Slowly migrating to KDE4

Recently the debate about KDE4 has reemerged with a lot of people complaining heavily - either their pet feature is not yet (re-)implemented, or they don’t agree with some fringe design decision, or something else is not exactly the way they like it. I could understand people had mixed feelings about 4.0, but now it’s getting out of hand in my opinion, while 4.1 won’t match KDE 3.5 in every possible respect, the improvements during the six months since 4.0 release are incredible - very promising prospects for 4.2 and I don’t even dare think about 4.3 - oohwee.

I’ve been running KDE 4.0 on my ancient Pentium III Compaq bricktop, since before it was released, and recently upgraded to 4.1 beta. I’ve also just played my first songs with Amarok2 on it, had a look at K3b and KOffice2, things are rapidly approaching a usable level. But this doesn’t really count as migration, as I rarely use this bricktop, and almost never do anything remotely productive on it.

I’ll definitely complete migration to KDE 4.1.x for my workstation when openSUSE 11.1 is released in December this year - at the very latest. Being a huge fan of KDE3 I’ll surely miss it, but there’s no use in clinging to the past - for the good of KDE and the free software desktop in general, it’s getting time to move on.

So far I’ve begun a slow, gradual migration of my workstation, from within safe and familiar KDE3 - moving to KDE4 components one step at a time.

I’m using KDE4.1 beta KWin with effects enabled (yes, in KDE3. seli made this very easy to do). I could never get along with Compiz for more than a few minutes at the time, but now I’ve used kde4-kwin for more than a week, virtually without a single issue. While it doesn’t have as many or as extravagant effects as Comiz, it does have the useful ones (scale windows, desktop-grid and such) and also wobbly windows and some pretty switchers.

I’ve also begun using a number of the great KDE4 applications as defaults for their respective jobs - using Okular for PDFs, KDE4 Gwenview for image viewing, using LoKalize for translations, Dragon Player for video, KTorrent, Marble, KDE Games… steadily adding more KDE4 apps to the bunch. Of course I’m using Oxygen icon theme for YaST2 and Firefox

I’m lovin’ it…

Bedre oversættelse af KDE?

Jeg har i noget tid gået og sparet sammen til et nyt indlæg om KDE-oversættelserne, så da Anders Lund i dag har blogget om emnet, er jeg faret i blækhuset. Det glæder mig, at Anders Lund synes, at det er rart at bruge oversættelserne, han påpeger imidlertid også, at der er en del slåfejl. Jeg vil ikke benægte at kvaliteten langt fra er, hvad KDE fortjener.

Sagen er imidlertid den at KDE er et enormt oversættelsesprojekt med mere end 115.000 tekststrenge alene i brugerfladerne - dokumentationen kommer oveni. Oversættelsesarbejdet er næsten altid blevet udført af en enkelt person med sporadiske bidrag. I mange år var det Erik Kjær Pedersen - I de sidste fem måneder har det været mig. Ikke alene skal jeg vedligeholde oversættelserne, men der er også et efterslæb efter næsten et halvt års inaktivitet, som gjorde at KDE4.0.0 blev udgivet uden danske oversættelser.

Derfor er kvalitetssikring en umulighed. Jeg er nødt til at forsøge at oversætte så meget som muligt så hurtigt som muligt. Selv hvis jeg havde tid til korrekturlæsning, ville det ikke hjælpe meget at gøre det selv, da man oftest overser sine egne fejl. Det er min forhåbning at kvaliteten senere kan hæves, når efterslæbet er endeligt indhentet. Jeg fokuserer i øvrigt ret ensidigt på KDE4.x, da der ikke er de store fremtidsperspektiver i KDE3.5.

Du kan hjælpe på flere forskellige måder.

Rapportér fejl. Skriv en e-mail til mschlander [at] opensuse.org eller kontakt “cb400f” via IRC i #dansk-gruppen på Freenode-netværket. Man kan altid finde e-mail-adresse til KDE-oversættere af hvert enkelt program på denne måde: Hjælp -> Om [program] -> Oversættelse

Fejl i KDE3.5 vil kun blive rettet hvis de har væsentlige forståelsesmæssige konsekvenser, men der er betydelig risiko for at de går igen i KDE4.x, så det betaler sig at rapportere dem alligevel.

Hjælp med korrekturlæsning. Alle oversættelserne jeg laver sendes til dansk-gruppens mailingliste hvor enhver kan tilmelde sig og give feedback.

Hjælp med at oversætte. Det er vigtigt, spændende og overkommeligt arbejde. Jeg vurderer at omtrent følgende kvalifikationer er nødvendige:

• Du æælsker KDE
• Du er habil til engelsk
• Du er habil til dansk
• Du er villig til at bruge et par timer eller mere i gennemsnit pr. uge over en længere periode

Det kan også være at du bare gerne vil stå for oversættelsen af et enkelt program, som du er ekspert i og som betyder meget for dig.

Kontakt mig via mail eller IRC som anført ovenfor.

Oversættelse af KDE

Efter adskillige års flittigt arbejde som nærmest ene-vedligeholder af KDEs danske oversættelser (formentlig verdens største FOSS oversættelsesprojekt?), trak Erik Kjær Pedersen sig tilbage i efteråret 2007 pga. travlhed i sit civile liv. Han overdrog opgaven til Keld Simonsen, som således nu er koordinator og har det overordnede ansvar for de danske oversættelser af KDE.

Imidlertid vil det være de fleste danskere, som interesserer sig for fri software, bekendt at Keld Simonsen er en travl herre med mange jern i ilden. Der er heller ikke sket ret meget hvad angår KDE-oversættelser i den senere tid, det står faktisk så galt til, at KDE 4.0.0 ikke er tilstrækkeligt oversat til officielt at blive udgivet med dansk sprog. Det kan vi selvfølgelig slet ikke have, så det vil jeg gøre noget ved og er derfor påbegyndt oversættelsesarbejdet.

Min plan er først få oversat de få uoversatte dele i KDE3 stable. Derudover påtænker jeg ikke at gøre meget mere ved KDE3, men hvis nogen rapporterer meget graverende fejl skal jeg nok kigge på at få dem rettet.

Derefter vil jeg fokusere på KDE4. Første delmål er at få oversat “essentials” således at dansk igen kan blive en officiel del af KDE 4.0.1. Næste delmål er at få oversat alle de centrale dele. På lidt længere sigt er målet komplet oversættelse og et generelt løft af kvaliteten, om alt går vel.

Hvis nogen støder på fejl og dårlige oversættelser i KDE4 eller voldsomt generende fejl i KDE3 er de meget velkomne til at gøre mig (cb400f) opmærksom på det i Dansk-gruppens IRC-kanal på freenode-netværket - #dansk-gruppen. Det burde ikke være svært at fange mig der. Alternativt kan dansk-gruppens postliste bruges. Samme fremgangsmåder kan i øvrigt anvendes hvis man har kommentarer til oversættelserne i openSUSE (dog kun de openSUSE-specifikke dele, såsom YaST, opensuseupdater, Kickoff, sysinfo:/, installationen osv.).

openSUSE and desktop environments

When GNU/Linux distributions provide both KDE and GNOME, it’s usually very clear which desktop environment (DE) is recommended, gets the most developer attention, etc. However there seems to be some confusion about where openSUSE (previously SUSE Linux) stands with regards to DEs. It’s been about a month since openSUSE 10.3 was released, and most tests and reviews seem to focus on GNOME. The Linux Action Show podcast even went as far as saying something along the lines of “openSUSE is obviously a very GNOME centric distribution”. Of course most reviewers are new users who don’t follow the openSUSE project closely and admittedly the situation is a little complicated, so you can understand they’re confused. This is my attempt to explain what is what. I’m a community person and thus don’t have to worry about corporate policy. I’m also a KDE user but I will be as objective as possible.

First things first, openSUSE has no default desktop. If you get the DVD the installer will force you to actively choose which DE to use. You can also download KDE or GNOME single CDs respectively - which again forces you to choose.

KDE in openSUSE
In 2003 Novell acquired the German GNU/Linux distributor SUSE. Historically a very KDE centric distribution, with a very large KDE userbase and community, with most developers also using KDE. openSUSE is still primarily developed by these people in Germany and Czech Republic. Former openSUSE project manager Andreas Jaeger uses KDE, and his successor Stephan Kulow is a very high profile KDE developer. SUSE still employs more high profile KDE developers than any other distributor. Novell are corporate patrons of KDE, and silver sponsors of Akademy.

openSUSE provides excellent service for KDE users. A stable and polished KDE, which is furthermore enhanced with components developed by openSUSE such as Kickoff menu, KNetworkManager, Sysinfo kioslave, Kerry beagle-frontend, KPowersave etc. Build service repositories provide lots of goodies - newest KDE applications backported to shipped KDE version, newest KDE3 builds, snapshots of KDE4 that are updated weekly, at least.

A survey done for openSUSE 10.2 with >27.000 openSUSE users participating, showed 71% using KDE and only 22% using GNOME. Naturally, this vast majority of KDE users also affects testing and the availability of online help and support. Currently kdedevelopers.org is running a survey about which distribution provides the best KDE, which openSUSE is leading at the time of writing.

GNOME in openSUSE
In 2003 Novell also acquired Ximian - a GNOME company with some high profile GNOME developers. However most of these guys have little or nothing to do with the openSUSE project. People like Nat Friedman and Miguel de Icaza take no part in the development. Their focus is steadily on SUSE Linux Enterprise where most development and marketing is centered around GNOME. Maybe this is why some people seem to automatically expect that the same thing goes for openSUSE. Novell also sponsor GUADEC, and various GNOME-related projects such as Banshee or Beagle.

Lately there have been improvements as for GNOME in openSUSE - both with regards to features and organization. There’s now a native GNOME updater applet, there’s YaST-gtk providing better integration with GNOME than YaST-qt. The slab menu and gnome-xgl-settings developed for SLED10 are of course part of openSUSE now. These things are mostly still rather immature though. GNOME also has some build service repositories providing latest stable versions and also unstable. They’re having frequent IRC meetings to improve things etc.

Final words
I hope this post has helped make things clearer. To sum it up, openSUSE puts a lot of emphasis on both large DEs, and it would probably be wrong to say it’s centered around either one, but a significant majority of its developers, active community members and normal users prefer KDE.

YaKR (Yet another KDE4 Review)

Since the release of KDE 4.0 beta 3 there’s been a lot of talk about KDE4, it even got mentioned in Danish mainstream computer media. So I thought I’d share my thoughts too.

I installed KDE 4.0 beta 3 on openSUSE 10.3 using the packages provided by the openSUSE KDE maintainers. Here’s the screenshot.

Stephan “Beineri” Binner’s openSUSE based LiveCD KDE Four Live is also available witk KDE4 beta 3+.

Besides testing beta3 I’ve also been following KDE4 pretty closely as users go - not following svn or devel mailinglists, but trying to keep up with the goings on. And I have also tried out some earlier builds.

Expectations
The first thing you need to understand about KDE4 is that it uses a little unconventional versioning scheme. It’s similar to the one Microsoft uses for Windows, so most people should be fairly familiar with the style. “Beta3″ actually means “early alpha” by usual standards, and 4.0 “final” will mean it’s actually a “tech preview” for the avantgarde to try out in real life. After a “service pack” or two, we can finally really judge the product (KDE 4.1/4.2).

This is actually not meant to be a rant against KDE, I believe in “release early, release often”, and I think it’s very important to get a release out there, to really get some traction going for KDE4 - but it’s also very important to adjust people’s expectations, or a lot of people will be extremely disappointed by KDE 4.0. Beineri already explained about KDE4 != KDE4.0 long ago, but I think it doesn’t hurt to remind people. Now that expectations are hopefully aligned with reality, let’s continue.

Platform
Most of the KDE4 development so far has been in the libraries. Not very visible to the user in the short term, but in the mid to long term I’m convinced that Solid (hardware), Phonon (multimedia), Sonnet (spellcheck), Akonadi (pim), Decibel (chat) and so on will help the creation of great applications to the immense pleasure of users and application developers alike. Considering that KDE3 is already excellent in this respect, the implications of a much improved KDE4 is almost scary. Not to mention all the goodness of Qt4 - such as great svg. Oxygen (icons, window decoration, widgets), will be a huge improvement. Note that my screenshot does not have oxygen widgets nor windeco, and the icon theme is not yet complete. While this all leads to very high expectations, in many ways it’s too soon to judge.

“User experience”
Some things are visible already though, and disappoint me. I thought that the most common complaints about KDE3 would be taken care of, but judging by beta3 this is not the case, au contraire. I noted at least three issues that I think are disastrous, no less.

Common complaint 1: “KDE gets in my way, it’s obtrusive, there’s too much going on”
What does KDE4 do about it? Well, it puts sound notifications on kwin events like opening, maximizing and minimizing windows. Konsole spews popup notifications left and right. Of course these things are configurable, but defaults are immensely important. The rule should be to “disturb” the user as little as possible, unless he asks for it.

Common complaint 2: “The panel takes up too much of the screen”
What does KDE4 do? It makes the panel even bigger (I didn’t pull out kruler to measure, but it’s certainly not smaller).

Common complaint 3: “KDE(3) is too much like Windows (XP)”
What does KDE4 do to combat this widespread misconception? It makes the panel look very much like the panel in Windows Vista. I know the artist says it’s purely coincidental that he happens to share tastes with Microsoft artists. But it’s still a bad idea - I think so partly for selfish reasons, I’m sick of hearing people say KDE is Windows-like, and now it looks like I’ll have to listen to it for years to come.

Apps
Not many of the applications are really anywhere near ready yet, but the few that are usable are mostly very nice. I was never a fan of the decision to replace Konqueror with Dolphin as the default file manager, but maybe it makes sense from a strategic point of view, maybe Konqueror is pearls for swine. But Konq in beta3 seems rather limited, having only one viewmode (dolphin-part), I think I have around 10 different view modes in KDE3-Konq. I hope this is a temporary situation, and that Konq will return to at least KDE3 level in 4.1 or 4.2. On a positive note the default toolbar setup for web browsing in Konq seemed very neat. The KDE4 games are very nice looking, always been a fan of kmines, and it’s great - though I miss the smiley face that was there in earlier KDE4 versions.

When KDE4 development began I was very worried about the prospects for KBabel which I more or less depend on for translation work, as it wasn’t maintained actively. Ironically, today it’s replacement KAider is one of the most usable KDE4 apps, and I’ve already used it for “production” a little bit.

KOffice2 released an alpha concurrently with KDE4 beta3. It also looks very promising. I especially enjoy that the KWord UI is starting to resemble Krita, making the suite more “uniform”, and simply making KWord look way cool. So far I haven’t tested the fruits of the Google Summer of Code ODF-project, but certainly ODF support, and compatibility with OOo must have improved significantly.

Conclusions

• KDE 4.0 might be nearing release, but KDE4 has a looong way to go
• With all the library work done and Qt4, there’s an amazing (unrivaled?) platform for desktop and application development available
• Developers need to be more careful about defaults. Do not let the default settings annoy people, for crying out loud
• Usable applications are still scarce, but many of them are looking very promising