Marek I disagree. What's the point in all these people working their butts off for KDE when it's narrowing the potential user base so severely? Why write software no one can use? Trinity exists b/c there is major dissatisfaction with the direction of KDE and the performance issues of Akandi are notorious. Do any search on how to disable Akondi and you'll see thousands of questions about it due to performance issues going back at least to KDE 4.
KDE 5 works for people with higher end computers, which excludes a huge percentage of Linux users. Akondi is the primary culprit.
Akondi is set up for people who use a very specific and limited subset of KDE features. PIM management in specific. I am baffled by how such a simple task can become such a resource monster. The only other thing it's doing is indexing files and that is actually not a big resource hog in KDE5. I didn't even have to turn that off. I DID have to turn Akondi off just to have a functional machine with KDE5. Though there were other performance hits and gotchas that were too aggravating. I am typing this on a laptop with 4 gigs of RAM. I have at least 30 Chrom tabs open Krusader open with a dozen or so tabs including 3 or 4 SSH tabs to other machines, Gthumb, Various taggers, converters, wave editors, text editors and long as I don't open FB it runs great. THAT is what Linux is supposed to do.
If we want wider adoption of Linux, and the support that comes with a larger user base we need desktop managers that are friendly and easy to use. KDE fits that bill except for being such a massive resource hog and some twerks that exist now in KDE 5 that don't have a solution like the focus insanity. People try Linux on older machines. If Linux struggles on that older machine they become disillusioned and stick with windoze. People try Linux when their machine can no longer run Windoze or people who get hand me down machines who cannot afford their own PC. If Linux doesn't work well they give up rather quickly and we lose another potential convert to Linux. To demand MORE resources than it takes to run Windoze 10 just buggers the imagination. Which is exactly what KDE 5 is doing. I can run Windoze 10 on a machine with 4 gigs of RAM. At least as well as windoze ever runs. I can't use KDE 5 even with Akondi disabled on a machine with 4 gigs of RAM. Something is wrong there. Especially since the features offered are so minimal and things most people do on the phone and cloud nowadays.
The PC is redefining itself right now. KDE can help keep the PC platform alive or it can help doom it, relegating people to phones, tablets and Android as a primary OS. The desktop needs to evolve and be good at things not easily done on a phone for it to survive. No desktop and Linux goes away entirely. Android rules the phone market. IOS a 2nd and Linux barely even a blip. Linux owns the server market but not many servers use KDE. They are running CentOS or RHEL, and usually headless or running a very minimal desktop manager. No sane person would put KDE on a production server. It's way too much of a resource hog.
So what niche does that leave KDE? What draw is there for KDE to attract new users? What roles can KDE fill? I'm not seeing any. It's too brutal on resources for low end machines and servers. KDE doesn't have integration features like smooth transitions between sound servers that would draw musicians to KDE. KDE has some great apps, but it's just easier to use XFCE, Trinity or something like that and add the KDE libs and apps. The integration between KDE apps seems focused on PIM and nothing else. So what exactly is all this work on KDE meant to accomplish? Who are they targeting as users? What market exists long term for KDE as 20 year users like myself migrate off KDE to other desktop managers and fewer people switch too KDE? KDE and Gnome used to run neck and neck in surveys. Today neither is likely to even make the top 3 in desktop user surveys.
You want to see KDE grow in users, make it friendly for low end machines. Add graphics & music integration that helps KDE standout above not only other desktop managers on Linux, but also against the Mac and Windoze. Make it smooth, pretty and efficient and people will flock to KDE. KDE has a lot to offer if you get past the performance hits and the focus issues. Just because something is traditionally a distro responsibility doesn't mean a desktop manager cannot provide something better. I've used KDE for 25 years now. I'd hate to see it die.
On Thu, Oct 15, 2020 at 12:32 AM Marek Kochanowicz <sirherrbatka@xxxxxxxxx> wrote:
This discussion is not a constructive one.
Draciron Smith pisze:
> Akondi is a monster that drove me away from KDE. If you do not have the
> latest greatest machine and multitask, Akondi will bring your machine to
> it's knees, then flog it, then draw and quarter your memory resources,
> grind your hard drive into dust, then chew up the remains. I had to switch
> to Trinity to get back a usable machine. So I will strongly beg to differ
> with your statement about Akondi being spoken of in any sort of positive
> manner.
>
> First on principle. The integrated PIM is mostly obsolete. Phones today
> have the CPU horsepower desktops had back when KDE 3 was embarked upon. A
> phone is a natural place to do contact management. Getting ISPs to continue
> support for POP3 & IMAP is like pulling teeth and when you do get support
> you gat ONE email address. So realistically if you want to do old school
> client server email you basically need a domain and a server where you
> control your email or you live alone without children. A dedicated domain
> and server is out of the cost range of most people. So webmail is really
> the only option for most people. I have email lists of 100+ people and
> might maybe have 1 or 2 people on those lists not using webmail. Those that
> do not are using work accounts usually.
>
> So what value is there in apparently doubling the memory footprint of KDE
> for something 90% of the folks do not even use? Doing so precludes using
> older machines. Which is the bulk of Linux users and one of the great draws
> of Linux. That is you don't have to go buy a new spiffy high end machine
> every couple of years just to do what you were doing just fine on an older
> machine. Akondi is increased resource demands with little to no value and
> turns a high end machine into a single tasking device that doesn't even
> match up to what your phone can do in those areas. I do not need 8 gigs of
> RAM on a phone to manage contacts. 4 gigs is simply not enough for a
> machine running Akondi. Not and actually multitask.
>
> KDE 4, I disable Akondi, 4 gigs of RAM on this machine. 2 gigs fo RAM on 2
> other machines running KDE 4 with Akondi disabled. Worked great. Ubuntu
> 14.04 LTS goes out of support. I have to upgrade to 16.04 and KDE 5 and my
> 2 Gig machines barely boot. My 4 Gig machine acts like I'm running WIn 95.
> It's constantly freezing. Locking up so tight I can't even SSH in. I have
> to power off at times to get it to come back after literally hours of just
> churning. I put Trinity on those machines they work great again. Hell even
> Gnome gave me better performance than KDE 5. I'm running 20.04 on one
> machine using Trinity and all good. XFCE also runs rine as a desktop
> manager. KDE has become the rich man's desktop as the poor cannot afford
> the hardware to use KDE anymore. I've been using KDE for my desktop
> manager since the 90s. I really like KDE, but it has become windoze like in
> hardware demands and performance. There's a reason I dumped windoze in 2000
> and never looked back. KDE is now causing the same problems that caused me
> to dump windoze 20 years ago.
>
> On Mon, Oct 12, 2020 at 6:16 AM Marek Kochanowicz <sirherrbatka@xxxxxxxxx>
>
> wrote:
> > Yeah, i actually have manjaro on the other machine and kmail works there
> > fine
> > as well. So it seems for me that questionable packaging techniques are a
> > factor here but I can't tell the precise details.
> >
> > As for the importance of the akonadi: it is actually a well designed piece
> > of
> > software architecture that simplifies all PIM apps drastically. Removing
> > akonadi from PIM is not only (IMHO) pointless but also prohibitively
> > expensive
> > endeavor. Instead I would try to investigate what is the actual problem
> > with
> > the packaging and try to seek some kind of remedy for it.
