Remember my last post about how Thunderbird didn’t have a proper import/export function? Since then I discovered, installed and used ImportExport Tools to allow me to import Mbox-formatted mail archives into my main Thunderbird installation.

I was able to easily bring all the messages from my Windows Thunderbird install into my OpenBSD Thunderbird, and now all of my e-mail is in one place.

The reason for this was switching over from IMAP to POP because I was (and am)  not comfortable leaving my mail on the server and want to have it all on my local drive (and backed up to CDs/DVDs).

To consolidate my two Thunderbird archives into one, I took the whole local directory from one, burned it to a CD, moved it to the other PC and then used ImportExport Tools to bring the folders into my current Thunderbird local directory one by one. Once I had the new folders in my new Thunderbird install, I then distributed the messages to the appropriate folders and deleted the temporary folders I created to hold the transferred mail.

I continue to think that this import-export function should be done wholly in Thunderbird with Mozilla-maintained code, but having an add-on that works is better than nothing at all.

Using IMAP, as I did for a few years, allows for quite a bit of flexibility, since the mail stays on the server and I can use any number of clients to access it from any number of machines.

To give me some added flexibility now that I’m POP-ping the mail down to one computer, I have Thunderbird configured to keep the last 10 days’ worth of mail on the server so I can use my mail service’s Web interface at any time to see all the mail in the past 10 days.

The one problem with this 10-day solution is that any e-mail I write in Thunderbird goes into the Sent file on my local machine and is never on the server. To get around this, if I write an e-mail that I’d like to have access to via the Web address, I BCC it to myself so it goes on the server as well.

I do the same thing if I write an e-mail in the Web interface, copying myself so I have it both on the server and in my local directory.

Before I close, I should say that I chose Thunderbird as my mail client because I’d been using it for a few years (and was comfortable with it) and because it’s a cross-platform app that runs in Windows, Linux, Mac OS and every BSD. I really like any app I can easily get in every OS I run (or might run).

Today in Click, I rant about the Thunderbird e-mail client, which is only too happy to import your e-mail from another application but isn’t quite so free and friendly when it comes to exporting that mail out of the Mozilla-based program.

I believe that it’s fairly easy to turn just about any Linux server into a WordPress installation. There is even a port of WordPress for OpenBSD. I thought that I’d try to run Movable Type on OpenBSD, but getting PHP and Perl working with the Web server looks incredibly difficult. If somebody else did this and laid out how they managed it (like How to Forge), I could probably follow along, but the difficulty of dealing with the chroot environment in OpenBSD’s default Apache Web server — which isolates the Web server’s files from the rest of the computer for security purposes — makes it extremely difficult for mortals to set up services in the Web server environment.

For a normal Web server with nothing but HTML pages (and no PHP, Perl/CGI), OpenBSD couldn’t be an easier system to use. It’s when things get more complicated that this that the non-OpenBSD geek is especially challenged.

But as I say, there is a WordPress port for OpenBSD, as well as WordPress packages for GNU/Linux systems such as Debian. I also seem to remember talk about a Movable Type package for Debian Lenny, and if it made the installation and configuration of the blogging system easier, I’d be all for it.

As it is now, between installing and configuring MySQL (or PostgreSQL), making sure PHP and Perl are running and getting all the directory permissions correctly set, putting together one of these blogging systems, even on Linux, is no trivial matter. The last time I set up Movable Type, all the MySQL issues I was having prompted me to dump it and use SQLite as my database software. At least MT gives you some options in this regard.

I’m using my WordPress blog to write about Movable Type because it’s time for an MT break. I’m suddenly knee-deep in reconfiguring a few dozen Movable Type blogs and find myself baffled by layer upon layer of templates and widgets.

I know there’s a reason why this project uses Movable Type, and I’ll probably figure it out eventually, but in the mean time, I remain baffled by all that is MT.

In an effort to have just a little more control over what I write, on which of the many PCs I use I write it, and where I post it (i.e. to one or more of the blogs and other sites to which I’m spreading the news), I started to use Google Docs again.

The problem being that I can hardly keep track of anything that I didn’t write and post instantly. I’ve got three notebooks, each a different size (one “moleskin” type book, though no skins were harmed in its making; one composition book; one steno notepad), and a number of computers ($15 Laptop, $0 Laptop, converted Maxspeed Maxterm thin client, Dell Optiplex GX520, iBook G4), have of which change Linux distros as often as most people change underwear (that’s daily, for those of you not following), with /home partitions constantly moving, being deleted and otherwise being ignored.

So the theory is that by centralizing everything on Google Docs, I can better keep track of what is where, what is going where and what went where. That’s the theory anyway.

And while I’m on the subject, it’s time for me to make complete backups of all my blogs, especially Click, which has the most posts and is on a server that is nowhere near as reliable as those of Google or WordPress. The great thing about WordPress, as far as backups go, is the ability to export the entire blog as XML. For Google, and probably for Movable Type, I’ll just have to save monthly archives going all the way back

But I haven’t been the most prolific blogger of late. It all began when the esteemed, highly qualified individuals who run the insidesocal.com blogs (including Click) decided that the best way to stop DNS-level spam attacks was to put an IP block on the entire European continent, costing me every link I could hope to get for Linux-related material (yep, Linux and FOSS is huge in Europe; where else would they even think of publishing “Debian Fur Dummies”?).

So I stopped pimping Click and started this blog, also throwing items to the great LXer. Losing Distrowatch as a source of links to reviews of Linux and BSD distros was probably the biggest blow. So I’ve pretty much been not caring about traffic on Click, which hovers at a steady 150-250 a day.

And I am using gOS’ version of Gparted to partition the hard drive on which I will eventually install gOS. I haven’t yet used the Gparted from a Ubuntu-derived live CD, since I have Puppy for that purpose. But since the version of Gparted on the last few versions of Puppy Linux have taken up to a half-hour to read the partition table, I’ve since turned to the Gparted and Partition Magic live CDs to do my partitioning.

But since this is gOS test, I figured I’d use it’s version of Gparted. It’s as lousy at reading the partition table in a timely manner as the version in Puppy. Has nobody but me noticed this? It makes Gparted all but unusable.

Not that commercial applications don’t have soul-killing bugs in them, but Gparted has been screwed up for so long now, won’t anybody fix it already? It’s the same thing as the Ted word processor in Debian. I’ve checked — all the dependencies are there. But you can neither open nor create a file in Ted. The RTF word processing app works fine in Damn Small Linux (where it’s the main WP app) and in Puppy (where it is an easily-added package). But it’s useless in Debian. Like Gparted in … just about everything.

But the bright side is that I discovered the Gparted and Partition Magic live CDs. I heard that development on PM is going to cease, and that would be a very bad thing, indeed. Hopefully somebody else will take up the mantle and either continue Partition Magic or start their own live CD focused on partitioning hard drives. That’s the beauty of open source: out of the ashes, a new project can always arise.

Anyhow, I deleted all the partitions from my drive, and I’m waiting the <em>next</em> half-hour for Gparted to scan the drive again so I can create new ones.

P.S. Even though Gparted takes so long to scan the drive, it makes changes to the partitions as quickly as it ever did.

After my less-than-successful forays into BSD, I’ve come to the conclusion that before installing, configuring and working with a BSD distro — whether it be FreeBSD, OpenBSD, NetBSD, and even DesktopBSD and PC-BSD — it pays to read up on in.

With Linux, you can fake it easier. Most Linux distros ship with everything you need, and even setting up a server can be as easy as clicking the right boxes during an install.

But BSD is a different animal. The hardware support is nowhere near as comprehensive as it is for Linux, but it is possible to get it working with the hardware you have (wireless notwithstanding — BSD and wireless is something I know very little about but want to know much more).

One thing people say about BSD — it can be faster than Linux for many tasks. I don’t know how that plays out on the desktop, but for servers and other processor-intensive tasks, BSD is an attractive alternative. And the security of OpenBSD especially is legendary. So for a server, I can totally understand it.

And the BSD that went into Apple’s OS X seems to be working very, very well — multimedia is probably better on the Mac than anywhere else, so for BSD and Linux, it can be done and can perform better than Windows by far.

But when rolling your own BSD system, I think a ton of reading is what’s required. And there really isn’t a book out there that focuses on BSD desktop systems, even though there are two distros that focus on just that.

FreeBSD is the most popular BSD, from what I gather, and that’s due in no small part to the excellent FreeBSD Handbook, a free 900+ page manual that is one of the best examples of FOSS documentation I’ve seen.

And new from NoStarch Press is “Absolute FreeBSD, 2nd Edition,” by Michael W. Lucas. Check out the PDF of the table of contents. At $59.95 retail, you better hope for a deep Amazon discount on that one. From No Starch parent company O’Reilly — and for $9.95 each — are two PDF-only books, “The FreeBSD 6.2 Crash Course” and “The OpenBSD 4.0 Crash Course,” both by Jem Matzen. Both of these PDFs promise some guidance on using BSD as a desktop OS.

As I’ve written previously, OpenBSD is the only BSD to install on my Maxspeed Maxterm converted thin client (VIA C3 Samuel processor), and while I was far from getting it set up for the desktop — there was just too much to do with ports and configuration — I now look favorably on OpenBSD as a great system on which to learn BSD. The OpenBSD FAQ is quite good, though not as extensive as FreeBSD’s Handbook . I was able to do the OpenBSD installation with no problems whatsoever by consulting it keystroke by keystroke. It’s also available in text form.

I’ve been warned away from BSD by people who know more than I do. And other way-smarter people swear by the use of BSD, even on the desktop.

From my point of view, having not just one OS alternative (Linux) but many — including the various BSDs — is vitally important to all of us, both on the server and the desktop — as we go forward. I hope the people behind the various BSD distributions keep the desktop user in mind more and more in the near and far future.

I’ve had trouble with Slackware 12 and the $0 Laptop (Gateway Solo 1450). Something happens when services load that prevent it from running. I’ll have to take a much closer look, but I’m ready to try Slackware 11 just to see if that makes a difference.

I see in the Slackware security page that patches are still being doing for version 11, and even 10.2 for Firefox issues, and back to 8.1 for things like CUPS.

But Slackware 11 is pretty well covered, and I recall having a better time with Zenwalk back in the 4.2 days when I was first getting into Linux.

Even so, Debian is running pretty well on the $0 Laptop. I don’t have Xfce installed, so I don’t know how much better it will do than GNOME. And in Ubuntu 7.10, I can actually make the touchpad do what I want, which is to run at the right speed and NOT tap-to-click unless I want.

And I have been running Puppy 3.00 from live CD (compatible with Slackware 12 packages), and now that I’ve managed to control the noisy fan with a few modeprobes set to run at startup and a cron job, I just might stick with it for … everything. It’s so fast. No tap-to-click in Puppy, but since I don’t want it, that’s fine.
I installed the SFS file for OpenOffice on one of my other Puppy setups, but not this one. I’ll have to try it. I actually like the way Damn Small Linux adds things like the GIMP and AbiWord — the filesystem in DSL (and, by extension Knoppix, which I’ve also been running on the $0 Laptop) seems more flexible than Puppy’s. I don’t think you have any size limitations. I think Puppy’s pup_save can only be 1.5 GB, but that still doesn’t stop you from mounting other filesystems and working with them, so it’s a six-of-one situation anyway.

And Damn Small Linux runs like crap on the $0 Laptop. I just can’t get the X configuration right, no matter how many different ways I try it. Puppy, of course, is perfect.

After not being in love with Debian Etch under KDE, I wanted to boot into Slackware to see if it was just me.

It’s not just me.

Since I never ran Debian with KDE before (nor had I run Slackware 12 in GNOME, with such an install being way beyond my capability at this time), I was surprised to see KDE running so much better in Slackware. I knew that Slack was quicker at just about everything than most distros out there, but I had no idea that it was so much quicker than Debian as well.

At this point, and on this slower-than-average box — the VIA C3 Samuel-based converted Maxspeed Maxterm thin client (256 MB of RAM) — I wouldn’t run KDE on anything but Slackware.

But … Slackware with KDE compares very nicely to Debian with GNOME. Makes me wonder how Slack would run with GNOME …

I’m a sucker for a Linux or BSD distro with a live CD. Even if you can’t do an install directly from the disc, at least you can figure out whether the damn thing will boot and how your hardware will react when it does.

One of my favorites, ZenWalk, just released Zenwalk 4.8 Live. Both Zenwalk and Vector — the top Xfce-based, Slackware-derived distributions — are very good, but I like the way Zenwalk looks and works just that much better. I was dismayed when Zenwalk 4.6 wouldn’t install properly on the $0 Laptop (the Gateway Solo 1450). Slackware 12 and Vector Standard 5.8 wouldn’t run once installed  either (I think I need to pass some boot parameters … which means I’ll either have to figure out how to do it in LILO or try to do it in GRUB), so it’s more than likely something that begins in Slackware that the other distros don’t clear up. I have my money on PCMCIA or SCSI services, and it is worth a try.

Anyhow, whether or not an Xfce-based Slack-derived distro can really “save” really old hardware remains an open question. One thing I do know, however, is PCs that do “OK” with standard distros like Ubuntu really do fly with Zenwalk or Vector (and more so with Puppy and Damn Small Linux, but that’s another story).

Before I continue, let me remind you about Zenwalk Live’s root password (you’ll need it if you want to do any configuration in the very-well-thought-out ZenPanel app):

ZenLive

And yes, it is case-sensitive (as are all Linux passwords).

Right now, in the live CD environment, I expect more speed from Zenwalk, but I don’t want to make a full judgment until I’ve done a complete install.

One thing (and it’s a big one) that Zenwalk does have going for it is the Net-Pkg package manager. It makes dealing with packages that much easier. And overall, Zenwalk’s ZenPanel is better in just about every way than Vector’s VASM tool.

And if you want to run Slackware but don’t want KDE — and want easier package management, you should test both Zenwalk and Vector before making any decisions on a permanent install. For me, the relative lack of time between releases for both distros — and what looks like the abandonment of updates for the older versions — gives me pause. Nothing a separate /home partition couldn’t cure, but I prefer the ability to stay up to date as long as possible without a full reinstallation of the operating system. (This is an area in which Debian and Ubuntu excel.)

Another thing before I go: Slackware, Vector and Zenwalk all run exceptionally well with the Fluxbox window manager. It’s included in the initial installs of Slack and Vector (the latter doing it the best, I think) and is readily available in the Zenwalk repositories. With Fluxbox, you get a lot more speed, and if Xfce doesn’t give you the level of performance you seek on an older box, any of these distros just might do what you need with this lightweight window manager.