Gnome 4.23...

If the GNOME people keep going on the track they are currently on...

... This is where they'll end up:

Grmbl.

I'm angry with Kanaal 2. They've broadcasted Star Trek series since they were founded approximately a decade ago, which was nice; not that I'm a fanatic who would never miss an episode, but I do appreciate a bit of science fiction now and then—and in my opinion, Star Trek is some nice SF. Well—most of it, anyway.

This year, they've started the first season of Star Trek: Enterprise as well... at 8 AM, on saturday morning.

Morons.

MythTV

Several people replied to my last blog post about a silly TV station scheduling ST: Enterprise at 8 AM, pointing me to MythTV. Well, point taken. However, to do that, you need a specific piece of hardware—one which I don't have. And a cable to my computer, which, I presume, isn't going to happen very soonishly.

The alternative would be to use the VCR to videotape it. Not as good a quality, but at least it should work, right?

Try again. I did try, but the VCR ate the fscking video tape.

Maybe I shouldn't be angry with Kanaal 2 and be angry with the fact that hardware will always let you down, but that's just too easy.

Scientific studies on Open Source

It's strange, but ever since I have been a Debian Developer, people start assuming I'll be interested in any random study. They'll put up a web page (which often doesn't even validate), and then start asking strange questions that have no link with reality whatsoever.

On the one hand, I have better things to do than to cooperate with random studies from people I've never heard of before and probably never will hear of again. On the other hand, only by understanding will people see the light, and only by cooperating with such studies will one help people understand. But seeing the same mistakes over and over again is annoying me. So, here's Wouter's HOWTO perform a study on Open Source or Free Software (and get away with it without annoying people)

• Make sure you tell people up-front how much time your questionnaire will take. It's annoying as hell to see that after pressing 'submit', there's a second page which is at least as long as the previous one. And a third. And a fourth. Especially if you didn't plan for that, and have to leave soonishly.

The best way to do this is to either have all questions on one giant page, or (if that makes it too long) test-drive your questionnaire with a chronometer nearby.

• If your questionnaire contains multiple pages, don't break the back button. One questionnaire I participated in did this, resulting in me going back to change the answer to a question, and then suddenly finding that I couldn't go to the final page anymore, because the system thought I was submitting that page.
• If you have multiple choice questions, make sure the answers make sense. Having a question like Where do you use Linux? with possible answers at home, at work and at school is nice, but not if you can not choose more than one of them at the same time.
• Additionally for those of you doing multiple choice: make sure you have at least one open question, so that people can leave feedback about things they consider you left out. You're free to ignore that feedback in the final results of your study, but it may help you in better understanding the FLOSS movement.
• If English is not your native language, find someone who is very fluent in English and have them check it out. This is important because if your question is unclear, I might give you the wrong answer—which will make the results of your study rather worthless.
• Test your questionnaire on someone else before you ask hordes and hordes of FLOSS developers to fill it out, especially if you're not too familiar with the FLOSS movement yourself. FLOSS is very popular these days, it must be possible to find someone in your vicinity who's working with it, or who at least has a feel for how the community works. Their feedback may help in avoiding questions that are based on incorrect assumptions, and/or get out little things you didn't see yourself first.
• Do not mail people in private, but instead ask people in public about filling out the questionnaire. Remember the "Open" in "Open Source".
• Finally, be sure to explain what the study is trying to discover, what is going to happen with its findings, and where I can read the results when they're there. Not that I want to see the results of any possible study I participate in, but one never knows—I might be interested.

That's the most important bits, I guess. I'll probably add more in the future, if I see more things that annoy me. But for the time being, if you break three or more of the above rules, don't expect me to participate—I have better things to do with my time than to help every undergraduate on this planet with his or her study

AOL

Since everyone else seems to give their opinion on DPL candidates over at Planet Debian, I'll join the chorus. Since I've met and know most of the candidates, I'll focus on what cannot be in their platform: their personality.

• Jeroen Van Wolffelaar: I've met him in person a few times (at FOSDEM, and at DebConf5 in Helsinki), and his dedication to and passion about Debian is... wonderful. His experience with being in the DPL team last year and his infrastructural involvement in the FTP team should mean that he's quite up to date on Debian's big picture.

That being said, I'm not entirely sure he'd be the perfect DPL candidate. All his passion and dedication aside, he's rather new(ish) to the project (having received his Debian account only one or two years ago, IIRC), and might not have the experience one would expect from someone who's been with Debian for a longer time. That's not a fatal flaw, but it does mean I'm a bit... reserved as to his candidacy. I'll still rank him pretty high, but not as number one.

• Ari Pollak: ha. ha. hah.
• Steve McIntyre: Also someone I know from both previous FOSDEMs and DebConf5. The impression I have from Steve is that he's someone who has a very (very) clear view on how Free Software in general, and the Debian project specifically, works. He's a very nice guy, easy to work with in the encounters I've had with him, and has the sort of calm but focussed attitude you'd expect from someone who's done a lot of things in his time, and therefore isn't all that easily impressed by new but short-lived manies anymore. Which, I think, is a virtue for a leader; a leader should be focused on the long-term benefits, not on the short-term fun. I expect him to be able to disconnect himself from immediate issues, instead having a look at the big picture when required.

I think the message is clear: Steve is going to get my #1 vote.

• Anthony Towns: a difficult one to form an opinion about. Aj has been a long-time valuable and dedicated contributor to the project; for that, he deserves (and receives) my respect and praise. He can be expected to know the project inside out, which is something very helpful for a DPL.

However, I don't think it's a secret that aj and I disagree on certain issues of the social aspects of the Debian project. Since his platform specifically mentions those things, I don't think I'll be ranking him very high on my list. I won't object to him becoming DPL, however, and I'll still be happy should he be our DPL. Just not as much as if it were Steve.

Additionally, Anthony has a bit of a tendency to be imposing his view on others, which I think is a bad quality for a DPL. He's aware of that, however, and claimed he'll try to avoid making that mistake, which is helpful.

Oh, and what's with the singing?

• Andreas Schuldei: When Andreas mentions in his platform that he's got great leadership capabilities, he isn't lying, as everyone looking at how DebConf is organized can see. However, the few times that I met him in person, there was something subconsciously that now makes me think twice to vote for him. I don't know what it is. I'll still rank him pretty high, but not as number 1.
• Jonathan Walther: This guy needs a reality check. No, really.
• Bill Allombert: Also someone whom I've met a number of times at FOSDEM. Bill is a bit of a modest, timid guy with a passion for Debian and Free Software. Though I respect him as a person and for the work he does in Debian, I don't think he has the perfect personality to become a DPL. No offense, Bill; but you'll end up rather low on my list.

"plain" CSS

If you've been reading my blog through my website, rather than through any of the places where I'm syndicated, You'll have noticed that I've recently created a new CSS stylesheet (which I've called "plain", even though it isn't empty) and made it the default. The reason for this change is that my previous default, the "hackergochi" CSS, was a bit of an angry fruit salad, and also required much of the browser to render it properly (firefox couldn't scroll it very smoothly, for example). This one doesn't change much colors (though it does change a few of them), and apart from placing the menu and having it stand out, there isn't much that it changes.

This will make my website work more efficiently, though not necessarily faster. The new CSS also doesn't waste as much screen real estate, which I think is a plus.

Only it doesn't look very good. But then, I'm not a graphical guy.

Ingram's website...

Silly people.

We don't support Safari; but if you use this little tool that allows you to change your user-agent and claim you're actually using Netscape or IE, it might work

Anyone else who thinks that's incredibly silly?

SF.net finally has SVN!

I don't believe it. It finally happened: SourceForge finally offers the ability to get subversion repositories.

The NBD tools are in the queue for migration right now... after I did some commits to the CVS repo. Oh well.

Employment law

Clint,

Please get some way to contact you up on your site. Having to reply through Planet isn't ideal.

Anyway.

I'm not familiar with employment law across all of Europe, but the claim that you have to employ them for a certain period after you actually fired them, or even that you can't fire someone without due process, is simply false—at least in Belgium, though I expect the same is true for most other member states of the European union.

First, there's the fact that every employment contract starts off with a "test period" which can take up to six months, depending on the type of employment (it is longer for hotshot management than it is for hand labour) and the length of the contract, if appropriate. During this period, an employer can fire someone without further explanation and without any process. This is to make sure that you can get rid of people who managed to sell skills on themselves they don't end up having.

Second, an employer can fire anyone right away for what is called 'urgent reasons'—things like employees stealing from the employer; basically anything that makes the employer lose trust in the employee (though this is strictly regulated, for obvious reasons).

Third, if you fire someone not during their testing period and not for urgent reasons, you do not have to keep them at your place; it is legal to tell someone that as of tomorrow, they no longer have to come to work -- you just have to give them a pay check for those three months (or whatever) you would have to employ them otherwise.

Finally, the thing also works in the other direction; if some important employee (say, the only guy with the root password) decides to quit, an employer can require that they stay at work for up to (IIRC) two weeks, so as to ensure continuity. But I understand you knew that part already...

/usr/local/sbin/mycpufreqd

#!/bin/bash

modprobe cpufreq_userspace
echo userspace > /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor
while true
do
	if expr $(uptime|cut -d',' -f3|cut -d':' -f2) '>' 0.75 > /dev/null
	then
		echo 1333333 > /sys/devices/system/cpu/cpu0/cpufreq/scaling_setspeed
	else
		if expr $(uptime | cut -d',' -f4) '<' 0.75 > /dev/null
		then
			echo 666666 > /sys/devices/system/cpu/cpu0/cpufreq/scaling_setspeed
		fi
	fi
	sleep 10
done

I tried using some of the standard cpufreq stuff that's in the archive, but none of them seem to work for me; they all make my laptop work at full speed, all the time. Which is silly.

Also, the kernel ondemand and conservative governors don't seem to work—when I write "ondemand" to /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor, it comes back with write error: invalid argument, or some such.

So, I hacked up the above script. It sets the speed at full (1.3Ghz powerpc here) when the load rises above 0.75 for the last minute, and sets it at the low speed again when the load drops below 0.75 for both one and five minutes.

Using the system load average as a metric to decide what speed you run your processor at isn't terribly state of the art, but it Works For Me(tm); my processor isn't put to full speed when some short cpu-intensive job comes up for three seconds, but if something decides to use it for a bit longer, it does turn it up after approximately 30 seconds. Once it reaches that, it stays there until (on average) a few minutes after the cpu-intensive job has finished.

I like it that way.

Update: I received no less than four replies to this post that I need to make sure that cpufreq_ondemand is loaded, which I can check by reading /sys/devices/system/cpu/cpu0/cpufreq/scaling_available_governors. Yes, I know. I did that. It's there. Yet, it still doesn't work, not even with echo -n.

Which is why I'm so confused. It's probably just a kernel bug, though.

Explaining Debian

Josh, I don't find such situations at all awkward. If people see me running around with a Debian T-Shirt and ask what that is all about, my answer is just Debian is a free operating system for your computer that is made by a couple of thousand volunteers worldwide, of which I am one. That is enough to quickly explain the thing; if they're further interested, they'l ask.

Occasionally, people wonder what an Operating System is. I answer that with You have a computer? Does it run Windows? Well, then Windows is your Operating System. If you would want that, you could replace it with Debian.

The point in such conversations is to make it clear how Debian is relevant to the people you're talking to. Don't start by saying stuff like Debian is a Linux distribution made by volunteers; because then you oblige yourself to explaining what a distribution is, and what Linux has to do with everything.

And about that Linux bit, I wouldn't worry too much. If the people you're talking to are business people, and they've never heard of Linux before, just get yourself a strange look on your face, and say something like you've never heard of Linux before?!? because frankly, if a business person in these days doesn't even have a vague clue as to what Linux is, they're not worth half their pay check.

Kerberos

I had nothing important to do this weekend, so I played a bit with kerberos. Pretty fun, that, and the Debian packages make it all so easy.

First, when setting up Kerberos, you should note that there are two competing implementations: one is heimdal, the other is the MIT reference implementation. The former has some additional features (like storing principals in LDAP rather than plain files), but I found that krb5 was useful enough to me, so I didn't investigate further. Also, storing principals in LDAP isn't as useful as it sounds—kerberos is designed to make password replication unnecessary, so you want to store it on a machine with little or (preferably) no other jobs to do. You certainly do not want to store it in a server which will put it out to the world for all to see.

Anyway.

It's pretty well documented in info krb5-install (in the krb5-doc package); and the Debian packages let you do most of the setup through maintainer scripts and debconf questions, except for creating a new realm, which you do through a provided script.

... and this means that setting up kerberos on Debian is peanuts. Really.

You need (at least) one KDC, which you do by installing the krb5-kdc package. For exactly one KDC (which will be your master KDC) you also need the admin daemon; you can get that one by installing krb5-admin-server. Then on the hosts in your network, you want to install krb5-utils, and libpam-krb5.so for authentication. After configuring them (as explained in the krb5-install document), you have a live kerberos system. It takes you about five minutes to do the initial setup (once you understand how the system works), and about 10 seconds to add a host to the system.

Of course, doing kerberos setup is useless if all you want is to be able to log on to one system. But if you have more systems, you might want to use, e.g., the Kerberos support that is in OpenSSH, which you can do by adding the following two lines to both /etc/ssh/sshd_config and /etc/ssh/ssh_config (though obviously not on the same hosts:

KerberosAuthentication yes
GSSAPIAuthentication yes

... and with that, you'll have a kerberized setup.

Fun.

Ultra 10

Uwe is having issues with booting Debian GNU/Linux on a Sun Ultra10.

Uwe, if your box has a Creator3D card, it might be that the Linux framebuffer is, by default, sending its output there. If you have no monitor, it stops when the screen turns black.

There's a kernel boot option to switch the default framebuffer from the Creator3D to the ATI something-64 (can't remember the name; the atyfb framebuffer is what drives it), so that you'll see something again; apart from that, it should work.

Interestingly, the 2.4 kernel works differently; if you boot with linux-2.4 instead of linux, you'll get a working installation.

rpmstrap

Andrew Pollock reports that there is (finally) an rpmstrap.

Finally, indeed.

I have set up RedHat chroots in the past, by manually downloading some file, manually filtering out the list of packages that are relevant for my architecture, manually downloading them, and manually running rpm --root, or whatever they use to set the root directory; but it wasn't pretty. Being able to say here's a directory, there's a mirror, I want this, go do it for me would be...great.

Hmm

I read quite some stuff through liferea: Planet Debian, Planet Grep, commit logs on the d-i SVN repository, and more. One of the things that fits in the "more" category is my ThinkGeek wishlist (hey, they provide an RSS feed, so why not? ;-). Which, obviously, does not often see any updates; I don't surf the TG website all that often.

Except that today, it did. And the best part is: I didn't even ask for it! Suddenly my thinkgeek wishlist contained some shirt that I did not ask for (and never will, since it's ugly). Strange...

Dear lazyweb,

I'm looking for a calendaring application with the following features:

1. Must have a (limited) command line interface. I want to be able to view my appointments of the day in plain text format, so that I can put that in my .profile; it'd be nice if it allows me to view my appointments on any random date, but that isn't required. Moreover, when someone sends me a meeting invitation, I want to be able to pipe the ical file into something which puts it in my calendar, with optionally also sending an email to the person who sent me the invitation that I'm accepting it (optionally, as in, it must have the feature, and it should allow me to decide whether to send the mail on a case-by-case basis, not as in it would be nice if the feature existed.)
   More command line features are nice, but not necessary.
2. Must have the ability to export ical files to a text file on my local hard disk; must not require me to set up a server with some weird login scheme and ditto protocol so that I can please put my webcalendar somewhere—and no, I won't sign up for one of those "free" webcalendar services. I'll put it online myself, thank you very much.
   This requirement rules out evolution, unless I missed some plugin somewhere (I don't think so).
3. Must have basic calendaring functionality. I.e., I should somehow be able to get an overview of my appointments of any given day, to get a weekly, monthly or yearly overview, to print out things, etc.
4. Must not require ages to start up. This, definately, rules out evolution. When I have a phone call, I want to be able to see within a few seconds whether I'm available at any given time.
5. The ability to synchronize my calendar with my cellphone somehow would be nice, but isn't absolutely required. I'm willing to jump through a number of hoops here. Note that gnokii supports my cellphone.

Suggestions are welcome.

ColdFire arrived!

The coldfire development kit of which I made mention in a somewhat recent -devel-announce post has arrived about a week ago. Mine was the last one to arrive; the other four have already arrived. It took a while, because mine apparently got lost in shipping, so they had to ship me a second one. That, and the fact that I said I'd ask around in my neighbourhood whether it'd been delivered at the wrong address (but then eventually forgot to tell the Freescale folks that I had done so and that, no, it wasn't) meant it took a bit over two months to arrive here.

But it finally has. Whee!

The box contains a number of things I do not and will not likely ever need (the CodeWarrior Development Studio, for example, or power cables for other places of the world, like the US, the UK, and Japan), but, well—that's not important. It also contains some things that are interesting, such as a serial cable—even if I already have a few of those.

The box contents also smells funny. Oh well.

In the mean time, however, I've come across a bit of a problem: I can't seem to boot the thing. Well—I can, except it seems to be mangling the serial output it's sending for some reason. I can see that some output is being sent to the serial console, but it's not readable. And I can also see that it doesn't find it's root file system—I'm doing NFS root currently, and there's no network activity.

Not sure what the problem is, though.

Update: someone suggested that it is the baud rate. Yes, obviously it's something like that; however, I did try every baud rate between 9600 and 115200, but none of them worked. Besides, I have my command line say "console=ttyS0,19200n8", and it's still not working. I suspect a driver problem at this point...

Now Reading: the UNIX-HATERS Handbook

... which is a free download, somewhere. No, I'm not a unix-hater. But it's fun to see unix being torn down like it is.

Sun Microsystems recently announced that it was joining with NeXT to promulgate OpenStep, a new standard for object-oriented user interfaces. To achieve this openness, Sun would will wrap C++ and DOE around Objective-C and NEXTSTEP. Can't decide which standard you want to follow? No problem: now you can follow them all.

Hope you don't have to get any work done in the meantime.

Mooha.

Why names are not very useful as a unique identifier.

No, I don't play volley.

Usability on the command line.

  -g, --generate            Unpack, but do not generate a new package.

Hmm.

Quickstep stable again

Quickstep.nixsys.be, the m68k/experimental buildd, was having issues the last few months, in that it intermittently dropped off the net. I couldn't quite figure out what it was, but seem to have found the problem now: I replaced the AAUI-to-RJ45 connector yesterday, and that seems to have fixed it. Did I say that hardware sucks, yet?

Anyway, quickstep is now happily building away at packages in experimental again. The current package happens to be gcc-4.1 (as I force-fed that to it), so it'll take a while; but apart from that, everything should be fine.

While I was at it, I changed the setup on jazz to point to experimental rather than unstable, too. Jazz is a 33Mhz machine (slightly faster than quickstep's 25Mhz), but only has 64M of RAM (which is a fair bit less than quickstep's 132M—don't ask). It also doesn't have much disk space (4G). I had set it up a while back when the backlog was huge and the need high, but its SSH key was never added to buildd.debian.org's config; as a result, it was just humming there, not doing anything useful. Since the experimental backlog is now larger than the unstable one, I just gave up waiting—it'll be helpful building stuff for experimental, too.