It seems that each distribution has found a niche: Red Hat and Ubuntu are the leaders in their markets, and SUSE is a comfortable runner-up. However, history has shown us that businesses are not content to stay still too long or play second fiddle. So, what will Red Hat, SUSE, and Ubuntu have to do in the new year to gain new ground?
The Big Three of Linux: Looking Ahead to 2008
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
129 Comments
-
2007-12-15 10:15 am
AlmindorPut up or shut up. Have you ever had to write a driver? Do you even know WHO writes the drivers for linux? Do you know if they even get payed? HOW F- DARE YOU DEMAND SOMETHING? You didn’t PAY for it, so be happy for what you got, or get down and fix it yourself, or get up and shout to the manufacturer to provide proper linux drivers.
The problem isn’t in linux, or the distroes, the problem is in people like you not knowing how things work.
The only thing you got right is the stupidity of xorg.conf which should’ve been dealt with 10 years ago, HOWEVER, you didn’t help doing it, and me neither, so we don’t have the right to demand anything here.
-
2007-12-15 10:48 am
agentjSo why people want to push linux everywhere e.g. on computers for ordinary users ? It’s better to pay MS and get everything working that receive “HOW F- DARE YOU DEMAND SOMETHING?” for free crap.
-
2007-12-15 12:54 pm
bert64Because linux provides choice, MS tries to take that choice away from the users by making their offering as proprietary and incompatible with everything else as possible.
I and many other people would gladly pay for convenience, if that convenience didn’t mean we were tied to a single vendor. And in that respect, i think Apple are a better choice than MS. They are closer to being open/standard, and their systems are more integrated and working right off the bat. MS is the worst of both worlds, lacks the flexibility/transparency of linux, and lacks the integration and “just works” of Apple.
-
Almindor
Push? Who pushed linux on you? Last time I checked 80% of laptops pushed windows on me, not linux…
You CHOSE linux, don’t try to pretend it was forced on you.
-
2007-12-15 11:58 am
renhoekwell, the nvidia guys definitely got paid. i’m trying to get xorg to display correctly on my tv-out. i do need to edit my xorg.config, i did it before, but i could not get it working. not with the ati drivers too, but it think that is because ati does not support tv-out (last time i checked).
the problem is the guys coding the drivers will easily fix their xorg.conf files, the guys who can not get their xconf correctly will not be able to fix the drivers. so the only thing left is complaining. getting comments like “rtfm!” or “don’t complain, it’s free” will not solve this problem.
for the record, i want everything autodetected! if you can write a driver, it should be just 10% more work to get everything working automaticly. why the hell do i need to specify the amount of ram, monitor size, refresh rates and all other kind of crap. i want to be able to override the autodetected values just in case, but i should not need to specify them in the first place.
ps. using caps/swearing won’t make you more right.
-
Almindor
for the record, i want everything autodetected! if you can write a driver, it should be just 10% more work to get everything working automaticly.
You just picked this number out of your ass. Did you write a driver? I guess not, so I’ll consider the 10% to be a lie.
I don’t know how much work it is to get it to work this way, perhaps it is 10% perhaps even less, but if so then you could you know.. try it..
-
renhoek
if we start nitpicking on minor details, i said should. and that 10% was an estimation based on the gus and vesa detection routines i wrote. please try to understand the point i’m making instead of looking at the unimportant details of my argument.
the point i’m making is that autodetection should be used where possible. these routines are relatively easy to code compared to the writing of the whole driver (happy now?). the reason this happens not very often (in the past it was a lot worse) is that the developer has his stuff working and correctly configured, and expects the endusers to do thesame. hmm, getting a bit offtopic now
-
2007-12-15 11:59 am
Joe User“Put up or shut up. Have you ever had to write a driver? Do you even know WHO writes the drivers for linux? Do you know if they even get payed? HOW F- DARE YOU DEMAND SOMETHING? You didn’t PAY for it, so be happy for what you got, or get down and fix it yourself, or get up and shout to the manufacturer to provide proper linux drivers.
The problem isn’t in linux, or the distroes, the problem is in people like you not knowing how things work.
The only thing you got right is the stupidity of xorg.conf which should’ve been dealt with 10 years ago, HOWEVER, you didn’t help doing it, and me neither, so we don’t have the right to demand anything here”.
Not everybody has 30 years of experience in Assembly and C development. You clearly mean that Linux is for developers. This utterly contradicts the “Linux for human beings” motto. Oh, and you bash the average user the same way this poor guy might have been in Linux forums. Good luck gathering more users and convincing them to use Linux this way.
-
Almindor
There’s a difference between constructive criticism and demanding. I also wrote some libs and programs but when someone goes on demanding something from me after all the hours I spent making the stuff and giving it away for free, I just give them the booty.
People don’t respect what they get for free, it’s sad but that’s how it is.
Whenever I hear someone demand something of linux, I usually send them to windows, since that’s what they want.
Don’t demand anything from a free OS.
-
Almindor
If you can’t help, you can’t demand it’s that simple. It’s a constructive community, once you did something to help you have some right to ask for something else in return. You don’t have to code, you can help in other areas like docs etc.
Anyone who just comes and demands something needs to be booted into windows.
-
2007-12-15 2:57 pm
raver31You sir, are retarded.
It is called FEEDBACK, and that guy was giving some.
Now, it might be that someone was already working on the problems that guy faces, and he might have got a reply.
What us in the “real” Linux community need are people who will test hardware and give feedback on things working or not, we dont need people like you with half a brain shouting at them to stfu or rtfm.
Go back to school.
-
Almindor
Feedback is “this doesn’t work for me, could you pehraps have a look at it?”.
Rude demanding is “this works in XXX and you suck because you can’t make it work”.
The latter warrants a kick from the community.
-
Joe User
“If you can’t help, you can’t demand it’s that simple. It’s a constructive community, once you did something to help you have some right to ask for something else in return. You don’t have to code, you can help in other areas like docs etc.
Anyone who just comes and demands something needs to be booted into windows”.
Who said that guy you shouted at didn’t help writing docs and testing beta versions?
-
2007-12-15 12:21 pm
MichaelThis has nothing to do with the guys who write drivers. It’s the job of the distribution to ensure that graphics, and all hardware for which drivers exist, is automatically configured so that, for the user, it “just works”.
In the cases of RedHat and SUSE, they do get paid, handsomely. If you want, you can pay Canonical for Ubuntu support. In each case, the consumer has every right to object when things don’t work, or when they’re asked to open up a text editor as root and edit a config file by hand. One typo and you’ve got a non-booting system. That’s not what you pay for.
-
Almindor
“Consumer”? I’m not a consumer of Canonical, I didn’t pay for it, hence I cannot demand anything.
Drivers need to support certain features before they can play well and be auto-used, it’s not as simple as just mixing it all right. If a driver is incomplete and doesn’t support eg: hotplugging, the distro people can’t do much.
-
2007-12-15 11:31 am
RIchard James13Compare that to the awful mess that is in windows where i *gasp* download the drivers from the manufacturers website (or the driver cd), double click it, click next-finish a number of times and my mouse suddenly works.
Perhaps you take that issue up with your hardware manufacturers and not Linux. Seriously you can’t expect Linux to magically work for all of your hardware if your hardware manufacturers are unwilling to work with Linux.
I mean if they treated MS Windows the same way they treat Linux would you be screaming why doesn’t Microsoft fix this? Or would you complain to the hardware manufacturer?
-
2007-12-15 6:43 pm
kristophPerhaps you take that issue up with your hardware manufacturers and not Linux. Seriously you can’t expect Linux to magically work for all of your hardware if your hardware manufacturers are unwilling to work with Linux.
I think the point here is that, even thought it is not the fault of Linux, the broad lack of such support will keep ‘average’ users from being able to adopt Linux.
]{
PS. I run Linux but, seriously, there is no way the vast majority of people could use it without geek support.
-
2007-12-15 8:44 pm
[email protected]Perhaps you take that issue up with your hardware manufacturers and not Linux. Seriously you can’t expect Linux to magically work for all of your hardware if your hardware manufacturers are unwilling to work with Linux.
it is not beeing unwilling, it is a matter of profit.
From my point of view, any concept of bringing Linux to the masses connected with profit failed since 1992.
-
2007-12-15 12:16 pm
VandersCompare that to the awful mess that is in windows where i *gasp* download the drivers from the manufacturers website
So what you’re saying is, it would be nice if hardware manufacturers supported Linux just as well as they supported Windows?
Given the level of zeal you wrote your post with I trust you’ve complained to your hardware manufacturer about their lack of support for Linux and that you’ll purchase your hardware more carefully in future?
-
2007-12-15 3:50 pm
google_ninjaIF your mouse is not set up automatically (it usually is), you can go into your xorg.conf file and change “Buttons” “2” to “Buttons” “3” or however many you have. It is simple and takes a few seconds, instead of trawling vendor websites.
-
2007-12-17 2:48 am
pixel8ryou think you need to do all that just to get 3 buttons working on a mouse? sure, you can make linux sound hard if you fill your post with a pack of lies.
I’ve not edited one config file by hand for quite some time. My mouse always works.
Lets look at what you just posted…
Compare that to the awful mess that is in windows where i *gasp* download the drivers from the manufacturers website (or the driver cd), double click it, click next-finish a number of times and my mouse suddenly works.
btw, good luck “clicking” next when your mouse doesn’t work.
So in windows you would blame the manufacturer if it doesn’t work yet in linux you blame linux? How is that fair?
Or maybe consider that my mouse works in linux with all buttons (not just 3) and I can scroll in any window without having to click (windows doesn’t do that), and all of this WITHOUT having to download anything…
Which was the easier one again?
Maybe you need to check your hardware before expecting a free OS to magically support all of your obscure hardware. Until manufacturers start producing drivers for linux, this problem wont go away, and neither will people’s ignorance on the issue.
Edited 2007-12-17 02:50
-
2007-12-17 6:24 am
lemur2I’ve not edited one config file by hand for quite some time. My mouse always works.
I had a logitec marble mouse. It worked automatically on first install of Ubuntu, but not all four buttons worked they way I wanted.
I did have to do a bit of research, and I did have to edit xorg.conf by hand. I found some posts on the web which showed me which text file to edit and how. I didn’t have to actually type anything … copy and paste from the web browser to the text editor worked just fine. It took about five minutes to research, and another five minutes to set up the way I liked. No command line involved.
Some bits of more obscure hardware take a little bit of configuration effort to accomodate on Linux. Nothing like what the OP posted, but all the same it would be nice if the manufacturer (Logitech) had included a Linux README file, rather than my having to research on Ubuntu forums what to do.
2007-12-15 11:45 am
Soulbender
Just up the “Buttons” value in your xorg.conf
Do you actually even have to do that? I know that my 3-button mice works without me doing any manual config in both Ubuntu and openSUSE.
-
RIchard James13
I know that my 3-button mice works without me doing any manual config in both Ubuntu and openSUSE.
That would probably be because most implementations of X Windows on various UNIX flavours has implemented 3 mouse buttons since the 80’s. I first used a UNIX system running X Windows on DEC with three buttons in 1991.
-
google_ninja
It depends on the distro. Most of them set it up automatically. If it doesnt though, its not exactly rocket science.
2007-12-15 8:15 am
2007-12-15 12:15 pm
netean
talking of package management.
perhaps a way to install and uninstall programs easily without being connected to a super fast broadband… maybe say the ability to install easily from usb stick, or linux magazine free cd-rom.
-
Michael
They got that. They all do.
But it’s different for each distro. So don’t expect a magazine cover disc to provide pre-packaged versions of it’s software for every major distro.
Instead, they (should) just install a generic i386 binary into a directory of it’s own, typically under /opt or $HOME, such as you might do if you downloaded Firefox from the website, or if you installed Unreal Tournament. Uninstallation then consists of simply deleting said directory.
2007-12-15 3:25 pm
Ford Prefect
Let’s see what we can do:
1. All mice more than 3 buttons must be supported
Arch Linux ships with a decent X server which supports your mouse via evdev. My Logitech Revolution MX (a monster of a mouse!) is supported flawlessly (13 buttons)
2. All Web cameras must be supported
There are enough cheap web cams out already that are supported. Hundreds! Arch Linux: “pacman -S gspcav1” (Reads: “A kernel module with support to 244 USB webcams”)
3. All Tablets must be supported
Dunno anything ’bout tablets.
4. All Advanced Graphics features must be supported (rotation, vivid colors like in digital vibrance,…)
Arch Linux with Xorg 7.3 gives you the new xrandr which supports output hotplug, rotation, etc. I’m also able to use colors, digital vibrance settings from nvidia:
pacman -S nvidia nvidia-settings
5. All Distros must have an excellent package management system to update and install applications (this moved me from fedora to ubuntu after 2years on fedora/redhat)
“all distros”!? .. Arch Linux has excellent package management: pacman!
6. All keyboard extra keys must be supported
No problem here, you can use xev to find and xmodmap to map even exotic keycodes.
7. All udf disks CDs and DVD must be automatically mountable
No problem here: pacman -S gnome-device-manager (for example using hal, gnome automounting)
8. All codecs must be installable with ease
pacman -S codecs
9. All browsers plugins are preinstalled (java, flash, …)
Preinstalled?! WTF?! Anyway: pacman -S jre flashplugin mplayer-plugin and done (last one is for embedded videos)
10. I will not be more demanding, the previous are enough for me.
So these are really the only demands you have? Have a look at Arch Linux http://archlinux.org
Edited 2007-12-15 15:28
-
2007-12-15 4:44 pm
Sophotect/me waggles indexfinger and says:
http://www.archlinux.fr/yaourt-en/
As to the Webcams, yes, i can confirm that. Just yesterday i went for the second cheapest cam in some shop which was a Logitech Quickcam Chat. Worked the way you described. Btw. for camera devices which aren’t covered by gspcav you can utter some yaourt -Ss webcam and see what else is there 🙂
To go back to topic. The OP is right there, despite my refusal to use the established OSes because of irrittion about their inner workings. They DO lead with regard to easy usage of peripherals. Therefore he gets my vote. Because one shouldn’t have to know how to search for driver/codec/whatnot to just use the system. Imagine you’d have to steadily supervise every individual ignition spark from the motor while you drive a car? Or take a fridge. Just plug it into a the powersocket and more or less forget about its inner workings. Linux (as of now) is for Techies, everything else is an illusion in my opinion.
Edited 2007-12-15 17:01
-
2007-12-15 5:39 pm
zombie processDear. God.
Assuming the OP is just baiting, please do not steer trolls and idiots to Arch.
Assuming the OP was actually sincere, please do not steer people who want/expect everything done for them to Arch.
Arch is a fantastic distro for people who like getting their hands dirty. Arch is a terrible choice for anyone who expects things be setup for them automatically, or who demand things be “simple” as in “I never have to think about anything again in my entire life.
-
Sophotect
Your wish is order to me ;->
Just to show what can go wrong with software installation in Linux:
http://forums.gentoo.org/viewtopic-t-597716.html
And please you Gentooers, don’t crucify me for this, because similar things can happen with derivatives of Debian the same way, where a simple “apt-get dist-upgrade” morphs magically into “apt-shred dist-chainsaw” if you choosen to have dared including some “unfree” stuff which is necessary in daily life.
And no, the “others” aren’t free of those hassles either!
-
Ford Prefect
I just wanted to show what is actually there. I don’t think this guy will install Arch Linux.
But just one additional note to you: Arch is not meant to and thank got not yet has become what I would call an “elitist” or arrogant community.
Whoever is seriously interested in it can expect me to guide him/her, regardless of prior knowledge.
Edited 2007-12-15 19:36
2007-12-15 5:06 pm
2007-12-15 10:45 am
2007-12-15 11:08 am
Xaero_Vincent
1) More stability. The experience is much better now but there still is room for improvement in say Xorg, which still enjoys crashing now and then.
2) Improved remote desktop. I mean a better terminal services client (rdesktop). Rdesktop is great and have worked around it’s issues but I want see more development: RDP 6.0 support and bug fixes.
3) OpenGL and Direct3D hardware accelerated guest drivers for virtualization software. This will virtually solve all the issues associated with playing Windows games that suck under Wine-type products.
I’ve just tried the latest VMware Workstation with experimental 3D support under Linux; let me tell you… it’s not even close to being useful for gaming when it cannot even refresh the desktop–you have to move the mouse to re-paint the screen! Attempting to launch 3D, full-screen apps just throws a BSOD.
Parallels and VMware should stop jacking around and port their Macintosh codebases to their Linux and Windows counterparts.
Edited 2007-12-15 11:10
-
2007-12-17 6:28 pm
gilboa1) More stability. The experience is much better now but there still is room for improvement in say Xorg, which still enjoys crashing now and then.
Can’t say that I remember when was the last time that saw X.org crash, but YMMV.
2) Improved remote desktop. I mean a better terminal services client (rdesktop). Rdesktop is great and have worked around it’s issues but I want see more development: RDP 6.0 support and bug fixes.
You -got- to be kidding me.
I’m using rdesktop (from Linux) and Microsoft’s own TS client to remotely manage a number of Windows XP/2K3 machines. (including my “Windows” development workstation).. and it sucks, badly. From lousy performance (on a 1Gbps network), to unstable clipboard and sound support…
RDP uses far-more bandwidth then, say VNC and lacks the just-connect-and-run rock-solid stability of X-over-SSH.
Oh, and then you have NX…
Granted, RDP 6 might be the best thing since sliced bread, but for now, I’m sticking to X+ssh.
3) OpenGL and Direct3D hardware accelerated guest drivers for virtualization software. This will virtually solve all the issues associated with playing Windows games that suck under Wine-type products.
Don’t hold your breath. You’ll need completely new guest display drivers and major changes to the host’s display drivers to make that happen.
… And even then, performance will be far from being host-like. (Memory needs to be copied from the guest’s memory space to the host’s and back – reducing the performance considerably.)
Para-virtulized solutions (such as Xen) might have luck at giving near-host performance, but I don’t see Microsoft releasing special versions of Windows Vista with virtualization-under-Linux support…
– Gilboa
2007-12-15 11:29 am
kragil
Mark is just a great down to earth guy! ( altho he was in SPACE! )
Steve is just a snotty egolomaniac who loves patents and is all about show.
There is absolutely no comparison in my mind.
Comparing a true philantrope with a greedy business man is beyond me.
Just my .02€
2007-12-15 11:41 am
superstoned
If Novell wants to be the future, they really need to start making some sane decisions on where to put their resources. Choosing the best technologies and not wasting time on stuff like rewriting Qt apps in GKT (Yast!) just because someone doesn’t like Qt. If they would’ve spend half the time improving the normal version they would’ve had a better Yast. Wasting time on rewriting a good application in an inferior toolkit, WTF?
After two years of no KDE release, Gnome is still catching up – the release announcements still read like a timemachine from the KDE 3.3 era (wow, Evolution warns you about a missing attachment?). When will Novell management realize they are wasting time and money? If you want to get ahead of the game, and be really competitive to Apple and Microsoft, KDE is your only chance. It lacks the polish, but is technically already ahead (and has been for a long time) while Gnome would need insane amounts of work to be lifted to even the same level, let alone be better.
Same goes for OpenOffice, btw – there is no ambition there! Just wanting to recreate MS Office 2000 or MS Windows 2000 won’t bring the Linux Desktop any further. Projects like KOffice and KDE 4 are innovating, renewing the underlying technology of the Free Desktop to enable real progress. Not joining them but wasting resources on obsolete technology is the biggest mistake one can make.
-
Joe User
“Wasting time on rewriting a good application in an inferior toolkit, WTF?”
LOL, no one cares about your philosophical opinions, what people care is GUI and visual integration, plus, using two different graphical libraries is a waste of resources. And it’s ugly. It’s all about polishing the desktop and bringing consistency.
“If you want to get ahead of the game, and be really competitive to Apple and Microsoft, KDE is your only chance. It lacks the polish, but is technically already ahead (and has been for a long time) while Gnome would need insane amounts of work to be lifted to even the same level, let alone be better”.
This is in your imagination. Both KDE and Gnome are great. Whichever I use: I don’t miss a functionality.
“Same goes for OpenOffice, btw – there is no ambition there!”
Have a look at the OpenOffice.org mailing lists, you’ll see if there’s “no ambition there”.
“Just wanting to recreate MS Office 2000 or MS Windows 2000 won’t bring the Linux Desktop any further”.
If you ask me, yes, I prefer MS Office 2000 to MS Office 2007. More simple, less bloat, less bells and whistles, cleaner interface, does what it should do, no more. OO.o is as good and it’s free.
“Projects like KOffice and KDE 4 are innovating, renewing the underlying technology of the Free Desktop to enable real progress”.
Gimme a break, KOffice is definitely not better than OO.o
-
superstoned
If Novell wanted KDE apps to visually integrate in Gnome, they would write something like the GTK-Qt theme which allows Gnome apps to integrate in KDE pretty good (fonts, colors, icons and style). So they wasted time, imho. Not too untypical, seeing projects like Galzium and that amarok-clone in GTK…
I didn’t say Gnome isn’t great – it’s UI is pretty good, and in many areas actually ahead of KDE. but in terms of underlying technology, I don’t think many knowledgable hackers would argue Gnome is ahead of KDE 3.5.x – let alone KDE 4.
And I also didn’t say KOffice is better than OO.o, I said it is more innovative. And that’s a fact nobody with a sane mind would deny…
Innovation is the way of the future – companies that don’t innovate die. In FOSS, the best technology often wins – thanks to mechanisms described in the famous “The cathedral and the bazaar” article described by Erik S. Raymond. So betting on Gnome might not be the smartest move around.
Of course, if you just pour in enough resources in Gnome and don’t invest in KDE, both will remain more or less equal in terms of user experience, and the Linux Desktop will always stay behind the non-free world.
-
2007-12-15 3:40 pm
sbergman27“””
“The cathedral and the bazaar” article described by Erik S. Raymond.
“””
“K” on the brain, today? 😉
-
2007-12-15 3:15 pm
shagaKOffice definitely is not better than OO.o, but:
1. It is developed by a very small amount of developers. Taken that into account, it is amazing what it looks like (and with KOffice 2.0 around the corner – I guess it will be available at the end of Q1 2008, this will be even more obvious).
2. It has the great advantage of not carrying the legacy of old code. OO.o is a horror to comprehend. KOffice is quite easy to understand, so for new developers it is much less a hassle to become involved.
3. It leverages the desktop underneath, so it is tightly integrated to the desktop. The experience is incomparable.
Again, as I wrote in another post here, just imagine a fraction of fundings that goes into OO.o going into KOffice. Like, for example, into doc filter and another “corporate” stuff.
-
2007-12-15 3:30 pm
RukkotYou’re right. Novell should bet for KDE, at least to distinguish SUSE from the rest. It’s important to have a different product to compete with Red Hat and Ubuntu.
-
sbergman27
“””
After two years of no KDE release, Gnome is still catching up – the release announcements still read like a timemachine from the KDE 3.3 era
“””
Superstoned, please stop trolling. I normally don’t feed the trolls. But I know from experience that you are not really “a troll”, but a regular guy, with some good ideas, who happens to be trolling right now. And who does so, frequently, on the topic of DEs. Please enjoy your desktop of choice without making gratuitous attacks upon others’ desktop choices.
-
shaga
I don’t think he is trolling. He expressed his opinion about what Novell made wrong and I second his opinion. If Novell after SUSE acquisition put more resources on KDE, and with SUSE it would mean just to continue with what they did, they would now be the driving force behind the innovation in the desktop (because KDE4 is the innovation) and would easily compete Canonical and RedHat.
Just compare the corporate funding of GNOME and KDE – GNOME is backed by RedHat, Novell, Sun, Canonical, the amount of investments into GNOME is much bigger than the investments into KDE. But still even in this situation the KDE desktop is still playing the same league as GNOME is. Why? Because of the technology. Now imagine that Novell put its force behind it (more than they do now, they still employ some KDE developers, but who is in the Novell top management?)
Maybe even I will be called a troll, but I can’t help not to think that the corporate move to GNOME was one of the most unfortunate decisions ever made in the freedesktop market. Be it KDE what was chosen years ago, the freedesktop would be miles away from where it is now.
-
sbergman27
“””
Maybe even I will be called a troll, but I can’t help not to think that the corporate move to GNOME was one of the most unfortunate decisions ever made in the freedesktop market.
“””
While I disagree with that assessment, I would not say that you are trolling. Trolling is not about what is said; It is about how it is said. It is not about the denotative; It is about the connotative. It is not about the rational; It is about the emotional. And it is almost always counterproductive.
Which is not to imply that I have not been guilty of it myself, on this very topic. 😉
-
2007-12-15 4:05 pm
HievThe founding Novell and RedHat put in GNOME or GTK is nothing compared to the money behind Qt, Trolltech have more men power on it, more hours per developers, etc.
GTK has some founding but not anought.
Edited 2007-12-15 16:17
-
superstoned
I doubt that – Trolltech is good and all, but it’s very small compared to Red Hat, Cannonical and Novell… Sure, there are other companies behind KDE these days, but most of those just started rather recently and aren’t big at all.
-
Hiev
Is not that TrollTech is smaller than RedHat and Novell, is that its spin as an enterprise is dedicated to the toolkit, Novell and RedHat are dedicated to services and most of the resources are for that,hence the support to GTK is not as good as TrollTech support for Qt.
Edited 2007-12-15 16:53
-
superstoned
Hmmm, that might very well be true when talking about the toolkits – but there is a lot more to a DE than the toolkit. KDE developers are notoriously anal about technological issues (has to be a clean API! must clean that API! can’t stand the pressure, aaarhg etc) which lead to a rather high quality standard for KDE 😉
My argument was that Novell, Red Hat and Canonical should stop wasting money on Gnome, and put their resources in KDE – I think the above only reinforces that…
So KDE gets more money poured into its infrastructure, Gnome into the bells and whistles. Wouldn’t it be smarter to all build upon that great infrastructure? Isn’t it a waste to build upon a more and more aging infrastructure which only slowly evolves?
Of course, this wouldn’t mean the end of Gnome – they probably re-invent themselves, restart, and become something better than they are now. I feel they are getting lazy – they feel confident about the money coming in and don’t work for it. A bit pressure would do good. The Free Desktop would be better of.
-
Hiev
My argument was that Novell, Red Hat and Canonical should stop wasting money on Gnome, and put their resources in KDE – I think the above only reinforces that…
So KDE gets more money poured into its infrastructure, Gnome into the bells and whistles. Wouldn’t it be smarter to all build upon that great infrastructure? Isn’t it a waste to build upon a more and more aging infrastructure which only slowly evolves?
That could be done but it would be necessary two dramatic changes:
1.- Fork Qt and don’t depend of TrollTech, the fact to have one corporation in charge of the toolkit is not atractive at least not for me.
2.- Change the license from GPL to LGPL, Not everyone is advocated to GPL and LGPL can cover both.
Meanwhile this two points are not reached Im afraid there will be room for GNOME and GTK.
Edited 2007-12-15 17:20
-
superstoned
The amount of ‘control’ Trolltech has over the toolkit (which is, after all, GPL) is limited. Forking isn’t necessary as long as they’re going in the right direction… Forking just because you’re afraid they might once do that seems rather silly to me. Let’s call that a Bogus Argument. We can fork anytime, if the need arises. Matter of fact is, the cooperation between Qt and KDE is improving (see the LGPL Phonon code dump from Qt in the KDE source tree as a good example).
And the change of licence – I wouldn’t want that anyway. I don’t want ppl to develop proprietary apps which take away others freedom without any contribution to the community. Same reason I dislike the BSD. For FOSS develpers, it won’t make a difference if the toolkit is GPL or LGPL, and for commercial developers – they can build anything they want for inhouse use or GPL. If they feel they need to be anti-social, they can pay TT (which still benefits the whole FOSS ecosystem).
In other words, I would really oppose the second thing you ask for, and the first one is covered already (there’s the Qt-KDE foundation which can release Qt into public domain if TT goes evil).
Maybe you should consider developing on the KDE 4 infrastructure 😉
-
Hiev
The amount of ‘control’ Trolltech has over the toolkit (which is, after all, GPL) is limited.
They control the versioning, the licensing, etc. that’s enouth for me to ask for an deattachment.
And the change of licence – I wouldn’t want that anyway. I don’t want ppl to develop proprietary apps which take away others freedom without any contribution to the community.
That’s your point of view and I respect it but not everyone thing the same, sorry.
In other words, I would really oppose the second thing you ask for, and the first one is covered already (there’s the Qt-KDE foundation which can release Qt into public domain if TT goes evil).
I wouldn’t consider any of both coverd meanwhile there is plenty of room for GNOME/GTK.
Edited 2007-12-15 17:43
-
2007-12-15 10:07 pm
segedunumThey control the versioning, the licensing, etc. that’s enouth for me to ask for an deattachment.
Novell control Mono (and you sign over copyright), and effectively, Red Hat controls GTK through their contributors – despite the license. It’s not completely black and white on that front. Yes, you can fork them, and you can fork Qt as well or create a completely separate Qt-compatible version (Harmony?), but there has to be some weight behind it for that to work.
However, I find this stuff quite interesting. Just how does a license affect the direction and quality of your code and how it will be used? When the people behind KDE chose Qt, and when Trolltech chose dual-licensing for Qt, they took the view that it was important to give free desktop software developers the best toolkit they could lay their hands on to produce good quality software. Proprietary developers’ considerations were way down in the pecking order, and if they want to develop good software for themselves then they can pay for a good toolkit so that Trolltech had the resources to put into keeping Qt ticking over at a fair rate of knots – benefitting open source developers further. That’s really how Qt has benefitted, and continues to benefit, KDE. If you want to create a quality free desktop to go up against what we have today, open source developers have to come first. That’s the basic premise.
It’s not about ‘non-GPL applications are evil‘. It’s about practicality for open source desktop development.
Don’t get me wrong. It’s great that we have development toolkits of different licenses for different purposes (one size doesn’t fit all), but creating a desktop is a complex thing where you have to take into account end users, and also creating usable APIs and libraries for open source developers to do their stuff. There’s a lot of reuse going on.
Because KDE developers have Qt at their disposal, they tend to be quite anal about having nice APIs for themselves. If they have to deal with something less than perfect then they tend to kick up about it. You can see this with stuff like Phonon, but many of the GStreamer folks couldn’t understand why they were doing it.
If I was to map out Gnome 3 today, I’d mandate a framework to develop the Gnome desktop on. This would either be the Mono framework, or a set of Java classes. My preference would be for Java, since you can compile to a JVM or compile to native through GCC should you want – unlike Mono. I’m not convinced everything can run in a VM, and language neutrality in Mono and .Net is worthless. Underneath this would be libraries such as GTK and Cairo, and to be honest, interfacing with this stuff directly as a developer fills me with dread. As a developer, Gnome just strikes me to have an awful lot of layering violations to get around some shortcomings, and put into that context, it has a long way to go.
Edited 2007-12-15 22:08
-
segedunum
The founding Novell and RedHat put in GNOME or GTK is nothing compared to the money behind Qt, Trolltech have more men power on it, more hours per developers, etc.
This story seems to change depending on the context.
From what we’ve been told, GTK and Gnome development in general has all these big enterprise players behind them with lots of money and resources to put in. Gnome is supposedly the default desktop everywhere, and this means that we’re all stuck with Gnome if you believe various people, with the big three of Red Hat, Novell and Sun behind it, with HP and IBM supposedly backing it as their default (that word comes up a lot) desktop.
I see absolutely no evidence whatsoever that any of that is true, and we’ve been told this for nearly eight years now. GTK is mainly maintained and kept ticking over by some Red Hat people, but as far as Gnome is concerned, pretty much all of the useful, different and innovative ideas are coming from individual contributors who just had a eureka moment and not from whiteboards in some company somewhere. That should tell you something.
-
Hiev
I see absolutely no evidence whatsoever that any of that is true, and we’ve been told this for nearly eight years now. GTK is mainly maintained and kept ticking over by some Red Hat people, but as far as Gnome is concerned, pretty much all of the useful, different and innovative ideas are coming from individual contributors who just had a eureka moment and not from whiteboards in some company somewhere. That should tell you something.
Sadly yes, GNOME developers have to worry about both things, developing of is own toolkit (GTK) and the desktop (GNOME), all that we so litle resources.
KDE developers by the other hand just need to worry about the desktop because the toolkit(Qt) is already been developed by a corporation like TrollTech.
Both have advantages and disadvantages.
Edited 2007-12-15 17:00
-
superstoned
I know I’m walking a thin line between expressing my opinion and trolling. But I’m not, as the Urban Dictionary states it, “Being a prick on the internet because [i] can”. I honestly believe this to be true, and I even think it hasn’t been said often enough. Most ppl don’t perceive a gap between the capabilities of KDE and Gnome, yet I’m sure there is one – even between the 2 years old KDE 3.5.x and the latest Gnome. Of course that goes mostly for the technical part of those DE’s, you can patch up the gui to make up for technical and architectural deficiencies – MS has been pouring billions in Vista to do exactly that, and I believe it is the same situation with what Red Hat, Novell and Cannonical are doing (though not to the same degree by far, obviously).
I of course do know many do disagree (though I rarely see someone with a good argument or example) – and saying this does often infuse some discussion – sometimes even bordering on a flamewar. But esp the latter is and has never been my intention.
Anyway, lets give an example. As a non-programmer, I’m not always right on technical matters, but I’m pretty sure DCOP, created by the KDE developers many many years ago and still in several aspects superior to D-BUS (only adopted by KDE because unification would be good for the Free Desktop) sounds like a prime example of something KDE has been years ahead in.
(of course, system-wide spell check is still something pretty unique in KDE, though I believe Apple has something similar as well. Poor them, still behind, KDE 4 adds grammar and automatic language detection to that).
Actually, just having a look at a recent ‘cool thing’ in the Gnome community often only shows how they’re behind – for example the ‘cool ARGB visuals’. Nothing new to see there, has been KDE/Qt technology for years (and is used in KDE 4 a lot)…
-
Hiev
The fact that KDE is better in some aspects doesn’t mean GNOME GTK can’t do the the job, GNOME and GTK are great for developing too, maybe w/o all the bells and wis. of KDE but GTK is really good, as good that has convinced many developers to work on it.
Another great toolkit is wxWidgets, is so easy to work with it. I recommend it for multiplatform support.
-
superstoned
I’m not saying GTK is bad – just that KDE is a more modern and efficient and complete development environment, and allows for quicker development and deployment of multiplatform applications than any other toolkit… How’s about that?
Of course it depends on the app you want to write – C++ isn’t suited for everything, sometimes python and java are better. But there is PyKDE and QtJambi. In other words, if you would decide to develop on something for solely technical reasons, I think Qt/KDE would generally be the smartest choice. If you think that’s not true, I would love to see some reasons why – as I rarely see someone write something like “hey, this is done much cleaner and faster in WxWidgets than in Qt” or something. And as such information after all forms the basis of my arguments (I can’t form an opinion by working with these toolkits myself) it would help me say the right thing.
Thanks in advance.
-
Hiev
if you would decide to develop on something for solely technical reasons, I think Qt/KDE would generally be the smartest choice.
That’s something I could agree, But I and my employeer have to consider other things of matter besite only technical reasons, if toolkit A and toolkit B take you to the same results then we need to evaluate other aspects where Qt/KDE not always win.
as I rarely see someone write something like “hey, this is done much cleaner and faster in WxWidgets than in Qt” or something.
It doesn’t mean it can’t be done in a clean way with wxWidgets, wxWidgets is easy to read and write, I develop with it and for my experience I can tell you is understandable and clean.
Edited 2007-12-15 17:14
-
superstoned
That’s something I could agree, But I and my employeer
have to consider other things of matter besite only technical reasons, if toolkit A and toolkit B take you to the same results then we need to evaluate other aspects where Qt/KDE not always win.
Of course, that’s obvious. May I ask what, besides costs, makes the choice for Qt/KDE worse?
It doesn’t mean it can’t be done in a clean way with wxWidgets, wxWidgets is easy to read and write, I develop with it and for my experience I can tell you is understandable and clean.
Again, of course. For many things I can imagine it doesn’t really matter what toolkit you choose – esp for small apps the influence on development time from the toolkit is not significant. But for a big application I would find it hard to believe someone did a honest check of all options and choose GTK over Qt… Unless of course the app needs to be proprietary and you don’t want to pay at all – even if that makes the whole thing in the end more expensive. In that case, you deserve a painful time, as you should’ve used the GPL 😉
-
Hiev
even that case, you deserve a painful time, as you should’ve used the GPL 😉
That’s a big problem, to thing that some one who doesn’t want to write GPL software and choose propietary deserves punishment is what is wrong. That0s what I like LGPL is fair with propietary and open source.
-
superstoned
Well, I agree with the FSF on the stance the LGPL is wrong and shouldn’t be used. Non-GPL apps ARE evil, they take away your freedom and should not be used nor written. I won’t shot someone who does so, but I would try to convince him or her that what he/she is doing is a BAD thing and hurt him/herself and others’ freedoms.
In a world where speech depends on Software, Free Speech depends on Free Software.
Writing non-free software only strengthens the proprietary ecosystem and makes it easier for others to restrict ppl’s freedoms. Sure, working on one piece of proprietary software for a proprietary OS won’t hurt much. Just as throwing one can of oil into the ocean doesn’t hurt it much.
Imagine everyone doing these and you know why I consider non-GPL apps bad by nature.
Again, I’m not religious in this, I won’t fight for it and I use flash myself. But I would never dare to argue for LGPL-ing Qt or any other library. As long as there is a way for a proprietary-software-developing company to use the software (eg paying TT) I would argue against it.
Of course, KDELIBS is LGPL and I support that. But if Qt was LGPL too, I might even argue for GPL-ling some fundamental parts of kdelibs and asking a (financial or otherwise) contribution for those who want to freeride and make money by writing proprietary, freedom-limiting applications using KDELIBS.
With the option to one day forbid them using it fully.
A world with only GPL software is what I would prefer, and everything which is a step away from that (LGPL GTK is one such thing) I’m against.
-
Hiev
Like I said that’s your point of view, but not all thing the same, like me and maybe others.
Besite that and other reasons we are glad GNOME and GTK is well suported in nice distros like Suse,Ubuntu RedHat and Fedora.
-
superstoned
Well, I would take any FOSS over any proprietary piece of software any time – in other words, of course. I’d rather see Maemo take over the mobile phones than MS, even though I would prefer Qtopia…
-
Sophotect
I know I’m walking a thin line between expressing my opinion and trolling.
I agree to every single point of your statements. Not so much from a technical point of view, but instead from the experience of using the different alternatives. After trying and using them (all?), which was a question of eating ones own dogfood, i settled for KDE which does more or less everything i need and want the way I want to.
Edited 2007-12-15 17:13
-
superstoned
Well, even though I’m happy you agree with me, it would still be possible that I was trolling even when you agreed 😉
Anyway, as for the final state of the UI, I can see why THAT would depend on preference. Both Gnome and KDE do things right and do things wrong, and both are generally as of now (KDE 3.5.8 and xxx) much better than win XP, imho. It’s the underlying technology, the ambition of the community and the speed of development they differ in – and in each of those, imho KDE clearly has a lead.
-
superstoned
(xxx should be 2.20, the latest right now. But I don’t think 2.22 would make that much of a difference)
-
Hiev
Im afraid you are wrong wit that. 2.22 will have some dramatic changes in infrastructure, the roadmap for 2.22 can give you the idea.
-
superstoned
(just have to say this: I read:
Printing multiple images per sheet
for Eye of Gnome. Again – wasted development time due to a sucky infrastructure, isn’t it? EVERY KDE app has been able to do something like that for like 8 years… blegh. suckers.)
sorry about that. There is interesting stuff in there (eg for evolution – if they also fix the bugginess everyone complains about, it might become usable).
gvfs is of course interesting, seeing Gnome catch up with KIO must be a godsend for many Gnome users. It of course screams for a ‘6 years late’ comment, but hey, let’s rejoice
I guess the Pango and Glade stuff is also interesting, but it’s not too much, is it… I mean, 90% of the changes are again in the apps. That might have been so too for eg KDE 3.4 and 3.5, but at least there was a lot of investment in infrastructure before that (and now, of course). I don’t see these things as getting ahead, it’s still catching up.
But yes, you’re right, 2.22 doesn’t seem to become as small an update as 2.20 was over 2.18…
-
Hiev
The “my toolkit already is doing it” is inmature, sheer your self that both can do the same now.
-
superstoned
It’s not that ‘my toolkit’ is doing it, it’s the KDElibs which have been doing that. There is much more to KDE that Qt, as I said before, even though KDE builds upon Qt.
Anyway, yes, it might be immature, and I said sorry already 😉
My point was and is that GTK/Gnome are way behind… But your point was that 2.22 is working hard to remedy some architectural stuff, and it indeed seems that way. I’ve heard and read more architectural stuff lately from the Gnome camp, like a Decibel-like thing and an Akonadi-like library etcetera. More attention to the backend, a good thing…
-
2007-12-16 2:43 am
anda_skoaMy point was and is that GTK/Gnome are way behind…
I am sorry, but I agree that phrasing it like this basically is trolling.
The GNOME software stack is following a different style, i.e. lightweight libraries, so on a first look from someone used to the KDE software stack style, i.e. lightweight applications, it might look weird, even insufficient, but it isn’t.
An interesting observation on recent developments is that implementing functionality in out-of-process services fits into both styles rather nicely, i.e. the client library can be lightweight since it is basically just an abstraction of some protocol and the application can be lightweight because a big portion of the resulting functionality is “below” the application.
-
superstoned
What you’re saying is that KDE has heavy libraries and light applications, and Gnome does the opposite. Isn’t that the same as saying Gnome doesn’t have as powerful a set of libraries as KDE has?
You claim it’s by design, Gnome WANTS small libraries, and apparently prefers to duplicate functionality in the apps, while KDE pushes functionality down to the libraries?
I think it’s just a lack of developers interested in working on the libraries and a general culture of ‘cooperation and sharing be damned’ or something.
It’s like saying the lack of features of some application is “by design, we wanted to make it usable”
– I call bullshit. Just say you’re not ready, you’re behind and you need more help from the community and other developers…
(now I know there are features you could drop to make the app more usable – but for many features that argument is bogus – for example, Evolution introducing a warning when someone forgets an attachment – that’s simply 4 years late, period. Now that’s a small thing and judging an app just because it was behind on one small feature is stupid, but I think my point is clear).
-
anda_skoa
Isn’t that the same as saying Gnome doesn’t have as powerful a set of libraries as KDE has?
No.
As I said, coming from a KDE point of view and its respective development philosphy, one can get that impression.
Being a KDE developer myself, I can’t really estimate how this model work, but obviously there are application developers who really appreciate it.
Probably more fine grained dependeny control or not getting complains when you do things differently.
You claim it’s by design, Gnome WANTS small libraries, and apparently prefers to duplicate functionality in the apps, while KDE pushes functionality down to the libraries?
I think one could phrase it like that. GNOME wants lightweight libraries and acknowledges the tradeoff of potential duplication in the applications. KDE wants to have shareable parts in their libraries and acknowlegdes that certain parts of them will probably only be used by a small number applications (e.g. the “at least two applications” rule for inclusion)
Anyway. The problem with using terms like “less powerful” or “way behind” is that there is no canonical method for measuring them.
-
superstoned
Hmm, well, I get your point – it might just be hard to understand. Having to do more work to get the same consistent environment with the same features – I guess it’s attractive to some 😉
I see there’s a tradeoff – if you only start 1 or 2 apps, KDE is rather ‘heavy’ – while Gnome does fine. As soon as you start a bunch of apps, KDE gets relatively smaller than Gnome. Depends on what you prefer.
Interesting, I never figured it was on purpose. Of course it is still possible it wasn’t on purpose and they just told themselves it was to resolve the cognitive dissonance – but hey, all the same to me.
Tnx for the insight.
-
2007-12-15 4:42 pm
DeadFishManI don’t know, but that didn’t sound like an attack to me. GNOME indeed is behind KDE in so many fronts that is not even funny and superstoned stating it is not exactly trolling. Perhaps, the only thing that GNOME has going for it is it looks.
I’d be glad to be proved wrong but wake me up when you can change all the shortcuts of all menu items in your GTK apps, use DCOP/DBUS commands to customize the behavior or create small “macros” to automate GUI tasks or even define application-specific window behavior on a per-app basis.
-
segedunum
Superstoned, please stop trolling. I normally don’t feed the trolls.
Ahhhh, the emotive question of open source desktops. You can only go so far before pointing out the truth, and when you do, others are simply not going to like it -).
There is a fair bit of substance in what he says, in that when you read the changelog you see an awful lot of stuff that has gone before. Novell’s Gnome isn’t even on a par with Ubuntu’s, and that has a lot of work in front of it. Apart from the SLAB menu that no one likes, there’s pretty much nothing coming from Novell on the Gnome front. The huge wave of Gnome/Mono development simply hasn’t materialised either. You can’t just wave that away as trolling.
In view of that, one is legitimately entitled to ask questions. We’ve had this before with Novell when they shafted WordPerfect, so this isn’t a new phenomenon.
Please enjoy your desktop of choice without making gratuitous attacks upon others’ desktop choices.
I just wish that people would stop telling me that because Gnome is apparently the default on all these distros, and it is supposedly this enterprise desktop, then it doesn’t matter if it is inferior. That’s certainly what is being inferred. Whenever you chat with someone about this, when there’s nothing more to say you get the inevitable comment “Gnome is the default” as a response. It’s pretty bizarre. Thom even made that comment to me once, just to show how endemic it is. Yes, when you look at the development tools, libraries, infrastructure, the applications built on them and the feature changelog, Gnome is simply inferior.
Sadly, an awful lot of people in the open source world, some of them developers, are just kidding themselves. Gnome is a good desktop and a great achievement, but pretending that you can persuade Windows and Mac users and developers to move based on it being the default on many distros (translation: you’re stuck with it), telling everyone it is simple (translation: we don’t have the features you have on Mac and Windows) and telling them they can develop for nothing (translation: our development tools and libraries aren’t good enough) is well, well, well wide of the mark.
The sooner we can all admit that, the sooner we can get past the mistakes of Unix on the desktop. Yes, Microsoft was monopolistic and aggressive about getting their desktop as the default for OEMs, but users and developers didn’t go for Unix desktops because they were crap and were bogged down in politics on whose standards were the default.
-
2007-12-15 11:58 am
RIchard James13
On the flip side of the coin, Red Hat has long since been usurped as leaders on desktops. There was Slackware, then Gentoo and now Ubuntu.
Should be “usurped as leaders in pandering to popularity of the in crowd”
I see Redhat->SUSE->Mandrake->Xandros kinda happening but not the distro’s the author says.
Apart from that the article is pretty good. I think that Red Hat has the power to take on the Desktop but I don’t think they see the money in it. Novell needs to get it’s marketing head out of the sand and Canonical just needs to keep on doing what it is doing now.
Though for most companies the Desktop is not really where the money is yet. It might be possible to improve on that but if I was running one of those companies I would not let the server/embedded/portable side of the business get away from me. There is too much money in those areas. That money can be used to support Desktop efforts but I would not invest too heavily in it.
As for what I think needs to be the big revolution in the Linux Desktop in 2008?
It is looking at a customer who does not get much attention. The person that builds integrated systems for businesses so that their business logic is automated. Microsoft is so entrenched in this area it would take a lot to change that. These people look at Visual Studio, MS Office, MS SQLserver, .NET, IIS and MS Windows and see how they can build a system for businesses. Unfortunately Linux can be quite hard to do some of the simple integration that you would expect. For example connecting a Database to a User Interface is hard to do in Linux but easy if you use MS Office or Visual Studio and MS SQLserver. I believe these areas need to be worked on in Linux.
Personally I believe that the way most Office Suites are built for Linux is wrong. They are monolithic items. It would be much better if someone wrote Office functionality that was made up of small components. Then you could pick and choose which components you needed to get a specific job done.
-
2007-12-15 12:25 pm
raboofI believe that the way most Office Suites are built for Linux is wrong. They are monolithic items. It would be much better if someone wrote Office functionality that was made up of small components.
I see your point – but this is not limited to Linux Office Suites, it applies to all Office Suites out there.
It seems to me that in this respect the Linux offerings are ahead rather than behind – but indeed still have a long way to go.
-
superstoned
Actually, you both are wrong. There is an office suite on linux which does what the parent asks – it is very modular. KOffice, the most innovative office suite in the FOSS world. Have a look at how KOffice 2 is turning out – and be surprised by what FOSS can do with amazingly little resources.
-
RIchard James13
I see your point – but this is not limited to Linux Office Suites, it applies to all Office Suites out there.
Yes that is right. What I was trying to say was to get more competitive the Linux Office Suites need to be broken down into components just like Microsoft’s concept of object embedding but far far better. With open source a programmer could just take the objects he needs from the Office Suite look at the source code and Documentation and then plug all those objects together in a fashion to make a new program that performs some business logic.
An example I think of is say you have Email come in that has written in it some important business information that you want to put in a database. So the programmer gets a Email object that can talk to the Email program and adds a GUI for the user with a object to interface into the database. Whenever the user receives such an email they just select something from a menu and it pops up the GUI which lets them choose to add it to the database. Currently that sort of integration is hard to do and is an area I think that Open Source could take on Microsoft.
Also currently there is no suitable Access replacement for Linux. Most people diss on Access because they think of it as a database but Access can be told to use another database backend. Access is actually a product that makes it really easy to create a database front end for end users. You can also achieve the same with certain widgets inside of Visual Studio. You can do this in Linux but at the moment it is much harder. Trying to run Kdevelop Designer and using the database widgets is just an exercise in frustration, it works in Visual Studio.
Another thing people forget is that the system has to be easy to program. C++ is hard for most people and most of the people doing the programming for SMB can’t and should not have to learn an esoteric language just to build a simple application. This is where Microsoft with Visual Basic is hitting Linux hard. Linux has the easy languages but not the integrated solutions for these programmers.
These programmers are locked into Microsoft’s world because Microsoft offers them a integrated solution package that Linux does not. These programmers make most of the worlds software and Linux needs to make more headway into their world if it wants to compete with Microsoft.
Not Games, Not Hardware. The poor widdle Visual Basic/ MS Office programmers. Win them over and Microsoft will lose badly.
-
DeadFishMan
The poor widdle Visual Basic/ MS Office programmers. Win them over and Microsoft will lose badly.
Although not as mature as Visual Basic, Gambas2 is an exceptionally good and OSS Basic IDE that is somewhat similar to Visual Basic in many aspects and it amazes me as how people tends to overlook it. The amount of widgets that it can use, including OpenGL wrappers, data-aware widgets that can connect to MySQL, PostgreSQL, SQLite and other databases through ODBC, is huge and it keeps increasing all the time and the development mailing lists seem to be reasonable active.
From the IDE itself, one can not only compile the executable file but also to create distribution packages. I’ve been playing with it for a few weeks and came out quite impressed with the results.
It is also worth to mention that GB2 not only produces GUI applications using Qt, but it also can be set up to generate GTK apps and even console applications so it has all bases covered. There are even people out there claiming to have successfully migrated small VB projects to GB2 with some effort.
Besides, there is also RealBasic which is another VB-like IDE but proprietary. I have no experience with it but heard good things about it.
Don’t know if Kylix is still around but Lazarus + FreePascal offer a good alternative for the Pascal and Delphi developers out there, although Lazarus unfortunately is stuck with GTK1 for God knows why. Those should cover the needs of developers that slap some widgets into a form, double-click them, pick an event and then enter some code into it perfectly.
Not to mention that both Eclipse and NetBeans are available for Linux for the Java nuts out there, and since Java remains as the most popular language these days, it is a good thing that good IDEs are available to its developers.
So in a nutshell, not only C and C++ developers can have their way in Linux as you can see. I think that inertia does indeed have a role into this situation and unfortunately that’s much more difficult to overcome than any technical hurdle.
-
lemur2
Personally I believe that the way most Office Suites are built for Linux is wrong. They are monolithic items. It would be much better if someone wrote Office functionality that was made up of small components. Then you could pick and choose which components you needed to get a specific job done.
Do you mean like this: http://www.koffice.org/
“The KOffice suite includes the following applications:
Productivity Applications
KWord – A frame-based word processor that can work in two modes: page oriented or layout oriented
KSpread – A powerful spreadsheet application.
KPresenter – A full-featured presentation program.
Kexi – An integrated environment for creating databases and database applications.
Creativity Applications
Kivio – A Visio®-style flowcharting application.
Karbon14 – A vector drawing application.
Krita – A layered pixel image manipulation application
Management Application
KPlato – An integrated project management and planning tool.
Supporting Applications
KChart – An integrated graph and chart drawing tool.
KFormula – A powerful formula editor.
Kugar – A tool for generating business quality reports. ”
No less than 11 components to get a specific job done. They use some common libraries, but you can install (or not) any of those 11, although the last three don’t have much point without the first four.
2007-12-15 12:40 pm
moleskine
It’s hard to see what Red Hat would gain by bringing Fedora back into the RH fold. My understanding is that the present system works pretty well while preserving Red Hat as a premium brand. Fedora keeps people inside, under the Red Hat system but for free. Red Hat itself keeps them but on pay-for. CentOS is a kind of halfway house for the smaller outfit: it’s free but support is DIY (which will still cost you in time if nothing else). So with this arrangement, Red Hat have all the bases covered.
I’m sure Novell think along similar lines with SuSE and OpenSuSE, though they lack a CentOS equivalent and the tragedy of Novell is that the market thinks it’s lacklustre (think stock performance) and their financial situation means that energies which could be used elsewhere probably have to go to firefighting. It’s still too soon to say whether Ubuntu will last the course. I hope it does, of course, but these are early days: come back in ten years.
We should be grateful to people who speak up when there’s a problem. If no one did, a lot of things would never get fixed. How they speak up really isn’t of much interest.
Those who think Debian is old and frumpy are talking cliches and marketing perceptions. There is no objective reality to this idea at all. Choosing or rejecting a distro on the basis of prejudice isn’t the way: stick to what works, in your experience and for you. The rest is hot air.
2007-12-15 1:36 pm
Googol
by that I mean: In commercial terms, RH and Novell make money on the servers, and to some extent on the enterprise desktops. But Ubuntu is only big with the free as in beer crowd. Does Ubuntu have any certifications on the server, does it secure noteworthy deals on the corporate desktop generating revenue?
You don’t just “penetrate” the server market by releasing a server version of your OS. There is a lot of expertise at RH + Novell and industry relations which you cannot simply buy with money. Let’s see how companies like IBM feel about supporting/certifying yet another flavour – I don’t think they will.
2007-12-15 4:19 pm
shaga
First, I am not sure about the “is nothing”. Trolltech is not that big. Second, it is a funding into toolkit, not KDE, and there is a long way from qt to KDE.
And to prevent one kind of reply – I know that Trolltech employs some KDE developers.
-
2007-12-15 7:33 pm
RHCE07
I believe Red Hat will release the ‘Global Desktop’ next year in which they should have been developing this all along.
I am not sure about Novell, it would be nice if they really marketed SuSE SLED preinstalled on Dell’s, Hp’s and other OEM machines.
Ubuntu already has a preinstalled config from Dell and they have taken the desktop market to heart. I tried out Ubuntu but I am not familar with Debian based distro’s. I felt out of place with it installed, since really the only distro I am familiar with is Red Hat since the 6.0 days to now Fedora 8 I may try it again in the future however at work we use Red Hat and I have Red Hat Server 5.1 installed on my laptop right now.
I really have become to like Fedora, yes it is ever changing and quickly but it seems to have some really cool stuff in it and most of it or some of it ends up in the RHEL release. The community of Fedora is growing and I forsee larger foundation of users with it. But like I always stated with OpenSource the ‘option’ is very powerful if someone wants to go one way with their choice of distro they can and come back to the one they left. I really like the fact more distro’s pop up and new ideas bring an entire fresh breath to the movement instead of focusing on ONE such as in Microsoft Windows world.
I would really like to see Red Hat PUSH and MARKET a desktop distro and make is widely known. Plus, revamp their website, I have to use the RHN on a regular basis and it is painfully slow and just hard to use. The product descriptions are vague and one would think they are not really interested in pushing their products. Oh well, nothing is not perfect but from a technology company one would expect a better site, FASTER and have bandwidth capable of handling the request made on it.
Linux Distro (x) will make more headway just by the sheer number of people converting over, which enables the desktop distro’s to grab hold of the market share little by little. Overall I am not sure what the stats are but the overall user base is growing, and MS cannot fight this user base, because it is not a static model, this is how attrition battles are won and lost. I am currently testing the Red Hat Cluster Suite at work, and I have to say some of the graphical tools they have evolved are really nice. Some people vow against them but when they work as well as the ones for Production in RHEL releases I have to say having X-windows is not such a bad option on install anymore…
-
segedunum
I believe Red Hat will release the ‘Global Desktop’ next year in which they should have been developing this all along.
I’m at a loss with this global and online desktop thing. It just strikes me as another fad for the sake of another fad. Either you have something to go against Mac and Windows or you don’t.
-
RHCE07
I’m at a loss with this global and online desktop thing. It just strikes me as another fad for the sake of another fad. Either you have something to go against Mac and Windows or you don’t.
It is not an online app, the way I understand it more like the SLED version of their Enterprise desktop offering.
http://www.redhat.com/about/news/prarchive/2007/global_desktop.html
The biggest difference in running a Linux distro (in 6 months the Linux distro is still running good). The Windows machine will have spyware/malware, running poorly all clogged up with junk and will need to be wiped and reloaded.
What is ironic is these companies are sitting on a goldmine if they would market their product to a wider range. In the Enterprise I can do everything at work running Fedora 8, Evolution Exchange connector (email), Krdc for remote desktop, FireFox, OpenOffice, Evince PDF, MPlayer, Xine, VLC, Real Player, Flash, I mean you name it and it works perfectly. I have RHEL5.1 Enterprise Server on my laptop and I can perform all of my administration from it. It is just good business sense, no viruses, SELinux (default), IPTables, SETFACL on file systems, Windows does not even compete in the same league for being locked down.
2007-12-15 10:38 pm
cmost
It seems to me that everyone keeps forgetting about this little distro that can! All one ever hears about is Ubuntu this and Ubuntu that… Ubuntu ain’t the king of distros anymore. Its slide from the top can be attributed to its accumulation of bugs that persist from version to version and that its devs are slow to fix. Canonical also has a fixation on adding ever more bling (read bleeding edge features that elicit oohs and ahhhs) rather than bug fixing and stabilizing the distro as a whole. It also doesn’t help that Ubuntu simply “forgets” about previous versions when a new version is released (sure it backports security updates, but one can forget about new versions of stock packages.) Fedora, SUSE and Mandriva seem to excel in all these areas. As does PCLinuxOS. I hope this distribution gets more cred. Before the Ubuntu fanbois mod me down, let me just say that I am an Ubuntu user (by way of Mint.) I just happen to have a lot of respect for PCLinuxOS because it’s a damn good distribution for home and small business users alike. It deserves more recognition as such.
Edited 2007-12-15 22:43
-
2007-12-16 1:58 am
aitvoIf all one ever hears about is Ubuntu this and Ubuntu that, it’s obvious that it IS the king of distros.
You yourself don’t even use PCLinuxOS, yet you call it king?
please.
Edited 2007-12-16 02:00
-
cmost
I never said I didn’t use PCLinuxOS. I have been using it on my file server for the past two years and it’s also on my laptop. Oh, and not to mention that it replaced your precious Ubuntu on Distrowatch’s number one spot awhile ago. Don’t assume anything honey.
Edited 2007-12-16 03:22
-
aitvo
ZOMFG teh page hit ranking, it must be true!
http://distrowatch.com/dwres.php?resource=major
No where close to first in the top 10.
Thanks.
Edited 2007-12-16 15:33
2007-12-16 2:56 pm
rklrkl
As others have pointed out, several references to “Fedora Core” are just 100% embarrassing to the author of this article. The last Fedora Core (version 6) was released over a year ago and has since been superseded by two Fedora “non-Core” releases (7 and 8) and this fact alone severely reduces the credibility of the article.
If the Red Hat suggestions by the author are correct, he seems to be implying that Fedora as a free OS is scrapped (at the moment Red Hat “sponsor” it and contribute significant developer time to it) and its name is changed (e.g. Red Hat Desktop or something) and it’s then charged for.
What this fails to realise is that doing this would surely reduce the non-Red Hat contributions to the new “Red Hat Desktop” as non-Red Hat developers will be less inclined to contribute to a commercial-only Linux distribution.
Also, what’s stop the CentOS team (or some equivalent of them) just repeating the trick they did with RHEL and produce a clone version of Red Hat Desktop? I found this article to be just another analyst-style musing with very little substance to back it up – barely worth linking to from OSNews, IMHO.
-
gilboa
/+1. The article, or at least the RH part (I didn’t really wasted additional time on reading it once I finished the “Fore Core” (?) part) was simply embarrassing.
As for CentOS, doesn’t it already include a mix of all RHEL server and RHEL desktop/workstation applications? I doubt that RedHat’s global desktop will be that different from their current desktop offering?
– Gilboa
Fedora Core was changed to just plain Fedora TWO releases ago. If the Author has not even looked at the Fedora site for the past 8 months then I guess they don’t really know what they are talking about. Most people I do business with know very well that Fedora is the test bed for RHEL and some are even glad that there are two brands. This means that the technically semi=literati people in their companies don’t press for the adoption of Fedora in productions systems.
As for Novell & SUSE. They have been very weak at marketing ever since the did the original deal with the Devil. Personally, I have never felt comfortable wih SUSE as a distro. There were just to many little quirks for my liking..
Then we come to *buntu. Here the author (IMHO) again showed their lack of research. Why should Canonical ‘Create an Active Directory like system’ when there are several GPL’d Directory Servers around that could easily replace the AD Monstrosity that many companies are lumbered with. Why didn’t the author suggest using one of those if it was so critical to the ongoing success on *buntu.
My own problem with Ubuntu & its derivatives is that its trying to be everything to everyone. RH found that this didn’t really work back in the days of RH 6,7,8,9. Canonical make a great desktop product. I think they should really stick with that but (personally) please change the default colour scheme of Ubuntu. Kbuntu had got it right.
I like this line:
Canonical needs to leverage this loyalty and create a vertical initiative that will provide even more features to its desktop users as long as the servers said users are connecting to run the Ubuntu OS.
I’m having trouble parsing what, if anything, the author means here. Maybe it’s some sort of Web-2.0 SaaS kind of thing. That stuff’s super-cool with the kids and their MySpace.
Regardless, this sort of uninformed, unsubstantial, and just plain incorrect drivel deserves not too much more than a combination of polite silence and perhaps a solitary simulated fart noise in the distance.
Edited 2007-12-15 11:17
Hey, the blog has the word “enterprise” in it so what did you expect? Research? Competence? Clue?
It’s not for nothing that the word “enterprisey” is a derogative term.
I expected starships, personally.
I think vertical initiative is marketese for expanding your existing market, instead of say breaking into new markets. This might imply that by further integrating their products (Desktop/Server) they can increase their user base, customer base, market share, profits or whatever you want to call it.
Being a highly community oriented Distro that sort of thinking at least from a marketing perspective makes sense.
Syntactically, “said” in this instance is an adjective, not a verb. So it’s not the servers said but rather said users, i.e., the “users just referred to”.
Semantically, the sentence means that Canonical should attempt to use their desktop success as a means of gaining a foothold in their users’ server rooms, and lock them in by giving their server customers additional desktop features not available to those who use the desktop version only.
Dubious advice IMO. Just leads to alienation of their existing base.
It’s pseudo-MBA-speak meaning, basically, that Ubuntu needs to leverage vendor lock-in if they expect to truly succeed as a free OS.
And as such, shows about as much insight as you’d find from the kids and their MySpace…
Ubuntu server? Isn’t that called Debian?
They already have a product out there. If what you really mean is a marketing machine like ubuntu’s for the server good luck.
I think the current situation is just fine. RedHat for servers and Ubuntu or Fedora for toys/desktop. whats wrong with that? Ubuntu had a different goal and they’re doing pretty well at acheiving that goal.
Debian is old and frumpy, Ubuntu is young and hip, like the apple dude in the commercials. I thought everyone knew that…
If you were to go get a cluebat then bury it six feet deep and forget about it for half a lifetime, then, maybe, you’ll be coming close to a level of cluelessness I imagined right there reading that.
Edited 2007-12-15 10:43
actually, I was trying to be funny.
nobody ever got fired for buying IBM.
I think you fell into the sarchasm.
l3v1, once you unearth that clue bat, hit yourself in the face with it about 10 times. Then perhaps you’ll recognize sarcasm.
Debian is old and frumpy, Ubuntu is young and hip, like the apple dude in the commercials. I thought everyone knew that…
Right. If you prefer your server software young and hip, funky and experimental, fully updated & reinstalled every 6 months with not much time spent for testing the stability and the quality, then Ubuntu must be your choice.
But seriously — Red Hat, SUSE, and Ubuntu are the leading commercial GNU/Linux distros and, hence, they always get most of the attention in the commercial media. But there are also several non-commercial distros that are very much worth checking out as well: Debian, CentOS…
I was actually just making a joke. The parent asked why in the world you would run an “Ubuntu server” when you have debian, unless you need the massive marketing and hype.
And I was just trying to clarify your joke. 😉
BTW, have you noticed that Ubuntu now officially acknowledges Debian’s importance to their own product: “Debian Unstable is the rock upon which Ubuntu server is built.”
http://www.ubuntu.com/community/ubuntustory/debian
I think it’s absolutely great that they’re so honest about it. 🙂
Ubuntu is leading and more or less a low-quality offspring of Debian.
Just a heads up, the parent certainly tripped my sarcasm filter, perhaps you should check you levels before replying…
I guess I need to be more clear in the future 😉
Or refrain from posting things on OSNews that might require a slight modicum of subtlety to “get.”
No, it’s just a packaging error. I often find that package sense-of-humor conflicts with package ubuntu-desktop, and must be removed for installation to proceed.
Uh-oh, did I actually say that out loud…
Exactly, Ubuntu is doing the latest desktop features for the new user or Windows users who want to switch. Debian offers nothing to them and it’s only till resent they’ve got a GUI installer.
It’s all about mind set, it’s hip and cool to have Apple stuff, Apple IS cool just like Ubuntu is getting that same vibe in the linux world. Novell is being seen as a Microsoft like company in with the corporates, yet not managing to distance itself somehow in the mind set of people.
Ubuntu seriously needs a cool look and then pretty much nothing will stop it and all the other distros will be left in it’s dust.
That the features of linux is almost complete for alot of people out there, the success of a distro would be in being more hardware friendly.
Eg:
1. All mice more than 3 buttons must be supported
2. All Web cameras must be supported
3. All Tablets must be supported
4. All Advanced Graphics features must be supported (rotation, vivid colors like in digital vibrance,…)
5. All Distros must have an excellent package management system to update and install applications (this moved me from fedora to ubuntu after 2years on fedora/redhat)
6. All keyboard extra keys must be supported
7. All udf disks CDs and DVD must be automatically mountable
8. All codecs must be installable with ease
9. All browsers plugins are preinstalled (java, flash, …)
10. I will not be more demanding, the previous are enough for me.
Edited 2007-12-15 07:47
1) They are. Just up the “Buttons” value in your xorg.conf, and use a mouse button mapping program like imwheel http://legroom.net/howto/mouse
2) many webcams (like winmodems) are impossible if the drivers are not available for it, due to the fact that they do alot of processing that is normally done with the hardware side on the software side instead. Vista and OSX are both in the same position as linux with this one, it is pretty much impossible to get 100% webcam support.
3) which tablets arent?
4) They are (for nvidia at least)
5) They do. Fedora has yum, Ubuntu has apt, Mandriva has urmpi, etc
6) They are, with the common keys auto mapped for you, and the option to manually map them to whatever you want.
7) UDF 1.02 to 2.01 work fine. 2.50 doesn’t yet, but work is being done to get it working
8) They are. FFMPeG is the multi-format decoder of choice on any operating system, and it is open source.
9) Browser plugins aren’t preinstalled on any other OS. Why is this required for distro success?
10) Good, cause other then then very new UDF formats, everything else is already supported
Yes, they are. On Windows they are installed automatically. On Linux you get “manual install” message instead.
Egh, no. That’s not the definition of ‘pre-installed’ last time I checked. They may auto-install easier, but they aren’t pre-installed.
Um, no. In windows you get please install latest version of flash on a default install. On OSX however everything is installed right off the bat. An OEM install of windows is not a default install of windows. Someone took great pains to configure your machine in a usable state. Linux tries to do that albeit with a bit of user input and mostly because of licensing issues rather than because it can’t be done.
Anywho. My issue with Fedora has always been that its too bleeding edge, RedHat is basically using its community as guinea pigs while those with the cash, get the nice stable well tested version. the only issue with this is that eventhough I like this pretty current I don’t want to have to deal with some of the issue that crop up with every install, Ubuntu si also buggy bu tina different way. I’ve yet to see a showstopper in Ubuntu, just relatively minor things that if given resources might work better.
preinstalled != installed automatically.
1) They are. Just up the “Buttons” value in your xorg.conf, and use a mouse button mapping program like imwheel