The Myth of the Bad Ubuntu Release

OK, so there’s something that always disturbs me when release time comes around. Here’s a rough chronology of every Ubuntu release (at least since I’ve been involved, so that goes back to Breezy Badger) and what the “buzz” around the internet says:

1. Alphas come out: buzz says,  “not much to see here folks, move along.”
2. Beta comes out: buzz says, “wow, great release, but where’s the new artwork?” and I’m thinking “How on earth can the pull this off?”
3. RC rolls around: buzz says, “new awesomeness right around the corner!” and I’m thinking “darn it, there’s a lot more to do.”
4. Release day: buzz says, “OMG I have to download this” and I’m thinking “phew, that’s over, I’m glad I rsync’d/zsync’d yesterday”
5. The week after a release: buzz says, “Noooooo, this is the worst Ubuntu release EVER!” and I’m thinking “wow, they really did pull it off”
6. Rinse and Repeat

Now, with that rough release chronology, I want to look at #5 a little bit more. I don’t consider myself particularly a diehard Ubuntu fanboy. I’ve been an Ubuntu developer for a few years, but I’ve mostly had to step back due to time constraints. I’ve gotten (much to my dismay and consternation) a bit of a reputation as being the “Devil’s Advocate” and a vocal critic of Canonical. This is mostly due to intense burn out, but that’s a subject for another post. My point here is that I’m not just blindly in love with everything Canonical and Ubuntu does.

I’m quite convinced, having participated in every Ubuntu release since Breezy Badger, that the “This Ubuntu release is the worst ever. <some previous release> was so much better!” sentiment stems from two factors:

1. it’s “true” because people say it is. If you read 20 “I’m not happy with this release because …” or “Argg, I’m going back to release X because Y is broken!” you’d get the impression that things are really in bad shape. The problem is that you don’t see the 20,000 people for whom the new release is just fine. People naturally expend more effort complaining than they do praising. This is similar to watching the news on TV. You get more ratings for telling people how awful the world is than talking about the good things that happen.
2. the same software affects different people differently. I don’t know how many times I’ve seen people say a new Ubuntu release is trash because one particular piece of hardware now doesn’t work right or behaves differently. What in fact is most likely occurring is that the absolute number of people having problems doesn’t change dramatically, but the parts of the user population that experience these “critical FAIL” moments shifts as code changes.  I’m willing to bet for every person that has more problems with Karmic Koala than Jaunty Jackalope, there is another who has the opposite experience. I know that for  me and my Intel video card, Karmic is a blessing. In fact, I’d be willing to bet that in a release or two people will be looking back at Karmic Koala as one of the best Ubuntu (and Linux) releases.

So my conclusion, for what it’s worth, is that while some Ubuntu releases are a bit better than others, this periodic buzz around the internet that the latest Ubuntu release is an epic FAIL is a self-perpetuating myth, mostly caused by people needing something to complain or write about.

Making money with FLOSS, really?

Martin just posted a lengthy critique of an article by Nick Fox on Linux business models. I’m not going to address Nick’s post but I do have a few issues with Martin’s.

Martin says regarding the idea that commercial software == closed-source software:

It also assumes that FOSS can’t possibly be commercial, a big mistake and a common myth. You can take a copy of a GPL licensed program and sell (that’s right, for money) the software to someone.

I would say it is a common FOSS myth that you can make money selling FOSS just because you are allowed to. This is why you will (almost) never find an company that actually sells just the open-source software. In an age when distribution of software is so cheap, why pay for something that you are almost guaranteed to be able to download for free?

Martin says:

The nature of the free market is that goods or services will be priced very close to the costs of replication and distribution.

In a free market the price is dictated by supply and demand. Supply depends on not only the cost of replication and distribution, but significantly in this case on the cost of production. It’s trivial give away something for free which cost you nothing to get! The reality is that software is not free to produce, even open-source software. You can ask Mark Shuttleworth about that. Additionally, there is currently high demand for closed-source, proprietary software products. This is mostly because this development model can still give good quality, professional software at a price that people don’t mind paying for. This is especially true in niche markets that don’t have a lot of interested FLOSS developers. The FLOSS development model can give better results, but that doesn’t mean it always or even often does, at least at this time.

I think there are a number of people out there in the Linux community that feel that open source would dominate the software world if only it wasn’t being “suppressed” by the big bad corporates and those odious, evil patents . I think that these definitely create an up-hill battle, don’t get me wrong. But I think our single digit market share still has more to do with the lack of quality software, quality support, and good money-making business models. It’s hard to be profitable in FLOSS (you essentially have to sell something other than the software), which makes it less appealing to big companies and mostly remains the land of “computer geeks”. I still love it, but it seems to me that FLOSS is more an academic and hobby enterprise that you’re lucky if you get paid for. Of course as time goes on, and the trend seems to support this, FLOSS could be come the dominate software development model for businesses. I sure hope so! But I don’t think it will come from selling software.

Edubuntu shout-out

I want to send a big thank you to Dave Morley for the testing he did for Edubuntu Beta. For Karmic Koala, Edubuntu has made the big jump to a full installation DVD. Since 2006 Edubuntu has gone from a single CD, to a 2 CD set, to an add-on CD, and now to a DVD that has both a Live installer and text-based installer and 3.5 GB of open-source goodness.

Anyway, back to the story. We’ve been struggling this release to get enough testing for the Ubuntu Release Team to officially put out ISOs. I was really getting worried that we’d miss Beta altogether due to a lack of testing. Downloading/rsyncing a DVD is no small matter and there are lots of things to test on a DVD. So imagine my surprise and delight when I did a last minute check of the ISO testing tracker to see that Dave Morley had completed all the tests for both amd64 and i386! Not only that, he’d filed some good bugs for us to look at.

So thanks to Dave, Edubuntu has Beta DVDs available for testing. Rock on!

MIT Global Jam

Tonight I went to my first Ubuntu Global Jam event. It was held at MIT (awesome!) in the Stata Center. First of all, as a chemist I was thrilled to be on the MIT campus. I often considered applying there as an undergrad but decided to stay on the West Coast. The Stata Center didn’t disappoint. Secondly, this was the first time since moving to the Boston area a month ago that I actually got into the city (yeah, technically Cambridge, but close enough). I took the Boston metro for the first time, which worked out pretty well.

Finally, I was able to get ~20 Edubuntu bugs squashed and seven packages completely triaged. I also sponsored an upload for one of the Jam attendees. All-in-all, it was a productive time and definitely worth the trip. I especially want to give a big shout-out to the MIT SIPB (Student Information Processing Board) for hosting the event. The pizza was very excellent.

Busy, busy!

The last two months have been quite hectic. I thought I’d give everybody a brief outline of what I’ve been up to:

• Finished my dissertation, defended it, and received my Ph.D. This is huge news for me. It took me 7 years but I finally got it done. For those that may be interested, my degree is in Chemistry and my dissertation title was, rather blandly, Ultra-Sensitive Detection Methods of the Orientation of Surfaced-Adsorbed Molecules. Basically I “developed” a new analysis technique that would most accurately be called Polarization-Resolved Normal Incidence Cavity Ring-down Spectroscopy (PRNICRDS, yes scientists love acronyms). The purpose was to use it to detect the rotation of an artificial, light-driven, molecular motor less than 1 nanometer in size.
• I got a job as a postdoctoral research working for the Space Dynamics Lab (run by Utah State University) and contracted to the US Air Force Research Laboratory at the Hanscom Air Force Base near Boston. If that isn’t confusing enough, I work in the Battlespace Environment division of the Space Vehicles directorate. I think this is probably the closest I’ll ever come to being a rocket scientist It’s exciting work even though working in the military has it’s own peculiarities.
• Given my new job, my wife, two pets, and I moved from Reno, NV to Boston. We’ve always been west-coast people so it’s been quite a different experience. Neither one of us had been to New England before and we’d only been west of the Rocky Mountains for family vacations. We love the weather so far but are bracing for the (in)famous New England winter.
• In between all this I seem to have been able, with help from Colin Watson,  to get the seeds for the new Edubuntu DVD working. This is a pretty big step for Edubuntu as for a few releases now we’ve been just an addon to Ubuntu. This meant our users had to first install Ubuntu and then grab the Edubuntu disk and install the educational packages from there. Additionally, LTSP, which is commonly used in schools labs, was shifted to the Ubuntu Alternate CD. The decision to go to a DVD wasn’t easy though. DVDs are a big download and we have to make sure that we maintain good sync with the Ubuntu installer, kernel, etc. Edubuntu 9.10 will be basically a Beta for Edubuntu 10.04 LTS as we’re trying to get the DVD thing going and revitalize development.

Well, those are the highlights from me, nothing too important but it’s been keeping me busy. I’m looking forward to seeing Karmic released and then focusing on the big push to the next Long Term Support release.

Microsoft and software morality

Martin Owens recently asked the question “Why do you like Microsoft?” It’s a pretty brave thing to ask an free/libre/open source Linux community like Ubuntu. While some have taken up the challenge, it made me pause and think for a minute. Do I like Microsoft? Do I hate Microsoft? I suppose the answer is somewhere in between. Here’s some reasons that I could like Microsoft:

• It was my introduction to computing. In particular Windows 3.1 was my intro to a GUI and Windows 95 was my intro to the Internet (where I first downloaded Debian on 11 floppy disks).
• It basically just works. There are certainly frustrations (why must I reboot so much, why is setting up a printer so difficult) but by-and-large Microsoft OSes work pretty well.
• There is so much software available for it and it’s a defacto standard. Whether it be the latest game, tax preparation software, or even state of the art scientific software, one can almost always find a Windows version.
• Microsoft Office is an excellent, if aggravating, software suite. PowerPoint and Excel in particular can provide good quality output with minimal effort.
• Windows is fairly easy to give support to for family, etc. because there is not hundreds of different “distributions” of it running around and a new release comes out every 5 years or so, not every 6 months.
• Despite some “interesting” methods, I find the story of a young “nerd” making it big a bit inspiring. These days it seems not uncommon to have < 30 year olds running multi-millon dollar companies but when I was growing up it was pretty unheard of.
• In the end, it’s kinda nice to be  a “consumer” sometimes rather than having to scratch itches to get what I want done.

Of course there are many reasons to not like Microsoft:

• When Microsoft products are frustrating there’s really not a lot you can do about it but hope the next version is better.
• It’s expensive. Windows alone isn’t usually a problem, but adding on MS Office and the “niceties” can get expensive for people on a student’s budget. This is the primary reason I haven’t purchased a Microsoft product in a few years and I’m relegated to using ancient software I’ve already purchased  the couple times a year I find I need it.
• For me personally I have some moral issues with organizations Microsoft has donated money to in the past. They’re certainly not alone but they are still on my naughty list.
• Microsoft products just are not that fun. I think it’s  fun to get to test, twiddle, and tweak my OS. Of course I pay for it but still, lots of fun.
• FLOSS is a more inviting, educational, and community-driven. Microsoft is pretty boring.

So in the end, Microsoft seems to work fine for ~ 90% of people and that’s OK. To me open-source software is a much cooler and hopefully useful way to develop software but I understand why it’s not for everybody.  I’m not sure that it will ever be the dominate or even most efficient development model, but I certainly wouldn’t  mind if it was

I think I view open-source mostly as a fun hobby that sometimes produces great software and can even be turned into a profitable enterprise occasionally. I just don’t see a moral or philosophical imperative to software. Proprietary software does not restrict or take away any freedoms or rights I might have, it’s just not as open and inclusive as it could be. I believe it is a grand mistake to try to cast proprietary software (and companies that produce it) as somehow inherently evil. Open-source software wins when it is the better, more useful, and more valuable product, not when we guilt people into thinking they’re somehow criminal or immoral for not using it. Attaching some sort of morality or idealism to the proprietary/FLOSS software struggle just seems silly to me. It is just software after all, it’s not that earth-shattering or deeply important in the grand scheme of things.

Edubuntu Meeting Minutes

The meeting was well attended with lots of new people (awesome!), attendees included:
bencrisford, nestor, Timequeezer, abruptus, Svenstaro, Lns, LaserJock, asanchez, pygi, Ahmuck, davidgroos, alkisg, jt4sugar, nubae, and some stragglers.

Agenda Item: Lns / nubae – Are they in Edubuntu-members yet?

Due to having only 2 out of 5 Edubuntu Council members still available, it is necessary to reestablish the Council, which is tasked with approval of new members. The Ubuntu Community Council, who has delegated the membership approval to the Edubuntu Council, should be consulted.

Action: LaserJock will send email to edubuntu-devel and Ubuntu Community Council to figure out how to reestablish the EC

Agenda Item: How to effectively advocate/market Edubuntu to schools/school districts

Because the future of what Edubuntu will look like is still in process it is difficult to have a specific discussion on potential marketing and advocacy initiatives. However the idea of collaborating with the Ubuntu Students team was brought up as a general strategy.

Action Item: bencrisford is to email edubuntu-devel with more details for the idea of working with the Ubuntu Students team on marketing

Agenda Item: How teams will be structured in future.

There are currently 14 non-LoCo Launchpad teams related to Edubuntu. Many of the teams are no longer in use or are owned by people who have left the project. A review needs to be done on the teams, and evaluation of what teams are still needed, and work with Launchpad administrators to do the cleanup. It was suggested by nubae to look at how Sugar has organized their teams.

Action Item: Lns will spearhead the Launchpad team cleanup process

Agenda Item: Attracting developers/contributors to the Edubuntu project.

There was a lot of discussion about the lack of developers and contributors in Edubuntu. There were lots of suggestions including:

• demo videos to show potential contributors what kind of things are available
• getting Ubuntu Students involved
• something like Google Summer of Code, but for Edubuntu/Education
• having more specific objectives and tasks for people to do
• better documentation
• mentoring

Action Item: pygi to start thread on edubuntu-devel about building the developer community
Action Item: bencrisford to look into collaboration opportunities with the Ubuntu Learning project

At the end of the meeting there was a general discussion of where Edubuntu is heading and how to move it forward. A major concern is that having Edubuntu as an add-on CD to Ubuntu has been a detriment to users and caused a decrease in interest and users. Anecdotal evidence indicates that a single installation medium is very desirable. Everyone was also encouraged to read the draft Strategy Document and comment on it.

Action Item: nubae and alkisg to look into the possibility of going back to single install disk/media

P.S. The raw log is available. The meeting starts at 19:00

Future of Edubuntu meeting

Tomorrow (Friday the 22nd) the education community in Ubuntu will be gathering at 18:00 UTC in the #edubuntu channel on the freenode IRC network.

The topic will be if or how to move the Edubuntu community forward towards Karmic and beyond. The Edubuntu community has had a rough go of it for the last year or so but it looks like some “fresh blood” is wanting to re-energize the project and get Edubuntu back to being a leader in the educational Linux market.

If you’re a K12 educator, an educational app developer, school IT adminstrator, or an Ubuntu developer/contributor who is interested in Ubuntu for education and/or young kids please drop by the meeting. The possibilities are pretty wide open and community is seeking input, feedback, and contributions of all kinds.

Landing on the GeekDeck

As Og has already pointed out, our good friend Pete has managed to rope us into another one of his crazy projects, GeekDeck. It’s basically getting together a group of computer geeks who like to write and putting together an eclectic “zine/blog thingy … to expand peoples geek palettes”. It’s going to have monthly issues covering a large range of topics and will hopefully have a nice looking PDF version as well soonish.

This month I talked about a couple personal examples of how I’m using open source software in undergraduate and graduate science (specifically physical chemistry) education. Og’s “Cherry Picks of the Month” is fantastic, a great read, as was Pete’s interview of our very own Alan Pope. GeekDeck Issue 1 has a total of eleven articles so go check it out!

Who’s in your Ubuntu fav 5?

So Jorge just posted about nominating people to the Ubuntu Hall-of-Fame. He also hit on something that I’ve been thinking about for a while. While the HOF is a cool place to so who’s rockin’ the distro, for me it’s still a little impersonal. Maybe it’s just me, but I love to share why the people who rock have made such a huge impact for me. I also like talking about the “unsung heroes”. So here’s a meme for you all:

1. Think of 5 people in the Ubuntu community who have influenced you the most, or helped you the most, or who you most idolize.
2. For each person write a couple sentence summary of why they are in your Ubuntu Fav 5 and post somewhere.
3. Nominate at least 1 person from your Fav 5 for the Ubuntu Hall-of-Fame.

Here are my Ubuntu Fav 5:

• Oliver Grawert (ogra) – One of the first developers I talked to back in 2005 when I first got involved with Ubuntu. A great motivator for wanting to make MOTU. We worked together quite a bit on Edubuntu and he’s a great mentor and reminds me that open source is FUN. He also loves cats, ’nuff said.
• Colin Watson (cjwatson) – Colin is like the epitome of an Ubuntu Developer, he’s smart, efficient, and helpful. He’s the force behind Ubiquity and much of what makes Ubuntu, Ubuntu. He’s been on both the Community Council and Technical Board and I trust him eminently to guide Ubuntu into the future.
• Barry deFreese (bddebian) – Barry is the guy that got me involved with Ubuntu development back during Breezy Badger. He was a fun part of the MOTU community, and even though he’s come and gone over the years as life dictates, he’s always just quietly done what he could to make Debian and Ubuntu a better place, with his typically self-deprecating and humble attitude.
• Martin Pitt (pitti) – Martin is pure awesome and is one of the most upbeat guys I know. What I like about Martin is that he’s very cheerful and friendly, even in the mist of imminent freezes and deadlines. And whenever I’ve had to ask him for help on something he’s always thanked me for my work, when he’s often the one that makes it really work.
• Steve Langasek (slangasek) – Steve is the Ubuntu Release Manager, a long time Debian Developer, and a really cool dude. What impresses me the most about Steve is that even though he’s like THE MAN and works for Canonical,  he participates in the, mostly volunteer, MOTU community.