Goodbye, Linux Journal

So my subscription to Linux Journal has lapsed, and I’m not remotely disappointed. Which is a little sad, since LJ used to be one of my favorite technology periodicals. For quite a while LJ and Dr. Dobbs were the only two journals I’d pay to receive: and now there is one. So what brought about this change? The new editor, Nicholas Petreley. Well, new may not be the right word — he’s been there for over six months now. But his column, /etc/rant, has driven me (and others ) away.

/etc/rant is, as the name implies, a place for Petreley to rant. And since he’s the editor, that’s his right. Here’s the thing: I don’t like rants. In particular, I don’t like absolutist positions that don’t allow for movement. I do like pluralism, and that usually fits right in with open source software. So while /etc/rant wore on my from the start, the February column, The Spirit of Open Source , is when I decided “yeah, I’m over it.” The Februrary column discussed a particularly silly flame war brought on when Linus Torvalds recommended KDE over Gnome stating (my words, paraphrased, possibly mis-remembered) that GNOME treated him like a child and didn’t let him twist all the dials he’d like. Even though I personally enjoy using Gnome and use it as my daily working environment, I can respect Linus’s opinion. After all, regardless of his contributions, Linus is just one man, and his choice of KDE doesn’t really matter to me.

But in that column Petreley took Miguel and GTK to task for using the LGPL claiming that it violates the spirit of open source. I have a few problems with this claim. The first (and one that really frustrated me the most) is the statement:

The whole point of the LGPL is to allow you to add something to GTK without having to compensate the GTK developers with either money or source code. No, it doesn’t allow you to add to GTK and not release your changes. I guess we can argue over what “add something” means; to me it means having your code distributed with GTK. I suppose Petreley might be using “add something” instead of “link to”, but that would just be sloppy. However, even assuming that is the case, so what? If a developer was OK with releasing their code under an open, freedom-granting license from day 1, why not on day 2? Just because it might be used by someone? If they were really concerned about that situation, they wouldn’t have (shouldn’t have) released it in the first place.

Petreley also draws QT into his argument, accusing “open source zealots” of mistakenly lambasting it for it’s poor license when (in his view) it is more free since it is available under the GPL. I get what he’s saying. But I don’t agree that QT is better or gets a pass because of it’s license, or even that it’s superior to GTK in this respect. My feelings on licensing have swung from GPL to MIT in the past two years: I like my freedom libertarian style (sunny-side up, natch), so a license like the MIT license (or LGPL) fits my mentality better than the GPL these days.

So my subscription has lapsed, and I’m OK with that. It actually took me a bit to realize it had happened. I was in Border’s and noticed a new issue on the newstand. And I was about to reach for it and I remembered “oh yeah, I think editor’s sort of a flake…” and put it back down. So maybe I’ll check it out in the future, maybe another periodical will catch my interest. Regardless, thanks for everything Linux Journal; have a good life.