Category: KDE

After having had a lazy week in Javea, in the east of Spain, where Nel, Christine’s mother has
been a great host, I arrived in Malaga with a slight delay (thanks, Iberia).

Having gone to bed quite early, I was quite fit to begin the first day [schedule]. We were welcomed by people of
the Andalusian government, the university of Malaga and the people of the Malaga Linux group.

The technical part started with a presentation by Pete Goodall
who would give an overview over the Novell Linux Desktop. Quite obviously, not
everyone was impressed. The software used on the desktop is quite old (running KDE 3.2). Also, Pete
showed a KDE desktop and didn’t make to clear that the standard Novell Desktop is gnome-based. The
rest of the talk wasn’t also to developer-directed, ow well …

Far more interesting was Till Adam’s
talk, who showed off kolab and kontact, the KDE approach to groupware. Pretty fun,
lots of cool features coming up, such as a webclient (which needs more developers, PHP horde people
might be welcome there!). Kolab stores all kind of information in IMAP accounts, which makes it
possible to use a couple of groupware stuff (shared calendars anyone?) with any simple IMAP server.
Kolab however adds management interfaces, user management via LDAP and loads of other stuff.
Impressing.

Mark
Shuttleworth from Canonical, the Ubuntu sponsors also did a very nice talk about
collaboration in the Open Source world. He talked about collaborated distributed
working, realtime collaborative editing (imagine working on the same document, at the same time,
discussing what you’re chaning …).

At the moment (uploading the blog might take some time, I forgot my wireless adapter at the
residence), I’m listening to Kurt
Pfeifle’s mind blowing presentation, demoeing NX technology. It’s far more
efficient than standard X11, as he and Fabian just showed by running an xnest session and a freenx
session in parallel and then increasing the lag on the localhost interface artificially. Nice.

… Kurt’s talk is finished, I got hold of a UTP cable and a dhcp lease and am uploading a
couple of photos now, enjoy
Aah … some sponsored social event for tonight has just been announced, is that the free
beer part? :-)

update: Yeah, free drinks.

The talks of Richard Moore
about Scripting with
KDE feel was quite interesting, especially for the power user who wants to enhance his desktop
experience.

Also Will
Stephenson had some interesting news to tell, skype integration into kopete being one of them.

I bake my own bread.

Normal dutch bread is pretty “fluffy” and light, it’s very easy
to compress but I like heavier bread much more. German bread is usually much heavier, and that’s
what I’ve grown up with.

Solution to the problem was the breadbaking machine I got from my mother, and I’m using it
pretty often, at least once a week. The fun thing is that this machine promotes creativity, plain
bread is easy to make, but it’s just as easy to put some other ingredients into it. My lasts breads
have been bacon bread (baked bacon in it, perfect match with cheese), fruitbread
(dried fruit, such as apple, apricots, prumes, best with plain butter), onion bread
(with fried onions in it, tastes good with everything except sweeter stuff, cheese does very
good, so does bacon). Today I visited the St. Annamolen, a clichee dutch mill where they’re actually
still milling. You can buy all sorts of flours right there, and they seem specialized in providing
bread mixes for bread machines. (And also pet supplies, hi Rinse.) They have about 40 different
sorts of bread mix there, I’ve chosen the “Pommodori bread”, which gave a really good
result, red bread with a strong taste of tomatoes, obviously.

It can be quite tricky to produce a good dough. The right amount of water is crucial, but the
amount of salt and sugar in your ingredients can have a lot of influence, too. Salt makes the yeast
rise less, sugar has the opposite effect. A good measure of thumb is that the dough should be a
little sticky. Well, it’s a science, but recent results show I’m advancing (ask [ade], he serves as some sort of Guinea pig for some
of the experiments). The fun part was trying to get an idea of the effect of sugar, when the dough
found its way outside the bread machine, cleaning the mess afterwards was less fun, though.

Friday, I’ll be leaving for Spain, where I’ll visit a friend near Alicante and head for aKademy about one week later. I’m preparing for the trip
which is finishing my paper, do other paperwork which otherwise might need attention through the
next two weeks, get notebook prepared for all kinds of conference stuff and hacking, wash clothes,
think about the content of my luggage, prepare list of people which I owe a beer… In short boring
and stressful.

I’m looking forward to meeting you in Málaga!

One of my special interests is power management (PM).

While discussing if a display powermanagement option really should be in guidance‘s displayconfig, I had a look where
PM-related options are living in kcontrol at the moment.
It was quite a shocking experience.

Assuming I want to optimize my laptop for battery lifetime, where are the places I have to go in
kcontrol, what assumption do I have to understand? Let’s have a look.

• display power management comes to my mind, go to Peripherals -> Display -> Power
  Control
• Most WiFi card drivers offer a powe saving mode, go to Internet & Network ->
  Wireless Network
• Power Control -> Laptop Battery offers actions for events when battery is
  running out and also some ACPI controls which make suspend modes and CPU power saving modes
  available – if setup correctly, also there’s a battery monitor
• My installation also has modules for IBM Thinkpad laptops (mostly for buttons / ACPI
  events) and a Sony VAIO module which aims to make the battery status of two batteries available (I
  have neither of those models, so these modules are of no use for me.)

I think the most easy thing for the novice user with respect to getting the laptop draining less
power is getting hopelessly lost.

The solution to this problem, however, is not trivial. Power management is related to a lot of
different areas, suspend modes such as suspend-to-ram, suspend-to-disk, power management of
different devices such as wireless adapters and the display play a role, but there’s obviously more
to it. I encountered that when designing the display powermanagement tab for guidance’s
displayconfig. It seems related to a screensaver (although a screensaver is no more than some eye
candy toy on more recent hardware), but DPMS is some sort of extreme screensaver. Then there’s CPU
states (C1 through C4 on some systems, although that’s really nothing the user should ever
notice).

The big question is: How can power management be handled in a sensible way?
That is without having duplicate configuration options.
The approach I’ve taken now in guidance
is offering a DPMS tab, mostly for practical reasons, with a button to configure the screensaver
which just starts the screensaver configuration module. I’m not sure if we’re going to keep it this
way, though.

If you’re also interested in this area, please drop me an e-mail at sebas@kde.nl. Maybe this
topic can also have profit by some discussion at aKademy
2005, which I’m going to attend.

So obviously I’m back from my vacation and already had my first (quite productive) working day.
CodeYard‘s infrastructure is nearly completely in place now (thanks to [ade]), and I wrote down some stuff for EuroQUIS,
mostly theories related to software quality assurance. With Simon having working internet access
again, KDE related stuff seems to have advanced a bit. After returning I also found Debian packages
of guidance including dependancies in my
mailbox. Next step is to test, review and merge the stuff. It’s probably a good idea to include the
debian/ directory in our subversion repository, so it’s easy for people to build updated
packages.

My vacation (together with $GIRLFRIEND) was really good, Berlin is still a great city. Lots of
stuff to do, such as checking out all sorts of new architecture. What struck me was that the “new
centre”, around Potsdamer Platz (where Sony, Deutsche Bahn, Deutsche Telekom and other big
businesses built their vision of a metropole) does have pretty much no charme at all. Good place to
take photos, bad place to have fun. Far too much commercial stuff. Bummer.
Much more enjoyable is Kreuzberg, which grew into a really good place to eat (especially the
delicious breakfasts here). Also the people there are
pretty down to earth, very direct in their communication, still friendly in most situations. And
terrific Currywurst at the Curry 36, probably the best one in Germany.
We also visited Potsdam’s famous castle Sanssouci, which really looked like “old Fritz” wanted to banish normal
life and create a place free of sorrows. Nice walks in the park though.
Checkout a selection of my photos
taken during the trip, partly even wallpaper material. :-)

I received my new displays today. Finally I don’t use CRTs anymore, my Samsung Syncmaster 710N
and my Iiyama S900MT1 have been replaced by two Samsung 173P 17″ TFT displays. The colors
are very bright, the picture is crispy, and the design very stylish. The first thing I did was
setting all my fonts 2 points smaller. :> Another technical highlight of these screens is the great
contrast (1000:1). More pictures can be found somewhere in my gallery pages. Although
these displays are not designed for gamers (response time is around 25ms) they perform reasonably
well with respect to ghosting even with a fast first person shooter. (But then I’m not a gamer-type.)
The screens have one button to switch them on and off. Switching on and off is confirmed by a
subtle beep, the button reacts on a slight touch, nothing moving.

The rest of the functionality is supposed to be setup via software. Of course the software
Samsung distributes with the panels is non-free and Windows only, which is quite lame. (They could
at least support free software projects by releasing specification to make it easier to support all
the nifty features for Open Source software developers. *sigh*) Ddccontrol, however, can do most of the stuff (changing
settings) via DDC so it’s quite usable under Linux, which is my OS of choice. I didn’t find a way
yet to use the pivot feature (both screens can be rotated a full 360º) and XRandR’s rotation
stuff does not work in dualhead mode.
Not at least, these screens save a lot of enery compared to my old CRTs, which is one the one
hand good for global climate, and on the other hand for the local one. My room can get pretty hot
during the summer, burning a lot of energy with old monitors did definitely not help that.

I already love my new toys. :-)

One of my favourite applications in KDE is digikam. I own
a digital camera, and I like taking photos. Nothing professional, though, but it’s fun and relaxing
for me. I use digikam to download the photos from my camera (a Canon Powershot A95). One of my
favorite features is that digikam automatically rotates the photos according to the EXIF information
while downloading them. That saves an awful amount of manual work. I also like the cool OpenGL
slideshow – the transition effects look neato!

I’ve been especially pleased reading through the dutch documentation of
digikam and the dutch
documentation of digikamimageplugins, because a lot of photos from my collection have been used
to illustrate the features of digikam. I’m of course happy to be able to contribute to this great
application in that way. If you’re translating the digikam documentation and you need photos to
apply all the plugins to, I can provide you with mine, the ones Rinse used to prepare the dutch
documentation. Just drop me an email in that case.

We have 75% of guidance pretty much
completed, userconfig, serviceconfig and mountconfig look pretty stable and are feature-complete.
The stuff that might need a little more work is the installation magic, which is done via pykdeextensions. Also the dualhead
configuration bits of displayconfig are not yet there. The abstraction layer is partly in place, and
we don’t seem to have broken too much (yet). Hardware detection (VGA and connected displays) is
still a bit tricky though. Furthermore I’m trying to get guidance fully reviewed for usability
issues before or at aKademy 2005, so I can try to fix
all existing issues at the hacking marathon there.

Yesterday I received an e-mail from a kalyxo
developer, who is – like the Kubuntu bunch – interested in
including guidance in their package archives.
Good to know that we’re obviously filling an
existing gap.

Ow, and I noticed syndication of my feed on planet KDE,
thanks clee and hello world! :-)

… that’s what I am. I’ve got a new job, that’s why.

I’m working at the subfaculty of computer sciences now. Two projects will keep me busy:

EuroQuis,
which has basically one target: preparing a catalogue of criteria that help assuring quality in Open
Source software. So I’m doing research on that issue.
CodeYard, the second project
I’ll be involved with aims to involve highschool students in Open Source development by providing
infrastructure – such as websites, version control systems, blogging facilities – as well as
guidance in the development process. So we’re preparing the new breed.

World domination, here we go!

The picture was taken yesterday at the KDE-NL BBQ in Bolsward in the north of
the Netherlands. Rinse had prepared a rich BBQ with all kinds of good food. The meeting was relaxed,
we discussed a couple of the current issues regarding our plans for world domination face-to-face,
which really helps from time to time. After having returned, we payed a visit to the 4-daagse-feesten (actually the zomerfeesten), which is a week-long
party in the whole city centre. That was fun, right, [ade]? :P

I’ve prepared a new release of kvandale
(TPFAK, “the program, formerly know as kgnuvd”), a KDE frontend for the online Van Dale, the
famous dutch dictionary. The changelog’s highlights of this release are

• a new name,
• i18n support,
• automated installation

… and a couple of small other improvements.

You can download it here, bugreports are
welcome, feedback is appreciated. :-)

Ow, and I got rejected at Google’s Summer
of Code. Bummer, but as I already saw, there’s a lot of
improvement to be expected from the accepted developers. Go go go!

Most of you know that under Linux, you can’t unmount a device if it’s still “in use”. Less
of you know that there are tools that actually deal with this. Most notably “lsof”, which
lists open files on a directory or devicenode. There’s also “fuser” which can even kill all
the processes that are using files on a specific device. The latter should be used with care, of
course.

Using guidance’s mountconfig,
you’re likely to encounter the same problem at some point – and that’s exactly what last week’s
coding efforts tackle. Umounting a device that’s still in use will now launch the dialogue shown in
the screenshot, identifying what’s preventing you from unmounting the device so you can deal with
these applications. The rude might choose to kill processes, but they have been warned.

Coming up next: Will the user interface survive Simon’s critical review? :-)