Rant - Frustrated and Disappointed

English Soapbox
#1

I’ve been a Linux user since 1998. Have been using OpenSuse and KDE for my PC since 2004. And please, before anyone criticizes my rant and points out that it’s all “free,” I’ve also been a financial contributor to KDE and Kmail in the past as well.

It used to be that most things all just worked - sometimes, you’d have to configure something - but generally, everything worked.

Not anymore. I’m not sure what is going on with the philosophy of developers, but it seems to be different than it used to be.

What’s the point of having a “beautiful” desktop with great effects, if it breaks things or other things don’t work any longer?

And Kmail - it used to work. Wonderfully. I used to use Kmail as an example of how great KDE and Linux was. Today there is no way I could walk into a client’s office and recommend KDE and Kmail.

Kmail and akonadi developers may be insulted by my post - but look - today, this email app is horrible. I am not the only one to complain; I’ve seen MANY complaints and the responses I’ve seen are head shake worthy.

Today, I am constantly trying to get at email in kmail - I delete 10 messages - and suddenly everything is stuck with “Retrieving Folder Contents” and some error message “Unable to fetch item from backend collection…”

This is not a “stupid user” issue. I am CONSTANTLY shutting down kmail, doing a akonadictl stop then waiting to then akonadictl start, akonadictl fsck, akonadictl vacuum - (if I don’t run those last two, there’s still no guarantee I’ll get to usefully use kmail), and then restart kmail.

Everything works again, until I delete 10 more messages. Then I have the same problem. I’ve posted elsewhere about this - solutions are temporary and I see others posting the same or similar problems.

Here we have this wonderful beautiful plasma - but it means basic vnc desktop sharing apparently no longer works. That’s… ridiculous. Why bother even including Krfb if it is not going to work?

These are just two issues out of many that I am seeing with recent distributions - but it seems there’s a lot of lost ideas of functionality and making things work with more of an emphasis on “effects,” and other things - that end up causing lost functionality. You can have all the “features” in the world - but if it breaks important things or causes things to be broken - then it seems one needs to revisit the future road map.

When basic functionality becomes a problem - that needs to be fixed before you add more “beauty” or “features.”

Please take this rant as it is - I’m not a desktop developer - but if you are one, usability ought to be the primary concern. New features that break or reduces overall usability and efficiency means that your new features and beauty enhancements are just merely toys and do nothing for real productivity.

After more than 15 years, I’m seriously considering going back to Windows - which I’m almost ashamed to admit.

I am sorry if anyone is offended - but the fact is, most things used to “just work” and they worked well. But now, many things don’t work, or don’t work nearly as well. If this is the direction of OpenSuse and KDE, well… you’re going to see a diminishing user base, which I’m sure you don’t want.

#2

Yes, blame is totally deserved for how software (and now other things) are created, built and modified.

Blame it on “Agile Development”
You can Google up the numerous interpretations and schools of thought today, it started… something like 10 years ago ago in earnest?
In “the old days” everyone used the Waterfall method… Everyone knows what a deadline is, what a product launch is, the long list of project objectives that is created when a project is first authorized, and from then to the last day before the end Product is launched, everyone is totally focused on The List.

For a number of reasons, no software is built that way anymore.
In particular, any long-running development effort that extends more than maybe 6 months would be doomed if it still used the Waterfall method and as the world turns and competitiveness continues to heat up at exponential rates, the time allowed from conception to delivery is getting shorter and shorter despite how big and complex a project can be.

I’ll let you read up on various articles that describe how Agile Development works and why there is no alternative to it today.
It’s also the reason why you see so many “artifacts” (a nicer name than “bugs”) and broken things.
But, if you can understand why there is no choice, then hopefully it’ll help to lessen the frustration… After all, it’s better to have buggy software than no software at all, and repositories like Github, Sourceforge and Freshmeat (yes, it was a major free software repo until 2 years ago) contain immense numbers of abandomware.

TSU

#3

Hi
We have a specific forum for rants and grumbles :wink: Soapbox… moving in a few temporarily closed…

#4

Hi
Moved and reopened… Use GNOME!!! :wink:

#5

Thanks for reading and understanding the frustration to some degree.

Although, I’m not sure I totally agree with you that buggy software is better than no software at all. I’ll give you some examples, and perhaps not really based on “buggy” software, but productivity:

I still use a paper based daily journal and planner. It’s the way I work best, and am productive. I am more productive spending 10 minutes, actually WRITING out my daily to do list, then peering at and using software, someone else thinks should work for me.

Maybe that is a poor example - but let’s get back to Kmail - I suppose having a really buggy Kmail is better than no Kmail - but there is huge frustration in using an application that is supposed to manage email, but I spend literally some days, two hours, trying to manage the application just to read what I should have been able to read, 2 hours ago.

That’s a LOT of time wasted over something that <b>used to work very well</b>. Productivity wise, there is a huge regression.

#6

Maybe, and that would be unfortunate. KDE used to be an awesome desktop. The idea of going back to something that I switched from two weeks after first trying it out years ago, and now relearning it because another one has become in many ways, useless, is a bit disheartening.

My complaints today about KDE and perhaps the direction of OpenSuse are the same complaints that Windows users had, 14 years ago. And those complaints were the reason some of us moved (and financially supported I might add) to OpenSuse Linux with KDE. And back then, developers took a lot of pride in making sure things worked.

Sure, sometimes there was a bit of a learning curve, or brand new hardware would not work - but generally speaking, you didn’t have to waste hours of your day on fixing up basic applications like email.

Windows was buggy. Linux distributions and desktops also had some bugs, but few were big showstoppers. That’s not the case anymore. Basic applications that one needs for every day productivity, like Kmail, are now worse to use than Outlook was to use 15 years ago. Manoman… that’s not something to take pride in, I don’t think. It shouldn’t even really be joked about, if there is any pride.

#7

I guess everyone has their own perceptions …

With every New Year comes reflections - and my own thinking about GNU/Linux an OpenSUSE and I confess I have not been happier wrt the state of affairs than I am today.

I like the stability / long life of LEAP and yet for another of my PC’s I have the cutting edge of Tumbleweed. We have a good forum with great people, and the same can be said about the Mailing List and other openSUSE communication channels.

I recently blogged about a number of my LEAP-42.2 installs. This is the first time since I have been a GNU/Linux user (where I started in 1998) that I have brought all my PCs up to date with the latest version of openSUSE and I have done that only because I am very pleased with this release. My blog entry, with the inxi input from these PCs here: https://forums.opensuse.org/entry.php/175-Installing-openSUSE-LEAP-42-2-on-two-EFI-PCs

So a very HAPPY NEW YEAR to all - and as opposed to the thread title, thinking of the past, and thinking of where we are today, and I am VERY HAPPY with the current release and state of our distro.

Best wishes to all.

#8

Happy New Year to you too. I trust the season won’t interfere with the comments made here however.

You’ve been around for a long time, and I’ve appreciated your help over the years.

But let me ask you a question - you may be happy - but how does a person, coming from a Windows environment - just a general user, who wants things to “just work” deal with this? Let’s say the moved to OpenSuse/Linux/KDE back around version 12.1. There were some bugs… but things worked. A few years later - they are spending two hours or more, trying to just read email, because the app they have been using for several years, under previous versions of KDE has changed? And is now wasting hours of their time every day?

Maybe if you’re a developer and/or have more than some basic knowledge of the workings underneath, you can figure it out. But there’s no way this is going to cut it in any kind of business office environment - but 12 years ago, KDE/OpenSuse could.

#9

You have to read what Agile Development is,
The idea isn’t to be permanently buggy with the same bugs, it’s to rush “usable” software out the door without waiting for a full QC which might not even catch all bugs. Early adopters essentially become the Beta testing, and there is supposed to be a full support framework in place to report, assign and fix bugs much faster than if bugs weren’t expected.

So, although there is a general expectation of bugs, that’s not to say that the bugs should exist long… if the bug reporting, fixing and re-release cycle is working properly. If any part of the cycle breaks down… eg bugs aren’t reported by Users, or is poorly managed so bugs are not assigned quickly or not enough people to fix bugs, then of course Agile Development breaks down.

If you also read WHY Agile Development has to be used, you will find competitive and economic reasons… Maximizing efficient use of Developer resources (everyone should always be busy), Users get new features faster, and costs of QC and other procedures are brought down which is supposed to deliver to you a superior result overall.

Yes, all in theory and depends on a lot of things working right.
But then, would you for instance give up mobile device integration if it had to wait 6 months or a year particularly if a competing Desktop is able to deliver while perhaps deficient in some other area? At least from a marketing and general perception view, if your product can’t be feature-competitive, Users won’t even try your product.

TSU

#10

For as long as I have been involved in computers (which is a long time), the basic underlying philosophy of developers has been “burn your bridges behind you”.

Well, they don’t say it that blatantly. But it has always been the philosophy that a major change requires throwing out a lot of debris.

So WIndows 95 is long gone. Windows XP is gone. Vista is about to go. Mozilla is already getting set to drop support for Vista.

Gnome 2 is long gone. It was replaced by an awful Gnome 3. But, in all honesty, Gnome 3 has improved and might even be worth considering again.

So KDE Plasma 4 is out, and Plasma 5 is in. Yes, some of Plasma 5 still has a beta quality. But it is improving.

I’ll manage to live with the changes. Overall, Leap has been a good experience.

I can’t comment on kmail. I barely used it with Plasma 4, and didn’t much care for it. I didn’t like it much back before Plasma 4. I mostly use web mail these days.

#11

… and I am happy enough with Thunderbird, for the little e-mail I bother with.:wink:

… but, I do understand the frustration, and I have also been feeling frustrated more often lately.

I am still, though, completely pleased with openSUSE and excited about the new LEAP, although I have been too busy (and KDE-reluctant) to move to it so far.

And Linux, IMHO, is still by far Superior to MS offerings.

In Windows, fixing a disk connectivity problem can be a hair-pulling experience, perhaps with no end in sight after many hours of scouring MS help and internet threads, since the offending code is buried deep in compiled binaries. In Linux, that same problem can be fixed in a matter of moments by simply adding a “#” in a script to comment out a single line of code.

#12

well wrt Windows - I know many more MS-Windows users than openSUSE users - and those MS-Windows can really rant wrt Windows problems when they want to … no version of openSUSE holds any special candle wrt dissatisfied user especially in comparison to MS-Windows

Like some - I’ve been using GNU/Linux since back in the late 1990s … and wrt SuSE-Pro, and later openSUSE, I can not recall one openSUSE version that did not have some aspect that caused one or more users to complain. openSUSE-12.1 was no special nor magical exception. It too had its problems. I recall rants of users claiming 12.1 would never work in an office environment. Purely impossible. Of course - they were opinionated. I had different views. I think many openSUSE versions (of which 12.1 was just one) could work in an office environment.

Now wrt users having trouble migrating - I don’t recommend watching me when I try to boot one of the MS-Windows partitions on my PCs … these partitions typically go for 6-to-18-months between boots - and its not a pretty sight when I try to boot them … and then if I try to do something new in Windows on my home PCs or even worse - at work where I have no choice but to use MS-Windows ? Its not a pretty sight … Thank God for the absolutely superb support we have at work for our office and operational PCs.

My view is one does better with an OS they know. One does better with good support.

Now I have friend’s with the MacIntosh who claim it just works - I recall one lent me his Macbook for 3 hours one evening to prove his point. When he returned after 3-hours, he could see my massive steam of frustration with the Mac come out of my ears … filling the room with smoke … I could not do many functions with the Mac and it boiled my blood … so he then spent 30-minutes showing the very-specific GUI actions to make the Mac way to work to do what I needed … and doing it “HIS” way (ie Mac way) it worked … only with ‘his way’ … with him claiming that way was most logical. When I claimed a different way looked more logical to me - he debated my view. So when I asked if I could change the Mac way to instead the way I wanted - he said no. Better people than I in Mac’s (and with 3 hours what did I know about Macs ? ) decided the Mac way was by far the best way. Very quickly it came clear - clear to me - that for Macs its the Mac way or the highway (or no way - dependent on how one wants to say things).

So other than Mac fans - I have very rarely met a PC user who did not have a gripe. In particular - some Windows users can really really rant.

12.1 ? Nothing special . I liked it better than Macs. Better than Windows … but back in the day of 12.1 I kept older openSUSE versions on all my PCs and I did not put 12.1 on all of them. As noted, LEAP-42.2 was the very first time, I put an openSUSE (or SuSE-Pro going back far) version on ALL of my PCs (with the exception of a Tumbleweed boot partition on one PC that also has LEAP-42.2.).

And as noted - most import for all users - and ESPECIALLY new users - openSUSE is more than just the OS running on the PC. It has the great community of support through a number of communications channels - the forum - the IRC chat - the mailing lists …

So what do I tell the new Windows users who are frustrated with openSUSE ? I tell them - go to the communications channels - ask your questions. … and if you will … and if I know anything about the problems you encountered … let me try to help. Most don’t listen to me - but some do.

#13

agreed but… kde have a lot of refactoring to do at the moment with move to kf5.
I do wish they would spend more time on central projects like kmail instead of spending resources on e.g. kde marble etc where free and better commercial alternatives exist.

#14

On Sat, 31 Dec 2016 23:26:01 +0000, WeeHughie wrote:

>> Blame it on “Agile Development”

Actually, I wouldn’t blame “Agile development”, Tsu - Agile as a
development methodology came about because software customers/users
wanted updates that were more frequent, and voted with their wallets for
companies that didn’t have long release cycles.

Waterfall is an inherently slow process, with gates for every step of the
release, and everyone has to agree. The longer-term planning has its
advantages, but a huge disadvantage is that when a customer or user says
“I want the program to do ‘x’”, it takes a long time for the feature to
get into the development cycle. It won’t happen in a month or three
months, it might take 18 months to be included in the next release.

In that time, a “competing” product might release the feature in their
product, and so people move. In the FLOSS world, that’s not a huge deal
(data portability notwithstanding), but in commercial development, it is
a huge deal.

Agile is used often in combination with some elements of waterfall (I’ve
seen gates used from waterfall with agile development processes with some
success), but at the end of the day, “customer demand” is what pushes
development to release smaller releases more frequently.

Jim

–
Jim Henderson
openSUSE Forums Administrator
Forum Use Terms & Conditions at http://tinyurl.com/openSUSE-T-C

#15

Thinking back wrt various openSUSE releases and IMHO often whether one likes an openSUSE specific release - or not like a specific release, really depends on the application(s) one will use and how well they are implemented. If one of one’s favourite applications has implementation problems in a specific release - then that problem can mean one dislikes the release. Yet someone else who does not use that specific application (or application(s)) may instead think its one of the best releases.

Sometimes - its not one or more applications, but rather a specific OS feature that has problems, that will turn one off for a specific release. If I recall correctly, openSUSE-10.1 had major problems with its package management - that resulted in many users leaving openSUSE. The GNU/Linux press in particular was very harsh on openSUSE and the package management functionality problems/limitations then. Some of us for the next few openSUSE releases instead adopted the Smart package manager, until zypper had a very good implementation. Smart worked well, although it did not take many releases for zypper to end up being a better package manager than Smart. But for those early 10.1 and 10.2 days of openSUSE, it was rough going for those who tried to stick with the package manager that was behind YaST.

Hopefully those who struggle with any openSUSE version, because of some broken feature or application, can eventually find a suitable and helpful workaround.

#16

I think that was SUSE 10.1 (opensuse did not yet exist).

I remember that I would boot up my computer. And it would be very slow. Almost all of the memory was being sucked up in package checking to decide if I needed updates. After around 30 minutes, that ended and the system began to work well.

My laptop had less memory. I turned off the automatic update checking, because it was otherwise unbearably slow. When my desktop found some important updates, I would then manually run the updater on the laptop while not expecting to do anything else on that machine until the updating finished.

Yes, that problem was eventually solved (maybe in opensuse 10.3), and today updating works pretty well.

#17

+1 with the KDE and Kmail rant. This is why I’m currently running on XFCE with Thunderbird (never was a fan of Kmail). The longer I am in the IT business the more I think vendor lock in is the worst thing that you can do. So right now when I’m using any desktop environment I try to use as cross platform/cross desktop applications as possible so that if something fails I can easily switch without too much disruption on my standard workflow.
This is also the reason (vendor lock in) that I believe iOS and OSX is bound to die sooner or later.

#18

I must admit that, for the case of Leap 42.2 I am also considering this route.

There is however, from my point of view, one issue with the “move from KDE to GNOME” solution:

  • digiKam – the version 5.x.y currently shipping with KDE Plasma 5 is very good: the digiKam folks seem to be achieving "constant improvement
    ", since more than a few years . . .

Yes, the integration of the KDE Plasma 5 Kontact PIM (KMail & Co.) into the Plasmoid world is not that wonderful – one could say abysmal . . .
I have raised several KDE Bug Reports with respect to the issues I have with the Plasmoids not doing what they were doing in the KDE4 Plasma world – as usual with Plasmoids, no response from the KDE folks, yet.

Given some of the comments being made about the new Kontact Suite, I’ll need to check the behaviour of the thing with my KDE4 Plasma e-Mail, appointments and activities collections, before migrating from 13.2 – I’m not amused by “nasty surprises”.

Are there any plus points with the Kontact PIM shipping with Leap 42.2?
IMHO, yes there is at least one quite nice new feature:

  • The KMail new message editor has automatic language detection – sentence for sentence – you can type in sentences in English, German, French, and the spell checking follows the language of the sentence being written – automatically (or is it automagically?).

There was another new feature which seems to have disappeared:

  • During the Release Candidate testing I did notice that, the “own” e-Mail address being displayed in received messages was being replaced by “Me” – I haven’t found this behaviour in the current shipped version.

Is there a very annoying “nasty” in KDE Plasma 5?
Yes there is: Dolphin:

  • In terms of it’s networking capabilities, it’s very good (once one gets used to a couple of “just do what’s intuitive” methods).
  • There is however a fairly major mistake with respect to File Management: the extended file permissions handling is very broken.

Several people, myself included, have raised “severe/critical” KDE Bug Reports with respect to this issue (T-bit [restricted deletion flag] and SETUID/SETGID bit handling); the answers that we’ve received are along the lines of “The Qt5 Framework is missing the capability and therefore this issue will not be worked.”.

Do things such as LibreOffice run happily on Leap 42.2 and KDE Plasma 5? Yes they do.

Do I like the “flat” (2-dimensional – looks quite similar to the Redmond Windows 10 appearance >:) ) default KDE Plasma 5 theme: “Breeze”? No, I definitely do not – I much prefer the default (3-dimensional – IMHO, nicer “eye-candy”) KDE4 Plasma theme: “Oxygen”.

Are the KDE folks lacking something?
IMHO, yes: looking at the financial statements of the KDE folks, the project seems to be under-funded. IMHO, if the KDE project can not attract an influential sponsor, this will not change – I suspect that “crowd-funding” will not be sufficient.

Is the project also suffering from a touch of “enthusiastic volunteers”? Possibly but, looking at the self-imposed project management and communal decision-making used by the KDE Community, it seems that the decisions made for the move from Qt4 to Qt5 were not that wild and, were in fact well founded.

Will KDE Plasma 5 improve?
I suspect, yes. The version shipped with Leap 42.1 was something I would place in the “nice legs; shame about the boat-race” class. The version shipping with Leap 42.2 is a vast improvement over the previous version but. in some aspects not as good as the KDE4 Plasma offering.
[HR][/HR]AMD Hardware.
Leap 42.2 on a Dual-Boot Windows 8.1 Laptop.
13.2 on a desktop with Windows 10 running in an Oracle VirtualBox VM.
Need the Redmond stuff for Sigma Photo Pro RAW image processing.

#19

> WeeHughie;2806290 Wrote:
> I’ve been a Linux user since 1998. Have been using OpenSuse and KDE for my PC since 2004. And please, before anyone
> criticizes my rant and points out that it’s all “free,” I’ve also been a financial contributor to KDE and Kmail in the
> past as well.
>
> It used to be that most things all just worked - sometimes, you’d have to configure something - but generally,
> everything worked.

Not true. KDE was and has always been a beautiful desktop environment, but the claim `everything worked’ could only be
truthful if ignoring the additional baggage (e.g. Konqueror, KNetworkManager, Nepomuk) that have always been flaky.
Fortunately the two essential packages (konsole and dolphin) have remained robust and stable over the years. KDE also
one of the most impressive text editors (Kate) and it is scandalous that it isn’t installed by default on many KDE
installations.

And Kmail - it used to work. Wonderfully. I used to use Kmail as an example of how great KDE and Linux was. Today there is no way I could walk into a client's office and recommend KDE and Kmail.

I’ve experimented with Kmail a few times and have always found it substandard, clearly not fit for purpose. The only
professional-grade' EUA for GNU/Linux I know is Mozilla Thunderbird, which works well but still has issues and not under proper active development AFAIK. I use mutt. I agree with the comments of its author: All mail clients suck.
This one just sucks less.’

Please take this rant as it is - I'm not a desktop developer - but if you are one, usability ought to be the primary concern. New features that break or reduces overall usability and efficiency means that your new features and beauty enhancements are just merely toys and do nothing for real productivity.

I’m sure the KDE developers regard usability to be their primary concern. It’s just that one user’s desktop experience
will be different to the next. For me KDE is perfectly usable but I choose not use many of it’s unnecessary baggage.

After more than 15 years, I'm seriously considering going back to Windows - which I'm almost ashamed to admit.

No reason to be ashamed. Nothing wrong with Windows. Every machine I own at least a dual-boot with GNU/Linux and
Windows. There are a lot things that Windows will always do a lot better than GNU/Linux (e.g. drivers, professional
Office applications, and 3D graphics games).

I am sorry if anyone is offended - but the fact is, most things used to "just work" and they worked well. But now, many things don't work, or don't work nearly as well. If this is the direction of OpenSuse and KDE, well.. you're going to see a diminishing user base, which I'm sure you don't want.

Remember openSUSE is DE-multivalent so you are not stuck to KDE. If you don’t want to use the unsightly GTK-based
equivalents (such as GNOME), there are Qt alternatives. I recently saw a post on the openSUSE Chit-Chat usenet from one
user enjoying LXQt on openSUSE.

I believe (rant follows) a lot of developer-time is wasted developing packages for different desktop environments in
parallel that do the virtually the same thing. I code, and hate having to check my Qt-based GUIs work equally well in
GNOME and KDE. It’s a waste of time in a way that Windows developers don’t have to worry about. KDE and GNOME have their
rivalling packages (e.g. Kmail/Evolution, Konqueror/Epiphany) but they are all just generally disappointing projects
(with notable exceptions including Kate/Gedit) that should have merged into something more promising (or died) long ago.
Unfortunately the fragmented GNU/Linux model with arbitrary divisions based on `religion’ (e.g. Qt vs GTK) lends itself
to producing a vast amount of mediocrity with only rare sparks of true brilliance (e.g. GIMP or YaST).

Nevertheless, KDE remains beautiful.

#20

There are rumours flying around here in the German press that, the company SoftMaker is currently involved in the maintenance of Mozilla Thunderbird – on their web-site however, they’re only admitting to integrating it into their Office product and providing a few add-ons: <http://www.softmaker.com/en/softmaker-office-linux-thunderbird>