What can Linux learn from Toyota?

Before I begin, let me set the premise for this article.

If you follow news at all, you’ve probably heard of the recent recall fiasco involving various Toyota vehicles.  If you’re not familiar with the situation, here’s a brief recap courtesy of the relevant Wikipedia article:

Three separate but related recalls of automobiles by Toyota Motor Corporation occurred at the end of 2009 and start of 2010. Toyota initiated the recalls…after several vehicles experienced unintended acceleration. The first recall, on November 2, 2009, was to correct a possible incursion of an incorrect or out-of-place front driver’s side floor mat into the foot pedal well, which can cause pedal entrapment. The second recall, on January 21, 2010, was begun after some crashes were shown not to have been caused by floor mat incursion. This latter defect was identified as a possible mechanical sticking of the accelerator pedal causing unintended acceleration, referred to as Sticking Accelerator Pedal by Toyota…

As of January 28, 2010, Toyota had announced recalls of approximately 5.2 million vehicles for the pedal entrapment/floor mat problem, and an additional 2.3 million vehicles for the accelerator pedal problem. Approximately 1.7 million vehicles are subject to both. Certain related Lexus and Pontiac models were also affected. The next day, Toyota widened the recall to include 1.8 million vehicles in Europe and 75,000 in China. By then, the worldwide total number of cars recalled by Toyota stood at 9 million.

Obviously this has become a PR disaster for the world’s largest automobile company, but the fallout has affected more than just Toyota.  The following commentary comes from one of many articles on the matter:

The Japanese government is getting increasingly worried that the Toyota debacle might turn into a worldwide backlash against Japanese cars, or even all Japanese products. As the world’s 4th largest export nation, Japan has a lot to worry about.

Prime Minister Yukio Hatoyama himself urged Toyota to ensure the safety of its vehicles and customers worldwide: ”When an event that impairs safety occurs, the initiative should be taken to work for the safety of people in Japan and worldwide.”

Land, Infrastructure, Transport and Tourism Minister Seiji Maehara was a bit less diplomatic. He said Toyota ”lacked customers’ perspective” and reacted too slowly: ”It might be that Toyota considered it a minor problem,” Maehara said and added that the company must deal ”quickly based on the viewpoint of customers.”

…Chief Cabinet Secretary Hirofumi Hirano said he ”would like Toyota to deal with the situation properly so that it can alleviate concern among users.”

…[Foreign Minister Kaysuya] Okada is concerned that ”this is a problem for the whole of the Japanese auto industry and it is also about trust in Japanese products.”

One final observation before I transition into the main point of this article: 9 million may sound like a lot of recalls (and it certainly is), but consider that Toyota manufacturers nearly 9 million cars per year – meaning there may be upwards of 75+ million Toyota-manufactured cars currently in use.  That perspective will become important in a moment.

So what does this have to do with Linux?

I believe the impact of the Toyota recall on the wider perception of Japanese automakers is hugely relevant to Linux.  Here are the parallels:

• Japanese automakers: Linux (for purposes of this discussion, desktop Linux)
• Toyota: Canonical (maker of Ubuntu – currently the largest, most popular distro)

(For the record, I’m not interested in debating whether or not Ubuntu is the most “popular” desktop Linux distro – it is absolutely the most visible, which makes it appropriate for this analogy.)

Let’s delve into this a bit deeper.

Why Canonical is like Toyota

In classic form, the launch of Canonical’s latest and greatest (Ubuntu 10.04) has garnered huge amounts of media attention – some good, some bad.  For purposes of this article, I’m going to focus on the bad.

Why?

Reason #1: when it comes to public perception, bad coverage ALWAYS trumps good coverage.

Reason #2: much of the overtly positive Ubuntu coverage comes from blatant fanboys and/or media personnel who fiddle with a liveCD for 5 minutes before writing a supposedly “comprehensive” review.  Reviews like this simply don’t deserve attention.

Reason #3: I am a disgruntled Ubuntu user, so I personally connect with many of the negative 10.04 reviews.

Early last year (2009) I wrote an article titled “The Only Feature Ubuntu 10.04 Needs“.  It has since become one of this site’s most popular articles, in part because so many people have read it and agreed (at least according to the article comments).  My point in that article is summed up as:

Two significant things need to happen between 9.10 and 10.04 if Ubuntu wants to stay relevant:

1) no new Ubuntu-specific features in 10.04 (new upstream features are fine)

2) make Ubuntu 10.04 a fix-only release

To save you reading the entire article, the basic premise was this:

Ubuntu is an excellent distribution, possibly the best general-purpose distribution of them all.  More than any other distro in recent memory, Ubuntu has the opportunity – and the financial backing – to carve out a real place for the Linux Desktop.

But there is one major, significant barrier standing in its path.  No, I’m not talking about the brown interface (which should improve in the future) – I’m talking about the perception that Linux doesn’t “just work.”

Yes, us lxers we know that it’s remarkably close to “just working.”  We know that if something doesn’t work there are a hundred different sites we can go to for help.  We know that driver issues are rarely Ubuntu’s fault – they’re the fault of lazy hardware vendors.

But everyday users don’t know this, and they arguably don’t care.  For them there are no excuses, only “it all works” or “it doesn’t work.”

And the sad fact of the matter is that it only takes one piece of hardware or one peripheral that DOESN’T work for the average user to throw Linux out the window.

But here’s the thing – it doesn’t have to be this way.

Up to that point, I’d had a pretty good experience with Ubuntu as a general-purpose desktop OS.  I first started using Ubuntu with the 8.04 release and a fair amount of work was required to get everything working.  The 8.10 release fixed many of my problems, and 9.04 was nearly flawless as far as my hardware was concerned (two desktops and a laptop).

When I wrote that article, my biggest concern was that Ubuntu’s reasonably good stability would lead to a change of focus from stability and performance to “features.”

I think it’s fair to say that my concern was justified.  VERY justified.

As I’ve written elsewhere, Ubuntu 9.10 was a disaster.  The given explanation was that a ton a features were crammed into 9.10 so that 10.04 – an LTS (long-term support) release – could focus on stability.

So WTF happened?  Ubuntu 10.04 was one of the most feature-heavy releases in recent memory, and its stability – at least so far – is not good.  Let me linkquote some of the most poignant reviews on the matter:

• Much as it pains me to say this, installing Ubuntu 10.4 on my desktop PC was a nightmare.
• I’m rather taken aback that it took this much work to get this going.
• I was content with 9.10. Why, oh why, did I upgrade?
• I feel that I can no longer support, advocate or promote Ubuntu as a distribution. (Tanner’s note: check out the many roles this guy had in the Ubuntu community.)
• I was really hoping that 10.04, being a LTS (Long Term Support) release, would have focused on supreme reliability and stability.  A sort of “9.10 without the bugs.”  Unfortunately this was not the case.
• While Ubuntu 10.04 LTS is scheduled for release today, development of this “Lucid Lynx” release has not been as optimal as many would have liked.
• “X freezes (GPU lockups) are being experienced on i845, i855 and other 8xx chips.” (Tanner’s note: Intel’s 8xx line comprises the majority of Intel chipsets made between 1999 and 2004 – the kind of chipsets Linux should rock.)
• Restoring my computer after a fresh install of Ubuntu 10.04… What doesn’t work: any gnome setting, evolution accounts (mail, calendar, tasks), telepathy/empathy and theme.
• Ubuntu users [are] having their desktops crashed as soon as they click the NoScript icon. Yes, your whole session gone, back to your logon page in one move!  Actually this may be triggered by different actions, not necessarily with NoScript installed, or reportedly just by having a long/complex enough interaction with Firefox.  It turns out to be a bug in the xorg-server package, which Ubuntu’s maintainers deemed not worth to get fixed before releasing Lucid Lynx.
• My overall feel is that this release was somewhat disappointing. Here’s a summary or reasons why: – Some basic functionality that was working just days ago on the Release Candidate does not today on the same hardware. – There is a bit of an overall “sloppiness” feel to Ubuntu 10.04 so far, which is specially significant given its LTS quality. (Tanner’s note: this is a very well-done review)
• When the system asked for reboot at the end of the sequence of package downloading, installation and removal, I was ready to leave it… However, I met with a hitch when the machine restarted: it couldn’t find the root drive.
• Recent comments from Mark and the way the community is headed has made me wish to part ways with not only using Ubuntu (in my case, Kubuntu) but also testing it.
• Yet again, with Lucid Lynx, Ubuntu has shunned a better technology for no good reason…
• Plenty of people appear to have had a great upgrade experience. Mine was a nightmare.
• I upgraded Ubuntu to 10.04 LTS (Lucid Lynx) last night and had an unpleasant surprise this morning – the fonts were upside down! (This link describes one of the most humorous bugs in recent memory :)

Some readers are inevitably going to accuse me of cherry-picking negative reviews, and that’s not incorrect – as I’ve clearly stated, I am deliberately picking out some of the worst problems with this release.

Thing is, I could have easily sampled another hundred of these.  If I wandered into the Ubuntu forums, I’d have enough bad 10.04 experiences to fill a hundred blog entries like this one.

I’m well aware that many of you will feel compelled to leave a comment stating “Ubuntu 10.04 worked fine for me.”  Good for you.  60+ million Toyota owners had nothing to worry about with the latest recall.  The vast majority of Toyota owners have had a wonderful experience with their vehicle, and the vast majority of Toyota owners are unaffected by stuck gas pedals.

But that simply doesn’t matter.  When it comes to public perception, one very negative experience will always outweigh any number of positive experiences.  Ask Toyota.  Ask Microsoft.  (Vista, anyone?) Ask any company that’s endured a PR disaster.

I’m sure Canonical has plans to gradually work out the kinks in their software, but that is exactly the problem: as an example of a “great” desktop OS, Ubuntu is way off track.

Consider this, as an example: what was the biggest, most publicized feature of the recent 10.04 release?

Remember?

The redone Canonical branding.

Seriously.

Yes, other changes got pulled in – but again, how many of those were merely ways to increase Canonical’s revenue?  (Further UbuntuOne integration and the 7digital music store partnership come to mind.)

How about the infamous 9.10 release?  Key non-upstream features included: a new installer slideshow, the Ubuntu One debut, and the Ubuntu Software Center.

See a pattern?

Let me clarify by saying that self-serving features like these aren’t inherently bad – they’re just so much less important than the basic pillars of stability, security, and performance.  Changing the branding is awesome… assuming a distro boots at all.

Sadly, the future of Ubuntu only gets more and more worrisome.  What’s next on the horizon?  How about Windicators – one of the worst UX ideas in computing history.

Ugh.

At any rate, let me pull this article back on-topic.  Two consecutive weak releases have certainly impacted Canonical/Ubuntu’s public perception, but what of the greater Linux ecosystem?

Again, let me invoke the Toyota analogy.  Toyota’s missteps in safety and reliability certainly harmed Toyota – no doubt there – but as I quoted above, they also sent shockwaves through the entire Japanese auto industry.  When the all-star member of a team displays weakness, the entire team suffers.

So it is with desktop Linux.  Love it or hate it, Canonical has been very successful in positioning Ubuntu as THE desktop Linux distro worth trying.  In 2008, when I again decided to try desktop Linux after several years away, the obvious choice of distro was Ubuntu.  Many, many blogs pointed to it as the best place to experience Linux, and since then Canonical and its throngs of loyal fans have only strengthened that message.

Frankly, I think the users of every other distro should be concerned.  When Canonical positions Ubuntu as the best incarnation of desktop Linux, then proceeds to release multiple problematic versions in a row, it reflects poorly on ALL desktop distros.

Take the following example from John, a new desktop Linux user who left the following comment on my aforementioned 10.04 article:

Couldn’t agree more.  I’m trying to move away from Windows so I thought I’d give the “best”? distro a try.

After downloading and installing (10.04) I get to the boot screen to find something about “gconf-sanity-check-2 exited with status 256″.  As a non computer geek (no offense) I am lost already.  Now I have the problem of getting rid of the untidy boot option screen or at least setting Windows as the default.  Maybe I’ll try again in 6 months.  Maybe….

Another perspective comes from that same comment thread, this time from Les:

Let me give you an outsider’s perspective. I’ve always been Windows user, but I love the concept of Linux and just installed Ubuntu 9.10.

My ipod won’t sync. My network won’t connect. I can’t get printers to work. Installing some programs is ridiculously complicated (tar.gz, then…..what?).

I’ve spent hours on forums, and yes, there are fixes, apparently, but none of them have worked for me so far. And I’m the computer guy all my friends come to, so it’s not an aptitude problem.

I’ve already fallen for Ubuntu, and I will persevere. But when I read that Ubuntu spent the last 6 months moving the window controls from right to left and making Ubuntu more social, I wonder if they will ever reach their goal of replacing Windows.

These are great comments from the kind of people Linux desperately needs – individuals brave enough to try an entirely new OS on their own, and individuals willing to persevere despite bad experiences.

What gravely concerns me is that for every John or Les who perseveres, how many others quit?  How many potential developers, designers, and translators has the FOSS ecosystem lost because people mistakenly think Ubuntu = Linux = FOSS?

Now in preparation for the inevitable onslaught of fanboy hate mail this article will garner, let me state that this same conversation could be had for any other distro.  Perhaps Fedora and OpenSuse and Debian have scared off their own share of potential Linux contributors, and that is also sad.

But because Canonical has positioned Ubuntu as THE desktop Linux experience, their responsibility is necessarily higher.  The more publicity Ubuntu garners, the greater its responsibility to deliver an optimal experience.

Before I end, let me state that despite its relevance, this article deliberately ignores the ongoing debate regarding Canonical’s contributions to Linux, Gnome/KDE/other DEs, and other components of the FOSS world.  This is a hugely complicated topic, and one I’m not particularly qualified to comment on.  For those who have never heard of this problem (or those who have but want to get re-angered ;), here are some excellent resources:

• First is a recent blog post from Keith Curtis. To quote Keith:

In my book I wrote that Debian has been terminally damaged by the split [with Ubuntu]. You could even say that Ubuntu is screwing Debian twice: once by stealing people away from what should be their natural home, and again by making them dissatisfied and causing them to leave this combined community. Mark Shuttleworth could hardly have conceived of a better way to kill Debian than what he came up with.

• Greg Kroah Hartman’s 2008 presentation is arguably the most cited source on this topic.
• Mark Shuttleworth’s ongoing attempts at “release cadence” often factor into the debate.
• A good general overview of the Ubuntu/Debian conflict can be found at Ars Technica.
• Finally, consider the irony of Canonical’s design team using – you guessed it – Macs.  (Makes you wonder how much faith they have in their own OS offering… )  Contrast this with Fedora’s approach.

Though I doubt it will help, I strongly suggest reading Is Linux Religion? before sending me hatemail.  ;)

Also note that all comments are moderated, so don’t waste your time insulting me or other commenters without some sort of factual basis for your insults.

bookmark and share this article

related articles

• Dear Ubuntu: I Have Some Concerns
• Linux, Day 1: Picking a distro
• The Only Feature Ubuntu 10.04 Needs
• 5 stars for this Ubuntu/Linux commentor