Linux why?

I want a divorce from technology. Somebody write up the papers… I’ll sign.

In all my messing around with Linux… it’s fun and all. I learn stuff. I can setup pretty much any of the various distributions with my preferred setup in a matter of hours. Sure they all have their quirks, but no big deal. I had my machine booting into six different Linux versions. All set up and working nicely.

Then I had to stop and ask myself, what’s the point? I mean really. Why would one run Linux instead of Windows unless they HAD to? Windows works so well. What’s not to like? I know that it’s popular or somehow “hip” to be anti-Microsoft. But I see no benefit in that.

Linux might make sense to me if my wife was open to using it. We have two desktops (her’s and mine) and a laptop. It’s just not that great to have one Linux system and two Windows machines. If they were all Linux I think it wouldn’t be so bad. But nothing cooperates with a Windows machine as well as another Windows machine running the same software.

I am quite torn over the issue really. Because messing with Linux is a lot of fun. And I mean a ton of fun. But if I want something more than a toy to play with… something that will actually do all the things I want it to do… I have to keep coming back to Windows. So it’s the difference between being fun or being useful.

Linux might be nice if I wanted to learn PHP, Perl, or certain other programming languages that are non-Microsoft based. But what’s the point of setting up something like dual-boot with Windows and Linux when I have to boot into Windows to do certain things? What’s the upside to that hassle? I don’t know that there is one.

It is kind of funny though. When I get Linux all set up nicely and working they way I want, I feel like I’ve really accomplished something. But when it comes to Windows… that sense of accomplishment is not nearly so pronounced, because it comes so much more easily. Maybe that’s why Linux is fun and Windows is useful.

Giving KDE a whirl

In my last blog entry about choosing a Linux distribution, I mentioned favoring the Xfce desktop environment. I mentioned ruling out Suse because it was crashing upon initial configuration. And I mentioned that I hadn’t really given KDE a try.

Well… since that time, I was able to get past the Suse instability and install a couple of different flavors of that. And I’ve tried Kubuntu, Suse w/KDE, and Fedora’s KDE spin. I also got a little tired of having 6 different installations of Linux. So I decided to settle on just one.

For now it’s Fedora with KDE. I’ve condensed my partitions and that’s what I’m running. So far so good. Everything works except for my scanner. Which is no problem since I can access that via my Windows 7 VM.

I’m leaving the door open to going with Suse in the near future. They are about to release their next version. And they are a distribution that seems to lean toward KDE as their default choice. So it makes sense that someone wanting KDE might choose Suse.

My current choice of KDE is a little strange. Because KDE is about as heavy as it gets when it comes to desktop environments. And because so far one of my favorite distributions of all has been Crunchbang which uses the extremely lightweight and barebones Openbox window manager with no DE at all! So I am torn between the two extremes. Typical.

In other boring news… I ordered two new hard drives.

I’m using an old spare 160GB HD in my laptop running Windows 8. I pulled my 128GB SSD and put that in because having only 12GB free space is a little too tight for my taste. In thinking about replacing the drive… I could have gone with a new 250GB SSD for about $180. Or I could get a 750GB 7200RPM HD for $73. I went for the 750GB drive. Space over speed.

The second drive is a “green” WD 3TB drive for my external Lacie enclosure. I have two of these enclosures. They are pretty spiffy looking. Although buying a drive with an enclosure was almost cheaper than buying just the drive… I went ahead and did the drive anyway.

Choosing a Linux distribution

The problem I get when messing with Linux is trying to figure out which distribution to run. They’re all free. And changing distributions costs nothing (except time and maybe sleep).

So I was thinking about the factors that I should consider when choosing. Trying out a bunch of different ones is fun. Been there done that. As I gain more experience with the various distributions, I have been able to build a much better idea of which one I want to run and why.

I might be tempted to rule out distributions that don’t support all of my hardware out-of-the-box. However, the only hardware I have that is potentially a problem is my wired ethernet and the scanner portion of my Canon wifi all-in-one printer/scanner. My wireless internet is supported on all of them. And I can use my scanner from within my Windows 7 VM. So these are pretty much non-issues.

Some distributions are heavy on eye-candy. Others not so much. While I like things to look nice, functionality and reliability are way more important than looks. A slick interface isn’t that great if it’s got bugs.

Then there is the matter of philosophy. Some distributions make it difficult to get so-called “non-free” software. To them the encumbrance of any closed code is something to be avoided (even at the expense of things not working). This is not important to me. I just want things to work.

One other factor is what I will call “conservative vs bleeding edge”. Debian is generally on the conservative side. They only release things after they’ve had a lot of testing. Fedora and Arch Linux are bleeding edge. If you want to be the first to get new versions of things, those are the ones to run. I think I would prefer something more conservative. It’s possible that a Ubuntu LTS (long-term support) version would be just up my alley. And that would include the officially supported derivatives.

However, the problem with “conservative” distributions goes back to lack of support for newer hardware and older versions of almost all software. While the Ubuntu folks recommend that you run the LTS versions of their stuff for stability reasons. The current LTS version lacks support for my wired ethernet and scanner. Whereas the latest Ubuntu non-LTS version supports both of those out-of-the-box.

Then we have the difference “mainstream vs obscure”. I’ve run across a number of distributions that have broken repositories and or download links that don’t even work due to some outage or something. I think with obscure distributions you may have some level of slickness and novelty, but you don’t have nearly the support structure that a mainstream distribution would have. Either in terms of available software or online help. And you probably have way more bugs due to a smaller test pool and fewer development resources.

So if I would limit myself to mainstream distributions, they would have to be Ubuntu, Mint, Fedora, Debian and Suse.

While Mint might generally be considered mainstream, I think it’s more of a Ubuntu knock-off. One may as well run the real thing and stick with Ubuntu. The whole Unity mess is easy enough to avoid by going with an official derivative. For me Mint keeps seeming like Ubuntu with extra crap on top.

I tried Suse last week and it crashed like three or four times just while I was trying to get it set up. A new release is in the works very soon. So I may give it another try then. When the OS crashes before you even have it set up that doesn’t give me a good feeling. Bodhi was another one that hung several times during the installation process.

So based on all this… my own choices seem to be narrowed down to the Xfce versions of Fedora, Debian or Ubuntu.

I like Xfce because it lacks most of the bloat of Gnome, Unity and KDE. Although I have yet to give KDE a try. That’s on my list of things to do.

For all this rambling… no conclusions yet. This is a work in progress.

Multi-booting with Linux

Windows 8 uses a GPT-formatted hard drive and an EFI boot partition. This is so new that Linux does not yet play well with it. While a few distributions handle it, I came to the conclusion that the state of EFI support in Linux right now is so poor that it’s just too much hassle. So I decided to change the partition table on my drive from GPT to MBR. Then I turned on “legacy” boot support in my BIOS.

Of course that means I will lose the ability to boot into Windows 8. But it also means that I can now boot Linux using the normal GRUB boot loader the way it has been accustomed to. As you can see above, I have SIX different Linuxes installed. And when I boot the machine I’m presented with a GRUB screen where I can choose which one I want. I have one swap partition and one partition for my data that I share between them all. I am NOT however sharing my /home directories between each one. That will prevent conflicts with user configuration files between the different distros.

I initially read that I should not install GRUB with the installation of each version of Linux that I set up. But many of those installations did not give me an option. Luckily each one recognized the other operating systems installed and added them to the new GRUB configuration being installed.

This is way better than messing around in virtual machines. I’d commented earlier that VMs take no guts to set up… because they can be nuked in a second and they’re not really running on the hardware. Booting directly into 6 different Linuxes becomes more real. Support for the actual hardware needs to be configured.

One of my goals is to test out which distribution will support my scanner out of the box. I already know that several of these are having trouble with my ethernet adapter. Though luckily they seem to support my wireless adapter with no problem.

System restore

In playing with my Dell desktop machine’s partitions I ended up deleting the various reserved system recovery partitions. I don’t really understand how they are organized or what pieces are required. But there seemed to be 3 or 4 of these reserved partitions when I decided to clean them up.

The only symptom I have now… is that the Dell Backup and Recovery app won’t run (or even install) without the “Windows Recovery Environment” which resided in one of these partitions.

I probably wouldn’t care about not having this except that this system came with some software that I would like to be able to reinstall if necessary (namely Microsoft Office and Adobe Photoshop Elements).

The answer, is to restore the system to “factory” using restore discs that I’ve already made. That should put back all those special partitions. Then I’ll use Acronis True Image to replace my OS and Data partitions. At that point I’ll be back to where I am now with the system partitions restored. In theory.

The main motivation for attempting a full restore of the entire system is to give my new version of Acronis True Image a good workout. I want to see how it works. And even the worst-case scenario isn’t that bad. If it fails I could always do it the hard way by building from a fresh install. But I don’t expect that will happen.

Windows 8.1

Windows 8.1 was released for download last Thursday. Of course I installed it on all of my computers. No real issues.

Unfortunately, Microsoft doesn’t provide any kind of install image or ISO for 8.1. So if I ever need to reinstall (which of course I will), I will need to install 8.0 again and then download the 8.1 all over again. That doesn’t sit well with me and probably a million other people who have a strong preference for a fresh virgin install.

There are some nice new features. I had a hack in the form of a shortcut that would take me directly to the all-app screen in Windows 8. No longer needed. Windows 8.1 has a setting that will give you the all-app screen in place of the regular start screen by default. And there is a setting that lets you shove all those useless “modern” apps to the end of the list. Yay!

I would have preferred a configuration option that would eliminate any possibility of seeing a “modern” app anywhere for any reason. But I guess that is too much to ask. Although I do hear that the modern apps have been enhanced to the point where one might actually consider using them for real things. I still don’t know why I’d want to. I like my desktop apps. That might possibly be why I run a desktop operating system! Ya think?

My wife is going out of town. And since I wanted to install Windows 8.1 on my laptop SSD, I pulled my Linux hard drive out of my laptop and put my Windows SSD back in. Then I upgraded it to 8.1. I did end up having an issue where I had to reformat. Only because of my lack of patience. I had a ghost printer that I could not delete. And problems with several versions of the same printer. They were grayed out, but would not let me delete them. After googling and trying some different things I decided a fresh install was in order. No problem.

In my playing around with Linux I recently came to a bit of a conclusion. Windows is a far more capable operating system. Not because it’s actually better, but because it has all the third-party support. Things just work. And while that’s true when comparing it to Linux, it’s also somewhat true when comparing it to Mac OS X. Although you’ll have a hard time getting Mac folks to admit that.

Linux as a primary OS

I’m totally loving Fedora Linux since I put it on my two machines last week. I have my desktop machine dual-booting with Windows 8… UEFI no less. And my laptop is running Fedora as well. And I have a Windows 7 VM in both primarily for running Quicken.

The whole UEFI secure boot thing is pretty new. Most distributions still don’t handle it. It came onto the scene with Windows 8-ready machines. But luckily Fedora does handle it without too much trouble.

I’m pretty sure I couldn’t run Linux as my daily-driver operating system except that anymore my PC’s are primarily “general use” machines. I don’t run any special-use software that I can’t live without. The only program I really don’t want to give up is Quicken. All the other stuff I depend on is pretty generic.

Dropbox is one program I deem essential. Other than that there are a couple of browser plugins that I use… like Lastpass and Xmarks. Of course Linux has Firefox and Chrome.

It’s weird. The longer I use computers the less software I seem to need. I’m becoming an “average” computer user. Other than enjoying the setup and configuration of operating systems and software, my needs are pretty basic. Of course there are a lot of people in the same boat. That’s one reason why Chromebooks are pretty strong sellers now. It might not be that long before all one really needs is a good browser with a few good plugins. (that’s essentially what a Chromebook is)

I guess one could argue that the lack of need for a real computer is partly to blame for dropping PC sales in conjuction with the popularity of smart phones and tablets. I personally can’t imagine that. I don’t think I could ever do without a computer. I don’t even like being confined to a laptop.

Dual boot

Ok… not long ago I decided to setup a virtual machine on my Windows box for every different version of Linux that I had. There were nine of them! And I had Linux installed and working on just about every one.

But then I decided to get some guts, and try setting up my Windows 8 to dual boot with Fedora 19. Of course when setting up dual boot you run a fairly serious risk of rendering your machine unbootable. But rumor had it that Fedora 19 played well with UEFI secure boot.

Before I did this I partitioned my Windows into four different partitions, mainly to aid in backups. I recently updated my copy of Acronis True Image so that I could take a snapshot of all the partitions on my machine so that if the whole dual boot thing went awry, I could put things back. The advantage to having multiple partitions was that I could put my huge stuff (movies, TV shows and music) in a separate partition that I could omit from the partition snap shot. That stuff is easy enough to put back that there is no snapshot needed. It’s just a lot of data.

One of my partitions was about 220GB that I was using strictly for above said virtual machines. So I backed that stuff off to an external hard drive, nuked the partition and set up Fedora for real on an EXT4 partition. Yay!

So it’s all working nicely. And I just now setup my Windows 7 virtual machine as a guest in my Fedora host! So I can boot Windows in Fedora and run Quicken.

I have a bit of a soft spot for Fedora. Back in the days when I was working for my brother, we administered a bunch of Red Hat servers. Big fun. Lots of memories.

And Fedora is pretty cutting edge when compared to many other Linuxes. Although I’m pretty sure it’s not as cutting edge as Arch Linux. Which is fine with me. I got tired of hanging out in the #archlinux IRC channel and seeing a steady stream of people coming in who had something break as a result of some update. I think that’s what you get with a rolling release that always has the latest of everything. Probably not bad if you like being a beta tester. Certainly it’s own kind of fun.

Fuse tap

I have only one 12-volt accessory outlet in my little Honda Fit. And my Garmin GPS needs one pretty much full-time. It doesn’t run very long on it’s battery. I don’t think it was designed to.

So the GPS ties up the only available outlet. Which is a pain when I need to plug in a mobile device for charging (like a smart phone or tablet).

To solve the situation, I wired up a female cig-lighter style plug to the fusebox using a fuse tap. I then ran the GPS power cord (with it’s male cig-lighter plug) under the dash and have it plugged in where it’s all out of sight. That frees up my one accessory outlet for various charging needs as they occur. Yay!

Virtual machines

I did end up installing Arch Linux on a VM using VirtualBox. Been playing around with it. VirtualBox has a cool “seamless” mode where I can have linux programs open on my windows desktop just like a normal windows program. The linux programs and windows programs are intermixed “seamlessly”. It’s really very cool.

Facilitating that was my latest installation of xorg-server, Openbox, and the VirtualBox guest tools on my Arch guest VM. Now that they are on and functioning the guest VM works as it should. I had strictly been using it for console sessions.

I’m am debating possibly creating a new VM with Fedora. I was inching toward dumping Arch Linux. But perhaps I will stick with it a bit longer. Now I realize… since all of my Linux stuff is happening in VMs, there is no reason I have to “dump” one to try the other. I can have a bunch of VMs and just continue to play with them all. I can try all sorts of things without having to nuke one to do the other. That’s pretty nice.

It’s still a little hard for me to see the point in all this. All this is running on a perfectly good Windows 8 machine. And Windows 8 will do anything/everything that I have configured Arch Linux to do. Sure I can IRC or browse the web using a Linux “machine”. And I certainly could set up other things to run on it. For what purpose I am not sure. Maybe just to say I did.

There might be some benefit to doing IRC and web browsing via a Linux VM rather than doing those things in Windows. In theory, I’d have less vulnerability.