Dear Mark Shuttleworth: please tell the truth

Damn, this early video of Gnome Shell from youtube seem so much more usable than current gnome shell. Normal tray and less insane paddings.

I kinda preferred the 'big wide Dash with apps and documents in it' style for quite a long time - I ran it for several months - though I'm pretty okay with the 'only application icons' style now, maybe I just got used to it.

Fedora has also been doing 6-month releases since before Ubuntu even existed. And Ubuntu's mission statement at that time looked almost plagiarized from Fedora's. (I haven't checked how they have evolved since.)

Kevin: I checked Fedora's history, but it's a less clear six month cycle - I think it was always targeted for six months except one cycle where it was targeted for nine, but there are rather a lot of cases where a release was significantly early or late, so it doesn't line up as nicely as Mandriva's. I wanted the references to be as clear and unarguable as possible, so I decided just to go with Mandriva's, to avoid argument.

Please reconsider what you are doing with GS and make something else for GNOME 4. This is not troll post i am serious. GNOME 2 was great but GNOME 3 + GS is useless and pleas admit this and move on.

GS: I'm not doing anything with GNOME Shell, besides testing it (as part of Fedora - I also test KDE, Xfce and LXDE). I don't work on it, and I'm not a member of the GNOME project. Having said that, I think GNOME Shell is great, I've used it since before GNOME 3.0 came out, and I think the benefits of the major changes are becoming clear in releases like 3.6 and 3.8. That is entirely a personal opinion, though. In any case, this post is not about whether GNOME Shell is any good. It's a simple question of factual accuracy: Mark is saying things that simply are not true. It bugs me as a historian (my degree is in history) more than a 'software guy', actually. The history of our community is still pretty disorganized and ephemeral, and when a big fish like Mark says something and keeps saying it, there is the danger that it becomes the 'accepted truth' despite not, in point of fact, being correct. His comment about Unity 'existing before' Shell got 31 +es, and I have since seen it repeated as gospel by other people. Despite the fact that it is absolutely, clearly, and incontrovertibly not the case.

"Having said that, I think GNOME Shell is great, I’ve used it since before GNOME 3.0 came out, and I think the benefits of the major changes are becoming clear in releases like 3.6 and 3.8. That is entirely a personal opinion, though." I loved GNOME 2 and i am not saying GNOME 3 should be the same as GNOME 2 but GNOME 3 with GS failed. No sane Microsoft, Apple or Android user will migrate to it and Linux mobile/desktop surely can't compete with GNOME + GS in the state it's now (3.8 included). "In any case, this post is not about whether GNOME Shell is any good. It’s a simple question of factual accuracy: Mark is saying things that simply are not true. It bugs me as a historian (my degree is in history) more than a ‘software guy’, actually. The history of our community is still pretty disorganized and ephemeral, and when a big fish like Mark says something and keeps saying it, there is the danger that it becomes the ‘accepted truth’ despite not, in point of fact, being correct. His comment about Unity ‘existing before’ Shell got 31 +es, and I have since seen it repeated as gospel by other people. Despite the fact that it is absolutely, clearly, and incontrovertibly not the case." Fair i have nothing against the truth/facts and if this are the facts i can accept them. P.S. From my point of view GS/Unity/Cinnamon are products of migration to GNOME 3 where a lot of distribution makers and users disagreed what GNOME did and didn't want to use GS and i don't know exactly when somebody start planing what will be GNOME 3 shell like i just know default one was not accepted and we needed something else...

@GS, I feel like Unity at the very least was more of Canonical's not invented here syndrome kicking in. Same with Mir, Bazaar, Launchpad, Quickly, and almost every other alternative that Canonical has come up with. They seem to never contribute to upstream, and always come up with competing inferior products. At the very least, Gnome Shell seems to be much better engineered than Unity. @AdamW, thanks for this post. It's about time someone step up against Mark Shuttleworth's ludicrous claims...

I am not advocate of anybody but for the reference: Ubuntu Netbook Remix was introduced in Hardy Heron release and was production ready 24 April 2008: http://en.wikipedia.org/wiki/Hardy_Heron#Ubuntu_8.04_LTS_.28Hardy_Heron.29 It looked like this: http://www.youtube.com/watch?v=6vcGvRs89KI We could argue if Unity is direct evolution from Ubuntu Netbook Remix or not and if GS does take elements from it. Looking at the video we could find some similarity with GS? But i think this is normal when it comes to software development and surely some concepts from Ubuntu Netbook Remix existed before 2008? About 6 moth release cycle i am not qualified to judge this but Ubuntu does/did until now stick with it? I don't know if this makes any difference but i think it's important deadlines are respected. "I feel like Unity at the very least was more of Canonical’s not invented here syndrome kicking in. Same with Mir, Bazaar, Launchpad, Quickly, and almost every other alternative that Canonical has come up with. They seem to never contribute to upstream, and always come up with competing inferior products. At the very least, Gnome Shell seems to be much better engineered than Unity." I do find some truth in this for sure but users don't necessarily put "better engineered" in front of "i like to use". I am not saying who is right or wrong i just wanted to add some comments/info as i see it and i wish you all well.

GS: we're discussing that in the Phoronix thread ATM. So far as I know, Mark has never suggested that Unity is somehow the same project as UNR. His first post about Unity - http://www.markshuttleworth.com/archives/383 - is very insistent about presenting it as something New, Exciting, and Different, and in fact talks about how it isn't yet ready to be *adopted* on the netbook edition - i.e. they actually wrote Unity separately from the development of UNR, and had to think about how to make it usable to replace the UNR's interface. We're not really talking about who inspired who here; that's a whole other can of worms that, as you say, is a really icky topic in computing. We're just talking about when two specific projects - Unity and Shell - existed. I don't think it's viable to try and argue that Unity and UNR are the same thing, ergo Unity existed in 2008. Mark doesn't seem to have thought of it that way. On the release cycle: check the link to the Mandriva page on distrowatch. MDK/MDV stuck very solidly to the six month cycle. Those aren't planned dates, those are the actual release dates.

@GS, I'm inclined to say that Unity is more of a natural evolution of Docky than UNR, though perhaps in some ways, Unity is a merging in concepts of the two.

@GS I'd go further with the historic analysis. I would argue that maemo 4 as implemented on the Nokia N810 in 2007 was the historic "innovation" predecessor for what we are seeing now in modern UI. The N810 was way ahead of its time as a "web tablet" device (not a phone!) and was touch oriented, well as touch oriented as anything could be at that moment in time given the available technology. And sadly, how Nokia approached creating the Hildon interface toolkit as used by maemo as an out of tree addon to GNome 2.x also speaks the same sort of historical lesson as Canonical's at arm's length approach to collaboration. Sadly Canonical didn't really innovate in "approaching collaboration wrong" either. Nokia was way ahead of them on that score, blazing a trail for Canonical to follow. Mark likes to retcon history a bit, its just a character quirk he has that we all have to live with. Its not appropriate to call it a flaw, it would make him make a great Marvel comicbook writer. But in the context of the ongoing discussion about Canonical as a corporate entity's decisioning making, his tendency to retcon is problematic. The best scholarly history of the unity/gnome situation was produced by Jeff Waugh, a former Canonical employee: http://bethesignal.org/blog/2011/03/12/relationship-between-canonical-gnome/ There are multiple blog posts in his series. They are all worth a read to get perspective. But really none of this matters. This historical navel gazing is fun, but ultimately a distraction for Mark and for Canonical. Like Adam here, it pisses me the hell off to see Mark engaging in it, but probably for vastly different reasons. Mark is squandering his leadership position by wasting time trash talking to absolutely no benefit to Canonical. What really matters is Canonical has until 14.04 to pull a fully working cross-device convergent platform out of their ass. They've wowed people at CES and at MWC with what is essentially a mockup of their deliverable built on top of Android plumbing. A mockup that doesn't do anything useful if you actually attempt to use it for more than about 30 seconds. They have set expectations extremely high to win the praise on their demos at those shows. They've screwed themselves and they don't even realize it. There is absolutely no way, based on their available talent pool, that they will be able to get this together by 14.04. No way. Even if they get half-way there, which would be quite impressive, they've shot themselves in the head by bringing up expectations so high with their demos. Half way there, instead of looking like a partial win, like progress, is going to look like failure when compared to the story they fed to the laypress at CES and MWC in the last few months. Seriously this will be the 4 unique implementation of the Unity UI. Instead of stabilizing and polishing, they are re-implementing a FOURTH shell in a FOURTH toolkit. 3 implementations built and scraped. Holy crap. I mean really build one and throw it away I get. But build 3 and throw them away? How many different implementations of Mir will they build and discard? How quickly are they going to hit limitations in the qt5/qml design and will be compelled to downstream patch-in features to get their Unity Next out the door and into the hands of OEM partners? You think the blood between GNOME/Canonical was bad? Just wait when they start screwing around with downstream patches for Qt and see how the KDE/Qt community reacts.

Jef: we appear to be locked in some kind of horrible death spiral, as Jeff's post is the very first link in *my* post :)

The six-month cadence (let's call it a 'release cycle') wasn't an innovation. Ubuntu was explicitly chosen to release a month or two after GNOME stable releases, which were on a six-month cycle. It wasn't a happy accident, and six months wasn't some magical commandment Mark carved on to a tablet after a blinding flash of inspiration. It was directly, 100%, following GNOME. And came as much from jdub as anyone else. (I was there when it was decided.)

daniels: I think Mark's implication is that they innovated in terms of having a distribution follow a six month release schedule with a predictable relationship to a desktop release. At least, that's the most generous way to interpret it. So I interpreted it that way, to be fair. Still doesn't work.

"We’re just talking about when two specific projects – Unity and Shell – existed. I don’t think it’s viable to try and argue that Unity and UNR are the same thing, ergo Unity existed in 2008. Mark doesn’t seem to have thought of it that way." OK but if we won't look back on how things evolved and try to put that into the context (video above) then this are the facts if Wikipedia info is true: GnomeShell http://en.wikipedia.org/wiki/GNOME_Shell Initial release April 6, 2011 Unity http://en.wikipedia.org/wiki/Unity_%28user_interface%29 Initial release June 9, 2010 P.S. I am not registered on Phoronix and to be honest i already spent too much time on this and i will leave further debate to others. But based on info i gathered both started thinking about it as early as 2008 (maybe before that year) and end users got Unity a bit sooner compared to GS.

GS: it's not about 'thinking about it', it's about two specific codebases, Unity and Shell. Unity is not the same codebase as UNR, it is completely different. You can trace Unity's existence with great certainty to late 2009. You can trace Shell's with great certainty to late 2008. Things are only getting more complicated because I'm entertaining this silly argument that Unity and UNR are somehow the same thing, when clearly they aren't. Mark didn't say 'we were thinking about netbooks before GNOME was', he didn't say "Unity had a stable release before Shell did", he said "Unity existed before Shell did". It's a very clear and simple statement that is clearly and simply not true. Really, this thread amazes me - it just teaches me that some people are willing to argue that 'up' is 'down'...

@adamw, up versus down... being intimately coordinate system specific, which are arbitrary constructions, leads to the unfortunate situation where 2 coordinate systems could disagree on the definition of 'up' but still be self-consistent. Making a convincing argument for 'up' being 'down' ultimately futile. But what I can do is argue quite convincingly that North is South , geomagnetically speaking. -jef"I re-added Jeff's blog link in the hope that a second reference to it would inspire GS to take the 20 minutes to read it"spaleta

jef: ...or that black is white, and get yourself killed at the next zebra crossing? :)

I tried to leave one more comment but it failed multiple times. Cheers.

Hum, I don't see any errors in my logs or anything in the wordpress admin interface. That's odd. Sorry, I'm baffled :/

OK one more try: "I’d go further with the historic analysis." Yes that is exactly what i said we could and probably both GS and Unity borrowed from each other and from the history and we could place that in context to bend it as we would like. "I’m inclined to say that Unity is more of a natural evolution of Docky than UNR, though perhaps in some ways, Unity is a merging in concepts of the two." I don't know looking at this image: http://i50.tinypic.com/z4b5u.png I do see launcher on the left, dash like functionality on top of the desktop and panel with indicators on the top. I would be confident enough to say Unity Current does have roots here (natural evolution). "GS: it’s not about ‘thinking about it’, it’s about two specific codebases, Unity and Shell. Unity is not the same codebase as UNR, it is completely different. You can trace Unity’s existence with great certainty to late 2009. You can trace Shell’s with great certainty to late 2008." Unity gets rewritten on yearly bases. ;) Unity Next could have new name again because it will use totally different technology to build it and it won't use Compiz anymore for example and the display server looks like it won't be X.org anymore but design wise we could argue looking at above image the roots are there? The code could switch to other repository and it would be dated to current date... Probably the name had to change too for practical reason it would not make sense to call it Ubuntu Netbook Remix/Edition anymore? "Mark didn’t say ‘we were thinking about netbooks before GNOME was’, he didn’t say “Unity had a stable release before Shell did”, he said “Unity existed before Shell did”. It’s a very clear and simple statement that is clearly and simply not true." It depends on how to interpret input data. If image of UNR does not convince you Unity has it's roots in it then yes it's hard to argue otherwise. If video above of UNR does not convince you that GS has similar layout for presenting apps... End users according to Wikipedia did get Unity first if we don't count UNR... But if only the date when GS and Unity official repository for hosting the code are important and nothing else then you are probably right one dates the other. But i don't know if that proves MS was wrong in what he said i will leave that for others to decide for themselves and for historians to determine the facts! Cheers.

If a thing 'has roots in' another thing, it doesn't mean that Thing A is the same as Thing B. If Mark had said something vaguer I wouldn't have had a problem. But he clearly identifies Unity as a specific Thing (and in the video you keep talking about, he does the same) which is a different Thing from UNR (he talks about "starting" the Unity design process *after* UNR happened), and then says very unambiguously that the Thing called Unity "existed" before the Thing called Shell. Not that it "had roots", or that he was thinking about it, or that it was based on older stuff, or anything like that. All of those are much messier concepts that this debate has brought in. It seems bending too far to suggest that's what Mark was talking about when he said "Unity existed before Shell". Not to mention that, as Jef says, he has considerable 'form' in retconning stuff. Like the *other* part of this blog post, the release cycle thing, which no-one has challenged.

I worked now great! Probably url to the image caused troubles and i uploaded the image to TinyPic and now it works! Anyway nice talk and to be honest it was fun and thanks for that but i had enough. ;) I hope GS and Unity will do good in the future... and that is it. Cheers!

"If a thing ‘has roots in’ another thing, it doesn’t mean that Thing A is the same as Thing B." Yes but as i said it would probably not make much sense to call it Ubuntu Netbook Remix/Edition anymore after it was decided it will become default shell on desktop/mobile? ;) Cheers!

Sure, but Unity wasn't presented as 'this is the new name for UNR'. It was presented as 'this is our new desktop'. Edit: and remember, like I said, I went to Launchpad and looked at the code. Unity does not re-use the UNR code. It is a new codebase. You can go check yourself if you like.

Yes code wise i do imagine once Qt/QML port is finished it will stay like that (code and technology wise) for a while.

AdamW is probably factually correct on both counts. My love affair with Ubuntu is long over, especially since I have just found out that I was naively accepting those non-functioning (but hella pretty!) Ubuntu for smartphone mockups as something that was ready to come out "any day now." That's a far cry from the truth. Shuttleworth is too concerned about how to make the news, and not nearly enough with running a upstream-contributing FOSS project with best practices in mind. Canonical/Ubuntu is increasingly a fountainhead of "our way, our brand (TM)", Mac envy, FUD and hype. Blah. GNOME 3 Shell, on the other hand, is something I use, support, and love, ever since it released. And if you really find the timely move to a modernized look, feel, and workflow so offensive, you have like half a dozen feasible alternatives at your disposal, including to install a particular combination of extensions on top of GNOME 3 to get back to the traditional experience. I'm sure a lot of old die-hards cried foul over the switch from text consoles to GUIs in the early 90s.

andrew: oh, sure, people cry foul over everything. People hated GNOME 2.0. I don't recall how, but your post got me thinking about how I was a big Gimmie fan back in the day. Google reminds me that it eventually got a fork, Mayanna, which ultimately merged into Zeitgeist, and so lives on in a tiny way in current GNOME...

Ain't going to happen Adam, as we have seen from his dealings with the Ubuntu community, he is essentially dishonest. For as long as he sees prevarication as useful to attaining his goals, that is what he will do.

There was also Ubuntu Light which was exactly like Unity without a name. It was demoed in may 2010, but there is no info when did the development start. So it might as well be true.

jimmy: Ubuntu Light wasn't 'Unity without a name', it was the first project built on Unity. The two were announced at the same time in the same post, discussed above - http://www.markshuttleworth.com/archives/383 . As you say, that was in early 2010: in the video that has also been discussed, which also dates to 2010, Mark says Unity has been in development "for nine months", which basically ties in with the dates on the bzr repo (October 2009, several months before the public announcement). I am giving them credit for the earliest date I can possibly find - there's just zero reference to "Unity" before the first commit in the bzr repo.

GS, the major point you're missing here is that GNOME shell's initial development (before it's initial release was 100% open and in the public and Unity's was not. I remember spending a couple of afternoons at the GNOME Boston Summit UX hackfest in October 2008 at Novell's Cambridge office (and Mark Shuttleworth himself was present, BTW) working on the initial shell design, which was pretty much the entire point of the hackfest. The first shell design document was written in July 2009 http://people.gnome.org/~mccann/shell/design/ GNOME shell was public 2 years earlier than Unity. Sorry, dude.

Ugh, it is exactly these types of battles that keep Linux form making inroads on the desktop. This whole "not invented here" crap that keeps getting slapped at Ubuntu/Canonical is really BS, and actually swings both ways. Someone above mentioned bzr as an example. It was developed in parallel with git and mercurial, all because of the flap with bitkeeper. Launchpad was designed not just as a hosting facility with bug tracking, but also as a complete integration system for maintaining, building, and publishing Ubuntu. Yes, it was closed for a long time, but other open source projects have been closed before going open (like java, solaris, etc). So what? As to the release cycle thing, I have used all of the major distros. Ubuntu at least sticks with their release schedules. Who else can claim that? I currently use several at work (CentOS, Fedora, Ubuntu Debian, OpenSuse). Of all of them, Ubuntu 'just works' 99% of the time (12.04 failed to install once on new hardware until I updated the kernel). Fedora 17 recently broke several of our build tools because of their constant moving target updates (kernel updates from 3.3, 3.6, 3.7 all in this release).

"GNOME shell was public 2 years earlier than Unity. Sorry, dude." End user got Unity before GS according to Wikipedia and UNR is predecessor of Unity if you like it or not. Look at the image i posted above if you don't believe me and in the end it reading the comments it just looks like somebody don't like MS and that is about it. And after sleeping it over if GS took 5 years to become what is it now i only hope GNOME 4 repository is already available somewhere because GNOME 3 + GS is dead end i only hope it won't take additional 5 years just to admit that and to do something about it and Unity starting with UNR did more in this 5 years it's actually usable ATM.

To me it matters not which was first, but which shell is more usable. From my use of both, I preferred Gnome Shell slightly more than Unity (55 to 45). However, I was and am frustrated by the constant resizing of windows and I switched to Cinnamon, based on Gnome's GTK. I have been a Fedora user since RedHat introduced Core, which is before Fedora 6 was released. Since the current release is 18, I have had the pleasure of having installed 12 releases, delivered about six months apart. On my system I setup three logons. One with Gnome, one with Cinnamon and the third with KDE. This way every few days I switch from one to the other to evaluate any patches or enhancements. Here is an off topic comment . "7 inch or 10 inch tablets are heavy items to lug around. Many do not have usb or memory card slots. A virtual keyboard is not the same as a true keyboard, neither is a Bluetooth keyboard that forms part of the tablet case. The Netbook with SSD and with a handle to make it easy to hold will replace the tablet. When the tablet includes the smartphone, then and only then will the desktop computer have competition. In the meantime, enjoy the fad."

PSA: Do you have a blog? Is it not hosted on static pages served through a CDN? You should take a few minutes to install Varnish and... hell, you probably don't even need to configure it - it comes with pretty sensible (and safe!) defaults. It's also an incredibly well-designed piece of software that has turned Reddit frontpage events from an incredibly busy day for me to completely non-notable. Sincerely, A Sysadmin

xiong: I have a PHP cache already (php-apc).

For Mandriva Linux, we're now on a "it's done when it's done"-cycle.. ;)

Oh look, they're at it again: http://www.pcauthority.com.au/Feature/345085,has-ubuntu-bitten-off-more-than-it-can-chew.aspx/2 “When we started Canonical and Ubuntu in 2004, and we said we’re going to release Ubuntu every six months, the reaction was ‘you guys are crazy, nobody can put out a full OS every six months – that’s just madness’,” says Silber. “Now, everybody does it. Fedora does it every six months, SUSE does it every six months; we showed how it could be done. Now six months seems a really long time.” I'm amazed people can lie that bare-facedly, quite honestly.

upstart - junk mir - stupid juju - stinks openstack from canonical repos - brokenstack, broken beyond belief Ubuntu LTS - 12.04 has had THREE kernels, 3.2, 3.5, 3.8. And to top it off, not bringing in 3.10, the actually decent one of the bunch. Mark is a dolt, a destroyer, a fragmenter