A shift in direction...

For years Jeff has been my little hole in the Great Firewall that allowed me to download pornography access the web unhindered while living in China. Basically the Great Firewall is a joke that only stops lazy people and stupid people (neither of whom you really want on the Internet anyway, so you could view it as a public service). Jeff, very kindly, kept a server in his basement hooked up that allowed me to redirect all requests for web pages that were deemed a danger to the state here through a Canadian server that allowed such things.

About a month ago this server's connection went flaky and died. Jeff, being newly married and kind of in a complex part of life, didn't have the time to check it out. I didn't mind, though, because very few sites I really cared about got blocked. That changed this week as Blogger turned out to be a threat to the Chinese government. It became imperative that the problem get solved and, for some reason, Jeff was incommunicado.

I decided that it was really unfair to have Jeff be responsible for my free (as in freedom, not beer) Internet access and embarked on a project to change this. So as of today I have my own tiny, cheap VPS in the USA that runs my little backdoor to the rest of the Internet; the stuff the Chinese government thinks is too dangerous to be seen. Like my blog here. The one I'm posting. Telling you what a bunch of utter shitheads the Chinese government is for being afraid of my little key-clickings telling you harmless, inoffensive things about China (for the most part). Apparently I am a danger to the state. Funny, I don't feel any different from last week this time....

Mr. Hu Jintao? I want the six hours I spent debugging this setup back. Please mail it to me you frightened little child.

The Many Faces of Lucas

I've been pestered enough, now, for pictures, pictures and more pictures. I've decided to do something about it. First I registered with a dynamic DNS outfit (if you don't know what that means, don't worry). Then I rigged up my laptop to be a server behind the firewall. Then I did a quick hack and put all of Lucas' pictures to date up on that server. Right now it ain't pretty, but it works. I'll work on pretty (and on videos functioning) later.

Now the pictures are VERY large. They're 3264 x 2448 pixels (which is just over 3.5MB each picture). As a result I do not intend to display the full-sized pictures on that web page. What I will do is take requests. If you like a picture, make note of its name (you'll see it to the right next to the word "Image" -- something like " p1000225.jpg") and send me an e-mail. I will, on a periodic basis, collect together all the picture requests and upload the full-sized images to a file-hosting site that allows larger files (but is a pain in the ass to use) and email back the access information. This is the way things are going to have to stand until I can get myself a proper Virtual Private Server host. (If anybody feels like donating US$20/month, to this end, drop by http://www.slicehost.com/ and set me up with a "256slice"—or better if you like!—running "Ubuntu Gutsy (7.10)" and I'll get to work on that right away.)

Until then, you'll have to make do going to http://ttmrichter.dyndns.info/lucas and looking at the smaller pictures and requesting the large versions if you really want them. Keep in mind that the pictures in question are hosted on my laptop in my house. If you can't get to it, try again later. Any one of a billion things might be wrong -- including my laptop being turned off or disconnected because I'm using it elsewhere.

Living the Life of Cassandra

I've always felt some affinity for Cassandra of Greek fame. (For those not up on the classics, she was given the gift of prescience by Apollo, who was smitten with her, spurned Apollo's advances and was then cursed to have her accurate predictions of the future never believed.) I have a variant of her curse, you see. I see something. I have a very good idea of where it's going to go. I tell people. I'm not believed. It comes true.

Lather. Rinse. Repeat.

I had this problem in my places of work, for example: a case in point being the company Entrust. The code name for one version of the software that was being pushed was "Project Banff". It was late and by all estimates was going to be slipping even further behind. Management came up with a "brilliant scheme" to provide "incentive" for getting it out on time.

"On time" for them was, if memory serves, the end of August of that year. Realistic estimates for completion put the real delivery date around October. This was unacceptable, so one of the VPs—the development VP—came up with this brilliant incentive scam scheme: if the product is shipped by August, the company would take all the developers and all the testers out for an all-expenses-paid trip to Banff for a week (or maybe two? -- memory fails from so long ago).

Now I was in an unusual (or, as it turns out, not so unusual) position in Entrust. I was a lowly software developer. Further, I had absolutely zero ambitions for a management role. Yet I think I may have been the only person in the whole company who'd actually taken, you know, honest-to-goodness management courses. Further, I was one of five people I knew—none of us managers (sorry, Jeff – you had no budget, so you were a supervisor, not a manager)—who actually read... well, anything, really, but especially books and articles on management and motivation. And what I knew from my training and my reading (as did anybody else in the vanishingly small group of us who knew anything on the subject) was that performance bonuses tied to a timetable failed. Always. 100% of the time.

You see, the management thinking behind such bogus schemes is that workers are malingering and grossly overestimating the time required to do tasks. If they are given an incentive, they'll stop goldbricking, put their noses to the grindstone blah blah blah blah blah blah. But this is, not to put too fine a point on it, total bullshit. At least in high tech it tends to be total bullshit. (It may be in other fields too, but I'm not in a position to knowledgeably discuss such.) In reality, in high tech the workers tend to be strongly self-motivated and, if anything, are too optimistic in their estimates. A fairly popular agile development process (or unprocess) has, in fact, as one of its techniques a way of calculating just how overly optimistic developers tend to be in their estimates and using these calculations to get a better approximation of the real amount of time required. So when you have an estimate for delivery in October, one thing is 99% certain: the absolute earliest that it will be delivered is in October. Further, any attempt to squeeze it out earlier without reducing the features to be delivered will have the opposite of the intended goal. You will delay final delivery.

I pointed this out to the VP in question. (Stupid me: I believed him when he said he was interested in employee feedback!) I further made the prediction that the actual delivery date, if this incentive plan wasn't unhooked from delivery date, wouldn't be August nor even October. I said that the delivery date would be more like March of next year. I was, of course, not believed. Because the VP in question, based on his almost months of experience in upper management, believed firmly that he could mutate reality just by wishing it so.

Fast forward to August. The release is a disaster. Only a small number of groups had actually delivered their stuff by the due date (the toolkit group I was part of being one of them – Jeff's influence here, and true to the "no good deed goes unpunished" adage he was viewed with suspicion for this). The product is nowhere near ready. The delivery slips past August. Past September. Past October. Slips all the way to April of next year (proving my point that developers are overly optimistic, seeing as I had predicted March). Because exactly what I predicted happened: an initial push of hard work started. Then people noticed that, despite putting in 12-hour days (testers especially), no real extra progress was being made. In under two months the whole company realised that the Banff incentive wasn't going to happen. In that time the developers and, more so, the testers had burned themselves out completely. Despair set in, followed by ennui. Developers didn't care any more, so the product slipped further and further and further behind. When it was finally shipped in April, this was with features scaled down on top of everything else.

So how was my foresight rewarded? With a rueful "I guess you were right", right? Wrong. My foresight was rewarded with an accusation that I had personally seen to the project's utter, complete failure. I was specifically named by the VP as one of the reasons for the failure. I guess my negative vibes (which didn't actually impact the productivity of the team I actually worked with, oddly enough, seeing as that team was one of the very, very few who delivered everything on time for the Banff trip...) were transmitted to the company as a whole—even people I had never met—and caused the project to die. Or something.

So why am I regaling the world with this tale now seven (or is it eight?) years later? Because this is only an example of what hits me every damned day of my life, practically. Joan giving me another perfect example of this.

On Friday Joan's laptop (my old Sony) starts acting up. The "L" key doesn't work at all and the "Backspace" key is flaky. I tell her, very candidly, that the computer needs to be repaired; that we should take it in on the weekend. Joan, of course, doesn't believe the only person in the household who knows anything about computers. Besides, she wasn't going to need the computer for anything in the near future anyway.

Well, the distant future of three days later, she's got a big task to do that needs the computer. (This being China she's given the big task with under 24 hours of notice before it's due, of course.) And not only is the "L" key not working now, nor is the Backspace working at all. Nor the delete. Nor the right arrow. Nor the shift key. Nor ... You get the picture. Too bad nobody warned her at all about having to get it fixed, eh?

This is my life, almost every day. See why I feel for Cassandra so much?

Printer Shafting

So, I'm getting a lot of good use out of my new HP printer (despite a few problems with the Linux driver that have yet to be fully worked out -- luckily I'm not printing photos!). Joan, when buying the printer, was already eyeing the price of ink nervously (the printer cost us 300RMB; one spare set of ink cartridges cost us 290) and got even more nervous when she noticed that I blew through the demo cartridge that came with the printer (3ml of black ink instead of 10ml, for example) with my first print job.

She was right to be nervous.

Printing at any kind of readable resolution sucks through ink at a prodigious rate. The printer manufacturers don't make their money from printers, you see. They make their money selling the ink cartridges. I have, since getting the printer, printed off five books on various things needed for my work (reference manuals in the main). Two books (of about a hundred pages each) is all I get per cartridge.

Luckily I live in China. In my neighbourhood you can hardly fling a brick without hitting at least three shops selling printer ink. And not just official cartridges, but, too, third-party cartridges (at half to a third of the price) and cartridge refill kits. These latter are the real life-savers.

Today I bought some ink refills. These are 30ml syringes (the black cartridges are 10ml, recall; colour cartridges are 8ml each colour) with the ink you need in the colours you need. Using them is simple: you peel back a sticker, insert the syringe, push the plunger and when the ink seeps a bit out of the hole you're using you're done. And they cost, literally, a tenth the price of the cartridge.

Going with the black cartridge (the ink I'll be using most often), that means that for one tenth the price of an official cartridge I get three times the ink. And refilling a cartridge is hardly difficult work! Fumbling with the packaging and tape of a proper cartridge means replacing a cartridge takes two to three minutes. Injecting the ink takes five. Hardly an onerous task when you consider that my print batches take hours.

Now sure, the ink quality isn't quite as good as the HP official inks. The black isn't quite so deep. The cyan/magenta/yellow isn't quite so vibrant. But it's still better than the official inks I used in my old Epson before it gave up its ghost and certainly more than good enough for the kind of printing I do (text).

So why would I want to be given the shaft by HP for its cartridges?

Well, I do blow my warranty away if they catch that I used an unofficial ink. On the other hand, if I refill my black ink cartridge three times, I've saved more than the price of a whole new printer....

Windows Security And Other Oxymorons

So, last time I installed Windows on Joan's laptop (my old Sony), I made a mistake. I installed Windows.

OK. More seriously, I made the mistake of installing anti-virus software after connecting to the network. Long before protection was in place her system had viruses up and running which could not be cleaned out with any anti-virus application. Still, the system was usable and there's no other Windows systems on the network for her to infect, so we left things lying.

Lying, that is, until her system was slowed down so much under the assault of viruses and adware that just minimising a window would take longer than 30 seconds.

So today it was "back up all your data so I can reinstall all the software in the known universe" day. This time, however, I did the smart thing and installed Windows, installed an anti-virus package, then installed the network. If this doesn't work, I'm going to tell Joan that she's got no choice. It's time to switch to Linux.