Slashdot Mirror
Ubuntu May Move To Rolling Releases
formfeed writes "The register claims that 'Ubuntu is moving away from its established six-month-cycle and potentially to a future where software updates land on a daily basis.'
While this sounds like a sudden change, it is apparently more of a long-term thought. The Register quotes Shuttleworth:
'"Today we have a six-month release cycle," Shuttleworth said. "In an internet-oriented world, we need to be able to release something every day. That's an area we will put a lot of work into in the next five years. The small steps we are putting in to the Software Center today, they will go further and faster than people might have envisioned in the past."' But given that many of Shuttleworth's thoughts became decisions later on, it might be interesting to see, where this one leads. Interestingly enough, five years is about the time when Ubuntu will run out of letters."
"changes land on a daily basis".
Excuse me, but they already do. What the heck is update manager, but a means for updates to land when needed?
www.eFax.com are spammers
I have used Arch Linux for quite some time,and I like it when I have the latest software without having to update to a newer release,or wait for a new one to be available
Lately, there has been a gradual shift in Linux hardware support where distros are limiting support for older hardware. I understand why they are doing it, but by doing what Ubuntu is [thinking about] doing, it could literally result in a situation where one day your computer is supported and the next day, it's not. That's not a good thing.
You go from one release cycle style to another. Periodic releases to constant releases. And then back.
Each style has its advantages, but in the end you just end up changing for change's sake and no real benefit is gleaned one way or the other. It's a lot like reorganizing resources in a company. You move some people here, you transfer some people there, you change from a horizontal hierarchy to a more vertical one. Then in 18 months you change it back.
In the end, the guys on the ground doing all the nitty gritty work do the same job they've always done and the company keeps chugging along.
That being said, it's usually a case of management losing touch with the guys on the ground that causes this kind of shakeup. I wouldn't be surprised if Shuttleworth is a bit disappointed in how the business is going and is looking to change the sales story for Ubuntu. From the "stable and great" OS it is now to "cutting edge and always up to date" OS it could be with constant drops.
I don't really like this news.
People like new things. I think you can get more people to change by telling them "The latest Ubuntu is out!" - which has some large changes which people can get wowed by and appriciate.
That said, with the update manager always updating at least SOMETHING every day, its sorta like this.
That said, I wish to add a question - What exactly is the difference between say Ubuntu 10.4 and 10.10 ? What exactly is a 'latest upgrade' ? When they change the way things look and a few policies (such as the default media players?)
While i understand that you want the foundation to be fairly stable that in itself creates a slew of problems. Foremost that stuff like Firefox, OpenOffice and other userend apps wont get upgraded to newer versions until the next rollover.
My personal dream would be a distribution where the user end is getting upgraded often and fast while stuff under the hood gets overhauled less often.
A suggestion would be major overhauls once every two years of the backend stuff while user applications is kept on newest stable versions. That way developers of backend stuff gets ample time to iron bugs out while users wont have to upgrade the whole desktop just to get a new version of an app.
HTTP/1.1 400
I know it sounds like a good idea but the Ubuntu releases also introduce some major feature or some behind the scene platform change. What new headache will each rolling release introduce to the user every day?
These comments are my own and do not necessarily reflect the views or opinions of my employer or colleagues...
Just add debian testing to your sources.list
Or if you are really curious how it feel when your userland changes on a daily basis, add debian unstable...
The biggest problem I've had with Arch is that changes are only tested on a few common packages before being sent out. If you use an obscure package routinely, it might not be tested decently with any given change, especially to libraries.
Once a change breaks something, you're left trying to install multiple versions, locking versions, modifying the source, or other such deep magic. Very quickly, the whole system gets to be too big a hassle to deal with.
You do not have a moral or legal right to do absolutely anything you want.
This was Debian's solution, still in development. No long cycles and it's great for the people who love the bleeding edge.
That'll keep your system fresh.
What is this fascination with change for change sake? What could possibly be so important that it has to come out each and every day? Then to use a silly statement like "in an internet-oriented world" to justify what sounds more like the capricious nature of an marketing executive then solid rational for updating software...meh!
As a recent (4 years) user of Linux I find the current release process more reliable then wondering what will happen the next day. Like another poster, I've had a "stable" release break what worked, causing me to wipe and rebuild with a previous version. These days I wait a few months for kinks to work out of a release before installing on any of my systems. Don't use the internet as an excuse to garner attention and push out sloppy programs just to say "we got there first". Promote and produce solid, dependable, and balanced releases that make the Linux (ubuntu) experience a positive one; you'll have a stronger following.
Life is a great ride, the vehicle doesn't matter
Ubuntu could release new and simple updates/upgrades regularly, and every few years, release a new version of ubuntu. Who would want to install then update when the most recent live disk is from 10 years ago? Ubuntu could get rid of incremental releases, and just release the LTS versions, then push updates and give users a choice "security updates only", "security updates and minor bug fixes" or "security, functionality, and major bug fixes - may be less stable".
five years is about the time when Ubuntu will run out of letters.
Looking forward to ZenBuddhist Zebra!
Smivs on the intertubes!
It's going to be a big enough job for Ubuntu to keep up with whether the latest daily change works with everything. Anyone using using packages not included in the Ubuntu distribution (for example: Boxee) is likely to wake up one morning, and discover their program doesn't work. Some new library will have replaced the old library that the application required. This will probably mean I change distributions, if they do this.
It also ignores the vast unwashed throng, of which I happen to be a member, who have limitations on their internet access, either due to speed, or daily quotas. It's not unusual now to find their recommended update exceeds my daily quota.
Please don't tell me I should upgrade to DSL. The only thing remotely resembling high speed internet here is satellite. It's that, or dialup!
This is one of the biggest features I miss about Gentoo. I like running the latest and greatest but stablity is also nice. Hopefully this would mean things like same day Firefox stable releases in Ubuntu and others. Not sure how it would handle big things like GNOME, X, or glibc but if they can do it well that would be amazing.
Linux is already much of a moving target when it comes to application development and getting some kind of a consistent environment, now it will be increasing harder (at least on Ubuntu). I can envision vendors spending more time updating their build environments than actually implementing their products.
I don't like sigs... I don't use it...
We have to run on "internet time", it's all about the schedule, we can't let the schedule slip. We have to release something, anything on schedule. Doesn't matter if it's crap, as long as it compiles, ship it! We have to have something new for the schedule, meet the schedule! Can't slip behind schedule!
Looks like Ubuntu/Canonical wants to become another Microsoft.
If you're talking about needing to "be able to release something every day" and you're talking about Ubuntu then the first few days are simple to sort, and I'm sure you could continue a pattern that would keep 90% of Ubuntu users happy:
Day 1: Lighten purple in default background ...
Day 2: Darken orange in default background
Day 3: Darken purple (but not enough to be back to original shade of purple)
Day 4: Put orange back to what it was
Day 5: Make all window buttons red instead of red and grey
Day 6: Put all window buttons back on the right side of the window
Day 7: Add a clock widget that uses bold
Day 8: Bundle a load of random pictures, slap Ubuntu logos on them and call them "The Ubuntu Desktop Pack" (see Gnome-Look.org for examples)
Day 9: Change the cursor theme so that it turns into an Ubuntu logo when hovering over a title bar - that's a feature, right?
Profit may be in there somewhere as well.
So, which one - if any - would be a LTS release ? I am really bothered about it. I am so far away from bleeding edge, that I want to change from one LTS release to another alone; let alone six-monthly ! And how about software developed for a particular release ? "Tested on Ubuntu release Nov 24 2010, 11 PM GMT+5" ?
I know lot of applications that run only on older versions of Ubuntu, so rolling releasess would break support for these kind of applications, because nobody is developing those applications anymore, but some people are using those. I actually use one program that runs only on Ubuntu 9.04 and I use it via VirtualBox from 10.04.
I admit that Ubuntu release cycle is too fast, because it makes lot of work for testing team. If you develop program for Ubuntu you need to test (and develop) your program several versions of Ubuntu. I think, one release in a year would be enough.
Rolling releases would be good for standard user, but in corporate environment it would total hell.
The last letter of the Latin alphabet is "Z", so "zen" sounds like a really good name for the final release.
They won't run out of letters, they're using Base64.
Since *nix tends to be case-sensitive, they can re-use the first 26 names without collisions, and it will still be in version comparison order. Then I expect to see "0-day 0liphant" and so forth. By the time we get to the plus, minus, and equals, Canonical will have sponsored the naming of 3 newly discovered species such that they can finish the cycle. At 2 per year, that gets them to (04 + 32) = 2036. That's enough time for John Titor to come back from the future to fix the 2038 bug once and for all, along with the Ubuntu naming conventions hopefully.
In other words, Don't panic.
Installing small daily updates is a breeze; I don't even think of it anymore because of automatic download and install options.
Upgrading to a new Version of the OS via many hundred megabyte download/CD/DVD is a pain in the ass.
Considering that it is possible to push out small incremental updates via the daily update system this seems like a good idea to me.
Also, you can already choose not to apply any of the daily software updates (esp. ones that might not be compatible with your hardware).
I look forward to the days where I'll never have to install an upgrade again...
it seems only yesterday that Mark Shuttleworth was demanding that the entire linux ecosystem move to a 6 month release cycle in order to match ubuntu
SURELY NOT!!!!!
This is what I have been waiting for. After my initial excitement about (k)ubuntu release updates to get all the hardware running and supported, I am now at a state where everything is fine. The ongoing new 6-month releases are more of an annoyance than a great feature. Having to upgrade completely every 6 months just to get access to the latest software releases does not seem like a worth while effort. Sure, you can say that's what the LTS releases are for. But while the LTS releases do enjoy long-term support for security-relevant updates, they do not get a lot of software updates. So if you favorite application gets a major update after the LTS release you are out of luck. (Of course, you can fiddle something together on your own, but that's not really a low-maintenance solution.) Also, from my experience the LTS do not have less bugs than regular releases.
So if they can make rolling releases working with a high level of quality and testing, this would be really awesome from my point of view.
Maintainability: ... Server Edition would probably move to publishing LTS exclusively).
It's hard to maintain a system that has billions of different iterations — "billions" might be an understatement; your typical system has one to two thousand packages installed. If half of them get a nontrivial update every month (which is a pretty conservative estimate), that's still 125 million (500^3) combinations each quarter. The easiest solution to this isn't a very user-friendly one; make internal 'milestone' markers and force upgrades that would push beyond them to incrementally hit each one before coming up to speed (this also has the pleasant side-effect of leaving a system stable at a more updated place than where it was rather than chunking through an upgrade that fails part-way through). Ubuntu could move to this model immediately (it's already there; each milestone would merely be an unannounced LTS (long-term support) release
The big hurdle is how to determine that an update failed; just because the software looks intact and post-install scripts succeed doesn't mean that hardware support wasn't retracted or that things don't play nice. Some smoke-tests are needed, compatibility must be itemized, and a big decision must be made about what to do with people who can't (or won't) upgrade past a certain point. I think the milestones are a good solution there, but that's assuming the LTS model will be continued.
Debian Testing:
Unless Debian is frozen for an upcoming stable release, Debian Testing is a rolling release. I often wish that Debian would fork the Testing repo with that freeze rather than once it's fully released as stable (which would create a new name, perhaps "beta"), though I understand the limited resources for both developers and testers (those of us using Testing). Given the relationship between the two projects, I wonder what Ubuntu Rolling v. Debian Testing will do, and how they'll differentiate themselves...
Use my userscript to add story images to Slashdot. There's no going back.
When will they stop pretending that apps are part of an operating system? Ubuntu confuses an OS with a kitchen-sink app. Desktop backgrounds, specific customized version of a specific browser, instant messaging, graphics editor -- none of these belong in an OS. None of these should be part of (depend on) an OS upgrade. An OS should provide a stable basis for all versions of all programs to run. If they could achieve that, there would be no need for frequent updates, except for the adventurous.
I've had my server running on gentoo since 2003 and never once had to put a CD/DVD in my drive to upgrade. Heck, I even started out initially on x86 and then went to ~x86 for some time and then back x86( lots of cruft to clean up but I managed ).
You just need to keep up with instructions given during emerge runs, running revdep-rebuild and perl cleaner also when instructed.
Objectively, I wouldn't consider myself in Ubuntu's main target group, but I really, really like it (despite stupid decisions all the time, and horrible, disastrous bugs such as https://bugs.launchpad.net/ubuntu/+source/mountall/+bug/616287).
I've been using Debian Sid for years, on all of my servers that aren't under contract.
I've also been using it on my desktops since *forever*
I've been using it on my laptops since about 2004.
I've had a total of a couple of hours of limited functionality, between my laptops and desktops. X barfs or my primary Desktop (Gnome or KDE) gets horked for a bit. I move to XFCE or andother Window Manager.
Servers have been rock solid except for a short time when the whole udev/hald/something changed its rules on how my NICs were named... no longer were they eth0/1/2 but eth4/5/6... Hurrah. Debian Sid is going mainstream.
greg, REMEMBER ED CURRY!!!
You know what is really starting to get on my nerves?
Updates and how.
Really, I am a grown-up. I do not expect perfection. I do not hold update counts against developers, in fact, the opposite, shouting huzzah whenever problems get fixed.
But here's what has happened in the past few months. There's something I want to do, and I fire up the software, and I'm greeted with an update available modal box which has to be addressed, probably with a "later" because at that moment I hadn't planned to be sys-admin dude.
With regards to Kubuntu, which I have installed in many places, I have had the occasional bit of bad luck in which a kernel updates seems to wack something for a few days until other packages catch up with their updates.
I contrast that with the iThings where updates are indicated quietly, download and install is neatly done, and I'm real sure that the update won't gum up the other apps or the gui.
I'm all for updates as long as they don't interfere with why I'm running a system, which is to do something that gets me paid or enjoyment, not administration and definitely not whack-a-mole dependency roulette.
It's not the frequency, it's the friction.
If ubuntu wishes to keep a presence in data-centers, they cannot move to a rolling release. Servers depend on a stable configuration. If a security update came along, most likely you would have to update many packages just to get the security fix and then be possibly be left with many broken things.I would however be supportive of a hybrid release with an LTS branch that only get updated for security and bug fixes.
Dear Mark,
Recently I tried to upgrade my Ubuntu system from 8.04 to 10.04 (LTS to LTS) by using the bundled distribution upgrade manager. The first upgrade, to 8.10, rendered my graphics card and video card useless. Since the 8.10 version was no longer supported, I had to jump through a bunch of hoops to get good drivers installed so the system was usable again. 8.10 to 9.04 went smoothly. When I upgraded from 9.04 to 9.10, upon system reboot, I was greeted with a message that one of my I/O modules had a memory conflict error at addresses 0x400 - 0x407 with some other module.After excessive googling and internet scouring, I found that this error had happened to a few other folks, who wiped their entire system and just used a live CD with the version they wanted to get their computer running. I also found three bug reports that had been filed as something along the lines of, "Put on the backburner because this affects an old distribution."
If updating my OS every 6 months requires a weekend long endeavor full of stress, strain, tears, and agonizingly obscure frustration, what in the hell makes you think asking me to upgrade my distro daily is a good idea? I don't have that kind of spare time. My computer is a tool, not a fucking hobby.
If Canonical moves forward with this idea, and implements in in such a way as to shave even more years off of my life through high blood pressure, then I will be forced to return to some other computing environment permanently. I don't have time to deal with that kind of crap daily.
Sincerely,
Me and probably a few thousand other Ubuntu users.
Motorcycles, Robots, Space Gossip and More!
Debian CUT == Constantly Usable Testing.
A recently started project in Debian with a similar goal of a rolling release (along with an idea of installable snapshots).
http://cut.debian.net/
http://lwn.net/Articles/406301/
"It is better to die on one's feet than to live on one's knees." - Albert Camus
The problem is that if you wait six months between upgrades then that means you spend 12 hours downloading and installing hundreds of megabytes of changes and then it crashes part-way through and your system is hosed.
It sounds an awful lot like you're installing new versions as in-place "upgrades". I've never had that work successfully, starting from RH 6.something or so around 1999. Your much better bet is to download the ISO, then install the new version in a fresh partition. Mount all your data like normal (you do have your data on a separate partition, no?), then give the new version a spin. If it hoses something, you've still got your old version on its own partition, and switching back is as easy as rebooting.
Keeping things in separate partitions and mounting as appropriate is one of the key advantages (for me, anyway) of Unix-style filesystems. An example partition list:
Create and use more or fewer OS partitions as you find useful. I have Windows XP on one (not used on the bare metal since shortly after buying the computer), Ubuntu 9.10 in the next (thinking about wiping this and replacing with 10.10), 10.04 in the third, and I keep the fourth around to play -- check out Fedora, Arch, Mandriva (when they were still viable), etc. In each OS, I just mount my data partition as appropriate -- generally just as /data, and then symlinked from the appropriate /home/[username]/data locations. (You could just keep all /home/[username] directories in your data partition, but I tend to find that this causes config file conflicts, so I just keep the equivalent of "My Documents" in the data partition.)
This way, "upgrading" is as simple as a full install in a fresh partition. This completely avoids the problem you (and I and many others) have run into: wasting time downloading and installing hundreds of megabytes of changes and then it crashes part-way through and your system is hosed. Install after a clean wipe -- avoid that "not quite fresh" feeling!
Cheers,
"What in the name of Fats Waller is that?"
"A four-foot prune."
So Ubuntu is turning into Debian unstable? Cool!
Here's a better idea - go for more stability, not less. If Linux is maturing as a desktop OS then there shouldn't be a need for 6 monthly, let alone daily, updates.
Here's a better idea:
1. Drop the 6 month release cycle and make LTS the default option. Then people can install an OS with a sensible lifetime.
2. Don't push any updates unless they are critical security vulnerabilities.
3. Offer optional upgrades to the major application packages, drivers etc. as they become available and where possible, and keep interdependencies to a minimum - i.e. compile them against the original distro + any vital security patches, not the latest everything (statically link them if you have to - RAM is cheap now).
The problem with the current system comes for the less technical users who want to (or are sensibly advised to) stick with the packages in the official repositories. Currently, you may find that the only "official" way to get the latest office software is to upgrade your whole fricking operating system. Its like having to take the back axle off your car in order to replace the radio.
Remember this is Linux - if we /.ers want to compile our own kernel, install the latest Firefox beta from a source tarball, reformat the drive as ext6 or scour the interwebs for a suitable .deb of the very latest LibreOffice then there's nothing stopping us. Or, we can switch to a more bleeding edge distro. However, that might work for us, but it won't work for others - and even I don't want to install a new kernel just to run the latest word processor unless it really, really needs it.
The problem is particularly bad with Ubuntu: it can't be "the Linux for the rest of us" and bleeding edge, because "the rest of us" don't want to be obliged to upgrade our whole OS every 6 months just to get the latest OpenOffice.
...its understandable with commercial software where the company depends on brining in the upgrade fees, but why should Free Software care?
In a survey of 100 programmers, 111111 thought that duck-typing was a good idea.
First thing I'd do is look for the off button when installing a release with this feature. Twice I've had an in place upgrade hose my Ubuntu system - and usually it results in quirky bugs if it doesn't entirely blow up. That's enough of that nonsense - every couple years I do a full fresh install and copy over all the important files from my old install.
It seems like Ubuntu is going the way of Firefox, Pidgin, and other open source software - making unilateral changes the users haven't necessarily requested or possibly downright don't want. Pidgin auto-resizing text boxes and Firefox magic navigation bar are easily on par with moving my window managers close/min/max button positions.
Lesson for open source: people are often happy with how something already is - put an option in settings to reset it to the old default when you make major cosmetic changes to your software. I wouldn't be using XYZ software if it wasn't already working for me. Thanks
Release early, release often... In my opinion its worked great for the Kernel folks since v2.6 was released. Does anyone remember the hell that was upgrading from 2.2 to 2.4, and again to 2.6?
The more things that change at once, the greater the pain will be, its as simple as that. Holding back all changes and releasing them all at once with a major version upgrade causes the most pain as possible, and people are reluctant to actually upgrade, so testing is limited.
Instead if they release small changes more often, they will get more testing as more people are willing to risk an upgrade if only a small number of changes occurred, and if something does break, its limited in scope. The key here is that you try not to upgrade too many important parts of the system at once. For instance Xorg should probably never be upgraded at the same time as KDE/Gnome if possible.
For example, if Apache releases a new major version, you can send that out and if something breaks, its pretty easy to roll-back or fix the issue, since only Apache was changed and maybe a couple other minor things.
Instead if you upgrade to a new version of the entire distro, if Apache breaks you don't know if its directly related to Apache, or one of the other 1000 packages on the system. It makes troubleshooting, bug tracking and quality control much more difficult.
Open Source Time and Attendance, Job Costing a
Stability and a predictable environment are more important to me and my work than having the latest bells and whistles.
Just because Windows is constantly slipstreaming updates doesn't mean Linux needs to do it too. If something is really really really important, by all means tell me about it, but let me make the decision whether to upgrade or not.
Upgrades for the sake of upgrades are not the answer. My main development box is Slackware 10.2, albeit with upgrades to many development-related packages (kernel, gcc, python, etc.). It works. It has to work, and it has to stay working.
...laura, Slackware fangirl
Yeah, this is a great "feature" to make me go to Debian or some other distro.
LTS is here to stay I hope, or I won't be using it.
Versioning is a feature, not a problem. You can't have your cake and eat it too. Introducing new features every day will just kill your stability, no matter what the marketing department says.
Also, versions give a definite reason to use Ubuntu. If it goes away, you'll never know what you'll get, and it is actually less marketable.
http://www.debunkingskeptics.com/
That's good news.
Often enough back porting of fixes result in broken software.
What must be improved next is that test suite for distribution bundled software must be mandatory _and_ repositories with QA based rating.
I really hated it when I installed a package in Debian the last time just to find out that xml-twig-tools is alpha quality -- which you can only read in the man page, but not in the package description -- and randomly omits values when working with big (xGiB) XML files.
Overall I like Debian and Ubuntu very much, but the number of packages is highly overrated because not few packages are just plain crap -- some of them might have worked a decade ago but seem be unmaintained multiple distributions in a row (not counting on maintainer that only care for their CV) for ages.
Get a test suite and drop that crap!
Right now they're using a system where the two words in the name always start with the same letter -- Feisty Fawn, Jumping Jackalope, etc. Once they run out of letters to use in this manner, they only have to use mismatched names. For example, the first version after Zaphod's Zipper could be April Bees. In this manner they should have enough letters to last several lifetimes at the least, even if they skip over combinations that don't produce anything you'd want to slap on a product.
How is the Riemann zeta function like Trump rallies? Both have an endless number of trivial zeros.
In 5 years this may be a quite reasonable approach. It might not. It all depends.
Actually, it depends on a lot of things. Debian testing is usually quite good, but I generally figure that at least once during every development cycle it will hose my system. It's a rolling release, though possibly a little bit less tested than he's thinking about.
FWIW, I've used Ubuntu, and it's not bad, but now I'm using Debian testing again. (There was a period a couple of months ago when it broke my system. So I figured that was a good time to test Ubuntu.)
Off-topic Addendum: Debian doesn't seem to work as well on portables as Ubuntu. They don't include all the necessary software on DVD-1 (wireless didn't work), and they don't seem to manage the battery as well. Didn't dig into the reason.
I think we've pushed this "anyone can grow up to be president" thing too far.
I think Rolling Releases will begin sometime after Quizzical Quail is released.
If this fixes the problem referred to in the subject line, it'll be worth it.
In Real Life [tm] the gap between LTS releases means an upgrade from one LTS to another rarely works. You are transporting configuration for a couple dozen things, but then you are eliminating a couple dozen other things in favor of newer or more functional packages, so now you have a few dozen things that have been precisely tuned trying to work without the things they were tuned for. Hilarity ensues, not.
If you make an ISO and install the new LTS clean, but preserving the /home partition (you did make a separate /home partition, right?) it works fine because you blow out nearly everything in system space, and these problems are relatively manageable in userland (where configs are likely stored in subfolders of /home/username). But if you try to upgrade rather than install you'll probably regret it.
Speaking as a person who supports a lot of end-users of Ubuntu LTS here.
Continuous Incompatability, aka Zymorgous Zebra
Great. Now shit will break every week as opposed to only when I decide to update (once a year or so).
Firefox and Chromium are the only applications on my Ubuntu box at work that get new versions instead of just backported security fixes.
"OH MY GOD, NOW THAT WE LIVE IN AN INTERNET WORLD..."
Seriously, come up with a better justification.
I understand that remote exploits and zero-day vulnerabilities emerge suddenly and without warning, but the idea of contantly yanking the carpet out from under your users is going to absolutely RUIN your following.
Automatic Updates, "ZOMG, IS IT GENUINE???" *Advantage* crap, and "PLEASE REPORT YOUR USAGE STATISTICS" spyware, are the reasons I fled Microsoft, and their invasive OS modifications which can suck up HOURS sometimes.
HURR DURR YOU NEED DOT NET AND SP3 NOW, HOPE YOU WEREN'T DOING ANYTHING IMPORTANT.
I mean really.
I already have to wait a week or so for their servers to die down after a release before I can install anything, much less upgrade or download the new release myself. Now, they will be slammed every single day.
I use PCLinuxOS primarily because I like that they use a rolling release model. Prior to that I was using Mandriva, which I actually think worked a little better. But I got tired of waiting for the latest Firefox, etc. Truth be told, recent Mandriva's were better about updating Firefox and other components, but you still had to do a full upgrade periodically. That should be okay too, except that I never got (or trusted) an upgrade to work, and usually did a fresh install to a new partition to check it out. And setting up all the 'non-free' stuff in a new install is a big pain - especially on a commercial distro like Mandriva that doesn't include stuff you want and sticks you with stuff like Fluendo that you don't.
Anyway, all that said, do rolling releases make it harder to support software that does not come from the distro's repositories? It's the old backward compatibility issue - if new libraries are likely to break apps, you've got to include new versions of those apps too. This would be less of an issue if they excluded libraries that break backward compatibility, but that would be limiting in its own way.
Posted from my Android phone. Oh, I can change this? There, that's better...
The problem is particularly bad with Ubuntu: it can't be "the Linux for the rest of us" and bleeding edge, because "the rest of us" don't want to be obliged to upgrade our whole OS every 6 months just to get the latest OpenOffice.
I think you're right in your hypothesis about non-nerds not wanting to upgrade the OS to update applications. Heck, ISTR upgrades not even being supported; you have to reinstall your OS every 6 months on the Ubuntu bandwagon.
Here's a better idea - go for more stability, not less. If Linux is maturing as a desktop OS then there shouldn't be a need for 6 monthly, let alone daily, updates. Here's a better idea: [plan]
Here I start to disagree. I think your goals are absolutely wonderful---stability for end-users is a great thing.
However, it seems to me one ought to ask oneself "how does software become stable?" in order to bring that about. It also seems to be that the answer must include---though this is not a complete story---testing on a lot of different hardware, and in a lot of different software interactions.
The best way I can think of doing this is by letting all the VERY_LARGE_VALUE_OF_N use the bleeding-edge less-tested versions if they want to, and somehow record for the packages how thoroughly tested the software is.
I mean... in order to fix software, someone has to first discover that it's broken.
In the Window world, Nontechs using windows don't care what version of MS Office they have because they're extremely interoperable. The difference between a company owning Word 2003 and Word 2007 is
Bad example - Word 2007 has a totally different user interface to 2003 and, by default, saves in a totally different file format.
Even a non-techie would be smart enough to realise that it doesn't make sense to have 2003 running on your desktop and 2007 on your laptop. Many relative non techies (not the total lusers) are just about smart enough to install the appropriate version of Word.
Now imagine that the only way to upgrade was to install Windows 7, and the only way to downgrade was to downgrade to XP... except your desktop doesn't have win7 drivers and your laptop doesnt have XP drivers...
In a survey of 100 programmers, 111111 thought that duck-typing was a good idea.
I think you got that wrong. An Ubuntu upgrade starts only after all (or maybe nearly all) the necessary packages are downloaded. As long as you're still downloading packages (which should be obvious from the progress dialog box), you're safe (from a nasty upgrade). I've managed to accidentally abort a couple of distro upgrades because my net connection died only to resume it later (even after an unhealthy reboot of the supposedly "upgrading" system).
As an Arch user, I have always wondered why there weren't any easy to use rolling release distros. For businesses that means never having to schedule down time for upgrading or reinstalling while still having the latest software, and I know it will mean less stress for many consumers also.
microsoft has almost daily updates my pc needs to download. is that what ubuntu is striving towards - making it a pain in the ass to set up a computer and just let it run?
i suggest monthly rollups of all the patches and improvements. because i don't want to dick with testing updates every day of the goddamned week
Remember kids, if you're not paying for the service, YOU ARE THE PRODUCT THAT IS BEING SOLD.
Ubuntu is not moving towards a rolling release.
http://theravingrick.blogspot.com/2010/11/ubuntu-is-not-moving-to-rolling-release.html
I will steer clear of Mint also!
Rolling updates would make a bad environment worse from the perspective of 3rd party application developers. Assuming you'll get all your apps from the OS vendor hasn't worked for Linux distros, and rolling upgrades would make the non-platform seem even less inviting.
The problem with the current system comes for the less technical users...
You left out application developers who tend to view OSes lacking feature stability as hostile environments.
Have the Linux desktop people suddenly forgot that apps are sorely needed? Have they forgotten the PC mantra that the apps sell the system?
"Desktop" is not a subset of "Server"... I wish geeks would stop assuming such a relationship.
... to see how many people think this is going to ruin Ubuntu. I would really like to see this kind of release schedule at least for the applications I have installed (such as Empathy, Skype, Firefox, Chrome, etc.) because I honestly believe that PPA's are the handy-work of the devil. Of course, I don't think they should make such updates transparent, but rather on an opt-in per application basis. I remember back when Pidgin was still in, changes to the protocols would often make it nearly useless (for me) and that's the kind of thing I think they want to stop. It is kind of ridiculous for a "normal" user to have to do some googling and command-line magic to get the latest version of an application that doesn't even demand new dependencies or anything like that to run. What would be great is if it offers to update the program when you actually go to open it, but will offer the option to suppress that particular update so that you don't get spammed by the offer every time you open the application. That's a pretty standard method for offering updates, right? I don't think that the Shuttle is talking about hijacking your DM via some ninja update and borking your system, it sounds more like he wants the users who say "Why the heck can't it just DO it?" to not have a reason to ask that question so often when using Ubuntu. This is the 21st freaking century and we still can't update to the newest version of our chosen browser without some serious user-fu? I certainly hope not.
A suggestion would be major overhauls once every two years of the backend stuff while user applications is kept on newest stable versions.
This is exactly what you'll get with Debian Stable + Backports. A stable base with regularly updated versions of popular software packages (eg Firefox, OpenOffice) to put on top of it.
And running this type of system now has more official support.
Last in-place upgrade I tried was 8.10 to 9.04 (done around July 2009), and that didn't go too well for me. Admittedly it's been a while, but the pain involved was enough to prompt me to go back to fresh-partition-installs. If things are working better now, I'm happy to hear it.
Cheers,
"What in the name of Fats Waller is that?"
"A four-foot prune."
I last tried an in-place upgrade around July 2009 for Ubuntu 8.10 to 9.04, and had all kinds of "fun".
Generally speaking, though, I've found that at least some systemwide config options generally change from version to version, so redoing systemwide config provides me a good opportunity to familiarize myself with the new options and make sure everything is just as I want it. Sometimes newer versions offer options that I'd like to use one way or the other, so redoing the whole configuration process ensures that I'm familiar with what's new and different.
But maybe that's just me. :)
Cheers,
"What in the name of Fats Waller is that?"
"A four-foot prune."
Hope it came in handy.
Cheers,
"What in the name of Fats Waller is that?"
"A four-foot prune."
...instead of spending half of every day installing Yet Another Slew of Updates.
On the other hand, just because they release them doesn't mean we have to install them on their schedule...
Better idea - add a rollback option to the bootloader - init callsapt directly - and fixes what is fucked up - Windows System Restore style.
Cool thought, but that requires either that the user have wizard-level OS configuration abilities, or that the group in charge of the distro change what they're doing. While both are possible over the longer term, over the shorter term, planning out partition use is an easier solution for more intermediate / not-quite-beginner users. :)
Cheers,