Mint 11 KDE development

[GDTR.ua]

For those who don't like migration to LMDE base

Ubuntu (Canonical) is focused on building their own money-making platform (say hello to Steve).
We can see the direction right now: IMO, much more "ubuntu-only" "ubuntu-proprietary" moves to come in future.
OSS community definitely can live without that sh**

IMO, LMDE will become a base for all editions in future. Its a matter of time.
The best is yet to come (c)

[JoeFootball]

Time for me to chime in. And from what I've read, one thing is for certain. We are all quite passionate about a distribution that we love.

Like others have mentioned, I personally would much rather prefer this be two separate versions of Linux Mint KDE. That said, I know that Boo is just one man without all the time in the world given his other commitments (e.g., a job), and while I'm sure he would love to accommodate everyone's needs & wants, he had to make a tough call here. Knowingly release a really buggy Linux Mint 11 KDE, or consider this an unexpected invitation to move to a Debian base with its own related consequences.

I don't like the way Canonical is heading, but I do want a Linux Mint KDE that's going to "just work" like I've known it to. In this thread alone, there have been a lot of great ideas stated for how this could work, and a lot of very good points on why this may not be a good idea. But I know that Clem, Ikey, and others are going to give Boo a lot of support on the Debian-base front. And lots has been ironed out on the LMDE side already. And there's the new update infrastructure for LMDE too. Hmmm...

All of this considered, I'm going to stay positive and optimistically welcome this change. I know things will be different. There will be bumps in the road, especially in the beginning, but I also know that this isn't an insignificant journey. And I'm sure the LM dev team knows this too.

I don't want a poorly performing Linux Mint 11 KDE, and a Debian base makes sense for the future of Linux Mint KDE. So, I guess the future is now, just here a bit sooner than expected.

This is going to be interesting. Let's find out.

Joe

[asymmetros]

In LMDE editions (Gnome, Xfce) mp3 files are playing out of the box.
For graphic drivers, the user has to install the packages from synaptic. There are some tutorials in community 's specific section. In any case, anyone is happy to help the others.
Regarding upgrades, a new upgrading system is already available http://blog.linuxmint.com/?p=1781

I think that moving to Debian direction, is a very practical move. Boo and the guys are spending valuable time trying to resolve upstream problems. Instead of this, they can this time in working with LMDE editions and adding new features (like "automatic" drivers installation )

[fraxinus_63]

I love Mint in all its forms and I am very interested in this discussion.

I am currently using Debian Stable (64 bit and Xfce) with backports and I am very satisfied with it. The problem with Testing is that at the beginning of the production cycle (right now) it can occasionally break (towards the end it tends to be rock-solid, though). This is not a huge issue for the experienced user but it can be a show-stopper for newbies and clearly it does not conciliate well with the original Mint concept. What I would propose for Mint would be Debian Stable with official backports plus extended backports from the Mint repos.

Am inclined to agree with this suggestion. I fully understand the possible benefits of moving to a Debian base but please let us not sacrifice the ease of maintenance that Mint has always offered - one of the reasons why I recommend it to many non-technical friends and family members.

[jlacroix82]

I'm interested in this discussion as well. I've always been a fan of Mint KDE, though I've never actually used it full time on any of my computers. Reason being, it's unpredictable release pattern. If Mint KDE switches to Debian, that would make the case for me to use it all the time.

Currently, I use Debian on my home server, Arch on my desktop, and Kubuntu on my laptop. I used to use Arch on my laptop too, but it seems like some component or another is constantly broken when updated, so I'm using Kubuntu for now. Other than that, I prefer the rolling release methodology, and honestly, I feel strongly that rolling releases are the future of Linux (though perhaps not as bleeding edge as Arch).

As for Debian, I don't use it full time on my laptop or desktop because I've noticed that pretty much every package is at least two or three versions behind. I think they're still using kernel 2.6.32 if I'm not mistaken but I haven't checked in a while. If Mint switches the KDE version to Debian, I'd be fearful that everything would just fall very far behind. As someone else said, I agree that Mint backported packages would probably fix this.

As a KDE user, having the latest version of KDE and related KDE software is important to me. Also, having the latest kernel is usually crucial because newer chipsets won't work without the latest kernel in most cases (I've found this out the hard way). In my opinion, going with the first maintenance release of each major kernel version is probably best, since initial kernel releases usually have issues from what I've seen (for example, 2.6.39 made X unusable for me, but 2.6.39.1 solved it). I also hope to have the latest Firefox and Thunderbird, but that is just a want, not a need.

I guess my point is that I feel that having a balance between a rollling release and stability is key. I don't really care if all the core components (samba, etc) are the latest and greatest, as long as I have the latest kernel, KDE and web browser. I'm very interested to see how Mint decides to handle this, and the decision could possibly win me over since I thought the last few Mint KDE versions have been stellar.

[dolphin]

Go for LMDE KDE. Some arguments:
1) Linux Mint 10 KDE has a year of support yet.
2) There is no LMDE KDE yet. Simple argument but important.
3) I installed kubuntu-desktop on top of Katya and it works nice, so anybody can do the same on her/his own.
4) For Ubuntu based Mint consider Xfce - maybe not for Katya but based on 11.10 . I would keep xorg instead of experimenting with wayland (at least until Ubuntu can prove it is a stable base for DE). Or even consider Gnome 3 with Shell (only if you find it ready), just as an alternative to Unity. This would definitely make Mint distinct from Ubuntu and attract new users to Mint. Anyway, you can wait with KDE flavor based on Ubuntu until 11.10. I think nobody will get hurt by this.

Whatever you decide, keep it stable. This will attract users who don't like to be an object of experiment.

My general thought is that Mint developers could make Mint even more Minty , I mean more stand alone distro than just a fork of Ubuntu. Which is already happening and it is good. So, you don't need release all DE flavors every half a year. Take your time and develop Mint in a minty way.
Although I still see some advantages of releasing Ubuntu based flavours, for example a lot of repositories in PPA. A lot of software developers release deb packages for Ubuntu.

I like the fact that Mint developers just ask community about our thought instead of just imposing whatever... I really appreciate it.

By the way, I don't understand this ongoing discussion about default applications as if it was something of primary importance. Whatever they are I always install, uninstall something, so no need to bother about it, really.

[asymmetros]

As for Debian, I don't use it full time on my laptop or desktop because I've noticed that pretty much every package is at least two or three versions behind. I think they're still using kernel 2.6.32 if I'm not mistaken but I haven't checked in a while.

LMDE uses 2.6.39

I also hope to have the latest Firefox and Thunderbird, but that is just a want, not a need.

It has FIrefox 5, Thunderbird 3.1.7


I guess they will go for KDE 4.6.5 but we will see.

For Ubuntu based Mint consider Xfce

Since March, Mint Xfce is Debian based.

Whatever they are I always install, uninstall something, so no need to bother about it, really.

That's my tactic too

[LifeInTheGrey]

As for Debian, I don't use it full time on my laptop or desktop because I've noticed that pretty much every package is at least two or three versions behind. I think they're still using kernel 2.6.32 if I'm not mistaken but I haven't checked in a while. If Mint switches the KDE version to Debian, I'd be fearful that everything would just fall very far behind. As someone else said, I agree that Mint backported packages would probably fix this.

You are somewhat right, however the kernel and packages you refer to are in Debian STABLE (currently Squeeze) ... that is updated every 2 years and changes little, so packages and kernels can fall way behind. Luckily, LMDE is based on Debian TESTING (currently Wheezy), which receives updates regularly that are now retested for stability by the new LMDE repo prior to being released for consumption (stable updates that occur frequently rather than in 6-month blocks). To show a couple of examples:

Debian Squeeze: 2.6.32-5 kernel
LMDE: 2.6.39-3 kernel

Debian Squeeze: Iceweasel 3.7
LMDE: Firefox 5.0.1

Debian Squeeze: Openoffice 3.2.1-11
LMDE: LibreOffice 3.3.3-4

As you can see, much more up-to-date than you imagined.

[jlacroix82]

What about KDE? What is Debian testing using as of right now?

This is really useful information, thank you everyone.

[cpatrick08]

What about KDE? What is Debian testing using as of right now?

This is really useful information, thank you everyone.

it is at 4.6 http://pkg-kde.alioth.debian.org/

[jlacroix82]

What about KDE? What is Debian testing using as of right now?

This is really useful information, thank you everyone.

it is at 4.6 http://pkg-kde.alioth.debian.org/

I'm starting to wonder if I should take another look at Debian. Is it 4.6.5? The page only referenced 4.6.3 but that was last month.

[MALsPa]

I'm starting to wonder if I should take another look at Debian. Is it 4.6.5? The page only referenced 4.6.3 but that was last month.

I think they must still be at 4.6.3 and working on 4.6.4. Check this page out:

http://packages.debian.org/testing/kde/

[jlacroix82]

Thank you for all your help. If Mint KDE switches to Debian, I will definitely check it out. Actually, I'll check it out either way but would be especially excited for a Debian base.

[fraxinus_63]

Whatever you decide, keep it stable. This will attract users who don't like to be an object of experiment.

Well expressed!

As mentioned by several posters here, it is great that this distro actually consults its user base before making significant changes to the user experience. Much appreciated, and very best wishes to all involved in taking things forward.

[rec9140]

Whatever you decide, keep it stable. This will attract users who don't like to be an object of experiment.

Well expressed!

That would be a good route to go...

Either :

1) Skip a full blown Kataya release, and create a ALPHA KDebMint for testing...

2) Create BOTH a STABLE/NORMAL Kataya KMint release, even if its later than normal due to work, life, and clean up of upstream, and a KDebMint ALPHA.

I think some of the Debian cheerleaders would soon find out its not so great in DebianLand, software wise... ANCIENT VERSIONS, lack of features in some programs, no builds in repos, source only, etc.. Debian is a great base... Canoncial made a good choice there for the base, the wheels started to fall off when they made choices on DE's, etc... but their inclusion of full featured and CURRENT versions of software in the repos makes the Linux experience 1000% better than a pure Debian. I am all for a KDebMint, but when its ready and MINTY! And thats going to take a LOT of WORK that I don't think Mint is ready for, yet.

As mentioned by several posters here, it is great that this distro actually consults its user base before making significant changes to the user experience. Much appreciated, and very best wishes to all involved in taking things forward.

I think your mistaken here, this is one of the biggest issues with Mint, community INTERACTION ie: involvement from decisions to testing to.... what ever... and I don't see any "consulting" of the user base to see what direction to go...especially on this change to Debian.. it was announced and that was the end of it. I don't call that consulting.

[mastablasta]

Well older programmes are not an issue if their bugs are getting patched. if not, then they are indeed useless.

Updates that break the OS are not good for sponsors/business environment. imagine having a bit larger system - loosing a whole day while you troubleshoot and try to solve th eissue can cost a lot of money.

for small users - if they are not really proficient in Linux it may take even longer and cost them their business (cost them much larger share of money). stability is one fo the key components of the OS. i won't use an OS for which i need to spend every weekend doing repairs to it, as it happened on one of my computers with Ubuntu 10.04. If i couldn't fix it with an upgrade i would have to get windows for it to get things done. luckliy 10.10 solved it.

[gosa]

I think your mistaken here, this is one of the biggest issues with Mint, community INTERACTION ie: involvement from decisions to testing to.... what ever... and I don't see any "consulting" of the user base to see what direction to go...especially on this change to Debian.. it was announced and that was the end of it. I don't call that consulting.

And this one is not one of the attempts at consulting the userbase?
http://blog.linuxmint.com/?p=1686

Just asking...

[rec9140]

Well older programmes are not an issue if their bugs are getting patched. if not, then they are indeed useless.

Well lets look at KDE to start, its way behind in Debian. 4.7. is about to be released, and any KMint or KDebMint released should wait till then, even if you wait till the .1 to clear up bugs. Too much looking at the calendar and its april or october...

As for other programs, some times older programs mean, option/feature you need to be of use and not being available because the version is not available without compiling. And for probably 80%+ of the *Mint community that would be an issue. I've had my share of compiling issues for bog basic programs which should, ./configure, make, sudo make install, and blow up, even with all dependencies, etc... thats one of the reasons for extreme vetting procedures on my part... test machines (physical) and test VM's.... nothing gets installed till I've run it through there. I need to know its going to install and not break, or compile and not break stuff...be it personal or agency.

Updates that break the OS are not good for sponsors/business environment. imagine having a bit larger system - loosing a whole day while you troubleshoot and try to solve the issue can cost a lot of money.

for small users - if they .....

Both of these are THE POINT and SPOT ON, ... look at the carnage going on in the DebMint... I can't have that on agency computers, let alone on my own personal ones. X and OEM drivers must work, period. And they need to install with out spending 2 months, a witch doctor, a chamin, offerings to Ra, etc.. Linux in 2011 is way better that this, the fact that the Debian maintainers see this as acceptable or allow it as acceptable is just another reason why Debian based is the way to go.. Canonical for all their idiotic decisions has made Linux useable from newb to enterprise.... yeah.. theres tons of things Debian makes choices on and the DFSG that I don't agree with...same with Canonical... but they clean up and package Debian to where its ready to be used, then Mint comes along and cleans and polishes it even more to create perfection... A plain Debian based Mint is not going to match that, with out a lot of heavy lifting, which I don't see the Mint development staff or resources ready to handle.

Any *DebMint needs to handle all these issues to be "Minty" otherwise it will be tarr'd and featherd just like a current release from upstream.

[LifeInTheGrey]

Well lets look at KDE to start, its way behind in Debian. 4.7. is about to be released, and any KMint or KDebMint released should wait till then, even if you wait till the .1 to clear up bugs. Too much looking at the calendar and its april or october...

So either wait for 4.7.1 to be released to release an operating system overall, or work with the incredibly stable 4.6.3 (.4 pretty soon) until 4.7.1 hits the testing repos? Sounds like you're waiting either way ...

As for other programs, some times older programs mean, option/feature you need to be of use and not being available because the version is not available without compiling. And for probably 80%+ of the *Mint community that would be an issue. I've had my share of compiling issues for bog basic programs which should, ./configure, make, sudo make install, and blow up, even with all dependencies, etc... thats one of the reasons for extreme vetting procedures on my part... test machines (physical) and test VM's.... nothing gets installed till I've run it through there. I need to know its going to install and not break, or compile and not break stuff...be it personal or agency.

Compiling IS scary as hell. Thank god as an LMDE user I've never had to use it ... that would drive me away. I just stick to apt-get and the Software Manager.

X and OEM drivers must work, period. And they need to install with out spending 2 months, a witch doctor, a chamin, offerings to Ra, etc..

I actually keep a witch doctor on retainer to get my OEM drivers to work. But then that's something I've done since Ubuntu 9.04 ... anyone remember installing the Broadcom STA drivers on that system? Awful!

Linux in 2011 is way better that this, the fact that the Debian maintainers see this as acceptable or allow it as acceptable is just another reason why Debian based is the way to go.. Canonical for all their idiotic decisions has made Linux useable from newb to enterprise.... yeah.. theres tons of things Debian makes choices on and the DFSG that I don't agree with...same with Canonical... but they clean up and package Debian to where its ready to be used, then Mint comes along and cleans and polishes it even more to create perfection... A plain Debian based Mint is not going to match that, with out a lot of heavy lifting, which I don't see the Mint development staff or resources ready to handle.

Any *DebMint needs to handle all these issues to be "Minty" otherwise it will be tarr'd and featherd just like a current release from upstream.

Ok sarcasm aside, it's obvious you haven't even tried LMDE yet. I've tried vanilla Debian ... and really, you're spot on about the analysis there. LMDE actually is DRASTICALLY better and more user-friendly. The fact that its based on Debian TESTING rather than Squeeze means much more frequent application updates (pretty much on par with Ubuntu), and as a guy who started his Linux voyage with Kubuntu ... it really isn't that much of a leap. You assume the heavy lifting won't be done by the Mint developers ... it already is being done man. You think Ubuntu was originally constructed from magic beans and unicorn farts to become this ubiquitous user-friendly system? Coders worked tirelessly on a DEBIAN base to improve hardware support and package management, and Ubuntu was born. Mint is just following those same footsteps, and they have the talent to do it right.

Making assumptions from fear and lack of knowledge on a topic rarely breeds good decisions. I suggest you try before you decry.

[rec9140]

So either wait for 4.7.1 to be released to release an operating system overall, or work with the incredibly stable 4.6.3 (.4 pretty soon) until 4.7.1 hits the testing repos? Sounds like you're waiting either way ...

At this point in the release cycle may as well wait for 4.7.0 to be released. Thats just another reason to get OFF this April/October calendar watch... so upstream released the base in April, if it takes KMint Dev (aka Boo) 1 month to clean up the garbage in, then so be it. If at that point KDE is ready to release an upgrade then we should hold for that release, that was a good call in the past to do so.. it put KMint ahead of the upstream base. I would rather get ONE good release a year versus two not so good ones.

I am in no rush to get Kataya, unlike some others. Its clear that work and life have priorities for KMint Dev (aka Boo) and thats A-OK with me, so take care of those first, then put the time into Kataya. I would rather see the work take priority and get a current KDE release in Kataya, and then possible a KDebMint testing release to see where things would stand.

Compiling IS scary as hell. Thank god as an LMDE user I've never had to use it ... that would drive me away. I just stick to apt-get and the Software Manager.

I have no issue with compiling, where the wheels fall of the bus is developers who do not document EXACTLY the dependencies for their software, a recent escpade with KMyPhoto or what ever its called, showed that the dendencies listed on their web site, were not all that was needed, and what certain pacakger(s) think was needed was not needed at all, made it far less than ./configure, make, sudo make install. This in LINUX OVERALL is the big let down, which will/does hinder its wider spread adpotion..

I actually keep a witch doctor on retainer to get my OEM drivers to work. But then that's something I've done since Ubuntu 9.04 ... anyone remember installing the Broadcom STA drivers on that system? Awful!

And thats where the OEM drivers stand on Debian, awful. Its 2011, not 1998. And don't even get me started on the the whole community driver(s) thing... thats a tirade for another day. I am forced to use that c r a p because ati .... URRRRRRRRRRRRRRGGGGGGGGGGGHHHHHHHHHHH!

Ok sarcasm aside, it's obvious you haven't even tried LMDE yet. I've tried vanilla Debian ... and really, you're spot on about the analysis there. LMDE actually is DRASTICALLY better and more user-friendly. The fact that its based on Debian TESTING rather than Squeeze means much more frequent application updates (pretty much on par with Ubuntu), and as a guy who started his Linux voyage with Kubuntu ... it really isn't that much of a leap. You assume the heavy lifting won't be done by the Mint developers ... it already is being done man. You think Ubuntu was originally constructed from magic beans and unicorn farts to become this ubiquitous user-friendly system? Coders worked tirelessly on a DEBIAN base to improve hardware support and package management, and Ubuntu was born. Mint is just following those same footsteps, and they have the talent to do it right.

Making assumptions from fear and lack of knowledge on a topic rarely breeds good decisions. I suggest you try before you decry.

There you would be WRONG. Tried it... but it had an huge uphill battle to start from its stinky foot DE environment, to lack of OEM drivers, to the nightmare updates. It didn't last long on the test machine...

when I went:

sudo apt-get install darkice, and then gave it my standard config, and it failed, just as I expected, not compiled with MP3 support. FAIL.

Disk wiped for the next test distro....

Mint as far as the DebMint version is doing SOME of the heavy lifiting, codec support, etc... BUT its NOT DOING the AMOUNT that is REQUIRED. what canoncial is doing for xMint right now, in repos, PPA's, the xswat PPA alone for OEM drivers. Just compare progam support in repos either from whats there to the versions that are there, ie incomplete due missing features, older versions. etc... Is Mint ready to RECOMPILE 100's, 1000's or more programs and provide the repos for them ? ? This alone is what makes Canoncials Debian versions worth while, lets dump in the whole binary blob firmware debacle... The developer and maintainers at Debian need to quit drinking the FSF FlavorAde and get back on track. The FSF and RMS are not helping Linux as much as they think!. And, NO putting them in repos or on an alternate CD/DVD does NOT COUNT, period. Newbs will have no idea why their wireless card or something may not work, just Linux failed to allow me to connect to the internet, and they run back to whence they came. The no net and no drivers issue to resolve is a big issue for newbs to resolve, especially when BOTH your wired and wireless don't work. Nor do most users care.. they want their hardware to work, period. Only the RMS zealots of Linux have their panties in a wad over OEM v. open source drivers... thanks I will take the OEM nVidia drivers, any day.

The standard 1A response when software is on the chopping block, is well just install it... well it won't be so simple to install on KDebMint, since it may not exist in the repos, or what does is outdate, feature impaired or what not. So sudo apt-get install xyz may not be so simlpe a solution. This then turns things in to a big ordeal... Is there a DEB? Are the dendencies met|meetable, or does this send the user to Compileville ? ?

Face it, Canonocial is providing a HUGE AMOUNT of upgrdes to software, v. Debian, that the Mint community benefits from in ease of use of installing software. *DebMints will have to take this up on their own. No pointing to Debians repos, just is not the same level of experience from version to feature to.... others have pointed this out as well. Thats a big investment in resources, servers & bandwidth, developers to recompile, repackage etc... Where will this come from ? The community? Who? And how? Just like the adage goes, No Bucks, no Buck Rogers! Same here.. some one has to pony up the $$$, £, € ,¥ and we don't have Shuttleworth to bank roll things... With out Canoncial there would have never been *Mint, period. And *Mint is riding the coat tails of Canoncial... now based on the direction Canonocial is headed... some of the free ride may soon be over for some versions.. how it effects KMint's base is open for interperation.. but some of the stupid decisions are being foisted upon it, so KDM, if LightDM proves unworthy, has XDMCP so it gets 1.5 stars to start, will have to be reinstaled for any future KMints past 11.10 upstream.

It means that a LOT MORE WORK is going to have to be done by Mint itself on things..