Ubuntu: 1/3 of Installs, Upgrades Result in Fail?
Inevitably, there will be what we call ’showstopper’ bugs in distribution releases. When this happens, in the vast majority of cases, those responsible for the distro get fixes out immediately. There has been some complaints lately about why Ubuntu should take a closer look at their quick release schedule. Take a look at these articles: “Early adopters bloodied by Ubuntu’s Karmic Koala” and “‘Ubuntu Needs a Longer Release Schedule!’” Whoa, now. Not pickin’ on Ubuntu. Fedora had one in its 12 release in November, as well: “Unpriviledged users could install packages.” Not good. OpenSUSE isn’t immune either. In the release of 10.1, the package manager was horribly broken.
Seems like we could tolerate such ’showstopper’ bugs once in awhile. But to have this many complaints about excessively-problematic releases of one particular distribution release after release seems a little off to me:
- Is Ubuntu broken?
- Tech-no-media blogger asks, ‘Is Ubuntu broken?’
- How to Fix broken Ubuntu Feisty Fawn
- Idea #15033: Don’t release 9.04 with broken major-hardware support
- Hey Ubuntu, Stop Making Linux Look Bad
- Advice Against Upgrading Ubuntu 9.04 to 9.10
Seems like I’m not being objective here, doesn’t it? Maybe I’m trying to get people not to use Ubuntu, or give Ubuntu a pounding. Nope, merely observing numerical data.
What data?
First, let me explain something: An LTS release is generally accepted to mean “a fairly rock-solid release.” The nonLTS releases are interim releases for those who want bleeding edge. By definition, they should have more bugs in them.
Secondly, these numbers might be taken with a grain of salt, as they did come from a support forum. How many people log into a support forum to say that everything worked perfectly? Probably not as many people as had problems.
With that said, let’s look at the numbers:
| Release | type | Upgrades w/many unsolvable problems | Installs w/many unsolvable problems | Flawless upgrades | Flawless installs |
|---|---|---|---|---|---|
| Gutsy Gibbon | nonLTS | 35% | 33% | 22% | 22% |
| Hardy Heron | LTS | 30% | 33% | 21% | 25% |
| Intrepid Ibex | nonLTS | 44% | 45% | 18% | 25% |
| Jaunty Jackalope | nonLTS | 31% | 35% | 29% | 32% |
| Karmic Koala | nonLTS | 33% | 38% | 31% | 34% |
Make an objective judgment of that for yourself. The LTS release had at least a 30% failure rate. Nearly the same failure rate as the nonLTS releases. The flawless upgrades and installs were nearly always lower than the number of seriously problematic installs/upgrades. Well, weren’t they? Where did these numbers come from? this spreadsheet, to which ubuntuforums.org linked.
One or two problematic distribution releases is forgivable. But an apparent failure rate of 1/3 on every release?
Mr. Shuttleworth… bro… ease up a little. Implement some improved testing or quality assurance policies or something. You have an incredible user base. Every time you release, you are seemingly causing measurable aggravation for about a third of your users. Each person that experiences a failed Ubuntu install/upgrade incorrectly attributes the problems to Linux in general. Try and help the Linux community out — not by sheer numbers, but by quality, polished installs and upgrades. Spread the joy. With all your resources, surely you can make some progress here.
While I applaud Canonical’s success in helping Linux become widespread, they also seem to be giving Linux a black eye amongst would-be Linux users.
And please, if you are one of the third that has a marvelous experience with Ubuntu, don’t shoot the messenger. I’m merely sharing data.
December 4th, 2009 at 5:23 am
How on earth can you even come up with those numbers? It is impossible. You must have made that chart up out of thin air. Nearly all installations go smoothly so we never hear about them; it is only the ones that go wrong that we hear about. I had a look at Windows 7’s forum: it makes win 7 look every bit as bad as ubuntuforums makes ubuntu look bad. Of course, the forums are where problems are reported. However, since with ubuntu, tens and hundreds of computers can have ubuntu put on them from the same, individual disk, it is obvious that no statistics can be made at all. So your numbers are obviously fiction.
-Yos
December 4th, 2009 at 5:57 am
You’re “merely sharing” vastly skewed data, from a support forum, which is frequented mostly by people who need support.
Doing a survey on release quality on a support forum is like doing a questionnaire on luggage loss rates in front of the lost luggage claim counter.
December 4th, 2009 at 6:05 am
So you’re basing your conclusions of ubuntu’s instability, by data collected in a SUPPORT forum. Do you honestly believe that regular and/or new users would visit the “community cafe” subforum where these polls were posted? And do you believe that users that had no problems would search for a way to say “yeah it worked flawlessly for me”? A vast majority of people who visit ubuntuforums do so to ask/find help with problems. If they didn’t have problems they would not visit.
If you’re willing to do a statistical analysis about something make sure that the population examined isn’t biased. I’m afraid your results are based on false assumptions.
Ubuntu wouldn’t be as successful if it wasn’t good.
December 4th, 2009 at 6:45 am
I am curious where on ubuntuforums.org you arrived at this data. If the data you are talking about is from this, http://ubuntuforums.org/showthread.php?t=1305924. Then you clearly missed the disclaimer at the beginning of the thread post. More people go to these threads when they have issues than do when they have flawless experiences. So these numbers are skewed, and not representative of the whole user base of ubuntu. Either way, if you want to post “facts” you need to make sure they are accurate and representative of the user base before you make large accusations.
December 4th, 2009 at 9:15 am
I would point out that in absolute terms, those numbers probably don’t apply to the entire installed base of Ubuntu users. They were gathered from ubuntuforums.org which is a site for helping people with Ubuntu problems, so you would expect people who participate in a poll there to have a selection bias toward, well, people with problems. So I strongly doubt that 33% of of *all* Ubuntu problems have failed.
That said, if the numbers reflect a similar sample then relative to each other that might say something. What I notice is that if you remove Intrepid as an outlier, the percentage of failed installs looks pretty stable while the percentage of flawless installs has actually improved over time.
December 4th, 2009 at 9:26 am
Thanks for linking to your data source, however, that spreadsheet only lists http://www.ubuntuforums.org as the source of the data and is not clear as to where on that site the data resides. That’s kind of like saying, “It’s on the internet.” So basically, this post is based completely on hearsay.
I won’t dispute the fact that a certain number of users have had difficulties installing or upgrading Ubuntu. I’ve had problems myself at times. However, I have no faith whatsoever in the accuracy of these statistics. As we all know, 92.3% of all statistics are made up.
December 4th, 2009 at 11:08 am
Ok, so… everyone is attacking my usage of statistics that I got from a help forum. The point seems to be, “Of course you are going to get a skewed result, because people who are not having any trouble won’t go there.” 100% valid point. Please understand, however, that I didn’t make them up, they came from the Ubuntu forums.
Also, no one really answered the question as to why there are so many articles out there discussing the issue. People appear to me to want to poke holes in logic and defend their religion. That’s ok, I get that way, too. But is there a problem or isn’t there? Based on the fact that I am not the only one making this observation, the only logical conclusion is that, if you read these other articles to which I have linked, there is a problem. And bad experience with *ANY* distribution of Linux gives would-be users a bad taste. Which gives Linux a bad name. Just clean up the releases. Ubuntu’s issues won’t get as much press as they have. If I were the only person saying anything, that would be one thing. But I’m not. That’s why I put the links to the other stories first. I actually happened to stumble upon the Ubuntu forum with the link to the spreadsheet, and included that after the fact because the numbers looked interesting. That’s it.
I hear many people love Ubuntu. Keep loving it. Just be careful when you upgrade, especially if your hardware is brand new.
December 4th, 2009 at 9:02 pm
My comment was mearly trying to state that if you are going to use these numbers you really should include the disclaimer from the thread that you are quoting, otherwise the data you are presenting is misleading.
However, to deal with the issue you are looking for an answer to. Of course nothing is going to be perfect. There are many reasons people would post threads to help forums or post to their blogs or tech sites about issues they have had or heard of from ubuntu, and those can be for many reasons. They can be for the same reasons you would see someone post something negative about windows or mac osx as well. Maybe the writer had a particularly frustrating experience, which of course happens. Maybe the writer has an unwarranted distaste for the operating system.
If you really want answers why people write negative posts to internet sources they would be the best reference.
Now, I love ubuntu, don’t get me wrong. I have had very few negative experiences, and have worked to fix the ones I do have. I also work to help my family fix their issues if I am able to. I would personally agree that there is an issue with upgrades. Upgrades are the easiest method for a new user to migrate from one version to the next one. They also seem to be the most flawed way of upgrading. After running into issues upgrading a while ago, I stopped doing it. Simple separate /home partition and fresh install gave a far better result and it was equally easy.
I have however heard that there are plenty of people who have good luck with upgrades. So of course, results may very. And that is the case with most software. I mean, one person might get never ending blue screens from windows, but then flawless results from ubuntu. And vice versa. The point I guess, is everyone should use what works best for them. Of course, everyone who loves ubuntu or linux as a whole would love to see more users use it and have the market for linux software become larger and entice more developers to the OS, but the reality is that right now that’s not the case. Hopefully we will get there in time, and of course when someone attacks something you are trying to build up a lot of people run to defend it.
Either way, I just request that you add the disclaimer. But, it’s your post, you can do whatever you want.
December 5th, 2009 at 3:32 am
I personally can’t understand all the bad press with Karmic, it bewilders me to be honest. I have done 5 installs and upgrades at home and they all work just fine. On one I lost a couple of desktop icons and I just replaced them back and no problems. CUPS works better, new laptop working much better than Jaunty, all machines booting faster and most importantly five very stable machines.
Wish I could say the same about the two Windows 7 installs here, one just decides to go on strike and take an eternity to load or do things and in both cases I had to hand edit a config file to get a network printer working on 7.
Personally I think your stats are way of the mark as generally only people that have issues got to the forums for help and ones that have either no probs and simple ones don’t. Getting reliable statistics on these figures is nigh on impossible. It’s just like trying to work out how many people use GNU/Linux on the desktop, it just can’t be worked out 100% right.
Thats my two cents worth !!!!
December 5th, 2009 at 2:26 pm
There seems to be a consensus that Ubuntu LTS releases should have fewer bugs than interim releases. But Canonical does not make this commitment. And why should they? Aren’t they going to make their best effort for every release? Canonical makes only two commitments for LTS releases. First, they will be supported for 36 months instead of 18. Second, users of the previous LTS release can upgrade directly to the next, bypassing the interim releases. Also, they have been known to make “point” releases when a sufficient number of bugs are fixed. I recall an 8.04.1 release. I recommend not being an early adopter of any release. Burn the live CD image to a USB stick, run from there, and don’t install until your sure that there are no show-stoppers for you. This applies to other distros too.
December 6th, 2009 at 4:55 am
@Doug
Fair enough, I’ll put up that disclaimer from the forums. You are totally right. Thanks for the clarification.
@Richard
Just thought it was interesting that there seems to be a disproportionate number of negative articles for Ubuntu releases than for other distros. I have already acknowledged that the stats could very well be skewed, though I did not come up with them. You’re right, it can’t be 100% right, but I’ve not heard as much bad news about each successive release of Debian, for example. The distro from which Ubuntu is derived. But yeah, just goes to show that different people have different results. Thanks for the thoughts.
@Don
Again, just seems like even with the LTS releases, there appears to be a disproportionate number of people complaining about problems. Great points, though….. especially about not installing/upgrading until you’re sure it isn’t going to break your systems.
OK, so the numbers are quite likely skewed. I didn’t compile them. Especially being based on Debian, one of the most stable distros out there, it just struck me as ironic that so many complaints about bugs in Ubuntu releases have been made. I have just not seen that type of consistency in complaints with successive releases of other distributions. Thanks for all the comments.