AUG
16
2007

Quickies: Amarok 1.4.7, MEPIS KDE 4, Desktop Survey, Lugradio Talk

Amarok 1.4.7 was released with improved collection backend, new streams, altered icon and bugfixes. *** The annual Desktoplinux.org Survey is under way. *** MEPIS released a KDE 4 Beta 1 live DVD using packages from Kubuntu. *** The videos from Lugradio Live are up including Ben Lambs' Conquering the Desktop with KDE 4. *** Finally, following the 10th anniversary of the free desktop last year, congratulations to another project which has gained double figures in age, but whatever did happen to those Scheme applets?

Comments

From the desktop survey:

"NOTE: Participants are asked to refrain from promoting or advertising the survey to mailing lists, or encouraging friends or co-workers to vote for specific software choices. This will ensure that the survey represents a broad sample of Linux desktop users rather than being used to advance a particular open-source software cause."

By posting a link on dot.kde.org, you are violating the spirit of the above request, and biasing the results in favour of the KDE project.

The PCLOS, Mandriva and Ubuntu forums have all had threads advertising the survey, and all have respected the wishes of the survey creators by deleting these threads. Please do the right thing and remove the link and all mention of the survey.

Many thanks in advance.


By anon at Thu, 2007/08/16 - 5:00am

The page; http://www.desktoplinux.com/news/NS7943272425.html has a link to post the story on dig. They *are* looking for more publicity. Asking the dot to remove it sounds quite hypocritical with that fact in mind.


By Thomas Zander at Thu, 2007/08/16 - 5:00am

While digg is a general news site, a posting on the dot is more like a request to vote for KDE.
Well i already voted anyway.


By Frank at Thu, 2007/08/16 - 5:00am

And I don't know what digg is and I'm not interested in. I'm happy that it was posted here because I wouldn't know that survey else.


By Tom at Thu, 2007/08/16 - 5:00am

Digg has no affiliation towards any particular software project, though, so the exposure from Digg will give them useful results. The same would hold for, say, slashdot if it hit their front page.

The exposure from dot.kde.org, however, which is obviously strongly affiliated with the KDE project, will taint the results, just as exposure from the PCLOS, Ubuntu, Mandriva forums (or from, say, gnome.org) would taint the results by over-representing their affiliated projects. The three aforementioned forums recognised this and removed the link of their own volition.

There's no hypocrisy at all, whatsoever, here: the survey creators want a large set of results from open-source software users, but they want them to be *neutral* with regard to particular open source projects.


By anon at Thu, 2007/08/16 - 5:00am

Digg has no affiliation toward any particular project ? that's why EVERY KDE NEWS ever posted on digg is FULL of flame against KDE. No criticism, pure flaming only. There's no arguments. Just insults and spitting.


By Anonymous Coward at Fri, 2007/08/17 - 5:00am

"The same would hold for, say, slashdot if it hit their front page."

Just as the other poster to your comment stated about digg, I would have to say that slashdot is definitely anti-KDE biased.

If it has more buttons than GNOME, its too complicated and they hate it. Just like they hate MS compared to Apple. They think that everything has to remove all buttons and we should all move to buttonless interfaces.

Seriously. Go read slashdot sometime and you will see for yourself.


By Forest at Sat, 2007/08/18 - 5:00am

I don't see how simply telling a group of people (dot.kde.org readers) that the desktop survey is running is biasing the results.

I, for one, didn't know the survey was on until reading about it here. I would also have submitted the exact same results regardless of where I had read about it.


By John at Thu, 2007/08/16 - 5:00am

"I don't see how simply telling a group of people (dot.kde.org readers) that the desktop survey is running is biasing the results.

I, for one, didn't know the survey was on until reading about it here. I would also have submitted the exact same results regardless of where I had read about it."

The logic is so simple that I'm astounded you cannot grasp it, but here it is, using your post as an example: because of the fact that dot.kde.org posted the survey here, KDE got one more vote (from you) than it would have done had it not been posted. Now multiply this my the readership of dot.kde.org. Even if you personally (contra to my assumptions) didn't vote for KDE, we can assume that a large majority of the readers who would not otherwise have voted at all if the link hadn't been posted here have now voted for KDE, biasing the results.


By Anonymous at Thu, 2007/08/16 - 5:00am

Well, it doesn't prove in the slightest that having so-called neutral sites like Digg advertising the Survey and not the dot will bring you results closer to reality, since you don't have the faintest idea of the general profile of Digg readers.
I'm also of the type of people who read only the dot and linuxfr.org, as far as OpenSource is concerned. I probably wouldn't have heard of this survey otherwise. Is my vote less interesting or valuable than any other Digg reader?
Maybe the results will be biased because most KDE users comes to the dot, while XFCE users got to Digg or some other so-called "neutral" sites. You have no way to know, but you assume that it is and you will claim that these results are more representative. And that's just obfuscation, in my opinion.
Why not just admit that this kind of internet survey is basically worthless, instead of trying to give it any kind of credibility?


By Richard Van Den Boom at Thu, 2007/08/16 - 5:00am

"Maybe the results will be biased because most KDE users comes to the dot, while XFCE users got to Digg or some other so-called "neutral" sites. You have no way to know, but you assume that it is and you will claim that these results are more representative. And that's just obfuscation, in my opinion."

Are you seriously trying to suggest that it's even *plausible* that the percentage of people who use KDE will not be higher amongst readers of dot.kde.org than readers of general news sites?
I'd bet anything you'd care to wager that there's a higher percentage of KDE users here than reading digg.


By Waywocket at Thu, 2007/08/16 - 5:00am

No, what I say is that you have absolutely no idea if general news site is in anyway representative of general Linux usage as a desktop. Nor Inquirer, nor Slashdot. So limiting the advertising to these sites is also biasing the results.
In general, limiting the voters to a certain category (in this case the ones that read general news site, which is not the case of many here) will bias probably as much if not more than what this limitation is supposed to avoid.


By Richard Van Den Boom at Fri, 2007/08/17 - 5:00am

But are slashdot etc a _closer_ representative of general linux usage than dot.kde.org ? Of course they are.

I second this request to remove the link.


By John Tapsell at Fri, 2007/08/17 - 5:00am

You have no way to prove the "of course". Maybe no KDE user goes to slashdot because they come here for news. That would bias, if true, results **against** KDE. Claiming that the dot is KDE biased (I don't argue that) is not an argument to claim Slashdot is not.
Can't you see that assuming big news site are representative of anything is already biasing results much more than letting all sites advertise this Survey?
That basically says : "I say Slashdot readers are representative of the average Linux user. So I ask them what they use and then I claim the results are representative of all Linux users". You're making an assumption, extract results from it then claim the results prove your assumpion.
Talk about a snake biting its own tail.

Every site (Slashdot, Digg, Dot.KDE, Gnomewhatever, XFCEwhatever) should be allowed to advertise this. That's the closest you can get to unbiased, IMO.


By Richard Van Den Boom at Fri, 2007/08/17 - 5:00am

Well, KDE are losing to GNOME by a comfortable margin now, so I guess posting it here has had negligible effects, after all. May as well just leave it up, I guess.


By anon at Fri, 2007/08/17 - 5:00am

Well, that's due to some script kiddie manipulating the poll.
Yesterday the poll was at about 23.000 votes with openSUSE at about 9.5 %. Now SUSE votes have increased to 16.5 percent. Do your math and you'll see that by a large margin most of the votes submitted since then must be openSUSE-Gnome votes. This is highly unrealistic given the fact that according to more scientific surveys done by SUSE KDE usage among openSUSE users is at about 75%.


By Skewed at Fri, 2007/08/17 - 5:00am

Indeed, there's a significant, improbably shift after 9000 more votes quite late in the poll. Since I also had a look yesterday, I can tell you the voted configuration of the vote manipulator:
openSUSE (from ~9,5 to 16,5)
Gnome (from ~37,5 to 42,5)
Firefox or maybe none selected (from ~54/55 to 58,5)
Evolution (from 14 to 21,5)
None -- I don't run Windows apps on Linux (from ~22 to 30)

If DektopLinux is smart and has a look at their logs (and the voter was dumb), they can deduct the forged votes.
Whatever it turns out, you can trust the accuracy of these votes only in a limited way. The results are skewed not only in a way which subset of a group can be reached (e.g. emailing - used app), but also which general kind of persons find their way to such a poll.


By Phase II at Fri, 2007/08/17 - 5:00am

You make a compelling case why dot.kde, Ubuntu, Gnome and others *should* post these links.

If you have one kde user who reads about it here and goes to vote kde, that's one kde closer to the target of perfect accuracy (everybody votes, honestly).


By ben at Thu, 2007/08/16 - 5:00am

In statistics it's called Sampling Bias. To ensure the accuracy of a survey of a population, the people taking the survey must be randomly chosen from the entire population. The survey is analyzing the a population composed of all free software users, so they can't sample primarily from one subset of the population.

For example, if you wanted to determine the rates of teen sexuality, then taking the poll in a teen pregnancy centre would put a significant bias on the results.

Here's a mathematical example: Say you have a population composed of
15% XFCE users
40% KDE users
45% Gnome users

You survey the population to try and find out that distribution. A certain percentage of each group fills out the survey. (XFCE had more people know about it, because it was posted on their mailing list, but not Gnome or KDE)
80% of XFCE users
30% of KDE users
28% of gnome users

Your final analysis would result in a population distribution of:
33.8% XFCE
33.8% KDE
32.7% GNOME (Yes, there's a rounding error)

As you can see, the results of the survey are nothing like the actual population distribution.


By Soap at Fri, 2007/08/17 - 5:00am

I am a strong Gnome supporter and I read KDE dot news. The Linux Desktop is a common plattform these days.


By bert at Thu, 2007/08/16 - 5:00am

Indeed, I read the Gnome planet. BTW is there a gnome news site like the dot??? Can't find any...


By Jos Poortvliet at Thu, 2007/08/16 - 5:00am


By cm at Thu, 2007/08/16 - 5:00am

aah, yes, I found that one, but it's pretty low trafic. Not bad thing of course... At least the information is of good quality and actually about gnome, unlike many of the blogs on planet gnome.


By Jos Poortvliet at Thu, 2007/08/16 - 5:00am


By David at Fri, 2007/08/17 - 5:00am

it very interesting to read about sawfish maintainer resign.

1. he said that sawfish is finished
2. he said so because the code become hard to maintain


By nick at Fri, 2007/08/17 - 5:00am

thank you!


By Jos Poortvliet at Fri, 2007/08/17 - 5:00am

Maybe, but I'd say the majority of people that read KDE dot news would vote for KDE, which makes linking to the survey here skew the results (unless an identical link is made on a similarly-popular Gnome news site, provided there is such)


By LiquidFire at Thu, 2007/08/16 - 5:00am

Might as well give up trying to balance it. The results of this survey will be wildly inaccurate anyway, no matter what you try to do to control where the survey is posted. With internet surveys like this, you have to realize that the results for each category probably have an uncertainty of +/- 20% or so. So unless the results really show overwhelming preference towards one choice over the other, you can't come to any conclusions from the data.


By Leo S at Thu, 2007/08/16 - 5:00am

Indeed. the bias is already in who read the news at all, who are interested etc. It measures more the size of the community, I'd say.


By Jos Poortvliet at Thu, 2007/08/16 - 5:00am

You've clearly failed to prevent it being posted to *any* of the desktop-specific websites so why not go for the obvious alternative and post it to *all* of them? Actually I think that would have been a better idea from the start.


By Adrian Baugh at Thu, 2007/08/16 - 5:00am

Yes, but would there be any disadvantage when the result would be biased? Or even, does the result of the survey mattars anything at all?

I don't think so.


By kde user at Thu, 2007/08/16 - 5:00am

It clearly matters to the people who created it in the first place, and it clearly matters to JR who posted it here.


By Anonymous at Thu, 2007/08/16 - 5:00am

as a Debian user, I found out about the survey on blog.thedebianuser.org, and I subsequently selected Debian as the distro I use. Does that mean thedebianuser.org is responsible for a bias in favour of Debian in the survey results? Not in my books - I use Debian/KDE, I voted for Debian/KDE. Should I have not voted because I learned of the survey from a biased source? It seems to me if DesktopLinux.org wanted to ensure no bias, they should have sent info on the survey to every community they included in it.

It's an internet survey (with no apparent mechanism for ensuring one person = one vote, beyond IP addresses, if that) - it is never going to be a balanced representation. Putting something on the internet and then saying "oh yeah, please don't link to this page" is by far the dumbest thing I've ever heard. You'd think as website operators they might have the foggiest clue about how the internet works...


By Geoff at Fri, 2007/08/17 - 5:00am

> It's an internet survey (with no apparent mechanism for ensuring one person = one vote, beyond IP addresses, if that)

No, they do not seem to check IP addresses. They allow any number of votes per person.


By Erik at Tue, 2007/08/21 - 5:00am

just imagine, if both camp,had worked together in this time, just imagine how much effort has gone in reinventing the wheel, how much lines has been written in this stupid flame wars, each one pretending he is the real desktop Linux.
shame on you, sometimes i feel like the developers has no idea of how people use linux, there is no gnome or kde fanatics, we use both of them, i am kde users, but i use openoffice, firefox (gtk based) and openproj ( java ), as i imagine people who prefer gnome may use krita or amarok or what ever software that fulfills their needs, people don't care about the framework used, we want applications well integrated that do the job that's all.
it's a pity that we piss on ourself and forget the real challenge, windows has 95 % of market share.


By djouallah mimoune at Thu, 2007/08/16 - 5:00am

I disagree, and I think both KDE and Gnome 'camps' get it quite well, along with the Firefox, Open Office and Eclipse 'camps'. Multiple GUI toolkits/desktops are not a problem if the various projects cooperate on infra-structure, as they do via freedesktop.org and other common interest groups such as for printing or document formats.

The people who are directly contributing to the KDE and Gnome projects are happily cooperating as well as competing. I think the problem is more with the so called 'fans' of KDE who bore everyone senseless on forums like OS News with the same tired old arguments about GTK LGPL vs Qt GPL dual license, or how Mono is the devil's spawn, or how Gnome or KDE's language bindings are much easier to produce etc, etc. I think we really need some kind of FAQ where we can put the various claims and counter-claims, and then just refer to the FAQ, which would allow use to discuss less boring things. Such as whether the best colour for widgets is blue or brown..


By Richard Dale at Thu, 2007/08/16 - 5:00am

take for example khtml, did you think, that we needed the intervention of apple to make webkit which is simply a khtml without the qt dependency, why you( kde developers) did not make the engine abstract in the beginning !! so others can used it !!

as a user when i used a (gtk, java, qt whatever) application i expect when i click on open file,or print, i'll find my kde dialogue ( and this true for gnome too), is this hard to implement, is this hard to make a fd.o specification !!!


By djouallah mimoune at Thu, 2007/08/16 - 5:00am

don't bother answering Richard, everytime djouallah is posting he needs to flame on something:

http://www.googlesyndicatedsearch.com/u/dot?as_q=&as_epq=%22Also+by+djou...

by the way, thanx for adding my ruby syntax suggestions to svn :-)


By patcito at Thu, 2007/08/16 - 5:00am

Patcito, me flaming !, if you found my post stupid, just don't respond, unfortunately we are not all that smart, this is life, so what! if you are not a super c++, qt hacker, you have not the right to write about kde !!!
the dot is a free forum, AFAIK ;-)

friendly


By djouallah mimoune at Thu, 2007/08/16 - 5:00am

But the problem is that you've put 'both camp don't get it ;-)' in the title of your question (albeit with a wink), which implies that you as a 'genius pundit' who somehow knows more than the Free Desktop developers do, and that if we only followed your advice all problems would melt away.

Of course there are people are putting in real *work* to make dialogs in the various applications fit in with the desktop they're being used with - they aren't just sitting there in an armchair telling others what to do.

KHTML was developed by very few people, and the reason that it wasn't developed without Qt dependencies initially, was that there was no need for that. And the extra effort would have meant that it would probably would never got out of the door in time for the KDE release it was aimed at. If you think your judgement of such matters is better than the likes of Lars Knoll, then I'm sure the KDE project would be happy to use your services as a release manager.


By Richard Dale at Thu, 2007/08/16 - 5:00am

why you are so aggressive, i was just asking questions,is this a crime !! i did not say to people why they have to do! i did not make a rant or a judgment, simply because i don't have the skills.

for khtml, i am just sad, that apple take a credit for webkit.

"then I'm sure the KDE project would be happy to use your services as a release manager "

that's bad, i did not expect that you treat users like that.( even if you think,they are stupid)

sorry, but as a non native English speaker, my writing perhaps seems rude.


By djouallah mimoune at Thu, 2007/08/16 - 5:00am

"i was just asking questions"

Well, no you weren't just asking questions:

"don't get it"
"shame on you"
"reinventing the wheel"
"stupid flame wars"
"each one pretending"
"gnome or kde fanatics"
"it's a pity that we piss on ourself and forget the real challenge"

You think these are neutral phrases, even if you can be given some slack for not being an english speaker? I have no intention of being rude to kde users or developers, but you must be more careful with your use of loaded phrases if you prefer to get polite responses to your questions.


By Richard Dale at Thu, 2007/08/16 - 5:00am

ok, take it easy, man, just take a look at the "Biasing Survey Results" thread, no comment !!


By djouallah mimoune at Thu, 2007/08/16 - 5:00am

I think you should take it easy and think twice before using such vocabulary such as "stupid" and "pissing on ourself". And don't use the "I'm not a native English" argument please because I invented that excuse :-) as I'm not a native either so I know how fake that excuse can be. By the way, not only is Richard an amazing programmer but he's also a very nice non-aggressive person ;-)


By patcito at Fri, 2007/08/17 - 5:00am

Well to answer your question, free software devs works on their projects for fun and are not being paid for it so they tend to only work on things that are fun to them.
Now think about it:
Making KHTML independant of KDE= not fun, boring and time consuming
Making a cool HTML engine for your favorite desktop= lot of fun

see what I mean?

When was the last time you did something not fun, boring, time consuming and difficult without being paid?

Ok, I guess you now understand why the KHTML guys didn't work on that first.


By patcito at Fri, 2007/08/17 - 5:00am

that's just not true... Djouallah generally is very sensible.


By Jos Poortvliet at Thu, 2007/08/16 - 5:00am

yeah !! thanks jos, i owe you a beer, oops, i mean, a coffee ;-)


By djouallah mimoune at Thu, 2007/08/16 - 5:00am

So we would be better off if we had just one team working on just one desktop? So, answer me these questions:

a) KDE-developers prefer C++ and Qt. Would they be as good in C and GTK?

b) GNOME-developers prefer C. Would they be as good in C++ and Qt?

c) Related to A and B: what language and toolkit should this unified desktop use? How would you convert the develpers of the "other" language and toolkit in to the new language and toolkit?

d) Both KDE and GNOME have very different design-philosophies that are quite incompatible on many levels. How would you go about combining the two without alienating users and developers?

In short: Combining KDE and GNOME would be downright stupid, since the two are different, and they attract different users and developers. If they were combined, many developers would be unhappy and they would start their own desktop-project, and we would be right back where we started.

Remember: open source is about scratching an itch. KDE does not scratch the itch GNOME-developers and users have, and vice versa.

To answer some individual comments you made:

"just imagine, if both camp,had worked together in this time"

you mean something like freedesktop.org?

"just imagine how much effort has gone in reinventing the wheel"

such as?

"how much lines has been written in this stupid flame wars"

99% of the time those flamewars are fought between users, not developers.

"shame on you"

No, shame on YOU, for coming here and dictating other what they should and should not do!

"people don't care about the framework used, we want applications well integrated that do the job that's all."

So this isn't about the desktop, this is about the apps? So why are you whining about the desktops then? And why are you whining since GNOME-apps work just fine in KDE and vice versa?

"it's a pity that we piss on ourself and forget the real challenge, windows has 95 % of market share."

Since when has this been about taking away market-share from Windows? I thought this was about creating the best possible desktop-UI for computers?


By Janne at Fri, 2007/08/17 - 5:00am

Janne, i know my English is a crap, but nevertheless i never said that.( did you read my posts !!)

by working together i mean sharing the maximum of technology not writing the super unified desktop environment.
things like dcop( yes now we have dbus but after 10 years), khtml,shared mime entries could have been adopted earlier, did we needed to wait 10 years.
take for example solid, phonon( oops i am aware that people will treat me as a moron ;) api to make access to hardware and multimedia independent from the OS or the engine( gstreamer, xine), which is a great.

the problem is as the others ( gnome, xfce, etc) will not adopt those library, because it depends on QT. as a result they would probably begin from scratch and create their "real, free" libraries, or in the case of phonon are saying gstreamer is the standard and we don't need another silly abstraction layer.

can you imagine, totem using phonon, or nautilus using solid !!!


By djouallah mimoune at Sat, 2007/08/18 - 5:00am

> by working together i mean sharing the maximum of technology not writing the super unified desktop environment.

Even if it might not be that visible to people not associated with the project or maybe not even to associated non-coders, this is what is happing.

Sharing technology is not as easy as it sounds, because for most options there already exist subsystems and any new shared replacement will have to be able to replace all of them, basically offer a union of all features.

Additionally switching to a shared implementation can involve a lot of work which nobody currently has the resources for. Maintaining even a less then ideal but working solution is often less work.

Not to forget that it depends on both old and new subsystem if the switching can be done without creating compatability issues.

Unfortunately KDE seems to get a lot of critique for not immediately "adopting" some software somebody else has just uploaded to freedesktop.org (which by itself does not make it shared technology, but merely a candidate).

It is so easy to forget that quite some KDE developers or developers close to KDE have refrained from adding Qt dependencies to their respective project, resulting in additional work, because at that time somebody from outside KDE indicated interest. In the end this additional efforts have almost never lead to adoption by other projects, probably because people just want to assume without checking that anything originating from KDE is somehow KDE dependent.

We have seen this reaction for things like aRTs, DCOP and others, we are starting to see it for things like Strigi.

On another front, KDE gets critique for actually using shared techology.
As you mentioned it yourself, a good example is Solid. Why do people assume that application developers should be doing low-level D-Bus calls to HAL instead of using a nicely designed API?

Not even the developers of the HAL project suggest one should be doing it, they even maintain and ship such a convenience API themselves. I am sorry but I fail to see why abstracting the D-Bus calls into a C-API would be OK and abtracting it into a C++-API wouldn't.


By Kevin Krammer at Sat, 2007/08/18 - 5:00am

Pages