Get yours!

I've published one solution on this site (the JobMonitor application), but that is only one facet of SysAdmin automation. Another project I've worked on is what I affectionately call "the Automator". It was written for a specific employer, so I don't know if I'll ever be able to give it the light of day, but it's basically a web-based console that allows for automating ssh commands across multiple servers. Sounds scary? Well, it's logging in as a non-privileged user with a private key that is password-protected. It can optionally su to root and run commands like that. It then puts the output into a database and optionally can integrate with the JobMonitor application to error check that output. There's also a scheduling system built into it so you can schedule up commands to be run across various servers using a syntax similar to cron.

It's been used to run over 700,000 commands in 8 months on over 300 servers. Would love to be able to open source it someday...

I remember reading about some deployment of SUSE Linux that I think Nat or Miguel had talked about a few months back. The gist of what they were saying was that there were some big deployments coming, but they couldn't be announced yet. Well, maybe this was one of those, and now it's got to the stage that it can be announced.

Will be interesting to see what effect this has on the IT landscape if it goes through. One successful large-scale deployment like this would make other organizations/governments/educational institutions sit up and take notice. I'm pretty sure that if it's done well once, the floodgates will open. What works to the advantage of FOSS is that something like this would make such big news if it is a success that it would be known about everywhere (in the IT industry). This means that even though there may be other projects that are quite happy sticking with MS, and even if a project like this is, in global terms of usage, not even a small blip, it would have an impact much larger than it's statistical importance because it represents the possibility that any large organization could do this.

By following the excellently simple instructions here I was able to get the CVS version of f-spot downloaded and building. I then took a quick peek in the src directory and saw a file called SlideView.cs. I edited the following section of code:

 private void ShowNext ()
                {
                        StopTweenIdle ();

                        if (current_idx != next_idx && next != null)
                                this.Pixbuf = next;

                        current_idx = next_idx;

                        black = false;
                        transition_interval = 75;
                        flip_interval = 5000; # Changed from 2000
 
                }

And then ran make all install. That's it! I can't believe how easy it all was. Since I'm partially interested in this whole Mono thing, and like the look of f-spot as an application, I've decided I'm going to try and enhance this a little by adding the ability to edit this value from within the menu. It would probably make the most sense under "Edit -> Preferences". If I can do this successfully hopefully I'll have learnt a bit of C#/Mono along the way, and maybe I'll even try and submit it as a patch.

Hard to believe, but I think this is the first time I've taken a look at the source code for a desktop Linux application, and it's been a very pleasant experience so far. Use the source!

Update: 2006-09-13

I've discovered on the f-spot mailing list that someone's already submitted a patch for this specific issue. Would be great to see it make it into the next release! 

Their basic argument is all things being equal, Microsoft fails to dislodge OSS altogether, and OSS fails to break MS's monopoly entirely. I'm okay with that, to be honest - I think that's a realistic point of view, and despite my bias towards OSS, I can believe that's a good model for how things will work out. However, while I agree with their conclusions overall, I disagree with some specific points:

"In addition, a larger installed base [for Microsoft] also implies that there will be more feedback on bugs and more suggestions for new features." I'm not sure I agree with this. In principle, yes, but in practice not everyone who uses an OS files bug reports and send feedback. Who does? An active minority who have the technical know-how to do so (therefore to some extent it's dependent on how easy it is to file bug reports and give feedback, and how useful/specific such feedback is), those who are very passionate about the OS, and those who have a serious pain point with a specific feature or features of the OS. Therefore, it's not entirely accurate to say that more users equals more bug reports and feedback - it's better to say that more of these specific types of users means more bug reports and better feedback. And I'd argue that Linux users as a whole tend to fall more into these categories than Windows users.

"The basic trade-off is the following: With a duopoly, more individuals and organizations use PCs because prices are lower, and this raises welfare. However, with a duopoly, no operating system ends up exploiting fully its potential because developers' efforts wind up divided between the two systems. However, with a monopoly, the efforts to develop new software and improve the platform are directed towards one system only and this may turn out to be better from a social welfare perspective." This takes a far too simplistic approach to the relationship between a software monopoly and software development. In practice, there are other issues to consider:

• How easy is it for developers to develop on a given platform?
  
• If there are cross platform development tools, efforts to develop applications for an OS aren't split between OSes - they benefit both
• A monopoly leads to software homogeneity
• From a security perspective this is not in the social welfare
• It can also lead to a lack of innovation because there are no alternatives to experiment with new ideas and approaches to things.

"we were also surprised to find that piracy may end up increasing Microsoft's profits". I don't think you should be surprised by this. I don't think Microsoft is and it's a fairly well known phenomenon in geekdom that MS would prefer piracy to using an alternative.

Other than that, I think this article has a lot of good points. Some interesting thoughts on how FUD (and other morally questionable approaches) is a strategic approach that can benefit MS. 

I fully support the project to move to ODF, and I don't want to sound like I'm souring on the poor souls who are trying to implement this, but what began as a promising project by one State (doubtless with many others watching) to move away from a proprietary document format to an ISO approved open standard document format, ODF, has deteriorated into using a third party plugin to Microsoft Word that "enables" ODF integration with Word.

Kind of.

Really what it does is allow a one-way transition from ODF to Word's DOC format. I don't see how Mass. can maintain that ODF will be their default format when the plugin in it's current state only allows you to import to Word DOC format and doesn't allow for native saving of documents from the "Save As" menu as ODF...

This is a big blow for ODF. It will show ODF to those who use it in the worst possible light. And rather than fixing the accessibility issues with OpenOffice (I really hope someone is doing this - this is Open Source after all, so implementing these kinds of things should be very easy), they've taken the easy way out and reverted to using Word with a crippled plugin.

I hope I'm wrong on this one. If progress away from proprietary Desktop OSes is to continue/start, the battle over document formats is only the first step. Or is it simply an indicator? Time will tell... 

I think there's a real opportunity here for AMD to step in and Open Source ATI's graphics drivers. It'd be great for the Linux community to know that they could get an AMD CPU and at the same time be confident of having an Open Source graphics driver.

This makes a big difference on laptops particularly, because Open Source graphics drivers make suspend/hibernate a lot easier to do. Which would be nice. 

 Without this, I'm very likely to make my next laptop purchase be an Intel chipset with an integrated graphics card. Enough to run XGL and a bunch of games would be fine for me, which seems very do-able.

Seems like it was a killer event. Would have especially loved to have caught the Django talk, have caught up with Paul Cooper and met the author of PiTiVi.

Unfortunately it wasn't to be, but great to see the momentum of community run events like this. Looks like Guadec is going to be in the midlands of England next year, so if we're still around for that, maybe I can go to that (and of course, LugRadio Live 2007).

One of the best things for me is the possibilities of where we'll be (I'm talking about the larger Open Source community when I say we - of course it's a gross generalization to imagine it's one group) a year from now. A year ago, Ubuntu was beginning to hove onto the horizon, XGL was a collection of letters that didn't belong together. It'd be quite interesting to chart all the developments in the Open Source community on a timeline so that you could see where things have moved. It'd necessarily be selective, otherwise you'd suffer from a serious case of information overload, but I think you could still make it somewhat interesting.