I was just browsing the web a while ago, when suddenly I got a popup window in the middle of my Firefox browser. It was a standard Vista popup, with the default colors and buttons and stuff, and a very authentic-looking MS Windows security icon, and it was churning through my filesystem doing a security check.

“Checking C:\WINNT\System\….”

The files were whizzing by so fast I couldn’t follow them.  After a minute, it gave me the bad news:  I had 19 viruses infecting my file system and over a hundred malicious entries in my Registry!  What to do? What to do !?

Wait!  I have no Registry.  I have no “C:” drive.  I’m running Linux, KDE4 desktop, with Ozone windows decorations.  If it looks like a Vista window, it can’t possibly be really on my system.  I was actually chuckling from the moment it popped up on my screen.  When it finished, and told me to click ->Here<- to download an updated “anti-virus” product, I just clicked the little X in the corner.  Bye-bye little malware.  I hope you had fun trying to infect a totally alien system!

As I’ve perused the various Linux news sites and blogs, one thing I notice is all the “How to Install <program X> in Ubuntu” – type articles.  Each time I see one, my first thought is, “don’t people know how to use Synaptic?  Or ‘Add/Remove Programs’ ?”

To be fair, Ubuntu and its brethren aren’t the only distros that I see with these articles, but they probably have 80 – 90 percent of them.  The reason, I’m sure, is that Ubuntu is the most popular distro among Windows-refugees, fueled in large part by the perception they have that Linux = Ubuntu.  Kind of a self-perpetuating cycle.

But, back to the point at hand.  Why do users need all the “How To Install…” articles?  Do they simply not comprehend how package managers work?  Are the articles simply catering to the common Windows perception that software must be obtained from random places on the internet, and installed separately?  Do major distro developers not know how to package updated versions for their current users?  Is the “fixed-point release” methodology of major distros to blame?

Almost none of the articles are simple “Open Synaptic, find the package, select for installing, and click apply”.  Sometimes they do point to a specific “ppa” repository, but it’s still followed up by a lot of command-line configuring to get it right.  Usually, it’s “go to this website and grab the tarball….”  I could understand if these were obscure, niche programs, but they’re almost invariably popular applications that I am sure exist in the primary repositories.

I know Windows users are used to installing software from random locations on the web.  I used to do it, too.  Maybe this just scratches that itch, to go searching the web for the software to install.  Maybe it makes them feel better about themselves, because they can do this “difficult Linux thing”.  You know, “Linux is different from Windows.  It must be hard.  See, look at how much work it takes to get a fully functional system.”

The distros themselves bear some of the responsibility, to be sure.  With fixed-point releases, new versions of software tend to go into the testing area for the next release point, just a few months away.  They rarely seem to get backported to older releases that people are still running (except maybe to Long-Term Support releases).  When people ask for newer versions of their software, they are told to simply upgrade their systems.  Never mind that release “Y” broke some functionality from release “X” that the user needs, and is why they never upgraded in the first place.

I’ll just stick with a rolling-release distro.  As new versions of applications are made, they get packaged and tested against the current system, and are made available for all users to install.  My nearly three-year-old install of PCLinuxOS2007, through the “magic” of periodic updates, is running exactly the same software as does a system that just had a 2009.2 install.

Maybe I’m just getting to old to understand.

I’ve been playing with KDE4 under Kubuntu, and generally liking what I have seen.  But Kubuntu itself has always felt … foreign … to me.  Maybe I’m just getting old and my brain cells are ossifying, but while KDE4 seemed decent, I found myself booting into PCLOS more often, because there, everything “just works”.

Recently, Tex released his current 4.3 implementation to the world.  This weekend, I took the plunge.First, let me say that KDE4 on PCLOS is not 100% ready for the masses.  While I haven’t seen anything major, there are some niggles and glitches here and there.  Many distros might be happy with it in its current state, but it still feels a little rough around the edges for PCLOS.  Also, the devs recommend that users running critical apps stick with KDE3 for their primary system, and install KDE4 (if they want it) in a separate partition.  As far as Tex and the gang are concerned KDE4 in PCLOS is in “Testing”.  Be forewarned.

Installing KDE4 is simple.  Open Synaptic -> Settings -> Repositories, click on your selected repository, and change “kde” to “kde4″ in the Sections box.  Close the dialog, reload the repository, and install “task-kde4″.  One point to note: this will remove all of your existing KDE3 apps and libraries.

“But wait!”, I hear you say.  “Other distros allow KDE3 and KDE4 to exists side-by-side!”  That is true.  But to do so, they have to either build KDE4 to live in /opt instead of /usr, with the necessity to rebuild everything to live in /usr when KDE4 becomes the default desktop environment, or they build KDE4 in /usr, and rebuild KDE3 to live in /opt.  But Tex is the only dev building KDE4 packages for PCLOS right now, and he has said he isn’t going to deal with changing build environments for either version, considering how many packages are involved.  You get either KDE3 or KDE4.  Just another reason to put KDE4 on a separate partition.

Anyway, I went ahead and installed KDE4, watching a huge list of KDE3 libs and apps get removed.  I then installed “task-kde4-extra” to get some more of the common libraries and apps.  That done, I did a full reboot to be sure everything was reset.  Probably didn’t need to, but, what the heck.   When it came back up, I was looking at the KDE4 version of KDM.  Cool.

I logged in, and … I am in KDE4, right?  Yeah, there’s the “cashew” up in the corner.  But, there’s the taskbar, looking much like I left it, last.  The usual set of icons on the desktop.  Wait, icons on the desktop?  I right-clicked on the desktop, and looked at Properties.  Yeah, there it was.  Tex has the default set to be “Folder” view, instead of “Desktop”.  And the default folder showing is the old “Desktop” folder.  Makes everything look familiar.  And the default “start” menu is the classic view — more familiarity.  Kwin now has a lot of the Compiz functionality built in, so I don’t need to add Compiz-Fusion to the system.  I can easily set the amount of “eye-candy” I want (and that my poor old graphics card can handle).  I spent the better part of a day playing with widgets and such.  I finally settled back on gkrellm, though, plus a nice weather widget that seemed as good as Liquid Weather.  I’m already familiar with Dolphin, from playing with Kubuntu.  Firefox 3.5 is still there (no kde dependencies).  KVirc went missing, but I’ll check out some KDE4 irc clients before I compile that.

I think I’m going to like this.

Okay, you all probably know I love PCLinuxOS. It has been, and will continue to be for the foreseeable future, my primary OS.  But I’ve been looking at KDE4 for a while, and with KDE4.2.3, it looks about ready to use.  So I cast about for a good KDE4-based distro to play with.

Generally, I haven’t had much use for the *buntu family.  Every time I have tried one of them, it has left me feeling, “Ho, hum. Another average distro.”  I haven’t bothered actually installing one for years.  But Kubuntu 9.04 was highly recommended by an online friend, so I thought I’d give it a shot.

I downloaded 9.04 – Jaunty, burned it to CD, and rebooted my system.  My first impression: seems kind of slow, but then, it IS running from the CD.  I played around a little, and KDE4 seemed to be in pretty good shape.  So I figured, I have a spare partition, let’s try it installed.  I closed the apps I had running, and hit the “Install” button.

Hmm.  Nothing is happening.  Neither the hdd nor the cd lights are blinking, either.  Maybe I didn’t actually double-click it.  Let’s try it again.

Still nothing.  Maybe I can open a Konsole and see what’s going on.  Hmm.  The mouse moves around, but nothing seems to function when clicked on.  Maybe I can just switch to a full console window: Ctrl-Alt-F2.  What?  That didn’t work either?  Try killing X: Ctrl-Alt-Bksp.  Nothing there, either?  Screw it.  Maybe the graphical installer from the desktop doesn’t work right.  I hit the Reset button, and waited for the grub menu to show up.

Yeah, there it is.  I thought I remembered an install option from the menu.  I choose that, and settle down to do this.  Sure enough, here’s a nice Kubuntu splash screen, with a zylon scanner running back and forth.  I wonder what’s going on behind the scenes?  I hit the Escape key, so I can watch the background processes go by in text mode.  WTF?  Nothing happens, just the little zylon going back and forth.

Okay, I’m not going to wait forever to see some indication of actual processing going on.  I’ll give it 30 more seconds to do something, before I shut down and throw this disk away.  Finally, just before I killed it, and over 45 seconds from the time I selected “Install”, the initial installation screen comes up.

Now I’m on somewhat familiar ground.  English language.  US keyboard.  Etc.  Etc.  Ah, here we go: where to install?  I will tell you where, so “Advanced”.  No, not that partition.  I want you to go over here.  A few more configuration steps, and “Finish”.  Go do what I told you to.

At least it’s telling me what it’s doing, now.  No need to just sit here, I’ll go help / talk with my wife for a bit.  A while later I check on my system, and it is finishing up with the install.  It took a little longer than my PCLinuxOS installs, but what’s a few minutes between friends.

All right.  Reboot, remove the CD, and wait for it to come back up.  Yes, there’s the grub screen.  Good, it found my existing PCLOS installation.  Let’s go with Kubuntu latest kernel.  Hmm, there’s that zylon again.  I’d rather see a verbose boot, just now.  I hit Esc.  WTF?  I can’t switch to a verbose boot?

It feels kind of slow to boot.  I didn’t time it, so I can’t say for sure.  But subjectively, it feels slower than PCLOS.  But up it does come.  Kde4 desktop.  I enable compositing.  Lots of shiny “eye-candy”.  I grab the latest updates.  I install a few things.  Turn on some widgets.  Visit some web sites.  Play some music.  This seems pretty solid.

Actually, my only real complaint, other than not being able to toggle to a verbose mode when booting, is that Kpackagekit, the default “Add New Programs” tool.  seems a bit unstable.  Every so often, when I went to install something, it would crash.  If I restarted it, and made the same selections, it would work fine.

I think I’ll keep it for playing with KDE4.  At least until Tex and the Gang put a version in the PCLOS repos.

The last few months have been … interesting.   Beginning about ten months  ago, Texstar “disappeared”.  He turned over the development of PCLOS 2008/2009 to the Ripper Gang, and didn’t say anything to the community.  A couple months ago, with 2009 almost ready to hit the streets, he popped back up, and delayed the release until he pronounced himself satisfied.

What happened?

First, my usual disclaimer that I am not a developer for PCLinuxOS, nor do I have any special inside track to what has been going on.  I’m just piecing things together from disparate public sources.  Since everything I have read is publicly available, I haven’t asked anyone for permission to discuss them, but I will exercise some discretion regarding details.  There is also some conjecture, because I have read some of both sides of the story, and while they describe the same incidents, they are so divergent that it is almost hard to grasp that.

Tex left because of some serious health issues.  He needed to concentrate on that, without the distractions of what is, to be honest, a hobby.  He turned the website administration over to Cindy, and left the Ripper Gang to continue developing the next release without him.  He left without an announcement because he didn’t want to disrupt the community.  If anybody asked, they were told he was on hiatus.

The Ripper Gang continued development, albeit a little more slowly than with Tex around, and everything seemed to go smoothly.  Until about two weeks before the 2009 release.  Then, as Tex puts it, “things started to get funky.”  While he wasn’t posting on the forums, or helping with the release, he was still keeping in touch.

Supposedly, some of the developers were taking the distro indirections they wanted, without input from the community, or even other team members.  They weren’t communicating well with the community, to the point many of us were begging for information.  This is not what Tex had envisioned for PCLinuxOS.

Emails and PMs started flying, getting more and more heated.  Many angry words were exchanged.  In the end, the Ripper Gang was ripped apart.  Many of the development team that worked on 2009 have departed from PCLinuxOS.  None were forced to leave.  Those who were most opposed to Tex’ approach to the situation felt they could not work with him anymore, and others, I believe, simply wanted out of the charged atmosphere that was left.  They have formed a new development group, named Unity, and plan to develop a new distro.  Some PCLOS remasters, e.g., TinyMe, have joined them.  Others, like the Gnome remaster and BEL have stayed with PCLinuxOS.

I know Tex was very hurt by the way all this played out, pitting friend against friend, and airing all the “dirty laundry” in public.  He still respects their abilities.  He admits that without them, there would be no PCLinuxOS 2009 in our hands right now.

On the positive side, those developers who left stayed with PCLOS long enough to wrap up loose ends.  MyPCLinuxOS.com and the Hardware Database have been handed off, and the development on their various packages was completed before they left.  A new crop of developers is in training, and a new selection of Moderators is managing the forums.  One of their promises is to be more open with the community, so we aren’t left in the dark.

I have put together this list of (nearly) fail-proof steps from doing my own updates, as well as the experiences of others on the PCLinuxOS forums. (Especially maddogf16: thanks, guy!)  These steps have not failed yet on a clean 2007/2008 install.  If you are working from a remaster, and/or have done a lot of system tweaking, including adding/removing packages, it might screw you.  I don’t really think it will, but you need to be aware of the possibility.

1. BACK UP YOUR IMPORTANT DATA! (That should go without saying, but I’ve found that usually means it needs to be said.)
2. If you are running Compiz, turn it off.  You can turn it on again after the upgrade.
3. Launch Synaptic.
4. Uninstall mozilla-firefox.  (The new FF is installed during the full upgrade.)
5. Uninstall Compiz. (The upgrade installs Compiz again.)
6. Uninstall Open Office.  (The size of this app really slows down the upgrade, and can cause problems, later.)
7. Click RELOAD.  DO NOT click Mark All Upgrades.
8. Find GTK+2.0, select Upgrade, and Apply.
9. When this app is finished upgrading, close Synaptic.
10. Do not try running the Update with Synaptic at this point.  Synaptic will freeze, and not finish your install correctly, leaving you with a mess to clean up.
11. Open a Konsole.  (If you’re Command-line-phobic, don’t worry.  This is pretty straight-forward and simple.  Just type carefully.  Or just copy and paste each step.)
12. Run the entire update from the cli.
    • type su and enter your root password at the prompt.  Your prompt should turn red, indicating you are running as root.  Then enter the following commands one at a time, waiting for each to finish before continuing.
    • apt-get clean
    • apt-get autoclean
    • apt-get update
    • apt-get upgrade
    • apt-get dist-upgrade
    • Type exit to get out of super-user mode, then exit again to close the Konsole.
13. Reboot
14. Select the new kernel at the bottom of the list 2.6.26.8.tex3
15. Launch synaptic. Reload, Mark All Upgrades, Mark, Apply.
16. Search for libmozilla-firefox-2.X (not sure of the name because it is no longer on the list on my system) and remove it if it still exists.
17. Install Open Office.
18. If Firefox did not reinstall do it now along with everything else you want to install or remove.

You should now have a complete, updated system, running on the new kernel.  You will want to check to be sure all your hardware is recognized, and functions properly.  So far this checklist has succesfully upgraded dozens, at least, of systems without a hitch.

I got involved in a “Happy Birthday, Linux” thread over at Civfanatics.  I had to stop myself before I got too long-winded and argumentative.  All because one of the posters asked a simple question, “why should I switch to Linux?”  I guess that made me a little extra sensitive to the subject, because I have noticed today a couple of items about switching to Linux, and MS “fighting back”.

First, Datamation has an article about The GNU/Linux Desktop: Nine Myths.  He discusses some of the most frequent arguments people have for not using Linux, and basically debunks them all.  You know the ones: “You have to use the command line”, “There’s no hardware support”, and the like.  He didn’t come right out and say it, but the most common source of these myths (FUD) is MS and their apologists.

While I was musing about that, I found a new blog post from helios, TOYA Boys Fight Linux Flow.  And he’s wondering about pretty muchthe same thing: why do apologists for MS fight so hard against Linux?  MS-paid astroturfers are one thing.  But why does the average “Joe Sixpack” fight so hard against the Linux tide?  Read his post.  Follow his links.  And ask yourself the questions he poses at the end.

“First they ignore you, then they ridicule you, then they fight you, then you win.”
—  Mahatma Gandhi

PCLinux2009.1 hit the mirrors today.  (The .1 because there was a wifi bug that Tex fixed last night.) There has not yet been an official announcement.

Update: The announcement has been made!

The Ripper Gang is pleased to announce the final public ISO release of PCLinuxOS 2009.1.

This release features kernel 2.6.26.8.tex3, KDE 3.5.10, Open Office 3.0, Firefox 3.0.7, Thunderbird 2.0.0.14, Ktorrent, Frostwire, Amarok, Flash, Java JRE, Compiz-Fusion 3D and much more. We decided to use kde3-5-10 as our default desktop as the we could not achieve a similar functionality from kde4. We will however offer kde4 as an alternative desktop environment available from the repo once we stabilize it.

Okay, technically, the 2009 iso isn’t out yet.  But the repositories have been unfrozen, and the “Big Update” was pushed to all the mirrors.  I updated a couple of weeks ago.  So how is it going?

To begin with, my first try at updating was a no-go.  It was all my fault, I admit.  I had been multi-booting, playing with other distros while I waited for 2009.  I knew I had screwed up my grub.  I had tweaked it enough to let me boot PCLOS.   I knew I had to go in and fix it before I did the update and reboot.  I knew I had to….

<sigh>

To make a long story short, I ended up reformatting my root partition, and reinstalling PCLOS-2008 MiniMe.  But from there, I re-did the Big Update, and everything was beautiful.

I’m currently running the 2.6.27 kernel, manually updated from the default 2.6.26.  All my hardware is correctly recognized (always good to not regress there!), and of course, there has been a large amount of software upgraded.  At this point, I have seen no problems with the upgrade, and definitely some improvements.

For instance, while 2008-minime recognized my ATI Radeon x1050 video card, and I could (and did) install the proprietary ATI driver, and it generally worked fine, it didn’t work entirely correctly.  One symptom: I could enter the PCC, and tell the system to use the proprietary driver, and click OK, and it would work.  But if I clicked the “Test” button, my X Server would go off to la-la land, leaving a gray screen, and I would have to kill X (Ctrl-Alt-Bksp) and re-login to get back.  In 2009, it works correctly.

In essence, PCLOS has gone from a solid, stable, “just works” system to a solid, stable, “just works” system.

Kudos to the Ripper Gang for all their hard work on this upgrade!

UPDATE: The Ripper Gang has just announced another ‘Little’ Big Update.  Nothing like the last “Big Update”, but still a lot of application fixes for things that have been discovered during TR6 testing, and in general with the Big Update.  Expect it by this next weekend, if not before.

I ran across an article this morning, Linux Versus the Microsoft Trained Brain Syndrome.  I thought it made some very good points.

The problem is that people who have lived, worked, and played in a homogeneous Microsoft computing paradigm are lost and confused when they encounter a different paradigm. These people have only seen the flawed Microsoft ideology for how computing systems should work and so have a difficult time with more elegant systems based on Unix. They see the Linux system with its’ own paradigm and ideology and try to force it into the only paradigm they know, which is Microsoft’s. This will always cause the user problems.

For myself, the first computer I worked with, way back in the pre-MS 1970s, was an IBM mainframe, I think even older than the 360 series.  I wrote FORTRAN programs on it.  Later, the college got an IBM 360, with (gasp) time sharing teletype terminals.  No more card decks!  I taught myself BASIC on those.

Professionally, I started on Honeywell mainframes, which used card decks and then video terminals.  The terminals gradually got replaced by DOS-based systems running a vterm emulator, and then by Unix-based terminals, also running a vterm emulator.  Only when I was stationed in Italy in the mid-90s did I work primarily with Windows-based systems.  When I retired and returned to the States, my first jobs involved replacing the old mainframe software I used to work on with Unix-based distributed networks.  I then spent several years in a predominantly Windows-based environment, but gradually was able to transition to more-and-more Unix.  Today, in a new job, I spend most of my time ssh-ed into a variety of Unix and Linux servers, writing shell scripts, perl scripts, html pages, and other programs using vi/vim.

At home, I bought an Atari 400 back in the early 80s, and other Atari machines up through the 65XE.  While I was in Italy, my wife picked up a cheap 486 running Windows 3.1.  MS had made its inroad into my house.

For years, I was an MS apologist.  What they produced was inexpensive enough for the home market, and they provided standards for GUI use.  (Their own standards, but at least they were internally consistent.)  Besides, it was the type of system that pretty much anybody who used a computer had.  But I noticed things.  For Win 95, the standards changed.  Okay, 95 was a major GUI redesign.  But then they changed again for 98.  and for Me.  And for 2000.

I was also getting frustrated.  I’m a professional computer geek, for crying out loud!  I should be able to fix my problems.  But all I could do was click a button on the screen.  If that didn’t fix a problem, I could sit on the phone for hours with a Tech Support person, who would tell me to click that same button.  And when that didn’t work, tell me to try reinstalling Windows.

I started dabbling with Linux in 1998.  This was a paradigm I understood!  If clicking a button didn’t work as expected, I could check the data behind the scenes.  More importantly, I could even check the code that worked with the data!  In early 2003, I wiped my personal desktop clean, and installed only Linux.

My wife, though, still suffers from MTBS.  Both her laptop and her desktop run Windows (XP and Vista, respectively).  But she is slowly gaining the appreciation that Windows is not automatically “good”.  She never really liked XP, and Vista can make her want to tear her hair out.  But the “pain” of switching to Linux still outweaighs the “pain” of using Windows, for her.