Submitted by Mark Shuttleworth today urged development of Linux models to rival what Apple has done on the desktop and mobile devices. Certainly on the desktop experience, we need to shoot beyond the Mac, but I think it’s equally relevant [in] the mobile space, Shuttleworth said, outlining the challenge as figuring out how to deliver a ‘crisp and clean’ experience, without sacrificing the community process. Key to this will be services-based mechanisms for creating revenue for free software that go beyond advertising, Shuttleworth said, adding that cadence in free software releases spurs innovation, and that a regular release schedule, as well as meaningful ties to Windows, will be essential to fulfilling the vision.
It’s nice to have goals but Apple isn’t standing still and I just don’t see it happening. Why not try set the goal to surpass Vista first? Or has it already happened?
Vista? Vista what?
I thought we were speaking Oses here, pre alpha are far behind now.
One thing that I strongly dissagree with this author about is that some vague thing called “services” is the best way to make money with Free Software. “Services” is a good model for the large companies like IBM, Novel, Mandriva etc. However it is not a good model for the thousands of one to ten person developer teams currently producing software onder the proprietary “shareware” model but may like to go to the Free Software model while still deriving some profit from their software.
For “shareware” people who would like to make Free Software out of their projects yet still profit from them I would propose a packaged software model similar to the “rackware” approach that put many of the “shareware” based games on the map in the late 1980s and early 1990s. Basically this would consist of building statically linked apps (these work under ALL Linux and Windows versions and distrubutions) complete with source code under the GPL or some other free software license, putting them on a cd or DVD and placing them on racks in stores willing to cooperate with this distribution model again or sell them on Ebay or similar services for about $5.00 to $10.00 a cd.
I believe such a free software “service” as the one I’m proposing here would be very profitable for small developers and developer teams who wish to free their software for whatever reason (the reason could be political or simply a wish to avoid the legal problems tied to defending copyrights and patents on a proprietary app in court) but still profit from it.
“I believe such a free software “service” as the one I’m proposing here would be very profitable for small developers and developer teams who wish to free their software for whatever reason (the reason could be political or simply a wish to avoid the legal problems tied to defending copyrights and patents on a proprietary app in court) but still profit from it.”
Personally, I think it would put the small software developers out of business. Shareware worked/works as it lets people try something, but then to get full functionality/game levels/etc you have to pay for it. Under the model you are proposing, the developers would be under. It would cost more to make the DVD’s or whatever and out on store shelves then they would take in. Unfortunately the majority want free as in beer, and can care less about scruples or free as in freedom when it comes to software. Why spend $5-$10 when, if the source is released and the program is GPL, I can legally download it elsewhere at no cost at all? The anti-piracy schemes we see today are the result of people doing just that, not the fault of the software developers that want to make money. It is all about the money my friend, nothing else matters to most folks. The way I understand your post, there would be hundreds of thousands spent for packaging and such, yet little to no income from the product. I would like to believe otherwise, however I do not from looking at normal human nature.
it’s about funding not profits. Part of the reason Microsoft can bludgeon everybody else is that they get pre-sales from OEMs for those 180 MILLION* copies at $50 a pop that’s money to do whatever they want, whenever and answer to nobody. Ubuntu needs to target OEMs to get them to see that for say $5 per unit they could use real Ubuntu and let Canonical take care of support and updates rather than each OEM copying it (badly, like gOS or eeePC) In a way it’s “service”, but it needs to be to OEMs that will value it and can pay, the only work needed from the OEM will be to send along patches for the latest hardware (much easier to NDA Canonical than random dev)… and maybe something to hook their proprietary bits for multimedia that can’t be shipped without licensing.
Ubuntu wants to be a “hub” for development rather than run the whole show. That’s what makes them different from Apple or Microsoft. They do need a way to point the projects in a unified direction rather than just “keeping up”. They’d take part of that money and put it up as bounty to get developers to fix things (get paid for boring, bu needed, stuff!) but the stuff would still be other people’s and everybody would benefit. Again, they don’t want to be Red Hat or Suse putting all the “famous” people on the payroll so they don’t quit projects, and many devs simply don’t want to work for the “man” anyway, so they want to find a third way of doing things.
The reason I believe it would work is this. Despite all this “Broad Band” and fast internet stuff we keep hearing hype about a substantial part of the internet is still on DIAL UP and have you EVER tried to download a major app on DIAL UP!!! Not only is such activity notoriously slow but very long dial up downloads tend to crash as well. And this can also even happen on faster connections like Satalite, DSL and cable modem when there is a large “free beer” demand for the product. Furthermore the “free beer” internet download products tend to be dynamically linked with all the “dependency hell” and “upgrade hell” consequences that accompany such linking. The CD free software products I propose would be STATICALY linked to their dependencies effectivley eliminating these two “hells”. The would be literally installable by simply copying the app to the hard disk just like most modern MAcOS apps. I think people would pay to get out of download frustration related to dial up internet connection crashes, high demand “slashdotting” crashes and frustrations related to dependency hell and upgrade hell. In fact I BUY all the Linux distros I use precicely because of these problems.
The Mac OS isn’t popular because it has a “pretty” screen.. It is because the whole underlying user interface is easier.
The Linux user interface isn’t going to be changing anytime soon as the average Linux user/developer lives in the command line.
Not sure why someone would mod you down – I can see disagreeing with you but it’s not like your post was a troll or abusive.
I’m curious how true that is. I’d be curious to know how many Linux users:
(1) use the command line for most of their file management
(2) use the command line for most of their package management
(3) use the command line to execute programs
(4) use a lot of custom scripts
vs.
How many people use Ubuntu basically like Windows most of the time. A friend of mine just installed Ubuntu and years ago I got used to doing things at the command line so I can’t even answer some of his questions without fiddling around with my mouse to see how certain things work.
For informational purposes, I’ve set up maybe a dozen people with linux and:
1. Nobody uses the command line for file management. Everybody’s used to the folder paradigm and understands it whether it starts at C: or /. Besides which, there are great GUI tools for that kind of thing (konqueror springs to mind; whenever I need to help somebody with something I just ssh with konqueror. Much easier than ssh on the command line and cp).
2. Package management has been difficult to make people do at all. This is probably what I get the most calls about (“I want a program to do X, can you do that?”) I show everybody adept but only a couple people use it. I just get a lot of calls the first couple weeks and then they have everything they need for 99% of what they do.
3. I install Kubuntu for most people and encourage people to use katapult (an amazing launcher that comes with kubuntu by default). Half of people pick that up and love it and the other half stick with the regular menu; it’s there, they’re familiar with it, it’s a no brainer.
4. Again, nobody who comes to me for help has ever needed any scripts.
What ends up happening as far as the command line is that I use it when they need help because I find it more convienent than the GUI apps (especially package management. Why open adept, search for a program, hit install, hit apply when that’s 1 line on a terminal?) but everything that they need has a gui.
Quag7s point is that if the developers use the system differently to how the users use it, the user experience is unlikely to be as smooth as it should be simply because the developers don’t understand how things could be improved.
If you spend your whole day on the command line you won’t really know how that one little file manager bug you’ve classed as “Minor” is really an irritating annoyance that ruins everything, for example.
Good point!
I mostly use the commandline when I’m moving a lot of files. Come to think of it, I haven’t even checked yet, if any graphical filemanager supports copying/moving based on strings with wildcards…
That’s one thing I couldn’t do without if I were to leave the commandline.
That’s one of the features that I love about OpenTracker (the BeOS / Haiku filemanager): pressing ctrl-opt-a will bring up a “select” dialogue. You can then type a text string (including wildcards), press enter, and Tracker will select all files in the current folder matching that string.
maybe because those users have discovered that the command line is, for a lot of tasks, both more powerful and faster than any GUI could ever be?
Both have their places. Most Mac users i know use the command line for a lot of things too.
“The Mac OS isn’t popular because it has a “pretty” screen.. It is because the whole underlying user interface is easier.
The Linux user interface isn’t going to be changing anytime soon as the average Linux user/developer lives in the command line.”
Actually Mac OSX is popular as it has a pretty screen. I find it’s usability to be negligible. Certainly no more usable than any other interface. The single menu bar sucks rocks, and makes it almost unusable IMO. I want my active application to be up top, not what OS X thinks I should have.
As for the “average linux user…”, most use the GUI these days. Please stop smoking crack before posting a comment. There is rarely a reason to drop to the command line. I drop to the command line more in OS X than I do in Linux to do basic things like ping, traceroute, etc. Unless there is a GUI tool I am not aware of in OS X that does that..which is more than possible. Please do not name one of the GPL tools either..as those are available on Linux without an extra install routine or a purchase.
/Application/Utilities/Network Utility.app
I’ve used Linux (dozens of distributions) for close to a decade and Windows for even more than that. Yet, last year I purchased my first Apple product, a Macbook. It wasn’t for the “pretty screen.”
I was curious and wanted to try a fresh alternative. I wanted to see an open source system (Darwin) with a unified interface and direction. Moreover, I wanted to get a glimpse into what Linux might be like if it had reasonable commercial support from Microsoft, Adobe and the other big companies as well as no hardware incompatibilities. Given that the price of a Macbook was identical to the Windows based models, from Lenovo, Dell, Toshiba etc., I really had nothing to lose. After all, I could always put Linux or Windows on it if OS X turned out to be a disappointing mistake.
While I still use Linux from time to time, my OS X experiment has been interesting and successful. Trying to get a hang of Applescript, Dashcode and Cocoa was fun and really forced me to learn new ways of doing things. Moreover, the interface is a pleasure to use once the learning curve is over. Having one menu bar for all programs gives a consistent target and saves screen real estate. Additionally, the Dock is a bit gimmicky but actually has some advantages over the taskbar. When you have ten or more programs open, it is much easier to distinguish the large icons. In a taskbar, the icons would be too small and the text almost unreadable. Moreover, it is useful in drag and drop operations. For example, I can drag a file to some program on the Dock and it will open it. Of course, it’s not perfect. Apple has made some really silly decisions like having the cut feature only work with text. Moreover, Leopard’s Dock has false perspective and the reflections are really just over the top distractions. That’s why I disabled them and switched to a cleaner 2D look. Just write the following in the terminal:
defaults write com.apple.dock no-glass -boolean YES
killall Dock
And here we come to one of OS X strengths and weaknesses. The customization options are more limited than in KDE, XFCE, or GNOME, if you want them through a GUI. On the other hand, this means consistent and well-thought out defaults which have been tested thoroughly. It’s a balance. Obviously I could write much more about this, but I don’t want to make a record for longest post. I hope my account was insightful.
most popular place for mac is usa
i know many university buy it…
but often mac lab is empty… example nebraska university…
Every 8 months or so, I try Linux. The last couple of times, one of the distros I’ve chosen has been Ubuntu. From what I can remember, during the past couple of days, I’ve had to use the command line for:
1 – adding software repositories
2 – installing sound card drivers
3 – installing nvidia card drivers
4 – compiling kernel (for sound card drivers)
5 – installing software (virtualbox, rar support, etc.)
6 – uninstalling sound card drivers (get Error 2 if done from the GUI)
I’m sure there were more things, but that’s all I can remember at the moment.
While some things could have been accomplished from the gui, the HowTo’s I found did things from the command line, so that’s how I did them. Some HAD to be done via the command line though.
I actually had to install it twice. The first time, sound worked for a couple of sessions, but stopped and wouldn’t work again. As I could find no way to make it work again, I had to start over.
You might argue that installing things is what required my command line usage, not using things, and you “typically” only install things once. However, I can install and uninstall anything in Windows without using the command line.
Argue it all you want, but to this day, Linux still requires use of the command line MUCH more than Windows does.
“Argue it all you want, but to this day, Linux still requires use of the command line MUCH more than Windows does.”
Out of the items you mentioned, all are done from the GUI with the exception of compiling a kernel, which you would not do on Windows. Though to be honest I don’t use *buntu at all, as I prefer openSuse. It could be in *buntu one needs to drop to a command line to do the things you mention. Again we were talking about the “average user”, which is not who comes to OS news. The average users are the ones that the majority of us IT folks support, not each other. I do have a couple of my users using *buntu and they have yet to require the command line for anything. Yes, I asked them if they needed to use it and the answer was no. Of course, they would not be compiling a kernel either for that matter.
This is hilarious and never gets old, no matter how many times I see it.
Somebody says, “I had to resort to using the command-line and a little bit of voodoo to get video/sound/whatever working in Linux.” To that somebody usually responds with, “Well, you were obviously using the wrong distro.” LMAO!!! Classic.
This little point underscores one of the fundemental reasons why the Linux user experience will never be as smooth as that found on OSX. It’s kind of hard to fix a problem when nobody wants to admit that a problem actually exists.
Edited 2008-07-24 18:17 UTC
“”Well, you were obviously using the wrong distro.” LMAO!!! Classic.”
Obvious misquote there. No one said anything or implied anyone was using a wrong distro. There are problems with all operating systems..OS X, Linux, Windows especially. All I did was say what I can do on the distro I use, not that they were using a *wrong* distro. I am pretty sure that can be done on *buntu as well, just needs the learning curve. Your user experience with OS X may be great, and mine is not too bad in actuality. Currently I have had some 3 software engineers that I support decide to go with OS X. The funny thing is that after 3 months they all wanted to go back to Windows and Linux, as OS X was not as capable for what they were doing so they claimed, and had nothing but issues. Is all in the learning curve, so YMMV.
Edited 2008-07-24 18:31 UTC
In Ubuntu you don’t need the command line to do any of those except maybe compiling a kernel. Adding repos in Ubuntu is dead easy to do from the GUI. The same with installing apps and 3rd party drivers. In-fact installing 3rd party drivers like nvidia drivers is much easier on Ubuntu than on most other distros. There is very little that needs to be done through the CLI. However most walkthroughs use the command line because its far less verbose and easier to tell a user to paste a command in the terminal, rather than actually explaining how to do it through the GUI. People looking for help mistake this and think that its the only way to do things, especially newbs. That is not the case and you hardly need the commandline at all in most distros.
I HAD to use the command line to uninstall the creative drivers. I booted into recovery mode, and that wouldn’t allow it either (still got Error 2). The only way to get them to uninstall was to kill the gui (gdm or something).
When I first installed ubuntu, the first thing I had to do was figure out how to get it to configure my monitor, which was giving me a headache due to the very low refresh rate because the system wasn’t smart enough to configure it automatically.
I still don’t know how to get the mouse buttons to work on my intellimouse.
The command line is fun. It’s like the good old days of MS-DOS : )
say…for example…blackbox? ‘any’ should be used with some caution there.
well, here’s the interesting bit: usability isn’t really about what you think, it’s about what the majority (in a group of users with several levels of experience) feels conformable with.
I’m still waiting to see the focus groups results of *any* linux window environment on a wiki somewhere. if not a pdf. anything really! are there focus groups? what are we letting bleed through as “usability”? randomness? a “hyea, that looks good there?” kinda’ thing? that’s not usability, that’s freeware back when Pascal was king and we used zmodem as a protocol to download something at 2400 baud.
“well, here’s the interesting bit: usability isn’t really about what you think, it’s about what the majority (in a group of users with several levels of experience) feels conformable with.”
Actually usability is about what I think, as it is what everyone else thinks. What is useable for one person is not necessarily useable by another. This is not a right or wrong issue, it is a persons preference. All the focus groups in the world can say something is useable, but that does not mean it is for everyone.
http://www.betterdesktop.org
You can also check out http://www.openusability.org for more info on the work being done.
sweet!
I stand corrected.
thank you.
You know… There is a way to make you understand ideas about OS X UI: RTFM!
I’ve been using various Linux distributions for about two years now (but before that I was toying with dual booting and never actually using it), and there are only rare occasions I use the command line.
The first is when a program I’ve just installed isn’t running. I run it from the command line so I can get some error feedback. Usually just to find out I’m missing a particular library, then it’s a simple matter of installing that.
Second, and rarest for me, is if I know exactly what I want to install, to the letter. It’s quicker to type ‘apt-get install’ than it is to open Synaptic and search, when I know what I want to install.
The third case is often coupled with the first. Compiling a program from source. I know, a lot of fanatics (and normal people) compile everything from source religiously, but when I can double click a .deb or use the package manager, I will. Since I’m currently using Mint, which is completely compatible with the Ubuntu repositories, practically the only time I ever need to compile something myself is when the latest SVN version of a program has a must-have feature. More often than not, the version in the repositories is good enough, or sometimes, the program’s site has a more recent .deb.
“The Linux user interface isn’t going to be changing anytime soon as the average Linux user/developer lives in the command line.”
Are you in the dark ages? I’ve been using PCLinuxOS for almost 3 years now and have yet to open a command line. Oh yeah, and last I looked, the command line was pretty popular in Mac OS X. They even use “sudo” just like Ubuntu does.
Edited 2008-07-24 20:36 UTC
According to your comment, “the underlying user interface is easier” on the Mac as compared to Linux. The underlying user interface of the Mac (e.g., below the “pretty screen” as you put it) is UNIX. The same (more or less) interface underlies Linux as well.
As for interfaces, where Mac has only one (Aqua) Linux has dozens. You can use Gnome, KDE, XFCE, and Fluxbox (just to name a few) or all of them! All of these interfaces are endlessly configurable to resemble Windows, Mac, Amiga, BeOS or any combination in between. Add the dizzying effects of Compiz-Fusion to the mix and configuration possibilities increases ad infinitum. In short, with Linux you can truly have your interface exactly the way you want it. It might take a bit of work on your part, or you can simply choose a distribution that shares your aesthetic ideals and comes pre-configured the way you prefer. Please don’t put down Linux because you think its interface is somehow inferior…that just doesn’t make sense considering the options available. Just my $0.02
The Mac OS is popular because it actually works, and the reason for that is that it runs in a controlled environment. Linux on the other hand is supposed to run on any odd PC that it is thrown at.
As for the Linux user interface not changing, I don’t know where you got that idea. Gnome releases new stuff every six months, and then there is KDE4 that is evolving faster than any user interface I ever seen, it may not be ready for the masses just yet, but when it is, Microsoft and Apple will really need to watch out. It’s not only skin deep, things like nepumuk will not only change how things look, but also change how we work with the Linux desktop.
Besides,whats wrong with the command line, it complements the GUI very well. Some things are better done in a script or a oneliner, even Mac users knows that.
“Actually Mac OSX is popular as it has a pretty screen. I find it’s usability to be negligible. Certainly no more usable than any other interface. The single menu bar sucks rocks, and makes it almost unusable IMO. I want my active application to be up top, not what OS X thinks I should have. ”
Here are the trolls!!!
Don’t you think that you don’t know what you are talking about? Or is it that your taste about good user interface is just horrible?
Anyway, that does not change the fact that what you wrote above does not make any sense….
” I drop to the command line more in OS X than I do in Linux to do basic things like ping, traceroute, etc. Unless there is a GUI tool I am not aware of in OS X that does that..which is more than possible.”
Oh yeah my friend, there is one, and it is called Network Utility located at /Applications/Utilities. Every OS X user knows that this tool is built in with OS X, that means two things. First, all your bla, bla, of getting more into command line in OS X than Linux is juste pure crap and second, you don’t have enough experience and knowledge about Mac OS X to make any judgement on it…. And so comes the question, what’s your point besides just trolling around?
Edited 2008-07-24 07:17 UTC
No, sorry. You may call me troll as much as you want but there’s plenty of people who actually *gasp* do not freaking like OSX’s interface. I don’t need any UI “fashionist” to tell me what’s good taste and what isn’t.
And don’t give me the “enough experience” (so it’s not so “just works”, eh?), I’ve been using OSX daily for months at work.
To each its own I suppose.
I honestly dislike the OSX interface too. Drives me insane when people with other tastes claim you are an idiot or don’t know what you are talking about because you don’t like something *they* like.
people, if we all liked the same thing, humans would be pretty f–king boring.
“Oh yeah my friend, there is one, and it is called Network Utility located at /Applications/Utilities. Every OS X user knows that this tool is built in with OS X, that means two things. First, all your bla, bla, of getting more into command line in OS X than Linux is juste pure crap and second, you don’t have enough experience and knowledge about Mac OS X to make any judgement on it…. And so comes the question, what’s your point besides just trolling around?”
I use OS X, though I did not know that existed. So think before you accuse someone of being a troll will you? Just because one uses an OS, doesn’t mean they know every little utility that comes with it. Thanks for pointing out that exists. I guess every user doesn’t know that is built in eh?
Well, to the person who couldn’t find it, I found it myself and have only used Mac OS X for a total in my entire life about 30min.
But the first thing I did when I saw the network utility was “They copied GNOME!!!!”
The Gnome Network Tools look almost exactly the same. The only difference was that the tabs were in a different order for Lookup, finger, whois, etc.
The only thing I was thinking was “Ok, which one copied which. The Gnome one has been out for a VERY long time now.”
The network utility app was a near direct port of a similar utility from NextStep. I’m pretty sure this would predate gnomes version.
It’s easier said than done. Moving the desktop forward to something that Apple provides requires focus and a lot of development time. It’s something that can only be bought through hiring full-time employees (nobody is going to build great frameworks in little time in their spare time, or do detailed UI/usability work).
The problem with this is margins. If you are going after the consumer desktop, you’ll have a problem, because the margins are very small. Very few consumers would be willing to pay more than say 50 dollars a year for a Linux distribution with support. Imagine how quickly that money is burnt on just support. And there’s an inherent problem here: people who don’t needs support won’t buy it since the code is free, people who do need support and will probably use it are the only ones to buy.
The advantage that Microsoft and Apple have here is bundling. They get a buck of every machine where their OS is used, no matter if you are completely incompetent in using computers or if you write compilers in your spare time.
Maybe one possible road around this is to focus on the desktop in businesses. But then again, Novell and Red Hat have provided this for years, but they are probably making their lion’s share off server usage.
The primary area outside the server where I can see Linux succeed in the shorter term is smartphones. Linux is open, has no vendor lockin for whoever produces or sells a phone/subscription, and usually no support is required. Besides that, the scope is much smaller, no one expects a full office suite or video editing program on their phone.
KDE developers are doing exactly that with KDE4. And apparently no-one pays them big cash for doing it.
You are right, though, that this kind of enterprise takes time. The user interface of KDE4 is not quite ready yet but the underlying frameworks look most promising.
True, many (if not like 99.98% of them I’d guess) KDE developers are not paid to work full time on KDE4 and its UI.
It goes without saying that OSX has had the consistent, smooth, and pleasing UI (and related application frameworks and services) for several years now while KDE4 is still quite far away from delivering the “real KDE4” experience completely bug free and speed optimized (OSX behaves and looks great on integrated Intel GPU’s).
I do hope that Nokia’s purchase of Trolltech helps Qt mature even faster and all that work can be reflected nicely in KDE (it would be nice if Nokia took interest in KDE’s development and hired a few programmers and artists to work on it full-time [KDE could scale to smartphones and MID’s quite nicely with all the work on SVG they are doing]).
Edited 2008-07-24 16:09 UTC
Maybe now they’ll start contributing more upstream?
Please explain.
I mean, hopefully Shuttleworth and Canonical will put more resources into upstream projects (i.e, helping with development instead of simply distributing) rather than just standing on a soap box.
And hopefully Steve Ballmer will GPL MS Office and encourage the community to help port it to multiple platforms.
If wishes were fishes… but I don’t think either is going to happen.
I think in the Linux world we need to fix the small things that doesn’t work before we jump on the new desktop paradigm.
Simple example is importing photos. I’ve managed to get my mum running Ubuntu for 2 years now. She uses Firefox for the WWW and gThumb to import photos. Every now and then the import of photos breaks. Lastly Ubuntu changed their default manager to F-Spot. F-spot refuses to delete photos on the camera.
I ended up creating my own script instead for importing photos.
If we manage to get all these small things working I think we would get a lot further.
Deleting photos on the camera would be a feature and GNOME doesn’t support features.
Agreed. I’d tell you about the fun I’m having right now trying to find a decent video capture program that can record from my webcam. So far XawTV and Zapping have an almost 100% failure rate. Zapping should be capable of doing it, but it’s been badly packaged (& the version in the 8.04 repository is downright broken!)
It looks like I’ll have to install a bunch of -dev packages and build from source to get a usable application.
Take that Apple!
Install digikam for your mom. It’s a much better tool than F-spot.
I can’t even figure out how to view the “Date Photo Taken” information without first pulling it into a photo app.
It’s nice to set bars and all that, as long as they don’t go ‘Beyond Apple’ in terms of price.
Edited 2008-07-24 11:20 UTC
If a Gnome 3.0 wants something to shoot at, Shuttleworth has just given them some ideas.
But one guesses that “Macbuntu” would be a vast undertaking, maybe too much for present resources in terms of salaries and devs across the F/OSS world let alone for a single distro?
One would need to sink all the plumbing below a “software abstraction layer” so that the user never needs to know there are file trees called /usr/lib or /var/www, or that what makes the mail server work is located in a file called /etc/exim4/exim4.conf.template, or that the program they see as a single icon represents distinct files all over the system. A vast amount of of work.
Colour me sceptical, but I wonder if there’s any point in sort of cloning the Mac experience especially since so much of that experience is locked to Apple’s hardware. Linux can’t compete with that in hardware terms and I wonder how many F/OSS folks would ever want to run something that’s “Just like a Mac only it uses Linux”. Surely they prefer “It’s just like Linux because it runs Linux”.
It’s good to be inspired, but if you lead where the crowd won’t go you end up as a prophet in the wilderness. In the meantime it’s back to sorting out some more basic things: wifi, interactivity with mobile phones, cameras, sat navs and a plethora of new smart devices. Sounds humble maybe, but without having all that licked 100 per cent, Linux will never be able to compete with Windows or Mac on the desktop no matter how fine the eye-candy.
I read the article you linked to and to be honest, I don’t even know what Mark Shuttleworth is talking about in regards to Apple. There are a few reasons why I think there is a perception that Linux in general falls down compared to Apple and even Microsoft:
-Computers are appliances. To you and me, they’re a tool with almost limitless uses. To 98% of your consumer computer users(here on out referred to just as users) out there they’re appliances that are turned on, some work done, and then turned off. They’re microwaves with youtube and email.
-Accessibility is an issue. And by that I mean if you buy an Apple, you get OSX. If you buy anything else, you get Windows. That’s pretty much the standard even taking into account recent deals between Novell/Ubuntu and PC manufacturers. Users don’t understand the concept of an operating system, live CDs, or dual booting. And for the most part, they don’t even care about the OS.
-Users don’t care about the details such as licenses. It’s the larger issues that matter to them. Can I get my email? Can I get onto the internet? Can I use itunes? Can I take my word documents home with me and work on them? Until they can’t do those things, or something else which Linux can provide over the competitor, there’s no real reason to change systems. When it comes time to get new hardware, the user is back at the point above where they have a choice of Apple(OSX) or PC(Windows).
From a personal view, I believe Linux is very competitive with Apple and Windows on the desktop. Technical reasons, aside from maybe package management, are not why Linux trips up in the race with Apple and Microsoft. Accessibility to Linux, forced accessibility really, is where the separation is. Relationships with hardware manufacturers are a great start and must continue to occur and be strengthened. I believe getting Linux into the hands of every days users is the key for market growth. Desktop Linux is progressing just fine from a technical aspect; consumer uptake is something that companies who want to sell Linux or Linux services need to make happen.
Edited 2008-07-24 12:51 UTC
This makes me think Mark has some surprises for all in the next Ubuntu releases.
I personally think Sun with its OpenSolaris and hardware experience could in fact build what Mark Shuttleworth wants. OpenSolaris currently has limited hardware support but doing an Apple style play but for the Enterprise desktop would work out well for them I think especially with x86 based hardware. Sun would need to put a lot more resources into the desktop but they could probably do it.
I wonder if they would be able to use an existing open-source solution or if they would need to make an entirely new desktop and toolkit chain in order to have the necessary control, especially since GNOME wants to prevent themselves from having too many features.
Sun has never been a desktop/usability specialist. I doubt they will get more off the table than just a slightly crippled version of Gnome.
Err, Sun is/was a major backer of the usability and accessibility work done inside GNOME.
It’s okay to talk big but considering how the last Ubuntu release went, I think Canonical needs to change their ways first.
GNOME could do just fine with some updates, but whatever distribution should have only production ready software. The beta software, the unfinished software, etc. is interesting, but if things don’t work, it gives a bad impression.
Any Linux distribution with good software could excel, but someone has to take the risk to make a quality release for someone who isn’t going beyond point-and-click. Ubuntu is pretty much there, if they’d just substitute software choices.
Why isn’t there a gamer’s Linux?
Lack of a real gaming API? (There is nothing comparable to directX)
Lack of games? (Ask your local store)
Because Linux users in general don’t pay for games? (Ask Loki Games or id Software)
Nothing like directX, luckily there is not, as it is crap. Linux has SDL instead…..
DirectX killed gaming on the PC, proper hardware access was always faster, I long for lost hours gaming on things that needed emm386 before they would even try to load
emm386, I hated that !
I remember asking all my mates how to get Rise of the Triad to work. I did not have access to the internet back then. In fact, a guy down my street used to log into some bbs.
How about ubuntu focus on getting basic functionality to work before taking on Apple in the eye candy stakes?
Seriously, I fired up KDevelop (Kubuntu 8.04) to write an application a few weeks ago, and found that cut n paste is not implemented consistently when highlighting text in its console.
I mean, its 2008, and users can’t expect a reliable cut and paste mechanism?
How about instead of ‘Competing with Apple’, Ubuntu work on competing with windows 95 levels of UI functionality and consistency. Its a far more realistic goal.
I use Linux as my main desktop machine, and for numerous servers – but Ubuntu compete with OS X for slickness any time soon? I don’t see it happening.
Not when you can’t even ensure that cut and paste works properly in the apps you ship in your distro.
How about putting the blame where it belongs. Ubuntu just packages stuff and their primary desktop is Gnome anyway, which is where most of their attention goes to.
So how is ubuntu going to compete with OS X if the ‘blame’ for any issues that make for an inconsistent user interface needs to go elsewhere.
Ubuntu can’t create something that competes with OS X precisely because of this reason – I mean its not like anyone at ubuntu is sitting there thinking ‘Hey, lets have inconsistent cut n paste behaviour’, but equally, they’re also not in control of any of the components of the system that lead to this problem.
The biggest issue with consistency is the ‘not my problem’ problem – e.g. the core X Windows application doesn’t think that the clipboard, fonts, managing window layout or doing pretty much anything except drawing rectangles on the screen is it’s ‘responsibility’ – everything else is supposedly best implemented as extensions or handled at a higher level by applications.
Which is one way to look at it, but its an example of the thinking that leads to the ‘nobody can fix this mess’ situation that Ubuntu is in today – When your architecture is based on a policy of ‘design in as little as possible, wherever possible’, its hardly surprising that other systems seem far better-designed.
Another example:
Using a KDE desktop, i can’t click on links in thunderbird and have them open in my browser, because theres no way to configure the default browser setting that thunderbird uses from either thunderbird itself or KDE. Is that Ubuntu’s fault? Mozillas? KDE’s? GNOME’s? X.orgs? POSIXs? LSBs? freedesktop.orgs? I don’t know. But its a ridiculous bug to have no simple solution to, in a desktop platform that supposedly will compete with OS X.
And the problem is hardly KDE or Kubuntu-specific – in firefox, under GNOME , try choosing a helper application when clicking on a link to an unreognised MIME type – instead of popping up the GNOME application menu with the list of installed apps, it gives you a stripped down filechooser widget starting in /home – from there youre supposed to intuit that you should navigate to /usr/bin and pick an application from out of the thousands of cryptically named apps in there. For anyone not familiar with the linux FSH, this means its practically impossible to acheive this simple task. This is not an issue that anyone who gave a crap about the usability of their default browser would let out the door unless they had no choice. And Ubuntu have no choice.
Sure, you can say these are all application problems, and shouldn’t reflect on ubuntu – but you will never hear that same line from Apple. What they package in their ‘distro’ pretty much works 100%, as expected, and they make it brain-dead simple for app developers on their platform to do things the ‘correct’ way easily – this is mostly by providing strong guidelines to developers and exposing simple ‘need a application picker? heres the way to do it’ functions through a consistent API.
Is the solution to dump KDE, and enforce a GNOME-only ubuntu? I guess i’m actually open to that, even though i’m a longtime KDE user because its quite clear there is no way to acheive a ‘Mac OS X like’ experience with a horrid mish-mash of APIs such as we have today.
And we, as a community, would pretty much rather face a firing squad that give up our mish-mash of APIs and work together. Thats why MacOS X is perceived as being far more easy-to-use, friendly and consistent than ubuntu, and why I don’t believe that a ‘OS X-like’ user experience is possible on Linux any time soon.
Well, I have just tested with plain Gnome Ubuntu. Works as expected.
Actually, I think that the best way out of mess for Ubuntu is just to drop Kubuntu. It was always inferior, and obviously, it makes just “bad press” for Ubuntu…
Clearly a firefox’s ToDo. They will get there. You can do this in Gnome.
Besides, this does not work perfect in Windows either (cannot comment about OSX).
Maybe. But maybe, who cares. As long time Win user, I have got an impression with Ubuntu 8.04 that it works quite well for most business oriented tasks I need.
In any case, compared to what Microsoft had “achieved” in 5 yers of development of Vista, I guess Ubuntu is “getting there” quite quickly…
Unless you have actual deals and relationships with OEMs to ship your operating system, you will *never* have the level of integration that is necessary to overcome Apple and Microsoft. That is why Linux has failed to build desktop market share.
Shuttleworth might know where he wants Linux to go, but to judge from the article, he has absolutely no idea about how to actually get there. Or if he does, he is keeping his cards very, very close to the vest, because the article, such as it is, contains not the slightest hint to suggest that he has worked out any kind of strategy at all.
Linux, as a community since it’s inception, to the present has seen over $10 Billion invested from Corporations to bring their needs to Linux and use it as a true testing ground for Operating System design while juggling relationships with Linus and his vision for his Operating System.
Linux has had billions and billions of dollars invested.
In fact, it’s had more money invested in it that OS X since the days of NeXT.
The beauty of endless distributions and niche versions has benefited and hindered at the same time.
Shuttleworth knows that the only way to expand into the Desktop realm is to combine the best of breed practices from within the Linux Communities and see if they can develop a unified offering.
By design, the Linux Community abhores Unity, other than the GPL, and relishes diversity by deflecting all criticisms with countless forked distributions and may the winner earn their place if it is to happen at all.
Just take a look at all the linux distributions and the many branches from a much larger trunk of few distributions we have in Linux.
Debian has more in it’s harem then the rest of the distributions on it’s level of history in the Linux Community, but you didn’t see it becoming the solution, you instead see it’s children vying for that spot.
Shuttleworth is living in a fantasy if he thinks LINUX will supplant OS X on the Desktop.
Hell, OS X is just about to start making in-roads in the Enterprise Markets when 10.6 arrives.
I love my Debian and it’s not going anywhere other than advancing on my general PCs but I’m equally fervant about my OS X and it’s future.
There is plenty of room in this world for both and the more they interoperate with one another the harder it is for Microsoft to dictate the markets.
I think it is wrong to focus on Apple. They are a small player in desktop, server, and mobile operating systems. They also have a very limited range of hardware. Linux’s strength is that it is highly adaptable to all types of hardware. Linux should be taking advantage of that by going places that the market leader can’t go with Vista. Personally I think Linux should continue to ride the UMPC wave rather than follow Apple.