As I’ve said before: A person’s experience “switching” to Linux will vary as much as their DNA varies.  Every person has their own unique needs, preferences and capabilities.  This is not lost on an operating system by any means.  If you’re the kind of user that is 100% happy doing nothing but Gmail/Hotmail, YouTube, reading news, and playing games, you will likely be 100% happy with almost any operating system, including Linux (whatever distro gives you wood).

If you are a seasoned Windows or OSX user, who has developed an intricate web of applications, scripts, utilities, and configuration settings to do an assortment of specific tasks, you may not be as happy… or happy at all.  You can still do email, surf the web, manage office suite documents, play (some) games and watch YouTube all day long.  But some of the dearest things you depend upon in OSX or Windows will not make you happy on Linux.

Am I picking on Linux?  Not really.  However, it brings up the viability of the argument that Linux is “ready” to knock Windows “off the desktop” of most computer users.  For some it is.  For some it is not.

I could cite specifics like lack of specific hard device drivers, flakiness of video options, limitations and complications of various FOSS applications, and nebulous support.  To be fair, Windows can suffer from those very same issues (except that FOSS apps are only one option, not THE option).  Case in point: Which is better?  Posting the same confusing question to dozens of discussion forums and getting flamed with answers like “RTFM!” and “stupid noob”, or pulling out the credit card to call for suppor, getting put on hold and then getting some half-baked goon in Bangalore who not only can’t understand the question but reads the “solution” from a queue card and closes the case.  Not much of a choice, is it?

As for applications… I got a ton of emails and blog replies suggesting this or that as a “perfect” or “near-perfect” replacement for “x” on Windows.  Almost NONE were “perfect” replacements.  In fact, only one was (GPAR2 in lieu of QuickPAR).  K3B is not a replacement for the entire Nero Ultra suite.  Not even close.  PAN is not a decent replacement for Newsbin Pro.  gtkPod, RhythmBox and Floola are not 100% replacements for iTunes (even though I despise iTunes).  Firefox and Opera still have problems (even on Windows) dealing with some sites developed only for IE.  OpenOffice still does not provide 100% compatability with MS-Office documents.  Gimp is a very worthy contender for replacing PhotoShop CS2 however (maybe CS3).

And that’s just the apps that exist for Linux.  Then there’s the ones that don’t.

There is no AutoCAD replacement for Linux.  There is no Inventor or SolidWorks.  There is no Macromedia/Adobe Flash MX.  Sure, there’s Wine, Parallels, VMware and VirtualBox.  But those are cop-outs.  They offer a half-way solution to getting the job done.  But that’s a lame excuse.  Until these types of applications have 100% suitable alternatives there will be a large segment of OSX and Windows users that simply cannot “switch” without paying a heavy price.  Even the price of searching for replacements, and then learning how they work, and then learning to how to tweak and optimize them, is a very expensive proposition for some.  I’m speaking of those who rely on their apps for their careers.  Downtime/learning time is expense time.  It’s time lost from executing work, spent on retooling, and with almost no garantee of a net gain in the end.  Only a sideways shift or lateral move.  For a business environment that makes no sense at all.  In many cases, that cost is not recouped by the elimination of software licenses, mainly because the OSX or Windows licenses have already been paid for.  If you dump them after that, you’re literally throwing money away.

What Linux and the surrounding “community” needs (and it’s not really a community at all, it’s more of a collective label of “people that hate Microsoft and Apple”), is a desktop environment, and a complete suite of all of the top 50 or 75 applications used by OSX and Windows customers.  That would be the starting point.  If there were a package like that, it would already be sliding into every corporate office everywhere.  It’s not.  It’s creeping into IT backrooms and server rooms and IT geek desktops.  Until this is resolved, Windows and OSX are only going to continue growing.  And growing they are.  If you doubt that, check the investor reports for each company for yourself.

I really don’t care who wins.  I really don’t want ANYONE to “win”.  I want to see intense competition.  It makes better products, and cheaper prices.  We all win.  I just think Linux and FOSS could do more to put pressure on Apple and Microsoft than they are.  The fact that Windows and Apple prices aren’t dropping drastically is a clear indication that FOSS isn’t eating away at their profit margins.

After running into some problems meeting an urgent demand to get work done for my brother and some friends, I found running the apps inside VMware unsuitable.  If not for performance reasons, well, for performance reasons.  And besides, as I said in an earlier post: Spending a significant amount of time doing work inside a VM amounts to stupidity.  It’s not a “native solution” as I also said.  If I have to explain that, well, you’re just not going to get it anyway.

I thought, well, this might be a good time to see if the “Complete Backup” I created within Vista actually would work.

I booted to the Vista DVD and ran the repair option.  I pointed it to the backup I created on the second hard disk and let it rip.  It restored alright.  Did it boot?  Nope.  Why?  Because GRUB was still sitting in the middle of the sandbox and not letting any little Windows kids come in to play.  My guess is either the boot record or MFT still had GRUB hooks.  “Sector Error 17″ is all I got.  Enter: Vista install take II:  A complete partition rebuild.  I blew away the partitions and created one big partition, reloaded Vista and was finally back on the evil dark side of operating systems.

Problems solved?  Not exactly.

After Vista, I soon found that I still needed to install some Dell updates to get things back in order.  I will say this: if you have a Dell Dimension E520/521 and you’re running Vista: get the latest updates/drivers for the BIOS, hard drive, video card (nVidia GeForce 7300), and most importantly: the DVD/RW drive.  Before doing that, the performance was horrible.  Operations would time out just trying to read from media.  After the updates, all is working great.

I’m not giving up on Linux.  I still have my lab machines at work running Ubuntu and the VM’s with Ubuntu and Fedora are still in use.  I will post my reflections on trying to get Linux apps to reliably replace Windows apps for my needs.  It was a mixed blessing.  I continue hoping that will improve with time.

I’m not by any means trying to convey or imply any significance to those specific “roles”, just giving them as examples. I would say that many “casual” computer users would be 100% fine with Ubuntu. I dare say many would hardly notice. I’m talking about the kind of people that are awe struck at AOL. Not that this bodes ill for Linux by any means. In fact, it’s very good. The major appeal of Windows for AOL-types is ease of setup and use. Ubuntu is damn good at supplanting those attributes. Maybe not the initial setup, but then again, most of those types of users buy a computer already setup. Let’s keep this Apples-to-Apples.

For many users however, life on a computer is entirely dependent on a select group of applications software. Some of these work on Linux, some work poorly, some not at all. I do not ever consider Wine or VMware a real “solution” because it’s a bandaid on a bigger problem: a dearth of equivalent native applications. Argue all you want, but that’s what it is. Running MS-Office in a Wine session, is still MS-Office. That’s not a “clean” solution in my book. OpenOffice is (or can be, sometimes). This brings into question another argument: Should Linux have to mimick every aspect of Windows to be considered “successful”?

No. Not really. But…

It all depends on what you define as “successful” for the Linux market. Is sales and marketshare the measure? Is it vendor support and product availability? Is it magazine articles praising it’s virtues? I’ve yet to get two Linux evangelists to agree on this subject. On one hand, many will blurt out boldly that Linux can easily displace Windows on the desktop of most users. Then they hem and haw on the particulars of exactly how that could or might happen. The problem is that Windows didn’t win by technical superiority, it won by clever marketing and lucky timing. There is no “clever marketing” that I’ve seen for Linux, because there isn’t enough money behind it yet. Luck hasn’t been too kind to it either. Sure, things are improving, but very slowly based upon the numbers. Linux needs a real ass-kicking, ball-busting sponsor to take the reigns and drive the herd to the creek (in one direction). Otherwise, it’s going to remain the sideshow freak for technoweenies to gawk over. (for the record, I’m writing this on my Ubuntu desktop)

Windows isn’t a technical total failure either. It has it’s share of merits. As Mark Shuttleworth of Canonical Ltd recently said, Microsoft did contribute heavily towards the trend of cheaper priced software. It’s funny how quickly we forget that and start beating them up for escalating prices. (In case you didn’t know, Mark is the founder of Ubuntu) Remember how much a UNIX workstation cost back in 1985? I do. We bought plenty of them and I was around the Navy when they were shelling out $30,000+ for each one, and getting truckloads at a time. A typical Dell Dimension E521 today will blow one of those away in performance and capability. Part of that was the dramatic drop in operating system and basic app prices. UNIX and the proprietary apps that we bought were outrageously priced, even then.

However, what bothers me about Microsoft today, is that they’ve given up on core innovation. They still innovate in very focused areas, and certain products. But as a whole, they’ve become a follower. Chasing AOL, Netscape, Real, and now Google and Adobe (with Silverlight). Everything they’re hyping with maps, web-apps, is all catch-up. Even the newer “server core” Longhorn is a catch-up to UNIX and Linux. Pure and simple. And the more recent legal attack againt FOSS is just too much. It smacks of desperation and angst. Meanwhile, their products are becoming infused and obsessed with “rights management” at the expense of complexity, licensing terms, increased restrictions and increased prices.

I had to give Linux another chance.

I’ve been a casual user since college (late 1990′s) with Slackware, then early RedHat, then SuSe, and Mandrake, then on to Fedora and now Ubuntu. Of course, I’m basing a lot of what I’m saying on the Gnome desktop moreso than the KDE desktop. For the record, I still think KDE is too much and Gnome too little. I wish someone would crank out a happy medium.

So, what am I trying to say? Good question. The answer is this: Whether you will “like” what you get when “transitioning” from one operating system to another is going to depend on multiple factors. One of the most significant of these is the types of applications you prefer and rely upon. If suitable replacements exist for the new operating system, you may be perfectly happy. Maybe ecstatic. Maybe not. Then there’s the tedious hunt for replacements when there really aren’t many around. Do I need to list specifics? No. That would simply be picking another argument to waste more time.

I can list quite a few apps on several platforms that are difficult, if not impossible, to replace on another platform. It’s all market-driven. That’s why the CAD/CAM world is still stuck with too many products and file formats. It’s why AutoCAD doesn’t run for shit on Linux or OSX. It’s why you can’t run OSX inside VMware on a Windows host. It’s why Microsoft is still building translators and converters for MS Office 2007 to XPS, ODF and OpenXML formats to make the EU and China happy. It’s why there isn’t a native iTunes client for Linux. Homogeneity is not only boring, but profit-challenged. Water, temperature and electricity flow best when two points in the circuit are not equal. Same with markets. As long as each platform has differences, there will be products and services sold to bridge those gaps.

Going back, sort of, to Clive’s article, he provides a very insightful list of Windows-to-Linux application comparisons.  It’s missing quite a few that I happen to really need (on the Windows side), but even a few which are EXTREMELY IMPORTANT to many Windows users besides myself.  Examples?

• iTunes –> Floola is the best Linux candidate I’ve found yet.  Though, still not perfect
• Autodesk AutoCAD –> ???
• Autodesk Inventor –> ???
• QuickPAR –> ???
• SQL Server 2005 Mgt Console –> ???
• IIS 6 Mgt Console –> ???
• UltraISO/WinISO –> ???

Clive also says “Nero” equates to “K3B”, but “Nero” is only one name for a SUITE of apps which are bundled together.  NeroWave, and NeroVision don’t have counterparts in K3B that I’ve ever seen.  Burning CD/DVD disks is mindless drivel.  Building DVD player menus, covers, and transcoding on the fly are a bit different.  I’m still hunting for the ??? items above.  (feel free to drop comments with suggestions if you’d like)
I’m digressing again.

So far, it’s all good.

Some interesting bumps along the way though, which I can work around for now…

- The NTFS daemon that comes with Automatix works as described, however, there seems to be some translation overhead that interferes with VMware if I try to use the 2nd hard drive (formatted NTFS) as my guest machine storage point.  It reads fine, but simply crashes out if I try to launch a VM from it.  I can create new VM’s and it makes all the files, but something happens at runtime that simply kills VMware when I try to launch it, even if it’s not been setup yet.  Once I moved the guests back to my system drive, all is well.  I will probably wipe and reload the 2nd drive with ext3 to make things smoother.

- Floola is very nice, in some ways nicer than iTunes.  However, it has some quirks with double-downloads on podcasts, and fails to eject my iPod when quiting and pressing Shift at the same time (as it says to do).  I can still eject from the desktop shortcut though.

- Picasa for Linux.  Google finally posted it in their Google Labs section.  It works nice, but I logged a complaint that the entire folder structure shown in any of the “Folder” dialog forms is all Windows-based.  Things like “My Documents”, “My Pictures”, etc.  Stooopid!  Makes no sense for Linux at all.  I’m guessing (hoping) it was simply an oversight and not intentional.

One interesting experience I have to share is regarding the installation of VMware Server 1.0.3.  Until now, I’ve always gone to VMware and downloaded the TAR source and done the brute-force install.  That’s usually worked, but I’ve always had to tweak and load and so forth.  Low and behold, I stumbled across something I should have seen before: Synaptics Package Manager.  It has a pre-built VMware Server 1.0.3 package, and it loads the required kernel modules.  And guess what else?  It works!  I tested it three times, twice inside VM machines and once on my physical machine.  No problems at all.  So, I have to ask Linux users why is that on all the discussion forums, blogs, etc. they suffer with doing the source install, but nobody every suggests using the ready-made packages?  Hmmm. I’d scratch my head, but it’s a little sun burnt today.

I still have more things to tweak before I can say my new desktop is equivalent to (or better than) my former Vista setup.  However, I’m happy with what progress I’ve made so far.

My first “mouse” was a 16-button digitizer puck. When I saw the first Apple mouse I fell out of my chair laughing hysterically. Mac nuts obviously take offense to that and start getting huffy and puffy about only needing one button. Now they act all heroic since the “mighty” mouse has finally caught up to Logitech (and in some ways surpassed them, but only for a little while). One button. Stupid. I don’t care what rationale you try to back that up, it’s stupid. I have four fingers and a thumb. I could “FLY” on CADDS and Intergraph with one hand on the mouse and the other on the keyboard. My typing rate with my left hand alone was close to 40wpm. Whatever. I made $6/hour back then. I was single and my only expenses were beer and gas.

In college I had to work on UNIX (Sun) in the CS labs, Mac OS7.x and OS8 in the math labs, Windows 95 in the English labs, and Linux (Slackware/X11) in the Physics labs. Each day I had to move from one lab to the next. Was I special? Hell no! We all had to do that. It was tedious but it was worthwhile. We eventually moved “up” to Solaris, OS9, Windows NT and RedHat 6/7 by the time I graduated. Along they way, we got to dissect and beat up NeXT, BeOS, and even Rhapsody (codename prior to OSX). Diversity was fun.

Why all the digression? Stay with me…

Most people that jabber all day about “switching to Linux” or whatever have only known Windows or (maybe) DOS their entire computer lives. DOS and Windows were second for me. I’ve been using Windows almost exclusively at work since 2000. Windows NT, then 2003/XP and now Vista and Longhorn are what I play with. I also have VMware, so I still stay familiar with various Linux flavors.

Still with me? You really have no life, do you? Press on…

I tried converting my home computers (note plural) to Linux to see how it would fare, but I ran into many of the well-known problems that other Windows boneheads run into. Especially as it relates to having kids. As it turns out, my kids have to be able to run their games, do their homework, and so forth. As much as I tried, some things just didn’t work. I’ve revisited the idea multiple times, and still do, as new Linux distribution releases come out.

So, what problems you ask?

• Games. Kids games don’t play for shit on Linux. Yes, some do. Most do not. Even with Wine, they play horribly. Kids don’t have patience for performance hits or graphics degradation. They aren’t bashful about complaining either.
• Homework. I would estimate about 75 percent of the documents they typed up in OpenOffice were fine for their teachers. 25 percent didn’t make the cut. The teacher would open it up in MS Office and immediately start nailing them for things OO didn’t do the way MS-Office does. Grades were duly noted.
• Media library. I tried Banshee, Amok, and just about every other media player/library manager I could find. The problem is that I have all my media on a server. MP3, AVI, MOV, MP4, WMV, etc. While they usually play things well, they do NOT like UNC pathing. They all seem to want a local drive mount. Trying to point them to a UNC path only caused problems (hangs, errors, crashes). Why can sucky-assed MS WMP deal with that easily but these supposed superior (however, uglier than shit looking) tools not?!
• iTunes. Yes, the dreaded iTunes. Possibly the worst piece of shit media app for Windows ever written (or hatched from a school of monkeys). As much as I hate using it, it works. There is simply NOTHING that suffices for Linux. Sure, I can move crap from my library or server to the iPod, but I cannot shop the iTunes store at all.  My kids would probably rate this item #1 on the priority scale.  It is a must-have.
• CAD. AutoCAD runs on Windows. AutoCAD is to CAD what Windows is to desktop computers: It simply dominates the world of CAD without question. I use AutoCAD a lot. Regardless of the promises competitors make, nothing replaces it with 100% true compatability. NOTHING. I know. I run tests all the time to make sure. I have not found a feasible, reliable, suitable solution for running AutoCAD on Linux. Period. Do not even attempt to suggest Wine or VMware to me. That is laughable for running things like AutoCAD, which is a floating-point and graphically intensive application. A native solution is mandatory.
• Printing. As much as like CUPS, it still sucks compared to native Windows HP desktop printer management. Printing from a Linux client, over a LAN to a Windows-hosted shared HP inkjet printer works “most” of the time, but for some things like photo-quality output, it lacks. The options are limited and not always recognized by the Windows service on the other end. It needs help.
• DVD authoring. I’ve been a fan of NeroVision for years. I’ve tried lots of other packages from ULead and others. Nero works best because it’s (a) reliable, and (b) simple. The built-in features are great. Not too much clutter, not too simplistic either. As advanced as you want to take it. I have yet to find anything for Linux that offers the same features. Not even NeroLinux. The ability to compile, transcode, merge, construct menus, covers and navigation controls, even fonts, colors, graphics, and burn a final disk (with 99.99% success) and fire it off in less than 5 minutes of work (the rest is transcoding and burning after all) is simply UNMATCHED by any Linux app I’ve found.

Sounds harsh, doesn’t it? It’s not meant to bash Linux. It’s simply a shopping list of what I need to bake my cake.  Some of you may think I’m baking an unrealistic cake.  Am I?  Is it wrong to demand more?  If we never demanded more, we’d still be using punch cards.

I would LOVE to not be tied down to paying insane licensing fees to Microsoft. I also don’t see any logic to throwing my hard-earned hardware out to go buy Mac stuff at this point. My wife would have me shot if I suggested that. “Honey, you know those five computers we have? What if I throw them out and replace them with Macs at roughly $1100 each?” (that’s a very conservative estimate too). Those would be the last words ever heard from me. “Yes officer, then I heard a loud bang, like a gunshot, from that house right over there!”

No. My optimal solution would be to (a) leverage the hardware I have already paid for, (b) replace each and every functional need for minimal or zero cost, and (c) remove the burden of upgrade fees for future software upgrades (at least for the operating system and basic apps). Like I said, I’m still looking. I have not given up. But I still haven’t found what I’m looking for. Oops. Bono sang that first. RIAA will be after me now I suppose.