Beta
×

Welcome to the Slashdot Beta site -- learn more here. Use the link in the footer or click here to return to the Classic version of Slashdot.

Thank you!

Before you choose to head back to the Classic look of the site, we'd appreciate it if you share your thoughts on the Beta; your feedback is what drives our ongoing development.

Beta is different and we value you taking the time to try it out. Please take a look at the changes we've made in Beta and  learn more about it. Thanks for reading, and for making the site better!

The Road To KDE Frameworks 5 and Plasma 2

Unknown Lamer posted about 2 years ago | from the greenspun's-tenth-rule dept.

KDE 89

jrepin wrote in with a link to an article about the steps being taken toward a Qt5 based KDE 5 and Plasma Workspaces 2. From the article: "KDE's Next Generation user interfaces will run on top of Qt5. On Linux, they will run atop Wayland or Xorg as display server. The user interfaces move away from widget-based X11 rendering to OpenGL. Monolithic libraries are being split up, inter-dependencies removed, and portability and dependencies cut by stronger modularization. For users, this means higher quality graphics, more organic user interfaces and availability of applications on a wider range of devices. Developers will find an extensive archive of high-quality, libraries and solutions on top of Qt."

Sorry! There are no comments related to the filter you selected.

Frameworks (0)

Anonymous Coward | about 2 years ago | (#42670161)

It becomes increasingly clear that frameworks are an abomination; a vile blight upon the hideous face of this godforsaken world.

Re: Frameworks (2)

Orcris (2652275) | about 2 years ago | (#42670511)

Why? All this is is modularizing KDE by splitting it into separate parts. This probably won't affect the end user at all, just the developers.

Re: Frameworks (1)

pallmall1 (882819) | about 2 years ago | (#42678043)

Oh shit, here we go again with KDE. Well, they can just ...

...K-K-KEEP IT!

1st (-1)

Anonymous Coward | about 2 years ago | (#42670235)

nt

two comments posted... (0)

apcullen (2504324) | about 2 years ago | (#42670293)

and is this site ALREADY slashdotted?

The Road To KDE Frameworks 5 and Plasma 2 (-1)

Anonymous Coward | about 2 years ago | (#42670321)

Balls in your mouth

Remote desktop? (4, Insightful)

Anonymous Coward | about 2 years ago | (#42670331)

NX is already pretty much unusable with compositing DMs, do they have a solution for remote desktop connections?

Re:Remote desktop? (1)

DuckDodgers (541817) | about 2 years ago | (#42670829)

Yeah, I think this is a fair question. I like having the pretty widgets on my desktop machine, but when I want a graphical interface on a server it has to be XFCE - everything else is too heavyweight to run well over a remote protocol like VNC or XRDP (which I understand just wraps VNC, but I didn't do a deep exploration of how it works).

I know it would be potentially a lot more work, but it would be easier if KDE and GNOME and Ubuntu Unity had a fallback 2D mode that was friendly for remote viewers. I understand Unity used to have that, but Canonical dropped it. (Cue flame war over Unity.)

Re:Remote desktop? (3, Informative)

Enderandrew (866215) | about 2 years ago | (#42670943)

KDE compositing can be turned on/off with a single key.

Re:Remote desktop? (5, Funny)

Anonymous Coward | about 2 years ago | (#42671511)

LIES!

It takes three keys: alt, shift and f12, and NOT ONLY THAT BUT YOU HAVE TO PRESS THEM AT THE SAME TIME! This is clearly far too complex a process, and this bug should be given highest priority.

Re:Remote desktop? (3, Funny)

Anonymous Coward | about 2 years ago | (#42671885)

Jobs. Is that you?

Re:Remote desktop? (3, Informative)

AJodock (1901718) | about 2 years ago | (#42672583)

The shortcut is assignable...

Who down modded this? Obviously a joke.

Re:Remote desktop? (1)

DuckDodgers (541817) | about 2 years ago | (#42676679)

Thanks. I didn't know that.

Re:Remote desktop? (4, Funny)

Tough Love (215404) | about 2 years ago | (#42672083)

...it would be easier if KDE and GNOME and Ubuntu Unity had a fallback 2D mode that was friendly for remote viewers...

A fallback text mode for remote access would be really sweet. How hard could that be? <runs away>

Re:Remote desktop? (2)

NotBorg (829820) | about 2 years ago | (#42672347)

K Desktop Environment. Servers really aren't their primary target. Also what can't you do for your server via a Bash login or other remote management protocol for a minuscule fraction of resources? What really requires a full on GUI desktop environment on your server? There might be some niche need for it, sure, but mostly I tend to think you're doing something wrong.

Re:Remote desktop? (1)

DuckDodgers (541817) | about 2 years ago | (#42676813)

When I started at this job we ran everything off Windows Servers, and our admins had learned to use Windows Server back before Microsoft had anything like PowerShell, so they were expert clickmonkeys that only occasionally resorted to cmd.exe.

I tried to sell them on Linux without X as a superior replacement, but I couldn't get any interest from the rest of the team. I set up a Linux server with XFCE and XRDP, the program menu in the bottom left, the clock in the bottom right, a taskbar, icons on the desktop, and Firefox, resource monitoring tools, and pgAdmin3 (a graphical tool for administering Postgres) in the menu. That got enough acceptance that we put it into production and our Windows team is becoming more comfortable in a mixed server operating system environment.

Re:Remote desktop? (1)

NotBorg (829820) | about 2 years ago | (#42677147)

pgAdmin has Windows binaries too. gah... I guess whatever makes them silly Windows admins feel special.

Re:Remote desktop? (1)

DuckDodgers (541817) | about a year ago | (#42679599)

I know it has Windows binaries. We were using it on Windows first, and once I could show them that it's effectively identical on Linux it was easy to make the switch.

At a big company with lots of resources, we could just take the Windows team and send them to training with the edict "adapt to the changes or be replaced". But this is a very small company, and just as importantly these people are pleasant to work with and intelligent. Even if I had the resources and authority to steamroll changes in place, I wouldn't do it.

Re:Remote desktop? (1)

melizeche (2576011) | about a year and a half ago | (#42737145)

A server needs a gui like a fish needs a banjo

Re:Remote desktop? (1)

cheater512 (783349) | about 2 years ago | (#42672643)

Why don't you use X instead of VNC?

VNC sends small bitmaps over the connection. X sends draw commands.

Re:Remote desktop? (0)

Anonymous Coward | about 2 years ago | (#42673225)

The user interfaces move away from widget-based X11 rendering to OpenGL.

This is why X won't work. VNC probably won't work either. I'm not sure what the solution is other than a KDE-specific remote desktop protocol or some of the magic I understand Wayland may make possible. (Remembering that Wayland itself isn't network transparent, just that network transparency would have to be at some other level than simply doing it the way X does it. Hey, we communicate locally over sockets, and TCP/IP is a socket, right? That means we're network transparent, right, right?)

CAPTCHA: miseries

Re:Remote desktop? (0)

Anonymous Coward | about 2 years ago | (#42678327)

Well, with modern toolkits and freetype, X also mostly sends bitmaps nowadays.

RTFA? (4, Funny)

pak9rabid (1011935) | about 2 years ago | (#42670415)

Perhaps everyone is reading the article before commenting? I know...+5 Funny.

Re:RTFA? (0)

Anonymous Coward | about 2 years ago | (#42670695)

" Fatal error: Out of memory (allocated 15728640) (tried to allocate 491520 bytes) in /home/sebas/www/blog/wp-includes/script-loader.php on line 248"

^--- That article?

Re:RTFA? (1)

jones_supa (887896) | about 2 years ago | (#42673511)

Kind of interesting that it's running out of memory. I believe most slashdottings are caused by network or CPU saturation.

Probably wasnt expecting traffic. (0)

Anonymous Coward | about 2 years ago | (#42670419)

Its a wordpress blog with 23.25 mb of memory, what do you expect. Should have hosted that on my phone, it would work better.

Re:Probably wasnt expecting traffic. (0)

Anonymous Coward | about 2 years ago | (#42671581)

No that sounds perfectly reasonable. If the web server is apache, and only has that much memory available, then you might be in trouble.

But that's not the issue. OOM is usually caused by infinite loops. I think we just have some shitty coding going on.

Oh no, not again... (1, Funny)

Bertie (87778) | about 2 years ago | (#42670579)

KDE moving to a new version of Qt? Abandon hope, all ye who dare to upgrade in its first three years of life...

Re:Oh no, not again... (4, Insightful)

Enderandrew (866215) | about 2 years ago | (#42671003)

KDE 4.0 was missing a lot of expected 3.5 features, but a lot of the bad reputation it received were from terrible Kubuntu packages. The Kubuntu maintainers admitted they didn't understand the new build process or what they were doing. Canonical pushed it way too early and pushed crappy packages, but KDE took the blame. Early KDE builds on openSUSE, Arch, Fedora, etc. didn't have the same problems.

And honestly, I think only a year later (KDE 4.2) they had a great desktop for everyone.

The shift from KDE 3 to 4 (and Qt 3 to 4) meant a massive rewrite, and having to reinvent most of their core features.

KDE 5 sounds more like an evolution than revolution, so it should be smoother this time around.

Re:Oh no, not again... (0)

Anonymous Coward | about 2 years ago | (#42671583)

So what you're saying is might be something I've thought for a while. It's the opposite of Star Trek movies. The odd numbered KDEs are good. Avoid the even numbered ones.

Re:Oh no, not again... (0)

Anonymous Coward | about 2 years ago | (#42671853)

In my neck of the woods this is the Dim Sum rule - only the odd numbered ones are good!

Re:Oh no, not again... (1)

Anonymous Coward | about 2 years ago | (#42671725)

This is pretty much on the mark, though it wasn't just the Kubuntu/Ubuntu camp that screwed up their KDE migration. KDE 4.0 and 4.1 were clearly alpha releases, neither stable nor feature complete. They weren't mean to be, they were developer previews. Yet many distributions migrated overnight without any thought to how that would affect the users. Kubuntu certainly made the mistake, Fedora did too. I think the SUSE developers were among the few to offer both the 3.5 and 4.x desktops side by side during the migration phase. A lot of distribution maintainers dropped the ball and a lot of users didn't educate themselves before jumping in. Otherwise most people would have started using KDE4 around the 4.4 release and things would have been smooth sailing.

Re:Oh no, not again... (1)

Tough Love (215404) | about 2 years ago | (#42672039)

KDE 4 is still missing the much loved KDE 3 feature of being able to slide the task bar off to the side.

Re:Oh no, not again... (1)

Anonymous Coward | about 2 years ago | (#42672275)

Not really, panels can be configured to auto-hide. Sure, a button is not left anywhere but do you really care?

Re:Oh no, not again... (1)

efitton (144228) | about 2 years ago | (#42673071)

I care that it is much harder to configure. I care that the taskbar didn't work as well as 3.5 until I finally gave up and went back to Windows. Remember Kaskbar? Square Icons with a preview of the window. Exactly like Windows 7 before Windows 7. I'm positive that MS stole it from KDE. Could I figure out how to replicate it in KDE for the first couple of years? Nope. Finally give up? Yup. And have you ever looked at the bug reports on the auto-hide panels? Ugh, especially with dual monitor.

So yeah, I still miss my KDE 3.5 setup and have not found a way to replicate it with KDE 4 and I get upset when people keep saying that it works the same because it flat out doesn't. Right click for stay on top? Nope. Pin the window? Not through KDE 4.4 anyway. Change window decoration by right clicking Kaskbar? Hah! And the panels never sized correctly with 4.

Re:Oh no, not again... (1)

Enderandrew (866215) | about 2 years ago | (#42673517)

There were several plasmoids that allowed you to replicate Windows 7 taskbar functionality in KDE, and they've been available for years. I used this one back in 2009.

http://kde-look.org/content/show.php/STasks?content=99739 [kde-look.org]

I think the current version is here.

http://kde-apps.org/content/show.php/?content=148813 [kde-apps.org]

Re:Oh no, not again... (1)

efitton (144228) | about 2 years ago | (#42673719)

So should I use the one that isn't stable or the one that isn't found (404)?

I am more than leery or using extensions as part of my core _anything_. Gnome Tweak means you don't have it. KDE plasmoids on kde-look means you don't have it. And this is an example of why, the software you gave a link to no longer exists.

Re:Oh no, not again... (3, Interesting)

Enderandrew (866215) | about 2 years ago | (#42673769)

I love that people are arguing that KDE 4 is missing a "core feature" that was actually a third-party add-on for KDE 3, but at the same time argue that installing a plasmoid means the feature doesn't really exist.

KDE 4 was designed to be extendable, and supports multiple methods of easily installing plasmoids. Installing content from the internet into KDE 4 was a core underlying technology since 2008.

http://userbase.kde.org/Plasma/Installing_Plasmoids [kde.org]
http://newstuff.kde.org/ [kde.org]

Re:Oh no, not again... (1)

efitton (144228) | about 2 years ago | (#42674973)

Kasbar was part of the BASE PACKAGE in KDE 3.5. It was included in at least three different distributions that I tried back in the day.

And now I'm suppose to install plasmoids that don't exist.

It does not exist. Honest. 404 (1)

efitton (144228) | about 2 years ago | (#42675735)

No seriously, it doesn't exist. I went to your link, I went to download it and got a 404. When I said it doesn't exist what I meant was: IT DOESN'T EXIST!

Re:It does not exist. Honest. 404 (1)

Enderandrew (866215) | about a year ago | (#42684193)

Most people don't install packages directly from the KDE Looks website I imagine.

My distro has a package for it in their main repository. I assume most do.

http://software.opensuse.org/package/plasmoid-smooth-tasks2 [opensuse.org]

And KDE allows you to browse for an install plasmoids right from your desktop.

Re:It does not exist. Honest. 404 (1)

efitton (144228) | about a year and a half ago | (#42707295)

If I still ran Linux, was on KDE 4.2 and trusted extensions maybe this would have minimized the pain. However, as I pointed out that this was far from my only gripe about KDE 4. I don't really want the Windows 7 Taskbar, I want Kasbar (which was included in base so please lets drop that entire it was 3rd party crap please). I dislike AutoHide, I liked the arrows or the option where it would drop behind the active window and reshow when the mouse was thrown in the corner. I liked the window decorations and options. I liked how easy it was to configure. In short, you might be convinced that I should have liked KDE 4.2, etc. and my complaints are minor and easily resolved. I, however, used 5 different versions of KDE 4 over 2.5 years and hated every minute of it. Telling me about settings I couldn't figure out and extensions I didn't find and might not have installed if I had known about them doesn't help me or get me back at all. It is like going to your restaurant that made some fabulous changes. However, your food was cold and the service was awful and your bill was wrong. And you didn't go back just once, you went back 5 different times and every time your service was crappy and your food sucked. Eventually you stop eating there but you still feel a bit bitter every time you drive by.

Re:Oh no, not again... (0)

Anonymous Coward | about 2 years ago | (#42678961)

Yep. This is GNOME 3's unsolved problem too. In order to address any of GNOME 3's shortcomings, its proponents link me to that extension site where I'm supposed to install countless hokey, random extensions uploaded by random people, with no guarantee that they'll remain compatible or maintained for any length of time. Sounds like a real usable and stable foundation for my desktop.

Hence why I don't use it. I've been back on Windows 7 for a long time now, which is a shame as a long time former Linux user.

Re:Oh no, not again... (1)

Tough Love (215404) | about 2 years ago | (#42676647)

Not really, panels can be configured to auto-hide. Sure, a button is not left anywhere but do you really care?

Bzzzt. No cigar. Autohiding is not the same as "sliding off to the side".

Re:Oh no, not again... (1)

efitton (144228) | about a year and a half ago | (#42707553)

I agree but it took me a while to figure out why I like the button and hate auto-hide.

With auto-hide the computer decides when the panel is visible. With the button I decide when the panel is visible. I prefer the latter and hate the former.

Re:Oh no, not again... (2, Insightful)

Anonymous Coward | about 2 years ago | (#42672081)

KDE 4.0 was missing a lot of expected 3.5 features, but a lot of the bad reputation it received were from terrible Kubuntu packages.

KDE 4.10 is still missing a lot of expected 3.5 features. I'm sure it's much closer to feature parity now (I'm assuming you can finally manage your system-level wireless connections in KDE, maybe fonts in KOffice even use decent kerning), but go back to KDE 3.5 and just look at the printing system. Seriously, just look at all of the things you can do to a printed document. Now look at the KDE4 one. Maybe KDE5 will fix this...

Re:Oh no, not again... (1)

efitton (144228) | about 2 years ago | (#42673085)

And where is my Kaskbar?

Re:Oh no, not again... (1)

Enderandrew (866215) | about 2 years ago | (#42673575)

Kasbar was a third-party app, not officially part of KDE3. And it wasn't maintained.

But Smooth Tasks is pretty nice today.

Re:Oh no, not again... (1)

efitton (144228) | about 2 years ago | (#42673739)

"Kasbar is included in the kdebase package" - http://xmelegance.org/kasbar/index.html

If it is in the official release I'm counting it. And no official or semi-official replacement.

Re:Oh no, not again... (4, Informative)

Anonymous Coward | about 2 years ago | (#42673299)

You're talking crap. KDE4 has managed wireless well for quite a long time. Not knowing this suggests you haven't used KDE4 in a long time and are just spouting some crap you once heard.

Re:Oh no, not again... (0, Informative)

Anonymous Coward | about 2 years ago | (#42672675)

Early KDE builds on openSUSE, Arch, Fedora, etc. didn't have the same problems.

Yes they did. I used all of those distributions around that time. The file indexer crashed horribly for about a year and a half and only recently became somewhat stable. The introduction of Phonon cause media applications to be gutted and preform badly for sometime. Many backends were written and half of them were crap and are now abandoned. There's only one that really works now and at the start it was nothing a mess of them that sort of worked. And don't even get me started on Dolphin and Konqueror who's features were missing or randomly broke every point release. Even Konsole was bad for a few rounds (it was stable at least).

As awesome as KDE 4.9.5 is right now you'd never know that 4.0~4.3ish were so universally weak. It wasn't just the Can-heads that had problems.

Re:Oh no, not again... (1)

Enderandrew (866215) | about 2 years ago | (#42673603)

I ran KDE from 4.0 on, running openSUSE and didn't have crashes. I had no problems with Phonon until they started integrating PulseAudio.

Konqueror still existed. Some people were butt-hurt that Dolphin was simpler and the default, but that didn't mean you didn't have the option of running Konqueror.

How was Konsole bad?

Re:Oh no, not again... (0)

Anonymous Coward | about 2 years ago | (#42674189)

I ran KDE from 4.0 on, running openSUSE and didn't have crashes. I had no problems with Phonon until they started integrating PulseAudio.

There's no PA backend for Phonon. It has to go through a higher level layer like gstreamer to get there. PA was a different problem set.

Konqueror still existed.

I never said it didn't. I said they both had broken or missing features. It was aggravating that I had to continuously switch back and forth between them. It practically chased people off to the command line.

How was Konsole bad?

It was horrendously slow at drawing text. It added minutes--not an exaggeration--to compile times.

KDE 4.2 == Awful Desktop (1)

efitton (144228) | about 2 years ago | (#42673125)

I made it until at least 4.3 and probably 4.4 before I gave up and went back to Windows. And this is from someone who hadn't used Windows in 4 years. KDE 4.2 was didn't hold a candle to KDE 3.5 for me.

Re:KDE 4.2 == Awful Desktop (1)

TemporalBeing (803363) | about 2 years ago | (#42673779)

I made it until at least 4.3 and probably 4.4 before I gave up and went back to Windows. And this is from someone who hadn't used Windows in 4 years. KDE 4.2 was didn't hold a candle to KDE 3.5 for me.

I stuck with 3.5.10 until 4.4 on some systems and 4.5 on others. But I haven't looked back.

Honestly, I looked at 4.0 through 4.3 from time to time but kept seeing that they were not production ready, and KDE themselves didn't call them production ready either - especially 4.0 and 4.1.

Re:Oh no, not again... (1)

timbo234 (833667) | about 2 years ago | (#42678775)

I was using Mandriva at the time and I can assure the early KDE 4 releases were just as unusable. I can't see how it was the fault of the Kubuntu maintiners, KDE 4.0-4.2 was half-baked, alpha quality software, full stop.

Luckily it's improved out of sight since then, but it's a worry that the KDE developers still don't acknowledge their mistake in releasing those early versions as releases, and not alphas or betas. I really hope they don't make the same mistake with KDE5.

dependencies cut? (0)

Anonymous Coward | about 2 years ago | (#42670683)

inter-dependencies removed, and portability and dependencies cut by stronger modularization

yeah, sure... My project went from depend of QtCore4.dll and QtGui4.dll to depend in those plus two Opengl dlls plus icuin49.dll plus icuv49.dll plus QtWidgets.dll plus the declarative one (i don't know why because I don't use qml at all, but without it it doesn't run) plus...

Supporting Legacy Graphics (0)

Anonymous Coward | about 2 years ago | (#42671339)

I wonder if they will drop support for low end hardware. I bet the opengl target is to be able to target the mobile platform.

Where are the screenshots? (3, Insightful)

loufoque (1400831) | about 2 years ago | (#42671363)

The summary essentially says "KDE is going to look better". Where are the screenshots?

Re:Where are the screenshots? (1)

rrohbeck (944847) | about 2 years ago | (#42673197)

I think it says "If you like KDE today, better watch out and don't upgrade until at least 5.2."

Re:Where are the screenshots? (1)

bill_mcgonigle (4333) | about 2 years ago | (#42674911)

The summary essentially says "KDE is going to look better". Where are the screenshots?

They're improving the graphics pipeline - no matter what the eye candy looks like, it's going to render to the screen better.

Organic user interfaces? (2)

hawguy (1600213) | about 2 years ago | (#42671517)

What does "more organic user interfaces" mean?

If that means that developers are more free to break with conventional UI's and come up with their own "innovative" controls and other UI interfaces, I don't want that - that sounds like when Flash designers started going wild on the web and each Flash web site had its own UI elements and the users had to figure out that flipping a virtual switch on one website was implemented as shooting an arrow into a target on another website and on another website you had to click the virtual LED light that was actually a button (but you don't know it's an active UI element until you discover that it's clickable).

Re:Organic user interfaces? (1)

ApplePy (2703131) | about 2 years ago | (#42671985)

What does "more organic user interfaces" mean?

Fewer pesticides.

Re:Organic user interfaces? (1)

hawguy (1600213) | about 2 years ago | (#42672005)

What does "more organic user interfaces" mean?

Fewer pesticides.

Now that's a UI I can support - I had to stop using Windows because their pesticide laden UI was giving me a rash.

Re:Organic user interfaces? (0)

jones_supa (887896) | about 2 years ago | (#42673599)

What does "more organic user interfaces" mean?

Fewer pesticides.

But then you get more bugs.

Re:Organic user interfaces? (0)

Anonymous Coward | about 2 years ago | (#42672147)

What does "more organic user interfaces" mean?

It means they'll grow on you!

Re:Organic user interfaces? (1)

rrohbeck (944847) | about 2 years ago | (#42673201)

Free range electrons and photons.

Re:Organic user interfaces? (1)

gl4ss (559668) | about 2 years ago | (#42673617)

it's supposed to mean that the interfaces spring up from genuine need for functionality at a certain place, not from some UI nazi deciding that all file dialogs shall be purple and have ability for previewing web pages.

Re:Organic user interfaces? (1)

Sigg3.net (886486) | about a year and a half ago | (#42690915)

An organic user interface means that the mouse is furry and the keyboard moans when you hit the "right" key, and when X is booted you'll be presented with the "operating system host" which is a Blender rendered ent that only take entish spoken commands (like Unity); the system will turn itself off at 8pm to sleep untilsunrise, and often not be responsive when there are other virtual events requiring system resources, like floods and firestorms.

Enlightenment (-1)

Anonymous Coward | about 2 years ago | (#42671861)

Once again, Enlightenment shows the way. Gotta love Rasterman....

Please don't screw up Kmail (4, Insightful)

Tough Love (215404) | about 2 years ago | (#42672009)

KDE devs: Please do not screw up Kmail more than it already is. In fact, please put serious thought into restoring the good old filesystem-based folder database and just do away with that horrible Akonadi mistake that is dog slow, when it works at all. Running critical apps on top of a full blown database may look like a good idea on a Presenter slide, but in reality it is just cruel and unusual punishment.

I know this isn't strticly related to Qt5, but just try to keep it in mind ok? Thxbai.

Re:Please don't screw up Kmail (2)

water-and-sewer (612923) | about 2 years ago | (#42673363)

Man, I totally agree with you. I've used Kmail since KDE2 and despite its flaws, I generally liked it better than the alternatives (and I've tried many). It's always annoyed me that while there are several GTK email clients, the KDE desktop really only has kmail, and they royally screwed the pooch with all this akonadi crap.

I just discovered an alternative. It's like watching a star form in the distance out of clouds and ether. It's Trojita (http://trojita.flaska.net/screenshots.html [flaska.net] ) and it's written by some Czech, mostly by himself. Now the project is picking up steam and is being folded into KDE.

It's still basic - no address book, and a lot of other stuff missing. But it doesn't require Akonadi or Nepomuk or any of that other stuff, and it is damned fast. Give it a look. I'm having fun experimenting with it and i'd be tempted to kick in some money if it helps the project advance in a way that makes sense.

Re:Please don't screw up Kmail (0)

Anonymous Coward | about 2 years ago | (#42674547)

I feed very much the same way. I have been using kmail since it's early days and ( after quite a lot of initial configuration ) i was always able to use it in a way that suited me best and i was really happy with it. But stability.. ooh my... Fortunately it always seemed to get just that little better with every minor release. Giving good hopes for a nice and stable future version. But then comes a major release and BAM, kmail and it's entire suite disappear from the world again. Both in Kde 2 -> 3 and in 3 -> 4. Only to come back in terrible form again. I understand the need for improvement but it just makes me sad. I guess kmail just uses to many kde components to ever truely become stable on such a fast moving target as kde. I sure hope they prove my fears wrong for qt5/kde5

Re:Please don't screw up Kmail (2)

tibit (1762298) | about 2 years ago | (#42675015)

I don't think that the database is the problem. The problem is that Akonadi is about as decent, code- and design-wise, as aRts was. Long on promise, short on delivery -- designed and implemented by people who demonstrably don't have enough software engineering experience. College kids may be bright and hard working, but that's no substitute for a certain amount of experience and understanding of the engineering side of things. Just look at this software engineering "recommendation" [kde.org] to give the idea what's wrong with the project. Give me a fucking break, this is supposed to be C++, RAII has been with us pretty much since forever, and those bastards pretend you need to have manual (and thus usually missing) solutions to problems nominally solved by designing for RAII? Seriously?

Re:Please don't screw up Kmail (2)

Tough Love (215404) | about 2 years ago | (#42676635)

I don't think that the database is the problem.

It's certainly part of the problem. The database may be fast, but it isn't anywhere near as fast as the underlying filesystem. And you are right, the basic design flaws are compounded by very obvious inexperience at designing and developing multitasking applications.

Re:Please don't screw up Kmail (1)

tibit (1762298) | about a year ago | (#42679577)

Huh? The database isn't there just to store data. It offers indexing, structured queries, etc. Usually, when people implement that functionality on top of a filesystem, it ends up as a yet another me-too half baked "database". It's like with any sufficiently large software project having half-baked LISP functionality in it :)

Re:Please don't screw up Kmail (1)

Tough Love (215404) | about a year ago | (#42680893)

Huh? The database isn't there just to store data. It offers indexing, structured queries, etc. Usually, when people implement that functionality on top of a filesystem, it ends up as a yet another me-too half baked "database"...

You sum up perfectly why this idea looks so good on a Presenter slide. Never mind that the reality sucks. Because in real life, the database is a whole lot slower than the filesystem and introduces a lot of new boundary conditions that tend to completely overshadow the shiny new functionality introduced. Sorry, but there's a big gap between the way you imagine the the universe, and how it actually is. You'd have done well on Bill Gate's Vista design team. Entirely designed on Powerpoint slides I understand.

Re:Please don't screw up Kmail (1)

tibit (1762298) | about a year ago | (#42682115)

You're crazy if you think MySql is somehow underpowered for Akonadi. It's not the problem. The problem is with how Akonadi's details were engineered, and how it was executed. If you seriously think a mainstream database is too slow to do relatively simple indexed queries on a *personal* email/contact database, you're crazy. The database is not a problem. That's about the only design decision I agree with. The filesystem does not offer the data indexing and querying functionality that you need in an email application. You're basically shifting responsibility for the index and query tasks from people who know their shit (database folk) to people who are comparatively clueless. My only gripe is that they should have used postgresql as that's a project with somewhat saner ownership and some nice technical benefits..

Re:Please don't screw up Kmail (1)

efitton (144228) | about a year and a half ago | (#42707263)

I'm a database guy so this hurts to write. JWZ wrote an email client. It worked. Netscape decided to switch to a database. It didn't work. JWZ wrote about it in addition to his line that: "To a database person, every nail looks like a thumb. Or something like that." http://www.jwz.org/doc/mailsum.html [jwz.org]

Re:Please don't screw up Kmail (1)

rdnetto (955205) | about a year and a half ago | (#42706517)

I'll second this. That said, I'd settle for the ability to reply to HTML emails without totalling the formatting, given that it's pretty much the only reason that I have to use an alternative these days.

(Not sure if this is fixed in a later version - I'm on Debian, so I'm still using 4.4 because of the feature freeze.)

Meet me half way (0)

ThatsNotPudding (1045640) | about 2 years ago | (#42672025)

Get rid of every program starting with a K and drop the omnipresent gear motifs and I might consider it.

Deduplication (is that a word?) (5, Insightful)

NotBorg (829820) | about 2 years ago | (#42672845)

One of the interesting bullet points:

Reduction of duplication with Qt by removing classes and using their Qt alternatives

A lot of classes were rewritten way back in the day when the licensing of Qt was under fire. Once those issues went away there really wasn't much point in continuing the duplication of effort. Bringing the two back together is long overdue. In the long run it could bring greater stability to KDE applications since more developers will be working on improving the same framework instead of two independent but close frameworks. This is good for both Qt and KDE.

Re:Deduplication (is that a word?) (1)

Anonymous Coward | about 2 years ago | (#42673111)

Ummm... this isn't correct. KDE developers never had a problem with the licensing of any of Qt classes. What this is referring to here is that either 1) something was written in KDE that wasn't available in Qt and Qt has since added support for it - might as well use the Qt version, or 2) a Qt class was extended with a KDE version to add extra functionality and Qt has added the functionality or the functionality is deemed not important enough for an extension.

Re:Deduplication (is that a word?) (2)

Kjella (173770) | about 2 years ago | (#42673597)

Ummm... this isn't correct. KDE developers never had a problem with the licensing of any of Qt classes.

Not the Qt licensing per se, but a lot of KDE developers have problems with the Qt contribution agreement. In order to support Qt Commercial etc. you essentially have to sign away all rights to the code, here's perhaps the most relevant part:

Subject to the terms and conditions of this Agreement, Licensor hereby grants, in exchange for good and valuable consideration, the receipt and sufficiency of which is hereby acknowledged, to Digia a sublicensable, irrevocable, perpetual, worldwide, non-exclusive, royalty-free and fully paid-up copyright and trade secret license to reproduce, adapt, translate, modify, and prepare derivative works of, publicly display, publicly perform, sublicense, make available and distribute Licensor Contribution(s) and any derivative works thereof under license terms of Digiaâ(TM)s choosing including any Open Source Software license.

As a result, a lot of code in the KDE libraries that could ideally could have been enhancements to Qt instead (both being LGPL) have remained outside Qt and will most likely remain so. Reducing the duplication has been a stated goal for ages but in reality not much happens since any non-trivial piece of code most likely has some contributor who won't sign.

Re:Deduplication (is that a word?) (1)

tibit (1762298) | about 2 years ago | (#42675067)

You've missed the non-exclusive part, for one. You're not signing away all rights to the code. Just the rights enumerated above. You can do with your code as you please, and they can do with their copy as they please as well, within the limits of non-exclusivity! You're not losing any rights to your own code! You can, for example, publish it under a license of your choice -- if the code is not a Qt-derived work in the meaning of copyright law. So, if it's not a derived work, you can publish it under BSD, or under a closed-source license, or whatever the heck you desire. You're only giving Digia a license, not outright ownership!

Re:Deduplication (is that a word?) (1)

Kjella (173770) | about 2 years ago | (#42675893)

True, you keep a slither of positive control in that you can still release it under the licenses you want, but you can't stop your code being used anywhere under any circumstances in any form since the license grant is pretty much limitless. Not exclusively so, but limitless all the same. I'll give you one Hermes Conrad point for being technically correct, though.

Re:Deduplication (is that a word?) (1)

NotBorg (829820) | about 2 years ago | (#42673919)

Qt licensing has continuously became less restrictive over the years. That fact alone surely has to have had some impact on where KDE libraries went. As I understand it there has always been this top down limitation where Qt changes (while they may have gotten their inspiration from KDE) always came from the Trolltech/Nokia side. The inability to mingle in the upstream (which wasn't helped any by the controversy and confusion in licensing) is what brought about the duplication of effort.

We're doomed, doooooomed.... (1)

aybiss (876862) | about 2 years ago | (#42674517)

" means higher quality graphics, more organic user interfaces "

How awesome! When I'm working with a spreadsheet I always wish it was more organic instead of just a boring old grid. I think I would be way more productive if it were organic.

And yeah, I'm sure we've all been there - you go to click a button or type into a form and notice how the graphics quality is so low - it's like they don't know that I have BluRay on my computer or something, even though VLC is installed and can do 1080p.

High, complex, high... ok, got it (0)

Anonymous Coward | about 2 years ago | (#42677819)

"Developers will find an extensive archive of high-quality, libraries and solutions on top of Qt. Complex problems and a high-level of integration between apps and the workspace allow easy creation of portable, high-quality applications."

Yes, someone is high all right....

Check for New Comments
Slashdot Login

Need an Account?

Forgot your password?