I’ve
written twice before about Charles “Cully” Stimson, the Pentagon official who urged corporate clients to boycott law firms representing Guant namo detainees pro bono. The latest,
via the legal tabloid abovethelaw.com, is that Cully has resigned over his remarks, a fitting end to the episode.
One of the things about the GNU/Linux system that I wouldn t want to miss anymore is the general flexibility: I had to send my main machine, a ThinkPad, for warranty repair (again...). But I wanted to continue to use my programs with all my settings and all my data. So I took my old desktop machine, upgraded the Debian to unstable (which is what my laptop was running), using dpkg --get-selections and dpkg --set-selections I made sure that it had all programs installed that were on my laptop and then I mounted the laptop s harddrive, which I put in a USB case, on /home. Et voil , I can continue to work just as before. I wonder if this would have been as easy with, say, Microsoft Windows.
It seems that I managed to send out bounces for some mails waiting for me
in my
TODO
folder... don't know how I did that. All those messages
where delivered succesfully (and most of them quite some time ago) and
where stored in a local Maildir-folder.
I don't know what went wrong... well... I was playing with
mbsync, but I was still in the
read the fine manual
state... so I still don't know what was going
on.
So: Sorry for the noise; as soon as I find out what happend, I try to
not do it again ;)
These are the current translation ratios for French and Czech in unstable, after 17 days of "blitz l10n NMU".
As expected, we are going down because of an uncoordinated upload of
viewvc. I even went a little bit more annoyed because the maintainer
did some changes to the debconf templates, updated *his* Spanish
translation...but didn't care to warn other translators. Next time,
I'll use my rubber cluebat, damog..:-)
We will be hit again becacuse the sing package was re-uploaded by its
maintainer iummediately after I NMU'ed it. Unfortunately, he
did then correct some typographical errors in the templates
and then fuzzied 9 formerly complete translations. Sad.
(update as of 09:32 UTC: this will be corrected by the maintainer)
Yesterday, I continued working on
Thomas Huriaux dynamic status page:
- bindgraph to be uploaded today
- openssl097 due out Feb 6th
- prc-tools due out Feb 6th
- ipsec-tools due out Feb 7th
- libpam-ldap dues out Feb 7th (with a call for updates running)
- php4-rrdtool due out Feb 7th
- glide still waiting for news
- ifhp needs the removal of debian/po
After
last post about a FHS amendment to address the structure of user’s home folders, I received a lot of comments and there is one very significant thing that can be changed in the proposal - instead of having $HOME/ .data .cache .config /appname structure, to change that to a mandatory $HOME/.library/appname/ cache config … . This version still has all the benefits of the first solution (configuration for an application can be easily identified and erase, and all cache can easily be excluded from backups using “$HOME/.library/*/cache” regexp) and also has additional benefits, main of which is the ability to later introduce the concept of user installed packages. The idea is that it would be possible to support having /bin, /lib and /share subdirectories in these application directories thus making an ability for the whole application to e packed in a single directory and allowing the application to be installed simply by unpacking this directory. I admit that much of this is glanced over from MacOS X world, but I do not think that it diminishes the idea itself.
Some problems appear there - support of these distributed bin folders, support of separate lib folders, handling of application plugins, handling of dependencies, handling of the application menu, upgrading notifications for the user software vs. system software. But nothing there that can not be solved. I feel that this can bring together FHS and LSB by providing something of an API for software being installed by users. Having no registry of the software in this solution allows for some interesting things, for example having multiple versions of one program just by renaming the application folder.
A lot of specification work is required here, therefore I proposed a workshop on this topic in Debconf7. I hope to have something that everyone can agree on and maybe even some code by then, so that after Debconf7 there can be a formal policy amendment proposal.
So there are two questions:
- Is there something better in $HOME/.cache/appname vs. $HOME/.library/appname/cache ? Any other trade-offs that I have forgot?
- What do you think about the further idea of users having the ability to install software in their home folders or remove such software in a very simple fashion?
On another ‘wild’ friday night I accidently spilled beer all over my X40. Panic. I SMS’ed half my phonebook in panic asking for advice.
After opening it and finding everything ontouched by beer, I noticed a nifty thing about ThinkPads I have never noticed before. Apparently they all have a spill resistant keyboard where the liquid gathers and drains through it, not touching the vital fine electronics inside. Probably old news to some, but new to me

.
Relieved to see that the only thing not working on my laptop so far is pageup/pagedown. They will randomly switch between working and not working, but switching keyboards is trivial and cheap. Yay ThinkPad!
mp3val is a tiny command line utility (for
Windows and Unix) that knows a lot about the MP3 format and is quite good at
detecting and repairing common problems. Depending on the encoder and tagging
tools you used, your MP3s might be more or less broken; sometimes it's just
an incorrect length information (which might then cause incorrect playing time
calculations). Mp3val can correct most of these errors.
Since mp3val operates on the MP3 frames and not on the decoded audio, it's
fast; the operation is mostly I/O bound so it's probably as fast as your disk.
Anyway, mp3val is a command line utility, and I know that some people aren't
too comfortable with command line tools. Here's a screenshot of the GUI:
Download source
code
As you can see, it's a three button application. Select the "Fix files" toggle
if you want files to be repaired, then press "Add" (while the added files are
processed, the "fix" button is greyed out). Then wait for the progress bar to
finish.
The scrolling window below will print the warnings and errors emitted by
mp3val. As you can see I have a few files with incorrect size information.
There are still some improvements that could be made. For example, the
application could keep track of the "broken" files, so that they could be
processed using mp3val in a second run without having to recheck all the "good"
files. Brokenness-statistics and an sortable list view of the detected errors
would also be interesting. But I consider none of these features to be
important; I'll just let mp3val run over my collection and fix the files. It
never trashed a file for me so far (but fixed some that didn't work), and I
can still re-rip the songs when needed (but I'll choose Ogg Vorbis then as
audio format).
The application is a tiny PyGTK script; that and mp3val is all you need. It
should work on Windows, too, but I can't help you with that.
Another "Pymp of the day" - "Python mini project", done in just an hour. :-)
Python and PyGTK are great for protoyping and mini applications. I love your
work. Thank you!
My project the last few weeks has been to start exercising at a gym regularly. While I could stand to lose a few pounds, I'm more interested in feeling less lousy than I have lately -- among other things, I've been feeling tired and having trouble concentrating, and my back has been hurting again. I don't know for sure if getting into shape will fix these problems, but I figure being more healthy couldn't possibly make them worse.
In other news, I've finally done what all the cool kids were doing in 2002 and started a
Flickr album. Right now it mainly contains cat pictures like this one:
A few things happen to remind us that we'll be parents soon...
* Angie just suggested that perhaps we could take the gin and vodka that we haven't been drinking out of the freezer to make space.
* We put together a crib yesterday. Anything else could've been random Ikea furniture. But y'know, there's just no other use for a crib. (Or a child sized one, anyway...)
* Angie turning around in what usually would've been plenty of space, and sending a glass flying with the end of her belly.
* The steady stream of (welcome) conflicting advice from people who are already parents.
There's less than 3 weeks to go to the official due date, less than 5 to the maximum date permitted (In Qu bec, they induce if we get 2 weeks past the due date). Midwife visits are now weekly. Birth bags are packed. We just need to finish assembling the phone and email lists and ...
... wait. =)
Probably time for more Angie pictures, though. I'll try to take some later.
These are the current
translation ratios
for French and
Czech in unstable,
after 16 days
of "blitz l10n NMU".
The re-introduction of debconf templates for viewvc with incomplete
translations will lower the stats as of today...:-(
I already notified its maintainer who is very l10n-aware so I'm confident we
will fix this soon. However, that's kind of a disappointment for me, of course,
as this will delay the day we will reach 100%.
Yesterday, I worked on the oldest bugs mentioned on
Thomas
Huriaux dynamic status page:
- courier has been uploaded, unfortunately without the Swedish translation
as the maintainer wrongly assumes this shouldn't be a target for Etch
- efingerd has been NMU'ed
- squidguard has been NMU'ed
- ssl-cert has been NMU'ed
- apt-proxy has been NMU'ed (man pages translations)
- defoma has been NMU'ed (man pages translations)
- qmail is ignored because of an obnoxious maintainer who is in my kill file
- gdm is ignored because #261086 is marked wontfix for reasons I don't want to dig
into now (I'm pretty sure that something better could be done for that
package, oh well)
- icedove is likely to be examined at some moment
- glide maintainer has been notified
- arb is ignored on request of its maintainer
- others are my targets for today..:)
PS: the title of this blog entry comes from a
bug report
from Joey Hess for ssl-cert. I didn't fix that because it certainly
requires too invasive changes at this moment but these debconf templates are
definitely not the best worded ever and the underlyign debconf code is
pretty weak.
As I've talked about in this space in the past, I've been coming to some conclusions about free software (open source, or whatever you want to call it): the economics of it, like anything else, are rooted in scarcity, and if you don't have a bit of that scarcity yourself, you have a lot less leverage
This line of thought comes from the age-old question "how to create free software and make money at it?". This is a very important question for me, because I enjoy hacking on free software very much, and would love nothing more than to spend my days doing so. Creating new things, though, rather than working as a "system integrator" and tying pieces together with the smallest amount of glue possible. It's obvious that there is plenty of money to be had for people who simply use free software for their own ends. But what are those ends? We live in a world of scarcity - not everyone can have a fancy car, the nicest house in town, or the fastest computer. Market economies work via the exchange of scarce goods and services. If you have nothing to trade, you have nothing. Now, free software is worth something. Worth a great deal - that is beyond the shadow of a doubt. But since anyone can make a copy, there is no scarcity - once it's out there, you can't trade for it.
So... where does this leave our potential free software creator? Let's think about some of the ways around this problem.
- You could, as some have argued, decide that since software has aspects of a public good, and decide that it should be provisioned by the government. Computer users pay taxes and there is a ministry of software that pays people to work on infrastructure type projects like the Linux kernel, Apache web server, and the like. Whether you like this idea or not probably depends on what you think of governments, free markets and so on, and thus becomes a much broader debate, but suffice it to say that it's not going to happen soon, because free markets do provide software, so most governments aren't likely to fiddle with the system in radical ways.
- Work on contract to create custom systems for clients. This works, up to a point. Since the initial system doesn't exist, there is your source of scarcity. The client wants it, and won't get it - free or proprietary - unless they pay you. It's also worthwhile to think about the "chain" of trade in this case.
Where does your client get their money? If you're writing free code that will simply become a cog in their proprietary system, the actual money comes from the scarcity they have created, so while you may be writing free software, in one sense, you're simply offloading the burden of creating scarcity to someone else, who is then able to pay you for your time. The other possibility is that your client works in the "real world" of scarcity directly - they sell books or beer or cars or something else where the product is, by its nature inherently scarse.
In this case, they still have the leverage - you work to help their business, but they are the ones in the driver's seat. Your work isn't worth much without their scarcity-based business. And they're not very sensible if they pay you for things that aren't directly related to their business. So it's going to be harder to create something new and interesting from scratch.
I've alluded to it in the past, but consulting operations have several problems, especially when considered from the point of view of someone who wishes to create software:
- The more time you spend putting existing pieces together, the better you are doing financially, because you can finish the project quicker, in less time, and so have more of a margin. These sorts of business will tend towards system integration, rather than the creation of new things. That's not a bad thing, and I mean no disrespect to those who are in that business, but it's not really what I want to do, and it certainly doesn't answer the question of how to write new free software and get paid to do it.
- When you create a one off solution that is not completely tailored to one client, if you make something that's proprietary, and know you can sell it to a few more people, perhaps you can beat the free software guy on price, because you can afford to sell yours at a lower price, knowing you'll spread the costs of producing it over a few clients. If the free software guy manages to sell the same free thing to several clients, you have to ask how 'free' the software is in the sense that it is apparently controlled tightly enough that these other clients can't come into possession of it on their own. Once, again, scarcity.
- It's not a business that 'grows' as well as having a proprietary product, because it's very linear. To make more money, you have to add more people (of uncertain quality, I might add). With the proprietary product, on the other hand, you could likely sell 100, 1000 or 1000000 copies with a workforce of the same order of magnitude. This leaves you a bit more breathing room - if you're a consultant, the minute you take a break, be it a vacation, or even go to sleep, you're not making money because you're not engaged. That's sort of a scary prospect as one ages, and needs to dedicate time to other things in life. At 20, I could spend all day every in front of the computer. At 30+, now I have a wife. Maybe there will be kids some day. A successful product-based business would allow me time for those things that a billable-hours business might put more constraints on (of course, there are obviously no guarantees that any business will be successful, but for the sake of argument, we'll treat failure as failure and not worry about it more).
So, there you have it. In one way or the other, to make money at something in the long run, you are going to have to find and sell a product that people cannot effortlessly get for free.
Junichi, qemuctl provides a nice control panel that contains, among others, a save state button and a resume state button.
As a nice addition you could use
qemu-launcher.
Not sure how kvm plays with qemu-launcher and qemuctl, but I guess it can be changed anyway.
I was interested to see the announcement on
Flickr of a new syntax called
machine tags (via
a story on machine tags from Programmable Web).
Machine tags are a way to provide structured information within a tagging structure. It takes the traditional tagging format (just words) and adds a little structure of the form "
namespace:
property=
value".
I like machine tags; they map cleanly into
RDF. I'll be interested to see how the
rel=tag microformat will grow or adapt to this new structure. My guess is that there will be some resistance.
tags:
microformats tagging metadata machine:tags=awesome
I am running through my post-SL TODO list, and I’m down to the item which reads “Write a report”.
We’ve had a very good time meeting users, prospective users and fellow DDs, as always. There were 21 people planned to man the booth, which is an all time record so far, and 18 out of those 21 showed up, which is really quite good :-)
I’m not going to list everybody here, first because the list has grown quite long now, and because all in all it’s the usual French team that has been at it for the past few years already. We had a couple of firsttimers though, including Pierre Habouzit and Yves-Alexis Perez. Everybody did great, thanks for helping out, guys!
Day 0: unpacking sl2007-booth… setting up sl2007-booth…
We had less hardware this year than we had the previous years, for a variety of reasons, so the booth setup actually went quite fast. There were no separations between the booths in the .Org Pavilion this year, so we put a bit of thought into that and figured that, oh well, it’d do anyway.
Pierre arrived at 7:30pm on the booth with some hardware from Sven and 1500 Debian flyers paid for by Linbox. We went for dinner, then Pierre and I spent the night hacking on the Babelbox scripts to get the damn thing to work with the current debian-installer. At 3:30am, we got to the point where the Babelbox would pretty much work, and couldn’t figure out what was going wrong at that point.
Day 1: Fixing the Babelbox, frying a graphics card
We ended up fixing the last bit of the Babelbox on the booth in the morning, and preparing the demo system (Pierre’s shuttle) at the same time. It all went fine and we ended up with a dual-head demo system and a beamer-equiped Babelbox in the afternoon after Niv brought a beamer. We used the back of Fr d ric’s Debian banner as the screen for the beamer.
We had a nice chat with the Linux+ magazine team in the afternoon, discussing the upcoming Debian release and a dedicated Linux+ issue. We’ll be helping them out tracking the Etch release in the next few weeks until the release.
Back at Pierre’s place, we found out that his shuttle apparently wouldn’t boot. Turns out that only the graphics card died, good news.
Day 2: Business as usual, Debian France meeting
The second day was really business as usual. We saw more people than the previous day, but still there were less visitors than the previous years. The exhibition was smaller than the previous years too, I just hope it’s only a temporary thing and that it’ll be bigger again next year.
This also was the “dinosaurs day”, the day all the old-timers pop by and discuss around the Debian booth. This is becoming a tradition or something :-) As Christian wrote already, we met HP people and talked about Linux support from HP on HP hardware.
In the evening, we had the first formal Debian France meeting. Julien Cristau and Christian Perrier got elected to the board of directors, replacing C dric Delfosse and Laurent Fousse. The board got reelected, so I’m still the President, Pierre is still the Secretary and Aur lien G r me is still the Treasurer. Thanks to Tarsus for lending us a room for the meeting!
Day 3: Packing up
Small day… Quite a few visitors in the afternoon, we talked about Debian France a bit more than we did on the previous day, mainly with other people from the .Org Pavilion.
Conclusion
Overall, a very productive edition this year, we talked about Debian France a lot, about Etch, about quite a few projects involving Debian too. Now we have a lot of work to do to release Etch and develop Debian France to the point were it will actually be really useful for Debian. Hopefully we’ll have things to show in a couple of weeks.
There is a feature in kvm, or rather qemu, that enables users to resume from a known state.I was using kvm for a while and got irritated by having to watch Windows booting.By using 'savevm' and 'loadvm' feature, it is possible to skip that step and always start from the point where you can start working.The tip here is to create a COW image from the HDD image that 'savevm' was done, so that you can re-run from the same state how ever many times you want.It feels better to be in a running Windows session within 10 seconds.
Ok, now that was lame.
Anyway, I’ve just upgraded
perlis to
Feisty Fawn at the wake of some 1169 MB worth of packages. An
apt-get autoclean is therefore in order. I’ve also yet to test hibernation and resume, but
knetworkmanager definitely impressed me. Too bad it doesn’t have ifupdown integration yet (which I’ve just done to
wlassistant now, pending testing and REVU.)
Just about the only snag I got caught while upgrading was a
file overwriting of some .pngs by
adept. Since I was upgrading using
adept-updater I had to work around using
dpkg -i --force-extract and resuming the dist-upgrade using
aptitude. I was told in
#kubuntu-devel that it will had to be fixed later, and even though I may be able to fix it, it’ll be hard finding a main component sponsor since it’s a weekend. :/
Oh well, life
exec()s on. Anyhow I’m too early for upgrade testing (which is on the 8th,) but the outlook for dist-upgrades looks good. And, I’ve other stuff to ponder about, like finishing Philosophy and Procedures 2 of my
NM application…
So, back to geekery after too many months away. While we
were in Thailand, I met
Theppitak
Karoonboonyanan and his friend Neutron Soutmun and a
couple of others from the Thai Linux community. Thep is in
the NM process for Debian, he maintains Thai support
packages in Debian and Neutron is a Debian user. Neutron
writes firmware for GPS receivers (IIRC) and other GIS
stuff, I'm hoping he will get involved in the debian-gis
subproject. I think I convinced Neutron to at least think
about applying to NM :D. We talked about a lot of things,
mainly about Thai localisation and the challenges involved.
He mentioned that the language barrier is a big problem for
Thai people, so their main focus has been firstly
infrastructure (text rendering, layout and wrapping, fonts,
input methods, locale, etc) and now translation (and the
associated, laborious localisation efforts). He told me a
bit about the writing system and how it is related to other
systems in the area. Thep also mentioned the possibility of
debconf9 being in Thailand, I recon it would be bloody
awesome to have debconf in Asia. At least one other Debian
Developer is interested in this, madduck is the
initial
instigator. I hope we both make it to debconf in the UK
this year. I also visited the open source lab at
NECTEC (the Thai
National Electronics and Computer Technology Center), which
is government funded. There, they develop LinuxTLE (an
Ubuntu based desktop distro), LinuxSIS (a simple internet
server for schools and businesses) and do lots of
translation and advocacy work within NECTEC and with
businesses and other organisations within Thailand. One
thing about LinuxPLE which I noted was that during the
post-install GUI configuration step, there is an option to
setup the system to use fonts from a mounted Windows
partition. IIRC, they explained that they found this was
important because of a reliance on Microsoft fonts in
Thailand. While I was there, I went to a couple of other
labs and saw a demo of a cool Thai OCR and car registration
plate recognition system,
English to Thai machine
translation (text) and direct English speech to Thai
speech conversion. They were also working on some medical
imaging and speech recognition stuff that I didn't get to
see. I also met the founder of
linux.thai.net, whose
company develops this
online map for Bangkok..
Also posted some
photos
from our trip through Thailand.