It's less than 2 days until the cutoff point for the decision, and I am still not sure if I'll be attending DebConf this year.
I had been planning to give a talk, and had travel funding, but I'm not currently really interested in giving that particular talk. Partly because I haven't done some needed reseach, and probably won't in time. Partly due to the long gap between submitting a talk and DebConf, which offers plenty of time to get into a different mood and get more interested in other things, which it's then too late to submit a talk for.
So I canceled my talk, and I assume this means I canceled my travel funding. Combined with the dreadful time I had on my last transatlantic flight, and the much larger sum I'd have to pay to attend now, clicking on that "buy" button is much harder.
Nice article. My favorite quote:
"If I don't install Debian on this, it will feel like a brick to me anyway!"
Google F 9 11 to find a documentary whose title derives from a famous book about a government banning and burning books.
Google F9-11 to find a recently famous number now being banned and burned by the governWMPAA.
I'm up at the top of Roan Mtn this morning in solid fog, all hazy white after 20 feet. No chance of the views I came up here for, but hiking up from the pass was still worth it, in the calm and chill. Nice end to a quick camping trip.
Pete Nuttall rants about how online shopping sucks. I agree that it has problems, but the core ones are perhaps not the problems he identified.
Pete argues that while "in a real shop, you hand them their goods and they carry it away", fulfilling an order is much harder for online retailers. It is, but it can also be much easier for a shopper, especially if you want to buy something not available locally. But a real benefit of online shopping is that delivery happens asynchronously; that you don't have to hand the customer their product on their way out the door. This lets my sister find native plants on her property, put them up for sale, and only dig them up when she gets an order. It lets Cafe press and many smaller outfits create customised products on demand. It lets me shop at 2 am. Asynchronousity is good.
Pete thinks that needing to have a paypal account is a problem, comparing it to needing to open a new bank account to make a purchase. Of course, this ignores the fact that once you get the account, you won't want to buy from many online merchants who don't accept paypal. As someone who's worked on credit card processing backends for random online merchants, I know that I don't want to give my credit card account to a random merchant if I can avoid it. This, coupled with the ease of use of not needing to enter all my purchase information every time, means I'll chose a merchant who accepts paypal over one who doesn't, nearly every time.
The real problem is that both online shopping and payment promote centralisation. There's a pull toward centralising as much shopping as possible in one mega-site, because customers come to trust that site, and understand how it works. That's why there is only one amazon, and only one ebay (effectively). Handling payment is even more centralised, with just a few credit card companies and paypal.
And I think that the problem underlying that is that our model of payment is all wrong. It's based on money immediately changing hands, with an overlay of a promise of payment; it should instead be based simply on a binding promise of payment. A merchant expects to get money transfered to their account as soon as the credit card processing goes through. That money really comes from the credit card company, based on a customer's promise that they will pay back that credit. The customer indicates this promise in the most stupid way possible: By entering a sequence of numbers that only they are supposed to know, but that is not unique to that purchase.
As I've said before, it should be technically possible to kick out the credit card middlemen, and make payment much simpler and secure by changing the buyer's promise of payment into a legally binding, unique, cryprographically signed message that's sent directly to the merchant. But for this to happen, we'd have to change the model of payment, so that merchants don't expect to get money immediately, but instead asynchronously (the same as they ship things out asynchronously). Even harder, we'd have to somehow interest companies in supporting that. Since companies almost always prefer making money by promoting centralised systems, I don't have much hope of this happening anytime soon.
Cowboy Junkies
Bijou Theatre, Knoxville
Just set up my outdoor summer office. Pretty nice amenities for hacking in the big blue room: wifi, power, shade, views, even music. This has to be in the top 1% of offices..
(Course, it's just a canopy and some chairs in my side yard, really..)
Today is tilt at windmills day. Alternatively, for Debian developers, it's check your package's Recommends to make sure they're sane[1] day. I did both.
[1] The 'Recommends' field should list packages that would be found together with this one in all but unusual installations.
I'm finally completely unpacked, it only took half a year or so. :-) I've turned the previously scary room-o-boxes into my server closet, with everything nicely shelved. Even my Atari is unpacked!
A while back I set up a d-i test lab, the dilab, a kind of automated mechanical computer torture rack, which has been trying (and frequently failing) to automatically install Debian onto a half-dozen varied machines every day for a couple of years.
Problem is that hardware sucks, even when it's tucked away in a friend's basement and so not blaring at you with its combined total of 100 fans and 50 disk drives. It has an annoying habit of dying, especially if it's oddball machines for different architectures. Cords have a habit of getting unplugged. It's hard to automate booting from installation media. And so on. A while ago I decided to go virtual.
Now, thanks to Dann and Matt and HP, I've gone from this:
to this:
HP donated hydra, a maxed out 8-way Xeon with 4 (soon 8) disks. The idea is that it can run 7 or so virtual machines, all doing test installs, in parallel. Fast.
I'm using kvm for the amd64 and i386 tests. Since Debian has mips and mipsel kernels specially built for running under qemu, I'm able to test the installer for both those arches. Hercules will cover s390, once the kernel gets fixed. This leaves some architectures still todo:
- arm should be supported by qemu, but Debian doesn't yet have a kernel for it. Please, please kernel team, add an arm qemu flavour!
- sparc's qemu support seems very rough. I haven't tried to get it working yet.
- powerpc probably has an emulator I can use. Any suggestions? I need something that can run headless. Qemu would work, if we had an appropriate special kernel for it.
- m68k I have not researched emulators for yet.
- hppa, ia64, and alpha all lack a usable emulator, AFAIK. Luckily I have machines for all of these, so I can pare the physical hardware down to just those 3 boxes.
I also still have to write enough installation test cases to keep this monster machine firing on all 8 cores all day long. It can probably handle around 300-400 installation tests per day.
Since hydra has the disk space, and these emulators are generating lots of disk images with Debian pre-installed, I have also set it up to retain the disk images. I imagine that some of these disk images will eventually be useful for devleopment, since it's much easier to boot a disk image rather than sit through a slow qemu install on an arch like mips. I haven't worked out how to make these images publically available yet, but hydra is located on the name network as gluck.debian.org, if anyone wants to get access to them.
Update: I also have qemu doing a 'savevm 1' before shutting down, so you can even restore the running system from the disk images. Especially useful if the install failed somewhere in d-i! :-)
BTW, I can give out accounts on hydra to any Debian developers who need a fast machine to run kvm/qemu on, or to anyone who is interested in setting up their own installation tests.
John posted some old pics from ALS, and that prompted me to dig this out of the depths of my svn attic:
This chat in the hotel lobby, or one like it, which also involved some developers from one of the BSDs, and at least one real Debian old timer, (Mike Neuffer?), is where we first discussed adding build dependencies to packages in Debian. It's stuck in my mind as one of the more interesting, and ultimately productive, late-might hacker talks I've participated in.
I try to avoid doing multiple things at once on the computer, since I know that focusing on one thing is when I do my best work (see fighting computer-induced ADD). But I've noticed that I can do two things at once quite well if they're the right two. Two of them are:
- listening to music
- reading (but not writing, or much programming, unless the music isn't based on vocals)
That's pretty obvious, but I was suprised by this pair:
- listening to podcasts, interviews, etc
- command-line drudgery (upgrades, sysadmin work, etc)
This really suprised me, since I consider the unix command line to be another form of speech. This is making me think I don't really process it that way. Perhaps my mind really considers it a wholey visual/manual activity.
It gets iffy when the sysadmin spills over into minor programming, and once I get into programming where I have to hold a lot of info in my head, I lose track of the audio.
I set up my Atari 130XE today, for the first time in eight years. Eight years ago I got it working long enough to make several backups of all the program disks, which are all going on 20 years old. Unfortunatly all the disks with the tetris clone I wrote fail to boot. All the other programs are still surviving. Playing with the Atari is still fun!
I need to get a SIO cable and a SIO2PC and transfer these programs off and put them in subversion.
Update: Tetris is on those disks, they're just not bootable, but it can be loaded by hand. Then it fails at line 10380 since the high score files are missing. (This is not the well-polished tetris disk I remember..) Fixable:
10380 RETURN
Then it works. Yay!
The Atari 130XE was my first computer. Not the first computer I used, but close. I got it around 1986, when I was in 4th grade. My dad and I bought it at some store, probably Sears. I paid part of the money for it, and he paid part.
I think he planned to use it for word processing, since he also got the one piece of commercial software I ever had for it, AtariWriter Plus. We also picked up one game for it, some kind of space battle thing on a cartridge, with a special controller, and a huge, clunky trackball. That game never worked (maybe it was for some other model of Atari?), so I was stuck with playing with programming in BASIC. The Atari had a nice manual that made this very fun to learn.
At some point in here, I think in the first year I had it, disaster struck. I had a cup of OJ at the computer and it fell onto the keyboard. I didn't know that you could just toss a computer in a sink and take it out and blow-dry it and it'd be fine, so I did nothing. Until keys began to not work, and others to repeat over and over. Not sure if the OJ etched the circuit board, or just gummed up the connections, but the Atari was out of operation.
A bit later, I'm not sure how long, "I" wrote to someone at Atari's offices in Sunnyvale, CA, asking if I could get it fixed, and "I" sent it out there and miracuously they sent back a completely new Atari. I knew it was new, because it was much cleaner than the one I sent out. I'm not sure why I remember doing this, since it seems likely it was really my parents writing the letters and doing the shipping. I am still grateful to that unnamed person at Atari, who may have known at the time he was at the last days of the company's heyday, and sent that replacement to an already out of date and out of warantee machine.
The several years after I got it back were probably my most productive on the Atari. I wrote a lot of games and other programs. Often I'd see some program on another computer, like the original Tetris, or Lode Runner on a C64, and write up my own clone on the Atari. I did a lot of semi-original stuff too.
I was a pretty horrible programmer, by the way. I remember whaling at code over and over, pretty much at random, until I randomly found something that sort of worked. Well, hey, it was just BASIC.. At one point I found a book at the library about Atari game programming. It had one thing that blew me away; some weird ATASCII characters that if entered just right, in the middle of a BASIC program, would let you redefine the Atari's character set, so you could use characters for sprite based graphics. That was some precompiled assembly code, and while I used it, I had no idea how to change it, or write anything like it. By the time I learned what assembly was and started looking for an Atari Assembler, it was too late to buy one.
Anyway, the key thing was that, if I didn't write the software, I didn't have any software. Especially a few years later there was no chance of even buying any, and without a modem, I wasn't tied into any BBS scene. The Atari was its own little island, with its own little colony of unique programs growing up without fear of competition. That, and the lack of an assembler to get really low-level (though I did plenty of PEEKs and POKEs), defined my early programming time, and have probably defined where I went and what I did since.
I remember when this ended. My dad got a second computer, a 286 laptop, and I started getting ahold of programs from friends at school and putting them on it. At one point I got ahold of Battle Chess, and I realized that here was a game that I couldn't hope to replicate myself on the Atari. For the first time I compared my attempts on the Atari to the wider world, and found them wanting.
The last program I wrote for the Atari was in 1992, when I lent it to Mary Calhoun to use as a display in the waiting room of her office and wrote a program to handle that. This is also the only program that has my name on it, and even a copyright notice. By then I had moved on and was clearly trying to do programs that were more like the stuff I was seeing on the PC.
I dug up an old radio and threw this together tonight. It's been done before, but the scripts out there seemed overly complex and didn't use alsa. Here's how I did it:
- radiorecord is a shell script to record a given number of minutes from the radio to an ogg file.
- crontab to schedule when to record shows
Of course, this assumes that the station doesn't move shows around and that the times are in sync. To do this really right, I'd need a radio programming data source, and some post-processing to identify exactly where the gaps are between programs.
I was out at the farm this evening and wow, Megan and her bf have been busy. The old studios, seen here dimly on a foggy winter day, are down.
Lots of other new activity; beehives, power preparations, garden, etc.
I'm happy that this change, unlike some other change, doesn't bother me one bit. Though the buildings did add a certian something to that first view of the farm, I won't especially miss them. It's great to see stuff being done.
PS, if it doesn't rain soon, I shall go insane. The driveway was all dead brown grass.
It will be interesting to see what new forms of social engineering will come about thanks to the neat new street view feature. To be able to check, from across the world, what car someone had in their driveway (or its license plate), or what a sign on a shopfront says, is sure to be good for something criminal.
The better google maps gets, the more places that don't have all the awesome data become in a sense second class citizens. With the exception of bandwidth, the internet has so far tended to flatten these differences out, not accentuate them. So this is also interesting, though it kinda sucks to be there (google can't even resolve my house on their satellite maps).