Archives for the month of: October, 2008

 About 3 or 4 months ago, I switched from Openbox to an XFCE desktop. Overall it’s been a pretty good experience. Speed-wise it falls somewhere between full-on Gnome and my Openbox setup on my aging P4-3GHz desktop. Feature-wise XFCE has been pretty much on-par with the full Gnome/Ubuntu desktop except I’ve been using the XFWM4 compositor instead of Compiz since I’m really only after window drop shadows.

However, this weekend on a whim I decided to boot into my Openbox session and found the relative speed boost VERY significant. So much so that it got me interested in switching back to my Openbox ways. One of my original reasons for switching away from Openbox was that I wanted to have semi-transparency for the key-status utility that I use during my screencasts. I had completely forgotten about the xtransset utility that can be used alongside the xcompmgr compositor to get exactly what I want within Openbox in this respect.

So now I’m back on the Openbox wagon with several things I’d like to improve or tweak from when I was last using it. Here are the things I’ve never had completely working to my liking in Openbox. I know these are possible, but I never actually got around to doing the requisite legwork to get them done:

- I was never able to shutdown my system straight from Openbox. I selected quit and then it always brought me back to the GDM prompt where I had to choose another menu selection to actually shut it down.

- I’ve still never understood GTK theming when it comes to Openbox. I have no problem changing my Openbox window decorations (and even creating modified themes on my own), but when it comes to changing the GTK controls and icon themes, I’ve never completely understood what I should change and where those changes should be made.

- I still sometimes get window artifacting when I drag some windows. Not always, but it’s annoying when it happens. I’m not sure if this is down to xcompmgr or something else.

- I’m using the Tint panel right now which is small and unobtrusive but it’s also a little low on features. I’m wondering if maybe a move to PyPanel or something else is worth looking into. I want something light however since this is a 5 year old machine and every ounce of system resources is appreciated.

Right now it looks like the ArchLinux wiki on Openbox is the best source of info on these things, but if you’ve got any tips or sources I should look at, I’d appreciate it.

Episode 266 of The Linux Link Tech Show promised to be a very interesting show, and it didn’t disappoint. Why was it interesting? Because Bryan Lunduke of the Linux Action Show came on to discuss his plans to introduce a couple of closed-source commercial applications on Linux. Talk about lots of fodder for discussion! Here are some of my initial thoughts, for what it’s worth.

First a caveat or two. I’m not a huge fan of the Linux Action Show. It’s always been a little too bombastic for me. Maybe I’m just too old, or too mellow. I listen to it every once in a while, but not on anything resembling a regular basis. But you know, I found Bryan to be an excellent listen on TLLTS. He had a lot of interesting things to say and some interesting viewpoints on things. I didn’t agree with everything he had to say, but that also made it very enjoyable. I highly suggest listening to the episode.

Before I completely step in it, let me give you my personal viewpoint on several key things, just so you know going in. I’m not a programmer (although I’ve done some programming). I’m a fan of the concepts of Free Software and Open Source Software, I run Linux at home and XP at work. I run a lot of commercial software for my job - structural engineering - on my work computer. I run very few if any, commercial desktop software applications on my home computer. I love Linux, GNU+Linux, or whatever you want to call it. And I think we all owe Richard Stallman a huge debt of gratitude, but I don’t share his views on a lot of things.

There were a lot of issues discussed during the 2 hour long show, but for the sake of brevity, I’d like to zero in on a specific issue that I feel is at the heart of a lot of commercial vs. free software discussions: whether or not free software developers should expect to make money at it. I honestly don’t think they should - expect it that is. If you’re approaching software development as a way to directly earn money, then free and open-source development is likely an avenue you should steer clear of - and that’s not necessarily a bad thing.

I think the FSF’s four freedoms are fundamental to the success of the whole free software community. However, the four freedoms and the GPL itself don’t deny the right of free software creators to earn money selling the software. But clearly, selling free software isn’t practical when anyone can get the source code and build the software himself. So in today’s world we are left with a lot of Linux developers who earn money in their jobs developing in-house software for companies and/or commercial software. A lot of these same people work on free software projects in their free time for little or no monetary gain. Is this a bad thing? I don’t think so.

There is something to be said for the word “amateur”. Looking at the French root of this word, it also symbolizes someone who does something for the love of it - not necessarily someone who’s skills were substandard. I think in many cases, the ‘for the love of it’ definition still fits. You have people who do things just for the love of it, and others who do things just for the money (and a whole hell of a lot of people in between). But I think getting a group of people together who are doing something primarily for the love of it (the money, as we’ve already discussed is not an option here) is a pretty compelling thing.

Maybe looking for the ‘commercially viable model’ for open source software that Bryan has yet to find is a pipe dream. Maybe end user software development is NEVER going to be a huge moneymaker if free software continues to expand and be successful. Maybe it’s something that Adobe or Microsoft can’t change even if they entered the Linux software market tomorrow.

Look at something like the GIMP image manipulation program. It’s about 12 years old (Photoshop is essentially 18 years old, by the way). And while you could say that the GIMP lacks some of the features of Photoshop, I would challenge anyone to say it doesn’t do far more than what 95% of users need. Remember, we’re talking about building perfectly viable tools that cost the user NOTHING. Lots of them. Don’t you think that at some point, at least some of these tools will equal or surpass their commercial counterparts? Don’t you think at some point people will realize that they can do every conceivable thing they need to do without spending that money? It’s happening now. Not on all fronts. Not all at once, but I have no doubt that it’s happening. And while this is great for the users, it just has to be all doom and gloom for those hoping to make a living creating end user software for money.

So what would we be left with? Maybe Eben Moglen is right. Maybe forty years from now, most software development will be done by young people in developing nations anyway, since the tools are becoming so cheap, easy and ubiquitous. Maybe we have to progress. Move on to other, bigger and better things. Maybe we make a business out of coming up with concepts and ideas while other developing nations do the grunt work. Problem is, we love some of the grunt work, and we don’t want to lose it. Maybe the only people left here doing development work are the ones who are doing it just for the love of it. Hmmm.

So you might ask, what the hell has this guy got against software developers? Nothing. I love free software and appreciate all the people who develop it. In fact I have tons of respect for commercial developers too. I can’t program (at least not in a way that wouldn’t be embarrassing), but I love to share knowledge and teach, so I do screencasts on Inkscape. I also contribute graphics to community projects whenever the opportunity arises. I do what I can. Am I making a living from it? No. Do I really expect to? No. Would it be nice to? Of course.

But we have to be realistic. It just may be that developing free software for the end user is not a money-maker. But that hasn’t stopped anyone so far.

I have been using the Task application created by Paul Beckingham for about 5 months to keep track of my todo list. Before that I was a big fan of the Todotxt.com method. Then I found out about Paul’s app and got it to build on Cygwin (which I run at work on XP). Both Task and Todotxt use a no-nonsense command line interface to create, modify and keep track of tasks. I found Task to be very much like the Todotxt system on steroids - still dead nuts simple but very powerful. If you want to see a good demo of what the program can do, check out this screencast demo.

Just a day or two ago, I finally installed Dropbox on both my home and work machines mostly as a glorified file transfer system between home and work. But the next day a light bulb clicked on and I realized that by keeping my task list file in the Dropbox folder, I could have syncronized access to my task list from home or work. Awesome.

Task keeps your task data in a hidden directory called, unsurprisingly .task (nifty how simple and straightforward FOSS programs can be). And the .taskrc configuration file in your home directory contains the path to that folder. So I simply moved the .task folder into my Dropbox folder on my XP machine at work and modified the path in the .taskrc files on both machines and Voila! instant gratification.

Now I know all you Linux tech-heads will think this is much ado about nothing (which it is) but damn it, I’m quite proud of myself! ;) I know I *should* set up a proper system and ssh into my system.. blah blah blah.. but this was dead simple and accomplishes exactly what I want. I’m willing to give up a couple of geek cred points for the sake of simplicity.

ps. A bit of a disclaimer here. I’ve also been creating .deb packaged versions for Paul ever since v1.1.0 (it’s now at 1.4.2). Task was good enough for me to want to learn Debian packaging, so that must say something!