Another rant

[dibl]

Re: Another rant

In a "directed" organization, whether corporate or nonprofit or whatever, it is possible (if not mandatory) to have a strategy. The senior executive can use one of multiple approaches to pull the needed planning and analysis efforts out of his/her senior staff, relative to whatever the service or product offering and target market are, and then through a process of analysis and consensus-building, reduce the choices down to a strategic mission statement, and use that as the basis for future direction of the organization.

GNU/Linux has no such organization, and nothing that I would characterize as direction, therefore it is not possible to generate a strategy or a mission statement. It is developed by volunteers, and volunteers are, by definition, not directed by anyone else -- at least not with any enforcement. Some pieces of the community, such as Canonical or Debian, are more organized than others, but even so they are totally dependent on the larger volunteer base.

So, what do volunteer software developers do, when the work is uncompensated? The answer: Whatever they wish. Something fun. Something that a developer would like to use. A new music player. A new text editor. How many of those are in the repos? I finally figured out what nepomuk and strigi are for when I read a detailed instructional piece about them, by a guy who does Linux support all day every day. He has a zillion e-mail messages on his system full of technical issues and solutions, and he needs to search them multiple times per day. Hey -- wouldn't a grand semantic desktop search tool be cool? Enter nepomuk!

Anyway, I think KDE 4.6x seems to be a pretty good piece of work under the circumstances. It seems less fragile than prior 4.x releases, and after getting the hang of most of the plasma bits, it no longer feels like it is controlling me. PIM still kinda sucks, to a large degree -- apparently developers don't find that very interesting work.

Today's two cents' worth.

[de_koraco]

Re: Another rant

Every program has room to improve, X is not any different... but to the throw out 20+ years worth of stuff, and features, and divide things off and convulet what many users already see as a mess with another layer of stuff to get back to what X already does/did is just a solution to a non problem.

If I've understood it correctly, X and all its protocols would still be able to run as one of Wayland's clients. In fact, the reason Wayland has been even considered as a replacement is the fact that it offers full compatibility (though full backwards compatibility is one of those myths like a stable API). The xorg devs have said that Wayland offers a chance to start anew and scrap a lot of the spaghetti code in X. Such an approach has been pretty successful for Gnome 3.

But I wouldn't worry, I'm sure the major players in the field like Red Hat largely have the same interests as you when it comes to maintaining systems. If anyone pushes Wayland, it will be Canonical, which does promise a disaster, but a Debian install is pretty easy nowadays.

[GreyGeek]

Re: Another rant

...
As far as the overall goals of the Linux community I would be interested in seeing the opinions on this. ...

I can't speak for "the community" but my understanding of the "goal" of Linux and FOSS was (is?) to "scratch an itch", except for RMS, whose goal was to circumvent corporate lock-ins and greed, hence the GPL. Frankly, without the GPL Linux and FOSS would be where BSD is today, used by few but exploited to the hilt by corporations who take and take and take but never give back. Profit through lock-in. To "scratch an itch". That someone would find any particular software usable was appreciated. That any would report bugs or pitch in with code was a benefit to all. Thus was born the FOSS mantra, "release early, release often", a necessity in a community of volunteers with no paid testers.

It has been my observation that from my first experience with Linux, RH5.0 in 1998, a Linux DE has always been as good or better than its concurrent Windows DE. RH5.0 was significantly better than Win95. SuSE with KDE was significantly better than Win98 (it wouldn't be fair to mention WinME), and was equal to NT 3.5 & 4.0. WinXP was a significant improvement and it took Mandrake/Mandriva to match it. But soon, a variety of Linux distros could match or exceed XP. Nearly ANY distro was better than VISTA. Win7 has returned to the stability of XP, with some improvements in usability, but nearly any distro running KDE 5.x or higher matches or exceeds Win7 in power, speed and usability. In my own case, running Kubuntu 10.04, running KDE 4.5.3, this desktop is significantly more powerful, IMO, than Win7 HE, which shares my wife's PC. I can't say it is more powerful than Win7 Ultimate because I've never ran that beast.

Linux has ALWAYS been more stable and less likely to crash, for any reason, than any version of Windows. That fact has remained true up until the late stages of XP and with Win7, both of which are nominally stable and rarely crash, at least in my experience. My XP VM never crashes (but, I rarely run it!), and the Win7 Home Edition on my wife's notebook has never crashed.

Crashing aside, the WORST problem that ALL versions of Windows still experience is that they are still, even with Win7, very easy to infect, especially if Joe and Sally do not keep their AV current, and many times even if they do. It's hard to explain a 4,500,000+ Windows zombie bot farm, many with active AV, other than to say the obvious: Windows running on home computers is still too easy to infect. Before I retired from the NE Dept of Revenue, our IT installed a Linux box running a $28K AV product to act as the gateway and filter to the Internet. That reduced the infections on our 400 Windows workstations from several per month to one a year. In fact, in the last year of my employment the only infections were those brought in on employees CDs and USB sticks. The IT staff quickly disabled CDs and USB ports on most employees workstations, except for developers

However, Microsoft has done something they should have done years ago: released a free AV product that is about as good as that genre can get, Microsoft Security Essentials. Microsoft has the advantage in that they have bludgeoned security houses, with a legal billy club, into given them the security holes they discover, and not releasing those holes to the public. But, that advantage does not go to the Windows users when Microsoft sets on the knowledge of an exploit which is active in the wild and does nothing about it, except to notify their big corporate customers, until such time as their bottom line dictates. Windows users are left exposed, hanging in the winds of adversity, the personal and financial information at risk.

Recently, FireFox and some other F/OSS projects have adopted the same tactic. This is utter foolishness. But, for the most part, the paradigm is to announce a hole immediately after it is discovered, along with a proof of concept code that would test any putative fix. Thus, the individual Linux user is allowed access the risk and to choose for themselves if they will suspend their use of the compromised application or not.

[barbolani]

Re: Another rant

Indeed, one of the best threads I've read in a looong time. You guys have made me understand what is behind the ranting, at least behind mine. Yes, at the end of the chain of questions there is ultimately one question: what are the goals of each party involved in (*)Ubuntu?

On a personal level, each of us have a different motivation to use and support Linux. While some may "scratch an itch", others just want "something that just works", others want to "move away from Windows", others want "freedom" (in whatever definition they happen to like) and others may want to "make money".

As individuals, each and everyone of us has different goals. Even conflicting ones. And our contributions, if any, to the community, reflect more or less those goals, if only because one tends to invest his time in things that he considers worthwhile doing.

That's all well and good on an individual level, but what about organizations? That's were I think I fail to understand Canonical.

If the goal were "Linux for everyone", then they would not trade off quality -defined as the size of the gap between the user expectations and what the product actually delivers- for goalposts or feature sets. Releases would be much more polished, even if late or not containing leading/bleeding edge versions. Yes, it is all awfully complicated by the dependencies on other projects (kernel, DE, HW drivers, etc), but that is what likely makes it difficult, challenging and beyond the reach of a volunteer organization.

If the goal were "put Linux on 200 million desktops", the emphasis would be on removing bloat at the expense of fanciness, and making sure HW compatibility was close to perfect. They would not even blink when deciding to remove features that were not robust enough for the general target user.

If the goal were "make money by taking away market share from MS", they would not invest in projects whose results (Unity, Wayland) are not likely going to provide a significant product advantage, and would focus instead on leveraging in as much as possible the free work provided by the rest of the community.

I could put more examples, but I think that the point is clear. We started using and supporting (*)Ubuntu somewhat believing that it was aligned with our goals, but more and more we fail to understand what its goals are by interpreting its actions. That's not the case with other companies or organizations around Linux. I can pretty much understand why IBM, Oracle, RedHat, SUSE, Debian or Gentoo are involved with Linux, and more or less their actions are aligned to their motivations.

Not so with Canonical, at least for me. Before finishing this post, I went to www.ubuntulinux.com and tried to get what the goals of Canonical are. I could not. Perhaps I missed something obvious. And without knowing what those goals are, I cannot truly understand the reasons for Canonical's actions, neither can I decide if they are aligned with mine. Hence my confusion and my ranting. And I suspect everyone else as well.

[Snowhog]

Re: Another rant

Not so with Canonical, at least for me. Before finishing this post, I went to www.ubuntulinux.com and tried to get what the goals of Canonical are. I could not. Perhaps I missed something obvious. And without knowing what those goals are, I cannot truly understand the reasons for Canonical's actions, neither can I decide if they are aligned with mine. Hence my confusion and my ranting. And I suspect everyone else as well.

Sustaining the Vision: Canonical and the Ubuntu Foundation might shed some light on this.

[bendy]

Re: Another rant

Jumping back in to the the thread I started, I must say I have enjoyed reading all the comments.

I think most of the problems people including me have encountered have been related to upgrading from a previous release and migrating from kmail to kmail2. I'm sure that most people doing a fresh install are having a smooth and rewarding experience.

Yes, Kubuntu is free (-$). Yes, it is put out there for people to use, help test, give feedback for development. Yes we are on a 6-month release (and we all know by now what that means in practice -- it means we have to continue to practice). And so on.

But what is all this about? What ARE the goals of Linux powers that be? From a marketing standpoint, of course the above critiques are correct. Hard to argue with that! In a commercial marketplace, you simply can not, must not, release a product before it is fully ready, lest you invite name-brand disaster. You can not release a product to the mass MS-Windows market that is not ready to go off-the-shelf (with respect to major apps like email, browser, photos, Skype-type-stuff, sound, and so on). But again, as I said, I really don't gather that this has anything to do with the focus and the goals of the Kubuntu developers/movers-and-shakers. So, back to the question: What is all this about (besides being, in random order, alternately fun and challenging and frustrating and satisfying)?

This is absolutely right. At the end of the day I haven't paid the kubuntu or kde devs a penny. And I haven't ever emailed the kde lists and expressed what my requirements are. So do I have any right to complain?

I guess my problem is that I have been using kde for over 10 years and until recently it has always been my perfect desktop that I was happy and proud to use. kde just looked and felt right for me, without me ever having to think about or define what "right" is.

But the direction kde is going now somehow doesn't feel "right" any more.

[SteveRiley]

Re: Another rant

As others have already written, this rant is indeed one of the more balanced and thoughtful threads carrying the title.

Regarding Canonical and their motivations, it's always struck me as a little self-destructive to build your primary goals around a complaint about your competitors. Namely, I mean bug #1: Microsoft has a majority market share. This kind of attitude smacks of an inferiority complex. Successful organizations rally themselves around something new, something different, something that (to continue the theme of my other post) confers selective advantage. To claim "the other guy is bigger than me, boo hoo" and then expect your own position to raise as a result of that claim indicates a failure to understand many things about what drives human nature.

[GreyGeek]

Re: Another rant

.....
But the direction kde is going now somehow doesn't feel "right" any more.

I have exactly the opposite feeling. I moved to Kubuntu from Mandriva specifically because they were offering the KDE 4.2 release, which Mandriva wasn't moving to for another 6 months to a year. As a bonus I got i915, which gave me faultless video acceleration, something Mandriva's i810 didn't offer.

My KDE 4.5.3 compares favorable to Oneiric's KDE version, running as a guest OS. Probably because I enable everything Dolphin offers. I'm getting antsy about the next LTS because I can't wait to jump into the version of KDE that it will be sporting.

[lcorken]

Re: Another rant

I use this computer the way I do but sometimes wonder how others use theirs. This discussion has given me some insight on that. We all have different needs.

Kubuntu 10.04 was the last Kubuntu that worked to my satisfaction on this aging hardware. I keep it on an extra partition. Still, being one of those who likes the latest and greatest I found that Lubuntu LXDE works fine and I can still install Konqueror, K3b and Kate on top of it. Can't live without those. That is, until Ornery 11.10. After seeing all the crap that Synaptic wants to install along with even just Kate alone I haven't yet had the courage to do it. May just do it and see. Worst case I can re-install Lubuntu.

Anyway, change will happen regardless of my opinion, so will observe and cope as best.
In nearly 50 years I still think the 1964 Ford Mustang was the coolest looking car Ford ever made. Later Mustangs just got ugly. Sorry, about the unrelated observation.

Ken.

[de_koraco]

Re: Another rant

Icroken
What does apt say if you run

sudo apt-get --no-install-recommends k3b kate konqueror

Still a bunch of dependencies or is it reduced?

[lcorken]

Re: Another rant

Thanks de_koraco.

I got this error.

ken@owl:~$ sudo apt-get --no-install-recommends k3b kate konqueror
E: Invalid operation k3b

I'l have to look into that --no-install-recommends thing.

Thanks.
Ken.

[lcorken]

Re: Another rant

Got it.

sudo apt-get --no-install-recommends install kate

Still wants to install 63 new packages. ??!!

What-the-hell. I'll go for it.

Thanks.
Ken.

[SteveRiley]

Re: Another rant

I'l have to look into that --no-install-recommends thing.

You can make this the default for apt by adding the following to /etc/apt.conf:

APT::Install-Recommends "false";

The semicolon on the end is required.

[de_koraco]

Re: Another rant

Thanks de_koraco.

I got this error.

ken@owl:~$ sudo apt-get --no-install-recommends k3b kate konqueror
E: Invalid operation k3b

Yeah, I forgot to add install, lolz
But yeah, the packaging is getting ridiculous lately, it's not just Ubuntu, my Squeeze box often wants to do horrible things. This is a scrotwm install, Openbox on there but not used, no KDE libs.

root@mrcbag: apt-get install kate
The following extra packages will be installed:
 kaboom kdebase-runtime kdebase-runtime-data kdelibs-bin kdelibs5-data kdelibs5-plugins kdoctools libattica0
 libclucene0ldbl libgraphicsmagick3 libiodbc2 libkde3support4 libkdecore5 libkdesu5 libkdeui5 libkdnssd4
 libkfile4 libkhtml5 libkio5 libkjsapi4 libkjsembed4 libkmediaplayer4 libknewstuff2-4 libknewstuff3-4
 libknotifyconfig4 libkntlm4 libkparts4 libkpty4 libkrosscore4 libktexteditor4 libkutils4 libnepomuk4
 libnepomukquery4a libphonon4 libplasma3 libpolkit-qt-1-0 libpulse-mainloop-glib0 libqca2 libqt4-opengl
 libqt4-svg libqt4-webkit libqt4-xmlpatterns libsolid4 libsoprano4 libssh-4 libstreamanalyzer0 libstreams0
 libthreadweaver4 libutempter0 libvirtodbc0 libxcb-shape0 libxine1 libxine1-bin libxine1-console libxine1-ffmpeg
 libxine1-misc-plugins libxine1-plugins libxine1-x oxygen-icon-theme phonon-backend-xine
 plasma-scriptengine-javascript shared-desktop-ontologies soprano-daemon virtuoso-minimal
 virtuoso-opensource-6.1-bin virtuoso-opensource-6.1-common
Suggested packages:
 khelpcenter4 konsole djvulibre-bin icoutils graphicsmagick-dbg hspell libqca2-plugin-cyrus-sasl
 libqca2-plugin-gnupg libqca2-plugin-ossl libqca2-plugin-pkcs11 gxine xine-ui libxine1-doc libxine-doc
 libxine1-gnome
The following NEW packages will be installed:
 kaboom kate kdebase-runtime kdebase-runtime-data kdelibs-bin kdelibs5-data kdelibs5-plugins kdoctools
 libattica0 libclucene0ldbl libgraphicsmagick3 libiodbc2 libkde3support4 libkdecore5 libkdesu5 libkdeui5
 libkdnssd4 libkfile4 libkhtml5 libkio5 libkjsapi4 libkjsembed4 libkmediaplayer4 libknewstuff2-4 libknewstuff3-4
 libknotifyconfig4 libkntlm4 libkparts4 libkpty4 libkrosscore4 libktexteditor4 libkutils4 libnepomuk4
 libnepomukquery4a libphonon4 libplasma3 libpolkit-qt-1-0 libpulse-mainloop-glib0 libqca2 libqt4-opengl
 libqt4-svg libqt4-webkit libqt4-xmlpatterns libsolid4 libsoprano4 libssh-4 libstreamanalyzer0 libstreams0
 libthreadweaver4 libutempter0 libvirtodbc0 libxcb-shape0 libxine1 libxine1-bin libxine1-console libxine1-ffmpeg
 libxine1-misc-plugins libxine1-plugins libxine1-x oxygen-icon-theme phonon-backend-xine
 plasma-scriptengine-javascript shared-desktop-ontologies soprano-daemon virtuoso-minimal
 virtuoso-opensource-6.1-bin virtuoso-opensource-6.1-common
0 upgraded, 67 newly installed, 0 to remove and 0 not upgraded.
Need to get 70.3 MB of archives.
After this operation, 180 MB of additional disk space will be used.
Do you want to continue [Y/n]?

[rec9140]

Re: Another rant

In a "directed" organization, whether corporate or nonprofit or whatever, it is possible (if not mandatory) to have a strategy. The senior

Linux has no such organization, and nothing that I would characterize as direction, therefore it is not possible to generate a strategy or a mission statement. It is developed by volunteers, and volunteers are, by definition, not directed by anyone else -- at least not with any enforcement. Some pieces of the community, such as Canonical or Debian, are more organized than others, but even so they are totally dependent on the larger volunteer base.

And is exactly what is one of its biggest weaknesses and strengths, but firmly in the weakness category and hurting adoption.

I finally figured out what nepomuk and strigi are for when I read a detailed instructional piece about them, by a guy who does Linux support all day every day. He has a zillion e-mail messages on his system full of technical issues and solutions, and he needs to search them multiple times per day. Hey -- wouldn't a grand semantic desktop search tool be cool? Enter nepomuk!

No! In a word. If its an option to add this cruft, fine, not the ram it down my throat option, screw up my email program, cause huge performance issues etc.. I have tons of emails too and search them just fine in KMail (not KMail2) with out any assistance from any thing else. One persons problem and solution is not a fit to a wider audience, except as an O P T I O N.