We Should See Other Blogs

Experience would dictate, having seen 99 white swans, that all swans are white. Experience is of course a ridiculous guide for making future predictions: One event can change your experience in a way that the preceding events could not possibly have foretold. The 100th swan, the black swan, gives lie to the statement that all swans are white; suddenly only most swans are white, or at the very least most swans in the observed sample are white.

Ubuntu is our Free Software black swan. How do we explain its sudden, rapid rise to international Linux stardom? How do we explain its overwhelming success in that domain?

Well, we don’t. You can attempt to apply narrative to Ubuntu’s sudden critical mass, but it doesn’t work. Ubuntu was there at the right time (whatever that means) with the right feature set (not much different from others) and the right community (a little less technical than others, perhaps, but not much) and the right backing from Canonical (though other Linux distros have their own sugar daddies).

All of those statements don’t really explain how Ubuntu came to dominate Linux mindshare. In fact, I don’t think there’s any real way to make a narrative out of it. The reality is probably more like the Linux community tinkered and hacked and scratched the itch and tried things until one of those things really worked. This is a signal, to me at least, that the Linux community is growing up: We now have, like any other domain, a winner-takes-almost-all distribution. For good or for bad, this is how these things seem to go.

I think sometimes that the world is like throwing things at a wall and seeing what sticks. Of course there seems to be some correlation between hard work and success: Those people who work harder are likely to try more things. But you can work and work and work and work (look at Debian and Fedora and Gentoo and Linspire and a hundred others) and not have someone else steal and eat your cake.

You draw your own lesson. Does Ubuntu deserve to be where it is right now? Sure. Maybe if you’re working on another distro it seems a bit unfair. But even the words “deserve” and “fair” imply you believe there’s some kind of narrative going on. I disagree. There’s no narrative. There’s a metanarrative, and that’s what matters.

On its own, Mozilla Thunderbird is a handsome, capable mail app. It does everything you would expect a mail application to do and a bit more. I would compare its capabilities — unfairly — to Outlook Express rather than Outlook proper, as it lacks calendaring and tasking capabilities. Outlook, though traditionally one of the major security holes in and attack vectors for Windows, is otherwise quite a functional application, though nothing particularly special.

What Outlook does, though, is easily plug into Exchange. Which just happens to easily plug into things like Sharepoint and Office. All of which rests on a foundation of MSSQL and Active Directory. Which only exist on Windows Servers. This is what we call an ecosystem. It’s one of the few things that Microsoft does right. Outlook is simply the thin end of the wedge, that little bit of lubrication that enables you to more easily give money to Microsoft.

And right now, there’s really no good alternative. Outlook + Office + Sharepoint + Exchange + MSSQL + Windows Server is damned expensive, (often) hard to maintain and administer, and hooked into a system of constant and unnecessary upgrades that ensure it will be expensive now and in the future, but it’s so easy.

Thunderbird doesn’t have that ecosystem. Evolution doesn’t have that ecosystem. Thunderbird is getting close with the Lightning calendaring application, a fine, even essential addition to the program. I can’t imagine installing Thunderbird without Lightning. But this is all frontend stuff. If you want to set up a proper backend for Thunderbird using, say, Linux + MySQL + Postfix + whatever, you’re in for quite a steep learning curve. Unless you have a lot of spare time, that learning curve will be almost insurmountable.

What the Linux business community needs, to penetrate the SMB market especially, is something along the lines of Exchange. Something like Zimbra, for instance. We need to cast aside this idea that a competent UNIX admin must be in charge of the Linux server. Most small and medium sized businesses simply do not have the resources for that. We need to be able to say, here, have this server. It will do what you need it to do.

Can you imagine a Linux-based server with a bunch of pre-built virtual machines designed to work with each other to provide a smooth computing experience for those of us who can’t afford to hire an admin full time?

You buy some iron, lay it down in the spare room, and say, okay, I need the “Storage” virtual machine and the “Mail Server” virtual machine and the “Web Server” virtual machine, and the “Collaboration” virtual machine. You install them, you click through a bunch of helpful wizards and boom, you’re done. Maybe it points you in the direction of a backup server for good measure.

You go to your Windows or OS X or Ubuntu machine and start it up. You install a couple programs on it that just work right out of the box. Could be Firefox, Thunderbird, OpenOffice.org, or whatever. You get to work and everything is exactly the way you want it.

Then one day when your company has enough money for an full-time admin, you separate that functionality into separate servers or whatever.

I guarantee that business owners will pay for that. Bundle all these free software ideas together and make a usable package out of them. I don’t care of you GPL your front end or not. I’m a pragmatist when it comes to things like that. But there is serious money to be made in the marketplace for a company brave enough to do just that. You can sell your product and the support of that product for far less than all that Microsoft software. You can undercut them and create a better, more secure product in the meantime.

A guy can dream, right?

I know, I blogged about Warsow and how amazing it was, but the game I’m most hooked on at the moment is Urban Terror. Seriously amazing game, great graphics, and some of the best gameplay I’ve seen this side of Rainbow Six: Rogue Spear. It combines some of the movement techniques of twitch gaming (powerslides! whee!) with the more measured combat of Enemy Territory and Call of Duty. It also focuses more on urban and suburban environments than most games emerging from the Q3A scene.

Download it; it’s big, so please do everyone a favour and use Bittorrent. (I currently have a share ratio of 11; you’ll probably be getting some of your bits from me.)

Urban Terror runs on Windows (of all stripes), Mac OS X, and Linux (I currently have it set up in Ubuntu). It’s free. You don’t have to be a gaming genius to play it, either.

I don’t play a lot of games. Really, I don’t. Granted, when I go over to my parents’ place, I might play a spot of Call of Duty 4. At home, though, there’s not really a whole lot of time to play, and I don’t have any disposable income with which to actually purchase any games.

That said, I’ve found a game to play lately, Warsow. It’s free, available for a variety of platforms, and breaks nicely from the traditional Quake-inspired Gothic darkness with cell-shaded graphics and an overall different appearance.

It’s also ridiculously difficult (for me) to actually play. Warsow is clearly geared to gamers and their trix. I, on the other hand, am just a lowly married man with bad hand-eye co-ordination. When I type cg_showSpeedMeter “1″ into the console (yes, it had a console, just like every other game worth its salt), I rarely if ever get over 400. I suck.

Yet I still have fun… try it out. Seriously. I double-dog-dare you.

Getting the perfect PulseAudio setup in Ubuntu 7.10 isn’t hard. It’s basically installing a bunch of packages and editing a config file, then making sure ALSA is using PulseAudio instead of ESD.

The problems are these: there’s no metapackage that installs all the needed utilities, modules, and libs; the GUI tools pretty much suck; and there is no usable documentation that I could find for the base install or the GUI tools. The metapackage should be easy enough to implement, but the GUI tools need lot of work. The documentation needs to be created.

PulseAudio is a great idea. It’s well-implemented (afaict) at the system level. It enables the user to do a lot of pretty cool things, like sending audio to different computers. I would very much like to see it replace ESD, but before it can really do that, it needs a lot of usability tweaks.

For instance, “sinks” may be a nice technical name for where the daemon sends the sound, but the difference between sources, server, and sinks will probably be lost on the target audience of the Ubuntu distribution. These differences should either be elaborated with end-user oriented language, or the less-accesed hidden from view in a less visible part of the interface.

Ideally, I’d like to see a graphical representation of the possible output locations, much like the graphical representation of filesystems and available drives in the “Computer” dialogue. I could select “Send sound to” with a check box, and select multiple locations if I wanted to, something that’s difficult to do with the available user tools. Or alternately I could select, “Get sound from”, which would pick up the sound on an available server. Volume controls, server management, RTP multicast, default servers, all these things could be placed in an options dialogue. Much like, for instance, when a user is selecting a Metacity theme. Padevchooser is too technical, t0o complicated, and it lives in the tray: stop making things too technical, too complicated, and for the love of all that is holy, stop making things live in the tray! I already have a volume control there: that’s enough, thank you very much.

With these usability tweaks, I think PulseAudio would be ready for prime-time. Maybe even for Hardy Heron + 1.

A nice front-end for Samba/CIFS included in the default install. I really don’t like having to alter a config file, and I’m pretty sure that’s a lot of people feel this way. This front-end could, for instance, manage Samba users, SMB/CIFS shares, configuration options, mountable network shares: the whole nine yards. It would be nice. I don’t really like having to open a shell and type in commands; I know, power user, whatever.

For me, a GUI is better. It is. It’s quicker, it’s easier to understand, it’s all in one place instead of two different places and three different commands, and it allows those who enjoy CLI work to continue using their computers the way they want.

• In exactly one week from today, I will be married. Well, okay, one week and a few hours. WOOHOO!
• I checked the uptime on my Linux file/wiki/backup/RAID/CMS server, and lo and behold, it has been running for 483 days straight. That’s awesome! Our Windows 2000 fileserver has an uptime of… one day.
• I have coffee in front of me, and it’s good coffee.
• Last night I watched the film “Paprika”. A bit of a mind-trip. But also okay. Not great, but okay.
• If you hear these words in the same sentence as the word Microsoft, you may consider yourself given a cue to laugh: Standards, Honesty, Ethics, Style, Taste, or Good ROI.
• Okay, I’m back to work.