Open Ballot: should KDE and Qt libraries merge?

TuxRadar

"Let's merge Qt and the KDE development platform. Let's put all KDE libraries, support libraries, and platform modules into Qt." So says Cornelius Schumacher, long-time KDE coder and the current president of KDE e.V. Such a bold move would be a "massive effort and require huge changes", he says - and the community would have plenty to talk about as well. See here for the full story, and then let us know what you think for our next podcast. Is this a good move to simplify the Linux desktop stack and eliminate redundancy, or are the projects simply far too separate and distinct?

Oh, and please give yourself a slightly more imaginative name than Anonymous Penguin if you want your comment to be read out. Don't just do what the man tells you to.

You should follow us on Identi.ca or Twitter


Your comments

Less is More

I think Linux should - in some cases - decide that less is more.
If we mix the power of 2 systems into 1, we have one powerful library, which is a good case (in my opinion of course).

I must however say that this is the opinion of someone without inside know-how so this might be just another worthless comment, but I just felt like saying that linux doesn't need 1000 options, but we need just a few GOOD options :)

Reduce Duplication

If kdelibs and QT would merge it would reduce a lot of duplication. Also it would mean that mainstream developers would learn kdelibs and not have to relearn stuff then they want to contribute to KDE which is awesome.

I find this a very good

I find this a very good move. Many times we have talked about consolidation in Linux and Merging of Qt / kde libraries is a good step towards achieving that. This if carried through would eliminate many duplicated effort and also make life easier for developers. It would also allow kde to tap into and benefit from the broader Qt ecosystem and also enjoy first hand huge development going into Qt. It would not be easy though and in many cases it would break with compatibility. But in the end I think the result would be worth the effort.

Sounds Promising

If it can be done and done properly without causing a horrendous mess of KDE in the process I think it would be excellent. As long as we can trust Nokia not to get bored with QT at some point.. they still seem to have a disturbing fondness for the junk that is Symbian. As someone said, lets not have a repeat of OOo, if Nokia really made it a foundation based approach I think both KDE & QT could really profit.

Given that Gnome and Canonical are having a bit of a Donnybrook at the moment it could be concievable they might use QT if it was seen as less of a KDE thing and the protection of a foundation was in place too.

While some would think choice is great it would be nice to not have 3 differing and entirly redundant implementations of the same stuff.

What to do?

What will KDE do concerning licensing? Doesn't this move mean that KDE source has to be donated to Nokia? For the application developer using KDE development libs this idea is great, but what about the KDE core developers?! I guess I only have questions rather than a real opinion.

Go for it!

And while you are at it, merge all the other libraries out there as well ;).

I don't know if I agree with the statement:
"If we mix the power of 2 systems into 1, we have one powerful library", it could just mean you get one monolithic bloated library.

No

My thoughts:

Nokia bought QT for its cross-platform use for smartphone applications.

KDE is DESKTOP software (although it is gaining some netbook and smartphone spin-offs) and this was the original starting point of the project.

Nokia is unlikely to want to support the Desktop classes long term even though there is short term benefit. Why would they? It is simply not their business. It is pointless work. Of course I have not read the article yet! (I am a Slashdot reader at heart)

As a long term KDE DESKTOP user who has a functioning desktop again at last (just) the idea of huge amounts of developer time poured into some sideways work like this is deeply annoying.

I would like a fully functioning KDE desktop, with search, multiple monitor support(!), a working email client, a Koffice I could replace OpenOffice with etc. I recall I had one in KDE 3.5 (sans search!).

A plea to KDE devs - don't mess with libraries AGAIN - deliver the best experience for DESKTOP users!!! PLEASE!!!!

hum

Qt = awesome, KDE = awesome. Awesome + awesome = probably less orgasmic then I'd like but still pretty sweet. What I assume would come out of it is more people would improve the libraries, at least from an KDE viewpoint, Qt will, with out a doubt, become more popular by its own right, with or without KDE.
I guess it's pretty much a piggy back ride on what I assume is the same assumption I have about Qt's future.

God please no

KDE just only got back to a decent state after two years in the wilderness, another big migration would alienate even more users and app developers.

And to gain what, exactly? Support from a company that *already* is the main sponsor of the project as it is, and which may or may not keep supporting it in the future? Should MeeGo tank in 2011, I bet Nokia will run away from the Linux world as fast as they can.

This is a terrible, terrible idea. Scarce development resources should be dedicated to more productive things (like fixing multi-monitor support, helping the poor PIM suite, improving usability -- you can't switch off bluetooth! --, etc etc).

In favour

KDE is all about integration... would welcome this move.

You're asking US?!

Like we would know.

The reason to do it is spelled: Meego

I am starting to think that Meego might be successful, at the very least MeegoIVI. One of the core technologies in Meego is spelled Qt. So if the KDE libs merges with Qt, it would potentially give the project access to a lot of new and talented developers.

So, in 5 years, Canonical will finally have killed of Gnome. The Linux Foundation, Nokia, Intel, BMW and the others in the GENIVI Alliance will have made Meego the greatest embedded Linux distribution and in the process KDE will have become the default desktop environment, again. All hail our new Qt ("cute"?) overlords!

Nooooooooooo!!!

I feel like I am watching a horrible accident progressing in slow motion, yelling "Noooooooo!!!" from the top of my lungs, but powerless to effect the outcome in any way.

KDE 4 is almost as usable as KDE 3 has been, just another point release or two, and I will be home again. This is the worst time to break it horribly again.

And as if we didn't learn anything form Oracle:

Step... away... from... the... large... corporation.
Slowly... slowly...

Now RUN!

YES

kdelibs is in real need of cleanup, and using QT interfaces wherever possible and adding to ones that don't yet exist will really help KDE be more portable, and make QT's base libraries more usable.

You're insane. But you could also be... a genius!

Couldn't combining the two provide a powerful relationship between Desktop and Mobile device? Couldn't such a project's focus bring the two together in a similar way to iOS and OSX?

Think about that and the possibilities for Linux in general! Only an idiot could find that a negative thing.

Advantages / Disadvantages ?

I'd like to see a dedicated page containing a list of the advantages and disadvantages (from a technical point of view) involved by this move. Are the benefits so big to deserve the effort?

Merge some things it makes sense to merge

That's it; I can't imagine merging everything will end well, but merging some of it will make it easier, I think.

Eh?

What on EARTH does all this mean?!

maybe...

maybe not merge them both into one big monolithic blob, unless it could be done cleanly, without much disruption. but they could certainly benefit form some form of co-operation/code merging in places

As long as it works

All I ever wanted from my computer is that it should do the things I want when I need it. As long as it works that's it! KISS.

if

Cornelius Schumacher thinks its a good idea then it probably is?
I'll vote for the man with the most experience, say yes and back away..

Just a clarification

Everything isn't intended to merge, just the parts where the Qt and KDE libs overlap, both KDE mailing lists regarding to the topic is quiet clear in the fact that it should be done right if done at all and that the absolute core functions that are to closely tied to KDE will not be merged.

It's very hard to decide. On

It's very hard to decide. On the one hand, if they merge, kde will have more backing and be unified, avoiding duplication. On the other hand, could we be seeing 'Librekde' sooner or later after this happens? Do Nokia really care about desktop development? I think they would starp it to symbian or something ridiculous.

However I would say that it would probably be the best thing for the kde desktop at the moment.

There's no Kia answer to this one...

As a lot of development work of QT is done by nokia we don't want to ride this horse into meego cul-de-sac of death!

having them separate means that KDE can go their own way to a point, and dare i say it fork QT if nokia (mee)go too wayward.

merging package managers would be time well spent, not this!

I'd bet my house on Sadken before this idea!

zingyyellow

convenience

yes, because then tuxradar will be able to kill 2 birds with one stone and, instead of slagging off the two projects individually, they will be able to slag them both off simultaneously!

This is difficult to know outside the community.

Cornelius does have a point.
It is just stupid and a waste of resourses to have overlapping libraries that are doing the same things.
It might also be so that the qt developers have more time and are more actively developing some of the libraries, in this case it makes a lot of sense.
I read some consern about Nokia and future need of forking but I do not understand how that would be easier with overlapping libraries, my guess is that in worst case itwould be equally difficult.

Cautiously in agreement

Like others have said above, if there's parts of the KDE libraries that duplicate Qt then it makes a whole deal of sense to merge these two. Especially if they use the Darwinian "most suited succeeds" method - so you get the best of the two alternatives each time.

imho the resulting KDE-Qt would be better than plain KDE because they'd then be able to access whatever high-powered talent Qt currently has in it's development team.

Oh, and to all those folks busy slagging off Meego - my advice would be to try it first! I've been testing 1.1 on my Acer netbook v's Ubuntu 10.10UNR (dual-boot) and Meego is just so much slicker than 10.10. So if that's an example of a Qt-based system, then getting KDE to use it would be a very good idea to me.

Disclaimer: I use Gnome on my desktops, so I've not seen KDE for a good while.

What he heck are you all taking about.

KDE QT reduction of duplication is already under way.

Massive effort only if they rush the process. If they spreed it out over 5 to 6 years no one will really notice the change.

Normal API deprecation.

Oh Puh-Leeze No!

Going from KDE3 to KDE4 was a marketing disaster. Major changes broke stuff big time. KDE4 still has an image problem because of this. And now some devs want to break everything yet again?

What person looking at KDE from the outside would ever consider using KDE if it earns the reputation of being in a constant state of transition?

I don't want my house or office to be a permanent construction site. Do you?

Break it, fix it until it's almost there, break it again. Please don't kill your brand that way! (The last big rebranding wasn't that long ago. Time for another one?)

Don't get me wrong, I love my KDE4 Plasma Desktop. I just want to grow with it some more before the next major break.

I'm the Anonymous Penguin

I'm the Anonymous Penguin from above, just forgot to change the name...

Well at least I didn't do what the man told me to :D

Please dont ask idiots online about things they know little abou

Listening to mouth breathers who know computers like I know japanese is tiring because the same crap gets rehashed that other idiots then repeat.

The "you got rid of my favorite 3.5 functions" is getting old.
Does anyone with even a remote understanding really think that the switch to v4 was done because they had nothing else to do? I agree that the message about 4.0 was missed by both those who know little and those morons who know a lot and STILL decided to ship v4.0 even though IT WAS WRITTEN THAT IT WASNT READY.
But that doesnt change the fact that staying with 3.5 wasnt an option if we wanted the desktop to move forward.
We were using OS8 and 9 at work (probably the two worse OS ever written) and they made a clean break.
Sometimes you have to do it.....

THIS IS NOT THE SAME THING....
I repeat.....
THIS IS NOT THE SAME THING....

I going to paste Nai's post above called
'Just a clarification' because many dont understand whats being talked about:
------
Everything isn't intended to merge, just the parts where the Qt and KDE libs overlap, both KDE mailing lists regarding to the topic is quiet clear in the fact that it should be done right if done at all and that the absolute core functions that are to closely tied to KDE will not be merged.
-----

Ok, have we got this?
>>>>>just the parts where the Qt and KDE libs overlap<<<

As a developer, I feel I dont have all the information necessary to make a decision. I know polls are fun but it youre clueless, its useless.
Im going to read up more and find out the pros and cons, the risks involved, the breakage involved and a host of other things.
If the reward is worth it after getting all the info, then I might vote yes because there are some very fine developers working on KDE who know more about this field than I do.

The move to KDE4.x was the right one. No matter what JoeBlow thinks. The technical reasons were sound and the benefits for both devs and users will be felt for years to come. Status quo/stagnation is always the easiest road.

Same with this, if the technical reasons are overwhelming, there still has to be a better job with the implementation and communication. It seems that telling people not to use something isnt good enough.

Independence must be preserved

As someone who believes in keeping some unity within Linux, it is important for development libraries, etc. to be independent of the desktop. If I came to power, I would have GTK+ and Qt living co-operatively with one another, independent of any environment. With so many desktop environments, it is vital to keep compatibility between them. Keep both GTK and Qt separate, make them both compatible is GNOME/KDE/XFCE/LXDE and watch both them, and Linux in general, thrive!

mmmm

No.

I don't know but!

NO fuck KDE, just die allready!

monolithic == MS != Linux

monolithic -> maintenance_complexity -> bugs

complexed != better
bigger != better

With the changes that may occur in development teams, the more monolithic the system, the harder it is to retain touch with it. So i would suggest 'no'.

jm

That's my comment above

That's my comment above ...

The login got screwed, i changed it to the same one that was being rejected --- wierd.

jm

more on monolithic == MS != Linux

To clarify a bit more.

Johnvile - wrote:

[quote]
>>
Cornelius Schumacher thinks its a good idea then it probably is?
I'll vote for the man with the most experience, say yes and back away..
>>
[/quote]

A bit of a silly comment really. Why not just go for win7 and be done with it then.

nai wrote:

[quote]
>>
Everything isn't intended to merge, just the parts where the Qt and KDE libs overlap,
>>
[/quote]

If there is duplication between libraries then isn't that saying there isn't adequate communication occurring between the development teams. Or that some structure to facilitate that communication is needed. The whole idea of modularity is that the unit is easier to maintain. As long as all the units agree on a form of communication so they may interface with each other. It can be correctly thought of as a paradigm issue. One that compares the already experienced disasters of the MS monolith concept with that of the Unix-like, modular approach.

In short, just check with the other library developments to see if that bit is already intended. And possibly work with them if one thinks their version of something may be an improvement over the others.

That does sound a bit like open sourcing though, in a way --- is that anti-nokia ?, or something ...

Easier said than done, but once set up and put into practice, all those things do become easier. Like the other lots office elevator install music never was.

jm

Somebody stole my air guitar,
but that's ok, 'caurse i've got a spare.

No

No

NO

This made me facepalm, it's absurd.
Why on Earth should they merge things that are not always going to be used together?
Qt applications are not always used in KDE.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

Post new comment

CAPTCHA
We can't accept links (unless you obfuscate them). You also need to negotiate the following CAPTCHA...

Username:   Password:
Create Account | About TuxRadar