Friday, February 18, 2005

Some Thoughts on Linux 

It's often said that Linux, due to its open source nature, can't really be innovative. I think Sabayon, a recently announced project, demonstrates that that's not necessarily true (the conventional wisdom that Linux project names suck still cannot be rejected, however).

It is a sysadmin tool for Gnome desktops, and it has a really clever idea about how to do this. Instead of a gigantic tool for setting all sorts of settings, which need to be specially programmed and documented, Sabayon simply opens a new X session in a window, pays attention to what preferences you change in that session, and then saves that as a configuration that you can then push out to users. Check out the blog post for a walk-through of how this works. This is a really clever solution to the problem, and I think it will ultimately prove to be very useful and attractive to administrators (especially amateur administrators).

I think it's also interesting that this is a solution that you pretty much just could not do on Windows without a huge amount of deep magic and/or expensive software. Perhaps there is something to the protocol-based nature of X-Windows afterall.

While I'm here, I'd also like to point out some promising Linux projects aimed at end-users. Some of this software looks so great, that it actually moves me towards wanting to run Linux to use it. For example, Tomboy is really cool, and I often find myself wishing I had it handy. I've also wanted a search tool like Google's search service recently, but I've found all of the offerings to be unattractive for one reason or another. Apple's Spotlight looks really good, but Beagle looks almost as good (I especially hope they implement the searching for control panel preferences that Spotlight does). Related to Beagle is Dashboard, which is a sidebar that constantly brings up information related to what you're doing. Some hackers are working on coding up an X server (Xgl) that runs on top of OpenGL, which would really speed up X and make a MacOS X-like interface possible. The recently announced Hula Project probably won't be of direct interest to most users, since it is just an email and calendaring server, but it looks promising and should make corporate users happy, if Linux ever gets rolled out there. iFolder makes it so that I can stop emailing myself files by setting up a centralized file server and service so that you can have a directory synchronize with other machines in the background. Currently, emailing files to yourself is just about the only good way to have your files in multiple locations, and it's a really archaic way of doing it. If Linux continues to accumulate desirable software like this...

One thing I should point out is that most of the software I mentioned above is written in Mono (not Hula or Xgl). I've written previously about Mono and whether or not the Gnome community should be embracing it (specifically, I thought they shouldn't). Looking back on it now, it's a different picture. By really embracing it, they have managed to push out some rapid and impressive software projects. Having that software available is doing a lot of good for Linux, and although I can't say for sure whether that is worth the potential legal problems with Mono (depends on whether or not the threats ever manifest themselves), it certainly puts more weight on the "yes" side of the scale.

Finally, I think it's interesting that several pieces of software I just mentioned are actually very near the "web services" space, despite being open source. Hula is really only a web service, and iFolder requires a server you can connect to. No one is going to run those servers for free, and few people will have the hardware, bandwidth, and expertise necessary to set them up for their own usage, but I'd pay a reasonable fee for a "Novell Web Services Suite" that will let me move my files around and calendar with my friends (or tolerate ads to use it). I don't know what this means in the big picture, but I do find the development of open source web service software interesting. Perhaps it points to a reason to pay for Linux distributions: pay for the support, get the web services along with it.

Afterthought: I started off talking about how open source can be innovative. I should also point out that both Beagle and Dashboard were projects that were started before Apple announced Spotlight and before Microsoft announced Implicit Query. Which doesn't mean the open source software started first, or will hit 1.0 first, but it does mean that they thought of these projects on their own.
Comments: Post a Comment

This page is powered by Blogger. Isn't yours?