC_MALLOC and AC_FUNC_REALLOC set that up if they detect functions that aren’t GNU-compatible. Unfortunately, there’s no implementation of rpl_malloc or rpl_realloc available, so the build died when the symbols couldn’t be found in the libraries.

For the moment I’m just keeping shell scripts that patch files (where necessary) and run configure with the appropriate arguments. I’ll throw them up on my website tomorrow when I get a chance. I’ll update this entry with the link. If anyone has any advice moving forward I’d love to hear it — especially if I’m going about things the wrong way. I just have no clue when it comes to autoconf and friends.

Update: The first packages are up at https://www.kjwcode.com/project/idev/. More will be on the way.

My first try is drudge’s HOWTO build the toolchain on Leopard, and it looks good so far, which is to say that I’ve got “hello world” running on my iPod. It seems to include everything you’d need.

The next big question is how to package things up for installation onto the device. For now I’m going to assume that people can scp things to their device and ssh into it (or use a terminal) to do the installation. I will likely get the Installer packages built eventually, though.

I also have no way of figuring out if the executable is able to find its shared libraries, as otool is conspicuously absent on the iPod, and I don’t know how to check out ARM binaries on my Macbook. My guess at this point is that the Aspen 1.2 SDK will only generate executables for OS 1.2 and higher, so 1.1.3 won’t work.

If this is true, it’s something of a kick in the teeth. I’m sure I’m not the only person who was expecting to be given an SDK that was immediately usable on my device. I trust that a lot of experimentation by a lot of people will turn up some goodies, but for now I’m guessing that I’ll have to wait to use the official SDK. There is the off chance that code that doesn’t touch the UI or specific features of the device will still work — perhaps well enough to port Perl or similar. I’ll be experimenting with that this weekend.

The first is that when I try to add an application to the iPod in the Organizer I get a message to the effect of “your mobile device has encountered an error” and the dialog specifies error code 34. I have no idea what this means yet, but I have a sinking feeling that it will only work with an iPhone, and not an iPod.

The second is that when I try to build the project for Aspen (as opposed to Aspen simulator) I get an error at the last minute about not having a code signing key. This I somewhat expected, but I also thought I should at least be able to put applications on my own device without having one. Typically, Apple is only allowing Americans to have code signing keys at this point, though they say that other countries’ developers will be included “in coming months”.

I’m feeling somewhat foolish for un-jailbreaking my iPod at this point. My only comfort at this point is that I’m not the only person struggling with it.

Update: The likely reason for this error is that the SDK builds code for OS 2.0.

Update: This seems even more likely, since programs built with the Apple SDK give bus errors when run on a device.

Update: If you want to get started building code that runs on a real device instead of the Aspen simulator, build a toolchain that will do the trick.

I also scrubbed my iPod Touch, effectively de-jailbreaking it, now that I can install applications on it through the SDK. I would imagine most people aren’t going to go through the hassle of jailbreaking their iPods (though it’s pretty much mandatory for iPhones in Canada) and most of what I will do will target iPods simply because I don’t have an iPhone to play with.

I’ll post more as I find it out, but my first goal is to build Perl 5.10 for the iPod.

I decided to do a new Perl 5.10 installation and take it from that point to my normal set of modules with CPANPLUS (conveniently packaged as the cpanp command). So far I’m completely won over. There are a couple of things to watch out for if you’re used to the usual CPAN shell:

• only the first letter of the command is checked, so if you try to “upgrade” you’ll wind up uninstalling
• upgrading happens using good old “install”, instead
• search results are numbered and you can operate on search results by the numbers (“i 4″ and so on)
• you can look at READMEs and so on without installing the package (yay!)
• as you might expect, it’s a dream to use with a local CPAN mirror

• Open Terminal.
• Press Command-T to open a new tab.
• Press Command-W to close the new tab.
• Lather, rinse, repeat, and watch your window get smaller and smaller.

Each time you open a new tab Terminal steals a line for screen real-estate for the tab bar, and each time you close the tab the tab bar disappears and shortens the window by a line. It would be much better if it returned the line to the terminal session or did just about anything other than what it does. Fortunately, iTerm has no such issue.

Stevan pointed out that I hadn’t made any attempt to get help from the community before giving up on Moose. If you read yesterday’s post, you may not be terribly surprised. The point is a good one, though, and one I’m taking to heart.

I’ll point to Pigeon source drops when there’s something to drop.

As Steven Weber observed in The Success of Open Source, the Open Source movement has become a kind of Internet era Rorschach test.
People tend to see in the open source software movement the politics that they would like to see—a libertarian reverie, a perfect meritocracy, a utopian gift culture that celebrates an economics of abundance instead of scarcity, a virtual or electronic existence proof of communitarian ideals, a political movement aimed at replacing obsolete nineteenth-century capitalist structures with new “relations of production” more suited to the Information Age.

While I nodded vigorously the first time I read the quote it hasn’t survived my thought on the topic unscathed. I’ve been a big fan of Open Source and Free Software (the last time in this article I will capitalise the terms) since before I could lay hands on it, I don’t see it as a political expression. Open source has always been a practical issue to me.

From the time I booted Linux 0.95a on my 386/SX-20 I have been ever-thankful that free and open source software is available, and like many people (especially younger ones), I did try to turn it into a political expression. Hating Microsoft was as much about political posturing as it was a protest of the shoddy treatment I had received as a user of their software. I had been ticked off when they took away the ability to switch the option character (from ‘/’ to ‘-’, which let us Unix-types use ‘/’ in pathnames), and I certainly wasn’t any more impressed when future DOS versions and Windows began to eat away at my (then-) copious free time, forcing me to spend more time kicking the OS and less time doing what I really wanted to do. As with any mistake in life, be it tight acid-wash jeans in the 1980s, strange musical tastes, or a penchant for too much beer, it’s something I look back on with a mixture of relief that I know better now and foolishness for having fallen into that habit.

Apparently not content to let Microsoft-bashing be the highlight of my political life, I also got somewhat into the free software vs. open source debate. The debate is far from over for a lot of people, but it’s not something I’ve paid attention to in a number of years. Other silly free software debates I was wise enough to be (mostly) quiet about were gcc vs. egcs, Perl vs. Python, Gnome vs. KDE, and quite a few more grassroots attempts to make mountains out of molehills.

Various other political scenes have come and gone in the past several years. “Forking is what open source is all about” vs. “forking is evil” battle, the “everyone should work on MyFaveProject and make it better” vs. “choice is good” feud, “$PROJECT is dead” chants, “if you don’t GPL $FOO I’ll call RMS on you” (as was the case with O’Caml when they decided on a new license), “ESR does not speak for me”, “Linus is $ADJECTIVE”, and the list goes on.

Where I first jumped in with both feet I quickly learned that I had to pick my battles. Before long I had realised that the best battle to pick was none at all, unless it was one that threatened me specifically. I also learned not to be too paranoid when assessing that threat.

The open source and free software community has been divided from the start. I think we got it from the Unix side of things. I remember being pretty young and my mom gave my brother a “Unix wars” poster for his bedroom wall. Now that which Unix you use is no longer politically interesting (you either use it or you don’t) the community has had to search for other divisions in itself to fight about.

I once lived in a housing co-op. The promise of such a thing is that everyone works together and makes the co-op a better place for all. Everyone joins committees that interest them and in doing so serves the community in the best way they can. Elders teach the younger folks the tried and true ways forward and… Well… It doesn’t quite work that way in real life, at least not in the co-op I lived in. Most people joined whatever committee they thought would be the least work. The elders and the younger folks may as well have formally declared war. Rather than work together, many people seemed to try to stab everyone else in the back so no one would try to take the title of “S/he who saved the co-op” from them. Dirty laundry was aired, rather than clean. People got so desperate for attention that they even started bringing their lawyer to co-op meetings. I’m not kidding. I left the co-op not long after.

In a lot of ways I think the nasty side of the FS/OSS community is the same as the co-op. While there are certainly political lessons I can tease out of the co-op aftermath I don’t feel a great need to do so. The same goes for the open source battles fought over the last several years. To be honest, if I’m talking about open source and someone starts up with those I either change the subject or leave.

So after all this rambling, finally my point — I don’t see any sort of political ideal in the FS/OSS community or in the act of releasing software under such a license. I strictly see a practical benefit, both for myself and others. No saluting Chairman Mao, no invoking Ché Guevara, no need to worship Linus, ESR, RMS, or any other set of initials. If you want to contribute, then go ahead. If not, then don’t. If you want to use the software, have at ‘er — and if not, then there’s a stack of commercial alternatives you can look at. The more that opinion and dogma are allowed to play a role in open source the more time is wasted.

These days the “in thing” is to dream up ways that open source is under attack. From websites that don’t support Firefox to perceived threats from Microsoft (who actually has a near-OSD-compliant license now) or Sun (who just bought MySQL and hired two big-time Pythonistas) or SCO (whose antics have not yet made any measurable impact on open source), everyone wants to rally the troops and storm the castle. If you doubt this, simply follow Slashdot or Digg for a little while. Once the troops are rallied, everyone writes an e-mail or floods a webserver or does whatever they’re told to do, and at the end of the day they claim victory, no matter the outcome.

With antics like these it is hard to take any sort of Rorschach test too seriously. A Rorschach has always looked a bit like goat barf to me, anyway.

What do I want to see in the FS/OSS community? Less opinion, less dogma, a lot more live-and-let-live. Tolerate others, even the ones who really seriously piss you off. Try to find the right times to watch something and the right times to turn your back. Keep a sense of humour — nothing pisses off someone who’s trying to get under your skin like being laughed at.

That’s as close to a call to action that you’ll get from me. I’ll just be over here, doing my thing.

The first is the new look. The UI seems to be lighter than before and uses a lot of translucent items, which seems to be all the rage amongst Apple’s HCI folks these days. It’s a good change, I think — I don’t have a copy of Aperture 1.5 to compare it with, but it is certainly easy on the eyes.

Something else I noticed were the speed improvements. Previews generate a lot faster, the UI seems a lot more snappy, and I didn’t feel like I was waiting for much. For general browsing about and checking out photos the speed seems to be on par with Adobe Lightroom. I think it may be a little quicker at start-up and browsing through all of my photos (5’500) in one go. That being said I don’t normally torture myself like that. I think it may get the speed gains through all of the files being stored in the Aperture Library, but I haven’t looked at the format of that file (or whether it is even a single file or one of Apple’s famous directories-masquerading-as-files).

I haven’t checked out its memory usage, but I am sure it’s on par with Lightroom’s. The hardware requirements say a minimum of 2GB of RAM, and I’m quite sure that’s the case.

The only annoyance I have had so far is that the loupe seems to have a mind of its own when it comes to the initial zoom percentage the first time you bring it up after selecting a new image in the viewer. Sometimes it’s 100%, sometimes 1600%, and sometimes just about anything in between — I was getting 165%, 473%, and quite a few other weird numbers.

I’ve got 30 days left on the trial and I’m going to try to use it alongside Lightroom and see which I prefer. Now that Aperture is available for a decent price I’m willing to consider switching. I know Lightroom well enough to teach it to Robynne, so I’m sure it wouldn’t go to waste — I’m sure she’d love to shoot in raw and get creative with it.

Though I’m sure we’ll have it in our hands in good time I can’t help but think that Apple should have managed our expectations a little better, and perhaps told people it would be late. Keeping us guessing isn’t going to win them any friends and will do little to deter the hordes of people jailbreaking their devices.

I’m now quite sure he’s an endocrinologist I can work with. He knows when to be hands-on and when to keep his hands off, and I think I understand how he thinks about some things. Taking numbers and charts in today made all the difference, I think — his annotations tell a lot and give me ideas how to improve my record keeping.

My data is currently stored in a spreadsheet. When I get it nailed down I’ll post an empty sheet here for any other Numbers users who would find it handy.

It’s funny how two tiny features with only the bare minimum of integration can push the relative worth of the subscription from “worth it” to “addicted”.

Am I sure that design patterns are the way to go? Heck no, but it doesn’t hurt to learn something new. It can be hard to tell the difference between fashion and function these days, so the best way to go forward is to read it and see what sticks.

So far (a couple of sections into chapter three) what’s stuck with me the most is the reminder that composition beats inheritance for many uses. I’d been (naively) constructing a class hierarchy for an application that was inheritance-based and it was the right reminder at the right time.

On page 225 I notice that he mentions:

A glidescope beam provides vertical guidance that directs the aircraft to follow the typical 3-degree decent path to the runway for a smooth, “glide-in” landing.

While three degrees may be a pretty decent path, I’m pretty sure he meant to be talking about a descent path in the sentence above. This echoes something I pointed out to Lorenzo last night — you can usually count on a spell-checker to tell you if you’ve spelled a word correctly, but you can’t count on it to tell you whether you’ve used the right word.

I wonder who David would deem responsible for the error above?

The first couple that come to mind are MD5 and SHA-1. MD5 is largely being replaced by SHA-1 due to concerns about some attacks on the hash, so it’s easy to choose between those two. UUIDs are also a possibility. Modern mail transfer agents (MTAs) like Postfix seem to use a much more compact value, so I should probably see what they do as well.

It’s also tempting to use time-based values, like the number if milliseconds since the epoch. One would only have to cache values used in the last second to make sure you didn’t reuse any IDs, and that would be a fair amount more compact than a UUID or a 160-bit hash in hexadecimal.

It seems that Postfix uses something based on the message queue ID it assigns to each message (src/global/mail_queue.c), and also uses millisecond-resolution time to ensure every ID is unique. I can’t find Sendmail’s Message-ID code after a quick look, so I’m guessing it’s something that can be defined in sendmail.cf.

I’ll need to do a bit more reading to figure out whether a SHA-1 hash of the current time with milliseconds combined with the sender, recipient, and subject or something similar will do the trick in a portable manner.

Another thing that’s changed is the price — just USD15 now. I remember it being a lot more expensive in the past. The 30-day trial will tell me if it’ll do the things I need it to do.

hiveminder windows mobile 6: I just got a new mobile with Windows Mobile 6 Professional on it, so I was actually able to try this out. I haven’t tried anything too fancy, but it seems to work just fine, so long as you are using the Hiveminder mobile interface. I haven’t tried the desktop interface on my mobile yet, and probably won’t.

opera 8.65 is too expensive: No, it’s not. It’s a very well-designed browser, and USD25 is a pretty good deal for software on most mobile platforms. I tested the software for over a week before I bought it and made sure it did all of the things that I wanted it to do. I ran into very few problems, which is another thing that sets it apart from much other mobile software — I seem to have a knack for killing pieces of software others have no problem with. I’ve been using Opera for months on my mobile with very few hiccups.

If you’re really dead-set on the opinion that Opera isn’t worth the money, simply use the browser that came with your device for a while longer. It won’t take that long to convince yourself. The only stock mobile browser that I have found to be worth its weight in gold is Safari as shipped on the iPod Touch (and presumably the iPhone), but even that has the price of the mobile device attached to it.

Finally, if you’re absolutely sure that a USD25 browser is a travesty, write your own. That’ll show ‘em.

There were a couple of photographers down there today who had Canon 1Ds and huge lenses set up down on the mud, off the pedestrian area. I find this behaviour annoying, not only because it makes other people think the behaviour is okay, but because you can tell a lot about a photographer by how they interact with the environment where they’re shooting.

A lot of people have heard me say that I love photography, but hate photographers. This isn’t to say that I am annoyed by every other person who picks up a camera. Rather, I can’t stand photographers who are entirely money-driven and don’t care about the environment they shoot in or those things that live in it.

Sure, the occasional photographer trampling the mud when the water is low won’t hurt much. The problem is that if people see one person doing it, they tend to think it’s okay for them to do it, too. I made mention of this to the photographers and pointed out a kid who had decided to go down there, and got a couple of lame excuses. They said it was the only kid they’d seen do that in four hours, but that isn’t really the point — one kid feeding the birds does little damage, but three kids running about trying to pick up goslings is dangerous for both goslings and the kids.

In the end I engaged one of the photographers in conversation, made my point, let him say his thing, and left. I was unconvinced by his argument, as I’m sure he was unconvinced by mine. I think that following a set of ethical field practices is a must, but I can certainly understand that it gets in the way of shooting for money. Oh well — it certainly wasn’t the first time I’d encountered parasitic photographers (street photographers are among the worst offenders, and I live in a city with lots of them), and it definitely won’t be the last.

Like anything I’ve only started to ponder, I don’t have any good answers. The subjects of my photography aren’t necessarily near and dear to me. I love birds, but they aren’t exactly my family. Industrial equipment, architecture, and industrial landscape don’t play large roles in my life. Indeed, I don’t spend any time around these things unless I’m photographing them. While several of the photographers featured were laying some aspect of themselves bare, there’s not much that can be told about me by looking at my pictures. My photography doesn’t necessarily show anything unique about my subject, either — there are thousands of photographers who can tell a story with a picture that I can’t even put in words.

So why do I do it? I think it’s a way of exploring a subject I’m curious about. Araki touched on the link between photography and memories, and mentioned that he doesn’t photograph the things he doesn’t want to remember. Perhaps it’s a way I obtain, organise, and keep information. That certainly seems like it fits well with the non-photography parts of me.

On a completely unrelated note, I just saw an ad for a nature show that used the song I Disappear by the Faint. Just when I thought I’d seen it all.

Instead, I’ve installed Fedora 8 and mach, which is probably for the best — it gives me the freedom to build for quite a few different distributions on different platforms. The default install on FC8-64 includes a target for CentOS 5 x86_64. So far it’s building cpanspec just fine. More useful packages will come soon.

Today was my first trip to a dentist since a 2006 root canal, and I was expecting to have a laundry list of “must-fix” issues to deal with. Instead, I have only one, and it’s one I’ve known about for a long time. I think this is my first cavity-free check-up since I was a child.

I’m relieved, to say the least. I’ll be resting easier this weekend.

In a fit of desperation I’ve reformatted my disk and reinstalled OS X. The OS is updating now, and Fusion is downloading. There were more problems than just Fusion that led to the OS reinstall so hopefully I can nip those in the bud, too.

In the meantime I’m going to try to blog when the urge hits, rather than waiting until the evening.

I was less than impressed at one point because the process of activating my account was acting wonky. I’d try to sign in and it would put me at the start of signing up for the current ten-days-free promotion. If I filled all that out and gave it my credit card it would complain that I already had an account and should sign in. When I clicked “Sign in” it would kick me back to the start of the promo sign-up. Lather, rinse, repeat.

I got on the support portal and tried chatting with the support reps, but no dice — they seemed to be completely ignoring me. After the second time I hit “Exit” to leave the chat and received one 400-class HTTP error or another I began to suspect something was up. I hunted down the phone number for their support and called. This turned out to be a good move.

A CSR answered the phone after about thirty seconds, and I explained I had just had a very unpleasant experience trying to get some assistance from the CSR chat on the portal. At that point the rep called me by name and explained that she had just tried to figure out what was up with that, as the folks in India had started checking to see if the chat system was on the blink. I tried to give some information that might be of use to whoever is stuck debugging whatever happened, then we moved on to the reason for my call.

After I relayed what was going on and a few minutes of debugging had passed the reason for the problems started to become apparent. My account had originally been an O’Reilly portal account that I had used for registering books I bought back in 2006. A little later that year I bought a book that had a Safari promotion and had transferred my account from O’Reilly. Apparently something went wrong there, because she described my account as looking half O’Reilly and half Safari. In the end, we agreed it was best if she just deleted the account so I could sign up again.

Three minutes later I was signed up again and was prowling through the book list. While I was angry at the beginning, I was won over again by the end. A big thanks to Diane at Safari for that!

(Disclosure: I work for a Safari Books Online reseller (you get a basic membership with your Pro Studio subscription), but this blog entry is about my experience and opinion, and something I bought on my own dime — not anything related to work.)

Typically for Beau the bag insert is around ten dollars under Crumpler‘s list price. I’ll throw up a pic or two if I work out a great way of setting it up, as I’m guessing I’m not the only D200 shooter who prefers to keep a 70-300mm on the body with a MB-D200 attached.

Aside from the mergansers there were a good number of common goldeneyes (Bucephala clangula), a couple of American wigeons (Anas americana), some American coots (Fulicana americana), and the usual smattering of mallards and gulls. The bonus birds of the day were some female and a male canvasback (Aythya valisineria) and various greater and lesser scaups (Aythya marila and Aythya affinis, respectively).

I got a few potentially good shots. I’ll put them on Flickr when I get the time.

I’m looking forward to exploring Ruby on this new platform. I don’t have any clue how to build binaries for the platform (or even how to set up a cross-compiler that targets it), so it’ll be pure Ruby for now. I’ll have to see if RubyGems will work for pure-Ruby gems.

I did have some problems trying to sign up to Hiveminder before, and that’s actually a good part of the reason I installed Trac in the first place. I don’t want the maintenance and security hassles, though, so it’s time to push those on someone else. Hiveminder looks like the best bet.

There’s some great technical stuff on the horizon — Apple is soon to release their iPhone SDK and I’ve scrapped my current set of personal projects to work on something a little more challenging and exciting.

I’ve still got around 60GB of hard disk space free, so I’m in no hurry to get a 250GB or 320GB drive for my laptop just yet. We’ll see how that number holds now that I have enough RAM to do some serious virtualisation. Tim Bray reports that new Solaris builds are looking pretty good, so I’ll have to give those a try.

When Jon passed away in 1998 there were about 2400 RFCs published. Now, almost ten years later, there are 5200 allocated RFC numbers with a very large number of those linked to finished papers. While this represents the work of many thousands of authors, it’s no less a monumental task than the Linux kernel. It’s just not as sexy and attention-getting. In many ways, Jon was the Linus of the RFC series, giving it a healthy start and shepherding it once it grew legs of its own.

Some reading of RFCs for a personal project or two got me wondering what the new RFCs are all about. I’ll likely touch on the ones I find interesting here.

Compressed raw logs take a lot of time to process. It’s not so bad if your site isn’t terribly busy or if you’ve got a lot of time on your hands. Granted, my site isn’t busy at all and I could wait a while for results, but I’m hoping someone else will find the program useful.

Intermediate storage is of course the key, but that’s where things get really tricky. It’s easy enough to store sums and averages of various numbers, but without storing stats for all 404s, how can you report on the top x of them? I’m half-way to an answer, but it’s probably not going to be pretty. It’s certainly not going to be elegant. But it’s also not going to be Webalizer, which doesn’t make sense to try to modify to suit my needs.

Another idea for a nifty tool is a web log playback tool — given an access_log, it replays the log, making the same requests at the same time offsets in the file. It may not be coming from all over the place (instead originating from one machine), but it may help debug some performance corner cases, for example.

I’m sure the folks in the social networking side of the office know I’m the loyal opposition.

Best of all, it seems to be really reasonably priced. I’ll have to contact my favourite photography store to see if they can get it in for me.

In the meantime I’m using my Crumpler Dreadful Embarrassment, a notebook bag, with a makeshift bucket liner (consisting of a Crumpler Three-Million Dollar Home with its strap tucked into its front pouch). It’s not too friendly towards the battery grip on my camera so I’ve got it off now, for the first time ever.

The backpack is fantastic for travel, but I think I will be looking for an over-the-shoulder style bag for local use. I may just get a Five or Six-Million Dollar Home and keep using the Dreadful Embarrassment if that works well enough. I Know that Lowepro has some shoulder bags with the AW rain slick, which is a pretty big deal for me. I don’t use the rain slick often, but I’m sure thankful I have it when I need it. I may also look at getting an attachable rain slick for the Embarrassment if I can find something that will work.

I somewhat like the bag-within-a-bag idea because it’s a bit more flexible — if I don’t need one piece, I can leave it at home. If I need both, I can carry both. I’ll have to see what works.

It’s interesting to see the collection of gems expand and the version numbers creep up. It’s hard to tell what’s stopped running up versions because it’s matured and in maintenance and what’s been abandoned. It’s even harder to tell what’s commonly used. The former can probably be found by lots of searching on RubyForge, but the latter is a lot harder to gauge.

Every day I spend working on the software is another day I have to spend using a spreadsheet to store the data that accumulates that day. I don’t know Java well enough to do incremental development — even the simple stuff seems to take an aeon and I feel like the resulting code is too brittle.

It’s nothing that experience can’t solve, but I don’t know if this is the right project to get that experience with. If anyone has a suggestion of what would make a good project to learn Java with — one that would include:

• essential Java classes and data types
• the Collections framework
• database connectivity
• XML

I’d certainly be all ears, especially if there were other Java newbies looking for someone to join their project. I have six or eight hours a week to dedicate to it most weeks — some more, some less. You can find me as user kjwcode at Gee Mail dot com.

This time around it’s web stats. I know the “modern way” is through Google Analytics or similar systems, but I’m something of a fan of static logfile analysis. A good part of the reason is that the data is mine — I can verify what I see and weed more data out of the noise at any time in the future, not just when I upgrade the JavaScript or when someone has it enabled.

I know there’s a lot that I’m missing out on, too, but that will be the interesting part of the project — figuring out the things that will set my pet project apart from the others.

Another gotcha I ran into was having a constructor that expects a “short” seem to be ignored because you have to explicitly cast to short for it to be recognised. Otherwise, Java thinks it’s looking for a constructor that takes an “int”.

There have only been eight or nine thousand invalid login attempts in the past couple of weeks. This is a big drop from what it was — over 110’000 in only three weeks! I only black-holed a couple of IPs for that today.

The new annoyance is people/programs that set their HTTP UserAgent header to random strings. Combine that with the tendency of these specific people/programs to access only data that has 404′d for a long while and you’ll find one more IP and a class C black-holed. I can understand changing one’s UserAgent to something different from what it actually is for privacy reasons, but cluttering my web stats output is just rude. Pick one and stick to it. I recommend “NotTelling/1.0″.

I have a feeling that the CLASSPATH is going to be my mortal enemy when it comes to Java programming. It’s not that it’s difficult, it’s that it’s horrendously nitpicky and important. It’s also not that it’s any different than @INC, it’s just that I’m not used to it. Bah humbug.

Ant is a far cry from “make“, as it is designed to be. I’m still not entirely comfortable with the XML syntax used in the build.xml files, but that will come with time.

By stealing sections of tutorials and cookbooks I’ve put together a build.xml file that almost works the way I want it to. It’ll be even more useful when I actually have a main class for my little project.

I’m most interested in JUnit at this point. I’m not expecting to go full-on TDD, but I do want to get tests in place to catch the early mistakes.

I started writing a tiny amount of code for my diabetic information tracking software today, so hopefully that will continue and expand. I’ll wrap that up in a few days when I’ve added a bit more and make another drop.

My first thought was Netbeans. I remembered Tim Bray mentioning the new version and the screenshots looked good. I was able to get it installed and configured, but it wouldn’t create a project. I eventually had to force-quit it.

Next up was JEdit, which had nothing wrong with it besides making my preferred fonts display badly. Unfortunately that’s not something I can live with, especially when I’m working with a language I’m horrible at.

I also gave Eclipse a shot. It’s a huge program written in Java itself, and it shows. It’s also likely overkill for what I want from an editor.

Textmate was a no-go from the start. I’m sure it’s a great editor, but it’s not my cup of tea. I know there’s value in simplification, but there are some things I need a bit more control over.

In the end I’ll probably end up using Komodo, Smultron, or vim.

One language that’s featured somewhat frequently on the site is Java, and with good reason — like any widely-used language it has complexity, warts, misfeatures, bugs, historical behaviour, and outright badness that makes for a good punch line or two. This puts it on equal footing with my admitted favourite language.

I was reminded of this today because I was language-shopping for a project. (Yes, another project.) I’ve been keeping diabetes data in a Numbers spreadsheet, but the charting I can do is limited and trying to do more than very simple functions would have me pulling my hair out if I had any. The criteria I had were that the language had to be cross-platform, handle graphical applications well, and have better integration with various operating systems than Perl does. Having a huge standard library to help cut down on dependencies also scored high.

My first thought was Mono, but it seems that project necessarily lags behind the current wave of C# and .NET from Microsoft. To go that route I would want to be on the bleeding edge, as learning earlier versions would probably frustrate the heck out of me. I did a bit of searching about and found a developer preview of Java 6 for Mac OS X, and I think it will do nicely. It meets my criteria, and has support for the things I’m likely to need:

• GUI (Swing)
• regular expressions (but not nearly as good as Perl)
• XML support
• sockets and SSL out of the box
• heaps and loads of documentation
• large installed base, not likely to be missing on many systems

A critical part of my decision was the availability of high-quality, well-regarded codebases to look at and learn from. Java won this challenge hands-down, as I have to look no further than the Apache project to get that.

Now comes the fun part — creating my very own WTFs. My Subversion repository isn’t public, but I will try to do (at least weekly) source code drops. Starting from the very first bits and pieces. You can see an abomination form in real time! Or not, if you don’t like looking at the programming equivalent of a train wreck.

Update: You can get the first source drop in the glutrack project directory.

It’s too early to tell for sure, but my day-one impression is that it seems to do a better job of getting my mid-to-late afternoon highs under control. This is my biggest hurdle in getting my HbA1c under control, as far as I can tell.

I’m also going to start dosing my basal insulin at bedtime instead of in the morning. This will give me better coverage while I sleep, and if things fall a little short it will be in afternoon or evening when I’m checking my blood glucose levels and can correct it.

Depending on how my experiments go I may keep the insulin aspart when I go back on the pump. It will mean rejigging my basal profiles, but if it means a more responsive course of therapy, then so be it — it’s not a lot of effort for so much benefit.

Atom is the new kid on the block and looks like a good bet, but RSS seems to be better-supported by Perl. I haven’t looked at the situation on Linux yet (only OS X so far), but the Atom module doesn’t seem to want to build (at least not using CPAN).

I got an iptables firewall figured out for my VPS, and also did some work on a spreadsheet for keeping track of diabetes info. Numbers made the latter task pretty easy.

Time to rest now — more details later.

Initially I tried VGA — no dice. The machine said the console was on serial and no further output would be sent to VGA. I went into the boot options and set VGA to be primary, but didn’t have any luck there, either. It simply did the same thing as before. I tried from the EFI shell and that was about as useful as anything else I’d tried up until that point. In desperation I enabled the legacy user interface and disabled the serial port consoles. This finally got me going on VGA in text mode.

Looking back I probably could have gotten the same effect by completely disabling the serial consoles in one of the other interfaces as well, but it didn’t occur to me. Why would I need to do that? What’s the logic in defining primary and secondary console interfaces if the machine isn’t going to use them in that order?

The good news is that the machine is up and running now, and I hope not to have to bother with it again for a while.

Fortunately, it wasn’t too hard to get the pharmacy to give me the insulins I needed. I’m using insulin glargine as a basal, and sticking with Humalog for bolus. I was tempted to try another rapid insulin, but I thought it best to stick with the devil I know for now. A couple of new pens, some pen tip needles, and I was set to go.

The first day’s numbers have been a little high, but they are roughly in line with what I have been able to achieve with the pump over the last couple of months. They can only get better from here, as I am able to hit far more injection sites with the pens, and it gives me a much wider site rotation.

I’ll probably use the pens for a month or two and then go back to the pump when my primary infusion set sites (within about six or so inches of the bellybutton) are healthy again. In the meantime it’s a nice change of pace, though I did find myself up a couple of times last night wondering where my pump was.

I was considering having a strictly-static site, but it makes sense to have some CGI there for full text and tag searches. It’s a little more useful and a little less 1996.

Since I reinstalled my VPS earlier this month I’ve chalked up over 110’000 failed login attempts. Some of them are understandable — attempts at root, mostly. Some of them are stupid — since when did the postgres user get a password? The rest are beyond moronic. What are the chances that a completely random server has a user named “avayo”? Do these people really think “frank” is a common name any more? I don’t personally know anyone named Frank who is under 60 years of age, and he certainly wouldn’t be on-line pretending to be a sysadmin in his mid-30s.

Days like this I wish I had a cattle prod that worked over the ‘net.

I’ll have to do a bit of work to weed out the existing (incorrect) APR installation, but that can go on the to-do list.

It’s nice to see that I’m not the only person to get to feeling that way. I often wonder where the real committers get the time and energy to do the work that they do and to deal with the people who seem to enjoytrying to waste their time.

Needless to say, I’m in complete agreement with brian. When people ask me which language to use to solve any given problem my answer is always the same. I tell them to use the language they know the best.

Tim Bray‘s Wide Finder project got some pretty good results with Perl, and Perl isn’t thought of as being a many-core friendly language. My guess is that when you start with a solid base you can get a lot more done than by trying to specialise too quickly.

His view matches my own — experimentation is good, and it’s something I used to do a lot more of when there were fewer books, websites, blogs, and frameworks to rely on. Since those have come about I’ve spent more time looking things up than I have working through them. There are good and bad points, obviously.

On the good side are the quick solutions or at least quick head starts on a solution that can be found from the huge number of resources. My job values TTR (Time To Resolution) above almost all else, and the great oracle Google can be trusted to help point the way. It’s also easy to figure out which of the many solutions is most likely to work. Simply seeing how many times it’s repeated and by who can tell a lot.

On the bad side, I have felt increasingly disconnected from my coding, even to the point of avoiding it altogether for a while. There wasn’t a lot of joy there any more. Ruby on Rails says to do a job this way, Django says to do it another. People stop coding, start drinking the framework kool-aid (and not just web application frameworks, either), and start bitching each other out and trying to claim superiority because their favourite framework/library/methodology is obviously the “one true way”.

This experience has led me almost full-circle. I jumped on the “modern programming” bandwagon, got filled to the brim with other peoples’ opinions of how things should be done, avoided doing much at all because some huge percentage of what was going on in a lot of communities was nothing more than advocacy, and then turned my back on those communities and their frameworks and opinions and started coding for myself again.

Sure, people writing code for a living can get pushed into believing that there is no excuse for experimentation or learning through experience when the quick and easy answer will do 80% of what you want 90% of the time. My guess is that if you look at two teams of programmers, one of each mindset (experimentation vs. copy-and-paste from their favourite code resource or going by the (cook)book) you would find that the copy-and-pasters will finish a task first, but the experimenters will do better work.

Of course, that’s just my opinion. I’m willing to discuss it, but I’m not willing to argue it. I’d rather be coding.

I haven’t decided which database to use yet. Berkeley looks likely, or possibly SQLite. It really depends on how I intend to use the tags.

Truth be told, it’s a great game. It gives my Macbook a good run for its money — it’s full fans during play. The graphics are pretty smooth, though. They keep the game interesting, and it doesn’t take too much to follow. I like the cartoons in the game, and it has just the right level of seriousness (that is, none at all).

The only down-side is that Lo wants to play it now, and keeps asking when we will get him a Mac to play it on. Serves me right.

I’ve got Webalizer on board now, and that will do the trick without any JavaScript muddling up the pages. It also means I can see results from users with JavaScript turned off, search engines, and other automated processes. The whole site is only a few dozen pages for now, but when it gets larger I’m going to want some insight into who’s crawling what and why.

This will be important especially once the birds section is up and running. I still don’t know what I’m going to do for site searches, but if there is a permitted/ethical way of using Google for that (besides hot-linking their search with a “+site:kjwcode.com”, which they may not like too much) I’ll want to make sure I’m getting good crawl coverage.

I got a good over-the-shoulder shot of her before the light left, though. There were a few other good shots there, too. Now I just need to get the time to get them put up on
my Flickr photostream.

The super shopper Robynne found it for $8 — a great deal!

Robynne has an infected tooth, so her dentist prescribed her some antibiotics and some Tylenol 3 for the pain. We took the prescription to Save-On-Foods at Park and Tilford and were given an unpleasant run-around by the pharmacist there. Robynne eventually took the prescription back and we went to her regular place where they actually know her well, and it was filled in five minutes with no hassle.

It’s unfortunate on many levels, because this isn’t the first screw-up by Save-On-Foods — it is, however, the last one I’m willing to put up with. I have had nothing but pain and bother in my dealings with their pharmacy. They take far too long to bring in my supplies for my insulin pump, they don’t interface well to my insurance company (and it’s not like Manulife or ESI Canada are difficult to find or get ahold of and work with to make sure they get things right), and their prices aren’t exactly competitive.

For now, Robynne will stick with her place and I’ll stick with mine. They both know us well, we have good rapport with the staff, and above all they have the knowledge to handle the complexities of their clients’ cases and the desire to work with their clients, rather than against them.

(As an aside to the SOF pharmacist in question, numbered Tylenols have 300mg of acetaminophen per tablet, not 500mg. Trying to say that taking 12 tablets containing acetaminophen per day is dangerous is ridiculous — try stating a limit in mg instead, like a professional, who would know that there are two or more common strengths of tablets. If one were to take 12 tablets of extra-strength Tylenol per day they would be taking one and a half times the recommended dose. And one more point — as a pharmacist, you are the dispenser, not the prescriber. There’s a good reason that the prescribers don’t dispense and there are fabulous reasons why the dispensers shouldn’t prescribe. Thank you for your cooperation in demonstrating this so successfully. You can’t replace years of education with a big, blue book.)

I didn’t do any real swimming, but instead chose to float and bounce about and talk to Robynne and Martyn. Lorenzo didn’t seem to have trouble finding a playmate, and had quite a bit of fun playing ball.

Robynne even bought me a bathing suit that fits, so I guess I don’t have an excuse now. Bring on the water!

The first is that there is a heck of a lot of data that I want to keep. The usual bird data like common name and species, the specialised data like a unique ID for each bird (often used in bird counts, for example), the official info for each bird (plus references and sources), my own observations on the bird, good web links about the bird… This smells like a job for XML.

In the “bad old days” I knew a bit of XSL and XSLT, but I have no intention of brushing the cobwebs off those skills. Even though I try to keep everything XHTML 1.0 compliant, I still think I can do more in less time with Perl scripts and the regular XML APIs than I could with XSL.

The other arguments in favour of XML in this case are that I can define a DTD or schema and use a validator to ensure that my data files are valid, and that I can easily transform one format to another as my needs change. This doesn’t save me all of the up-front work of figuring out an appropriate format for my data, but it does mean that it will be easier if I have to change down the road.

My next job is to figure out whether to use a DTD or schema to describe the XML application for the data. I know DTDs much better, but namespaces could be useful, and I think those are a schema-only feature. We’ll see.

I didn’t do Macheist last year. To be honest, I can’t even remember which programs were offered. This year’s offerings are great, though. There are some that I probably won’t use much, if at all (iStopMotion, Awaken, and WingNuts 2 come to mind).

There are some that I’m really looking forward to, though. Snapz Pro X (which I’ve used at work) is great piece of software, and worth $20 more than the bundle price on its own. Pixelmator looks like it’ll be a hoot to play with, and CSSEdit looks like a good way tweak CSS if I’m not currently in Komodo IDE. 1Password is probably going to be my favourite, as I have become quite addicted to Keychain and know that not everyone wants to support features only present on OS X.

The “looks interesting” list includes TaskPaper (appears to be a graphical front-end to text-based to-do lists), AppZapper (appears to be a graphical front-end to Archive.bom files), and Cha-Ching (appears to be a budget planner/bank account balancer).

Head on over and give it a gander. 25% of proceeds go to one (or all) of ten different charities, which is a nice touch.

Today I bit the bullet and got credit reports from both Equifax and TransUnion. I was in for a pleasant surprise.

I’m in the last six or so months of payments to cover old debt. That feels pretty darn good, because I’m coming down the other side of that mountain and picking up speed. The mountain doesn’t look any smaller from the other side, but it is nicer to look at when you’re beyond wondering when you’ll hit the top.

The best part is that I’ve got new spending habits that I’ve developed in the intervening time. I buy stuff when I can afford it, and make do with what I’ve got until I can. I don’t buy the cheapest thing that will “do for now”. I spend the money and buy what I want and need, and it’ll have to be pried out of my cold, dead fingers (or I’ll have to see something even better, like the 14-bit-per-pixel image engine on the Nikon D300 — but I’m not spending $1900 for two bits of colour space and being stuck with live view and a bunch of other features I don’t care about).

There’s something to be said about what makes that the case. Sure, there are words that come up, like “responsible”, “organised”, and “planning for the future”. Perhaps that’s true, but there’s more to it than that. A big, important change is that I’m no longer the novelty fetishist I used to be.

When I bought a laptop in the past it was usually some hunk of equipment that would do about half of what I needed and I could barely afford. Most of the time I would be eyeing up other systems before I had even bought the one I did. These days I’ve learned to appreciate what I’ve got. Heck, the last laptop I bought was almost identical to the last one I had, but had a 64-bit processor and could handle 4GB of RAM instead of 2GB. If there hadn’t been a virtually identical system available that could handle the extra RAM I would almost certainly still be using my old laptop. I just like it that much.

If I think about it, I could point fingers at all sorts of things. Self-confidence and self-esteem are good candidates. After all, if you don’t have those traits in good doses you’re more likely to judge yourself by what you own instead of who you are. I can think of quite a few other facets of one’s self that can be related to these things.

I intended to spend time today looking at my financial situation going forward, but wound up looking deeper than that. Fortunately, I like what I see.

The inconvenience instead pushed me to install a private Trac instance on my VPS. I’ve done a few Trac installations before, so it wasn’t a big deal. It’s locked right down, with no permissions granted to the anonymous user. I only want the ticket tracking and SVN browsing abilities. I’ve got no use for a wiki now.

I’ve got a ‘kjw’ component set up as a sort of personal to-do list. I’m actually learning something about creating reports in Trac, so this is likely to be useful.

Next up are monthly and yearly indexes, then creating and updating the symlink to index.html so it points to the right place. I’m also going to refactor some common functionality out of the blog generator script and stick it in a library. There’s certainly lots to do before I’d be willing to let anyone else see the code.

After the blog end of things is complete I’m looking at doing the birds section. That should be a lot of fun. I’ve got a lot of planning before I figure out the raw format, though.

I was courting a headache at the time, but it wasn’t fully-formed. I was able to take something for the headache after the procedure, so now it’s just on the periphery. Hopefully I’ll be asleep before it hits.

There were some great birds down at the lagoon. The usual smattering of mallards, Canada geese, and scaups were in evidence. There were also a metric heap of common goldeneyes and even a male common merganser who was kind enough to pose for a few pictures.

We certainly weren’t the only people taking advantage of the non-sucky weather. There were some other photographers down there, including a couple that resembled Revs and Audy from Flickr. There were also the crazy swan lady and crazy swan man, who we kept a wide berth from. There were the usual legions of people feeding bread to the birds, but thankfully the idiots with the carmel popcorn and Doritos were absent.

After the lagoon and the restaurant we headed to London Drugs where I picked up a Bluetooth mouse to go with my keyboard. It’s approximately a zillion times better than my USB-based wireless mouse, and all told I’m totally won over — Robynne will inherit the USB-based wireless stuff if she wants it.

Today was my brother’s birthday, so we went to his place for a yummy fondue. There was a cheese fondue followed by a chocolate fondue, and I will admit to having a few chocolate-covered marshmallows rolled in coconut. Yum! Several of us got into a conversation about the worst fashions of the late 1980s and early 1990s, and the idea of a theme party came up. First episodes of 90210, 21 Jumpstreet, and so forth were threatened. Too bad I don’t still own any of my clothes from that era!

The benefits are many — it parses simple files and outputs proper XHTML (provided that the files it sources don’t break the rules). It’s perhaps a bit pedantic for general use, but I like it just fine. Another benefit is that I maintain all of the expressiveness of XHTML with CSS2.

I implemented a small templating module to make this run, and may actually release it to CPAN as Text::Template::Trivial or something like that. I’ll have to get Module::Starter involved.

I may also look at latex2html, but when I saw it last there was a problem with node renumbering ruining bookmarks. I’ll get it figured out soon, I’m sure.

The highlights are necessarily personal, but one of the goals I’m looking forward to is releasing some open source software to the community. ActiveState is a great place to work, and I’m lucky to have a supervisor like Gozer.

In other vaguely performance-related news I’m happy to see that mod_perl is easily supported on CentOS. My plans are something like this:

• automatic POD-to-HTML
• automatic LaTeX-to-HTML
• photographs and other large files stored on S3
• as much as possible, make everything as simple and efficient as possible

It seems possible to do — now it’s just a matter of a strategic tuit shipment.

There are enough people using them on public production servers that the security problems are largely known and documented. If I had to put forward a guess as to where things went wrong it was probably with TWiki. It seems to have a tendency to be used behind corporate firewalls, and has a much smaller installed base than Linux or Apache.

Of course, it’s entirely likely that it was a more familiar demon (not daemon) at work — inexperience. That was perhaps my second TWiki installation, as opposed to the Linux or Apache installations I’ve done, which number in the thousands. It’s easy to imagine that I missed some important step in securing the installation. It may have even been one of those things that would have even a moderately-experienced TWiki-er rolling their eyes and saying “oh, you didn’t miss that huge, important step with all the blinking lights and police tape”.

That being said, I’m not too terribly interested in seeing how my third TWiki installation attempt would stand up to the world. I’m looking at some alternatives that I can vouch for. I trust my building blocks. I certainly don’t think that I can do large-scale software security better than the TWiki folks — indeed, I’d argue for their abilities over mine any day of the week. I don’t intend to do anything large-scale. I intend to do it as small and simple as I can get away with.

Someone who is a lot smarter me said something to the effect of there being two ways to build software. You can build complex software that has no obvious defects, or you can build simple software that obviously has no defects. I’m going to make a play at the latter. I’d sooner risk taking my lumps along the way and learn as I do it, rather than fiddle with a huge piece of software that I will use 15% of the capability of and don’t have the time to fully learn and understand.

So what’s next? Some software, obviously. But not before I bring back the blog entries that were on the wiki and make sure I can keep writing them. Where once there was TWiki there is now vi, and I can handle that. So first things first — the archives!

So what’s next? Re-installed CentOS 5.1, uninstalled pretty much everythingg that isn’t a dependency of something I use, and I won’t be bothering with a wiki any more. I’d sooner do it by hand that worrying about whether someone or another decides to try to pump their porn site.

As for how the intruder (if there was one) got in — no clue. I didn’t have a known-good set of tools handy, and I wasn’t in the mood to do a whole lot of forensic work that would likely turn up nothing. He (I’m assuming that women can figure out more useful things to do) was probably logged in from a similarly broken-into machine, and those trails don’t lead anywhere unless you’ve got the cooperation of pretty much the whole
world, and that doesn’t happen.

There are also too many unknowns for me to make too many guesses — could it be a violation of Xen paritioning? Could it be a bug in Xen itself that showed me something that wasn’t really there (in my instance, at least)? The connection was to one of my spare IP addresses, and those don’t fall into any sort of order that I can make out. Indeed, the address he was connected to is in a different class A than the rest of the IPs on my instance.

Oh well — there’s naught that can be done about it. I took SHA-256 sums of every file that matters and will be installing a little thing or two to keep an eye on the system for me. I’m also going to follow the KISS principle when I get more dynamic web stuff up. No big wikis (as much as I do like TWiki), and I’m probably going to have a bad case of “if it’s not Apache and not Perl and
NIH, then screw it” for a while.

My girlfriend asked what was going on and I explained it to her. I likened it to going to your cottage and finding out that someone had broken in and shit on your couch. In a lot of ways it’s true — my next job was to change the locks, clean up some stuff, take pictures of my tools to make sure they stay as I left them, and a bunch of other things that one shouldn’t have to do but has to do because there are still people out there who get a rush out of breaking into peoples’ cottages and shitting on the couch.

And that, my friend, is sad indeed.

The second is not having a dedicated Control key. The balance of the paragraph above applies to this, too.

The third is that I can’t seem to find out where to manage imported certificates. I imported the CAcert root certificate, but I still get a warning when I access my website. Now it just claims that I have asked to be warned. I want to convince it otherwise. Hopefully I’ll get it figured out.

I am using Set::Object a lot in some code I’m writing, and I can’t get Moose to work well with attributes of this type. Moose complains that instances of Set::Object don’t pass the type constraint _ANON_. I got it partially working using lots of absolutely horrendous code, but that’s not how I want to spend my time.

For now I just used Module::Starter and regular Perl OO to get me back on my feet and got the code that failed to work with Moose rewritten and working, with POD and tests. It wasn’t a lot of code and certainly wasn’t complicated.

Hopefully one day I’ll find the time to re-explore Moose. But until then, the Moose can go to pasture for me.

Today I got an e-mail message from Brian, my former boss and an all-around good guy. When I was moving from there I was single and his wife was expecting their second child. Today we chatted briefly about our families and he mentioned enjoying playing with their two-year-old. It is funny how life changes!

We also ordered a couple of books on-line. An etymological dictionary and a book on practical Gàidhlig will certainly help our learning along. There is a lot of material out there that’s geared at teaching one to communicate with speakers of a language; however, my interest is learning to communicate as a speaker of a language. There is a lot of idiom and other important details that are hard to learn, but I hope to learn them as early in the process as is practical.

Our company controller pulled a rabbit out of her hat and got our T4s ready for today. I picked up a copy of QuickTax and ran my numbers. With the information I’ve got right now I should see a return of CAD150 or so. Not a great amount — just enough for a night in Harrison with Robynne sometime soon.

Performance reviews are coming up at work. I know I don’t have anything to worry about, plus Gozer is a pretty straight shooter. I’d probably know about it by now if I did. I’m still conditioned to fear performance reviews, though. I’ll be relieved when they’re over.