FOSS Week in Review
Yep, it looks like the end of the week is upon us once again, and with it there has been a lot of news in the FOSS realm. What you might have missed, if you weren’t paying attention, is the following:
Happy 13th, Tux Paint: There was reason to break out the candles this week — 13 of ’em — and put them on a cake before saying “Happy Birthday” to Tux Paint. Tux Paint was first released to the wider world on June 16, 2002. Now that it’s a teenager, we can see what fantastic progress the New Breed Software folks have achieved in this time.
FOSS Force — or rather, I — wrote about Tux Paint back in August of last year, and on its birthday back on Tuesday, there was a note on the site that said, and I quote, “We’ve just learned that Tux Paint (based on what will become the 0.9.23 version) has been created for Android, and is available for free in the Google Play Store!”
Androids notwithstanding, if you haven’t tried Tux Paint, stop what you’re doing right now and download it. If you haven’t tried Tux Paint and you have kids, you’ll actually be doing them a favor by downloading it and letting them go to town with the software. Don’t worry — it’s multi-platform and multi-architecture, so chances are, if you have the hardware there’s a good chance there’s a Tux Paint version for it — kind of like Debian, in that regard. Happy birthday, Tux Paint. I hope your teen years are happy ones.
SeaGL Keynoters: Meanwhile up in the Pacific Northwest, the folks at the Seattle GNU/Linux Conference — that’s SeaGL (pronounced “seagull”) for those of you keeping score at home — have wasted no time in selecting and announcing its two keynoters for the October event, which essentially wraps up the FOSS expo year in North America. Keynoting SeaGL this year are OpenHatch’s Shauna Gordon-McKeon — a freelance programmer, researcher, organizer and writer who has worked for the MIT Media Lab and Civic Commons, among others — and Richard Stallman.
Yeah, that Richard Stallman. You know, GNU’s project leader and the inventor of all those great tools that are essential to Linux or GNU/Linux, depending on what you call it — like the original Emacs, the GNU Compiler Collection, the GNU symbolic debugger (GDB), GNU Emacs, and a wide range of others. Oh, and his advocacy of free software, as well as campaigning against software patents and copyright laws, are the stuff of which legends are made.
Incidentally, the Call for Papers will open soon. One more thing: While she joins RMS as a keynoter, are we now going to have to refer to Shauna as SGM?
Elementary OS Bails on SourceForge: From the and-not-a-moment-too-soon desk in the newsroom comes this report that elementary OS, a Linux distro with an ever-expanding user base, has packed up and skedaddled away from the malware-ridden SourceForge as a source for downloading the distro. In a blog item on their site, elementary OS explains where they’re going — Digital Ocean — and why, although they are certainly more cordial toward their former host in leaving SourceForge than many would have been, given the reason they had to leave. Good luck in your new home, elementary OS!
Help Us Cover FOSS: Though our IndieGoGo campaign is over, as we mentioned this week on our own pages, I would be remiss if I didn’t mention that there are still ways to support FOSS Force. How? Glad you asked. In a word, subscriptions. You can become a one-time subscriber. Or you could become a sustaining subscriber. Remember when we mentioned that you could give up a sandwich a week (or one venti mochafrappiatto from Starbucks) and give the money to your favorite FOSS web publication on a regular basis? That’s how sustaining subscriptions work. Options can be found at the link above, and if you like what we’re doing, give us a sandwich.
One More Thing: My fellow FOSS commentator Bruce Byfield submitted for your approval this interesting opinion piece posted on Datamation this week on the feeble objections to systemd, which seem to be getting more feeble by the minute. At the risk of being a spoiler, Byfield said that systemd opponents “are doing more fearmongering than raising constructive doubts. Systemd may not be ideal, but most users are unlikely to make a fuss so long as their systems continue to boot and function the way they are supposed to.” Amen to that, Bruce.
See you Wednesday.
Larry Cafiero, a.k.a. Larry the Free Software Guy, is a journalist and a Free/Open Source Software advocate. He is involved in several FOSS projects and serves as the publicity chair for the Southern California Linux Expo. Follow him on Twitter: @lcafiero
I read Bruce’s article on systemd. In my opinion it was a very poor and purely opinion driven article with little grasp of the technical aspects of the systemd debate.
He dismissively insists there are no real reasons why people don’t like systemd and insists it must be because they don’t like L. Poettering, or because they might have some kind of shadowy agenda. He goes on to repeat the same tired refrains that because systemd is FOSS licensed it can be ignored simply by forking, and that it is not bloated or monolithic.
All of this is nonsense.
Sure, some people do object to Poettering, but if you think that’s the driving force behind a lot of anti-systemd sentiment, you’re very much mistaken. I personally think he’s a bit of an ass, but as far as I’m concerned, he can be whatever he wants as long as he writes good code. Unfortunately for systemd, he’s an ass who writes bad code.
Forking systemd is not as simple as forking, say, a paint program. As a core system component systemd extends an extensive API to higher level components (like Gnome) which come to depend on systemd being present to function. This, combined with systemd’s ever growing reach and swallowing of other projects means this API is 1) Ever changing. and 2) only replaceable by re-implementing systemd. So to get rid of systemd you are forced to recreate systemd because of the API dependencies now integrated into other applications. Despite this, there ARE several forks of systemd which are trying to work around its spreading influence through the OS. But Bruce would rather have us just focus on Devuan, which is a fork of Debian in its entirety and much more problematic because of the scope.
Bruce’s article also argues against claims that systemd is bloated and monolithic by saying, as we’ve heard many times, that it is a bunch of tools, not one. The truth is, systemd is an ever growing collection of tightly coupled utilities,libraries, and daemons that depend on one another and replace many parts of the system typically handled by independent projects. Why does systemd need a time sync daemon, chron replacement daemon, a web server, and more? Perhaps the most hilariously damning thing Bruce says is in response to arguments that systemd violates the unix philosophy of having each program do one thing and do it well, which mirrors a similar philosophy of good software design: The SRP – https://en.wikipedia.org/wiki/Single_responsibility_principle
Bruce says “Many desktop applications, including LibreOffice and Firefox, violated this principle long before systemd.” He picked perhaps the two best examples of bloated desktop software in existence, bravo. Now that’s what I really want in a low-level core system component which has ties to every other part of the operating system: Something as bloated as a browser or an office suite. Think about that for a bit and then think about how stable and secure your system is likely to be under systemd – the Internet Explorer of init systems.
Too bad Bruce’s article was not a real attempt to discuss the debate behind systemd.
I want to clarify that my comment “He picked perhaps the two best examples of bloated desktop software in existence…” was not meant to be directed against LibreOffice or Firefox specifically, but rather office suites and web browsers in general. No offense to those projects was intended.
“No offense, LibreOffice and Firefox, but your software is bloated.” That’s what I hear you saying in your clarification, Mike, but never mind.
Much of your objection to systemd sounds like the guy who goes to the LUG meeting with a 20-year-old laptop that maxes out at 256MB and complains he can’t run anything new (believe me, they exist — our LUG has one).
I’m still waiting for a valid argument outside of the subjective “single responsibility principle” that keeps coming up — a valid point based in philosophy more than anything concrete. Let me know when you find something legitimate.
Also, I think we need a new definition of “bloat,” since what was bloat 10 or 15 years ago may not be now. Here’s why — storage and memory capacities are much larger than they were then, so if my 8MB or RAM and a 500GB drive can handle a program and what it produces, what do I care how big the program is?
Here’s the Devuan link: http://www.devuan.org/ — if you really can’t live with systemd, then go for it. Knock yourself out.
No, I don’t feel LibreOffice or Firefox are especially bloated compared to other products in their particular realms. It was the category of software he chose to compare to, rather than any particular product, that caught my attention. Office suites and web browsers in general tend to absorb a lot of “optional” functionality which makes them big complex beasts. That’s why there are so few of them around. I’m a software devloper and I’m nowhere near crazy enough to try to write a web browser. My point is that is not the type of model you want for a low level system component.
There is no doubt systemd is many orders of magnitude more complex than the software it is intended to replace. That is an empirical argument that can not be disputed. You may think of bloat in terms of hard drive space, but I (and many developers) see it more an issue of complexity: https://en.wikipedia.org/wiki/Programming_complexity
In layman’s terms, more code means more possible branches in behavior. It tends to be exponential with regard to the size of the code.
Regarding the SRP: It is not a philosophical argument. It is a well established principal of good software design. Complexity breeds mistakes and there is no avoiding it no matter how skilled the developer.
There are plenty of other alternatives for people who don’t like systemd besides Devuan. Gentoo, Slackware, and LFS for example. I’ve switched to building my own custom linux systems from source after starting from LFS: http://linuxfromscratch.org/
I really don’t know why you think my objections sound like some luddite trying to run 20 year old hardware. I tend to run a lot of new stuff – newer than much of the people around here certainly. I’m watiing on a new laptop as we speak- https://www.crowdsupply.com/purism/librem-15
I’m a software developer and I’ve written for a number of platforms including Windows, Mac, IOS, Android, Linux, and even the lowly Windows Phone. I’m fluent with about a dozen programming languages and know something about software design. I’m certainly more current than some crusty old unix curmudgeon. Isn’t that what anyone who dislikes systemd is typically pegged as?
I don’t need the link to Devuan. Do you really think I would read and comment on that entire article that compained endlessly about Devuan without being to find it myself? Here’s a link for you: https://en.wikipedia.org/wiki/Passive-aggressive_behavior
Comments are closed.