Submitted by “During this years GUADEC Red Hat developer Havoc Pennington proposed his idea of an ‘Online Desktop’ to the developers of the GNOME project. Through deep integration with web services and ‘zero-maintenance’ the Open Source client aims to get the ‘perfect window to the Internet’. During GUADEC Andreas Proschofsky had the chance to talk to Pennington about advantages and possible problems of the Online Desktop concept, the necessity of Windows-support and about Red Hats ‘return to the desktop’.”
… then count me out. At the moment I am using GNOME and I am fairly happy with it. However, since the inclusion of Mono I am watching GNOME and I have become very worried. This new approach to desktop computing is – at least in my eyes – another serious mistake: I will never ever deposit my personal things online – no way. What’s more, I see absolutely no need at all to move all this online. What for? The browser does everything I need.
Edited 2007-07-29 19:29
Since when did GNOME include mono, if your talking about Tomboy it’s optional.
When all gnome-distros install tomboy as default, mono becomes a defacto dependency. Also, lots of gnome code is written by ximian (now novell). It will be interesting to see if gnome adopts the GPL3.
Thats the distros choice, it’s got nothing to do with gnome itself. If distros choose to go out on a limb and install mono thats there choice, yes choice you’ve heard of that right?
You don’t have to be so harsh, now do you? Not many regular users now what apps are actually part of GNOME and what aren’t. The fact just is that nowadays all the major distros (and many minor ones too) do compile Nautilus with Beagle enabled and so forth, so of course it’s very easy for one to assume Beagle is part of GNOME, is it? And if the distro compiles it with Beagle then the user doesn’t really have much choice about it.
It’s just a backend, just like it has a tracker backend, so you do have a choice. Use the regular search, use beagle search or use tracker, it’s that simple.
It’s called having the integration there ready so it can be used.
Edited 2007-07-30 17:51
It’s just a backend, just like it has a tracker backend, so you do have a choice. Use the regular search, use beagle search or use tracker, it’s that simple.
It’s called having the integration there ready so it can be used.
Either you’re confusing two different things or we’re just not talking about the same thing. You do know that it’s a compile-time thing? If you compile Nautilus with Beagle enabled, then it depends on Beagle. If you compile it with Tracker enabled, then it depends on Tracker. You can’t switch them at run-time. And f.ex. if Nautilus is compiled with Beagle enabled and you remove libbeagle then Nautilus will stop working..
People rail about distros pointing to proprietary codecs. Pointers to proprietary codecs don’t bother me much. But pulling a great big, friendly looking, Trojan Horse through our front gates *does* worry me. Yes, I know all about the ECMA standard and all that. And I don’t trust Mono for infrastructure. As The SCO Group has demonstrated, you don’t have to have a case to generate a law suit that will endure for years.
I welcome Mono… sort of in the same way as I welcome Samba. I wish to the gods that we didn’t need it. But in Samba’s case, we do. In Mono’s case… we’ll see.
Sometimes you’ve got to keep dangerous stuff around just in case you need it. Like insecticide. And while using it to combat a roach infestation might be necessary, if distasteful… I would not use it to season my lasagna.
Count me as a “yes” vote on Mono’s availability. Count me as a “no” vote for including it and Mono apps in distros. At least for now. I favor the inclusion of Samba and related software because, gosh dang it, we need it.
Edited 2007-07-30 18:14
Very likely just a temporary problem. The indexer/search engine projects have agreed on a neutral D-Bus based interface for desktop search, called Xesam: http://www.freedesktop.org/wiki/XesamAbout
Once the projects have implemented it, clients such as file managers can use either engine without directly depending on it.
Libbeagle is just a C library, Actually no, if you remove beagle it should default back to the default search for nautilus. libbeagle talks to the beagle functions and nautilus talks to libbeagle, if you remove libbeagle nautilus will still work.
Nautilus still works without libbeagle in Fedora and if by your method how is beagle supposed to work in Fedora now with nautilus?
If an application is linked to a library then it will stop working when the library is removed. Doh. I removed libbeagle*, did “killall nautilus”, and guess what?
werecat@kinuski ~ $ nautilus
nautilus: error while loading shared libraries: libbeagle.so.0: cannot open shared object file: No such file or directory
It depends on the circumstances, that really shouldn’t happen, libbeagle should sit on it’s own ready to interface not do that. Maybe Fedora dont have nautilus compile beagle at compile time.
But that would mean you can’t use beagle in nautilus, which makes no sense.
If you had read all the comments relevant to this discussion you would know that libbeagle is not beagle. It is in fact a 81K library written in C that doesn’t depend on anything other than libc, libxml and glib.
No, it’s called bloat.
There is already enough backend crap running when using Gnome.
Seems like Gnome needs a server running in the background for every tiny feature these days.
And it’s hard to controll it from the end user’s side because they don’t provide any tools for that.
Yeah, they do actually, Gconf, that one has to run from cli, great usability! Way to go Gnome!
Don’t kid yourself, if the big guys want something in, they get it in. Those who didn’t like sun’s ‘usability’ approach just had to walk away. Same will happen with mono, and java. It’l happen with the embedded stuff, and will happen with the online desktop.
That happens in other projects too, Im
sure some of KDE developers wanted to use Cairo but the big guy aka TrollTech forced the use of arthur, or maybe some didn’t want to see KDE apps. ported to Windows, but again the big guy pushed for it, or maybe they simple wanted to use Compiz fusion, but again, the big guy pushed for its own toolkit to show it off to its comercial customers, ist a petty. I know.
I second that. Personally I do not object to an optional web based backend (with an open source server that I can host myself). However as I understand they’re going for a mandatory path, which doesn’t seem to be acceptable.
Edited 2007-07-29 19:58
No they are not, and Pennington even clearly states that in the interview
I took you seriously until you said “since the inclusion of Mono”. It never happened.
Despite the fact that Ubuntu, which uses GNOME by default, doesn’t list Mono as a mandatory dependency (and therefore proves that Mono is optional), people keep spreading FUD like this. What ridiculous claims will we see next? “Since the inclusion of MS Office in GNOME I’ve been worried”?
Edited 2007-07-29 20:11
Ubuntu installs Mono by default as a dependency for F-Spot.
“””
Ubuntu installs Mono by default as a dependency for F-Spot.
“””
Is F-spot part of standard Gnome? Or just an application available for Gnome? I think it is just an application available for it, like beagle.
I am glad that we have mono for compatibility purposes and perhaps to lure .net programmers. But I would never want to see us depend upon it for infrastructure, for a number of reasons.
Yes, I know that people have different opinions on the matter. But that’s mine. And it is firm.
FooBarWidget was argueing that Mono is not default in Ubuntu. Blablabla disputed that.
“””
FooBarWidget was argueing that Mono is not default in Ubuntu. Blablabla disputed that.
“””
Yes, you are right. I was preparing for a reactionary attack upon Gnome and mistook that attack upon Ubuntu for one. 😉
Edited 2007-07-29 21:41
Hmm when I try to remove libbeagle from ubuntu, it wants to remove the whole of gnome, nautilus and everything. I’d say that’s pretty much dependant wouldn’t you?
libbeagle is an optional Nautilus dependency. It’s a small library written in C, unlike beagle.
You can check its dependencies on http://packages.ubuntu.com/gutsy/gnome/libbeagle0
+1 Improper use of “FUD”
He claimed that Mono is “included in GNOME”. This is false. By spreading false information he makes people fear GNOME. The false information and fear, together, cause people to doubt the GNOME project’s progress. People become uncertain about whether GNOME’s goals will benefit them.
So how again is it improper use of “FUD”?
it’s not in gnome because Red Hat and Sun didn’t want it at first. Now they don’t disagree, and it’ll be in there some day. And that makes sense, it’s practically in anyway cuz the distro’s ship it that way.
I agree, if we are talking about the Internet. This doesn’t mean that similar tools and techniques could be used in e.g. an intranet, where storing information on centralized servers actually could increase security, e.g. by not being on laptops that gets stolen, and being backed up regularly.
However, I think that it would be much better if they spent more time improving gnome-vfs, and make sure that ALL applications use it. Or, even better, they could join forces with KDE to create some toolkit independent replacement for gnome-vfs/KDE kioslaves.
It should be just as easy to edit something stored on a samba file server as something stored on NFS or your local file system.
that GVFS thing might indeed happen, some guy is working to ensure the KIOslaves will support it. Won’t be a 4.0 thing, but it’ll be possible. If GVFS turns out to be good stuff, why not?
However, since the inclusion of Mono I am watching GNOME and I have become very worried.
Well, Mono isn’t a part of Gnome, but the Mono framework is something that Gnome could do with. Gnome lacks a general purpose, universal and consistent looking API and framework for developing applications, and the Mono framework or some Java classes would sure help with a lot of things. Something like it anyway.
Now, the questions then become:
1. How comfortable are people with it? Are people comfortable with a .Net clone, and are there any issues?
2. Can you really develop a lot of applications and a lot of running infrastructure that make up a desktop, and then run it in a VM? Does it have to be native?
3. Maintenance. The Mono framework, or Java classes, would make things infinitely easier from an application developer’s perspective, but what about the libraries it wraps? You’d still have GTK underneath, as well as a plethora of other libraries that go to make up Gnome.
4. What will the distributions do?
If they will all ship it, it should get in. I’m not a big mono fan, but it just makes sense…
Havoc talks a lot about not forcing users to change their favorites applications to his new vision of online desktop in one hand but at the other hand he keeps saying that default matters and that by having the online apps already available, he would like to make them the default apps thus making new users getting used to that model.
Personally, I don’t think that this will fly. I think that this could be done already on an application level instead of on a desktop level. Why not add the ability to interact with Flickr to F-Spot, predefined online radio sets to Rythmbox, Firefox extensions specially tailored for the social online communities, blogging, etc, etc? In fact, I believe that this is what they are going to do as this is the only feasible way that they can take this thing.
Not to mention that none of the darlings web apps are OSS and therefore one may need to take a risk by trusting their personal data to these online services. Privacy will be a huge a huge concern at a later stage of this project. Or yet, that one of these services might go down with all your data. There are lots of things at stake here…
Disclaimer: I am not a GNOME user but I don’t think that this GOD
is in their best interests.
Why not add the ability to interact with Flickr to F-Spot
That’s been done for a long time now. I have been uploading my pictures directly from F-Spot to Flickr for at least a year or two.
Firefox extensions specially tailored for the social online communities, blogging, etc, etc?
Umm, Firefox isn’t a part of GNOME.
Not to mention that none of the darlings web apps are OSS and therefore one may need to take a risk by trusting their personal data to these online services. Privacy will be a huge a huge concern at a later stage of this project. Or yet, that one of these services might go down with all your data. There are lots of things at stake here…
I’m not so worried. So far mugshot is the only application to come of this and it is great and doesn’t require you to store anything online that already isn’t online. It’s really just a web community aggregator.
When the GNOME devs have an epiphany like this and find that there is a major need for extensive internet integration in their API and applications, and precious little support in place – perhaps its time to consider putting their efforts into a different project.
After consciously abandoning the ‘NOM’ part of GNOME back when the switch to 2.x was made, there is now a vague plan to rebuild the ‘NOM’ around, well, from what I can tell, a bunch of proprietary and constantly changing protocols that will be built into a new GNOME API layer.
This, of course, will leave all existing GNOME applications out in the cold (does anyone remember how long it took for gedit to get ‘save to gnome-vfs location’ support?) for years, and will probably lead directly to a GNOME 3.0 that, like GNOME 2.0, was a complete break with the past and took literally years to achieve performance and feature-equivalence with 1.x
The retarded thing about this, is that had the GNOME devs regarded the ‘NOM’ piece of GNOME as useful and important as they seem to now during the switch to 2.x, this type of ‘internet desktop’ functionality would be pretty much a no-brainer.
For example, KDE has a native web browser and robust support for inter-application communication as well as complete support for kioslave-based VFS by default in it’s apps.
These factors alone make an ‘internet-enabled’ desktop simple – Konqueror (as well as desktop shortcuts etc.) will seamlessly access remote storage via DAV, FTP or SSH, as well as having all my KDE apps accept those URLs for opening and saving documents – theres also integrated IM, calendaring and email. My applications can invoke actions in other KDE appsbased on context – e.g. send an IM to someone in my contact list from any app by invoking a KParts call.
In fact, i’d say if you want an internet-centric desktop then KDE has 95% of the infrastructure in place.
And yet, it hardly seems that having this kind of ability has spawned a ‘KDE Online Desktop’ – i guess when the core capabilities are already in the base API theres no need for a new project to provide it.
However, the major stumbling block – even with KDE, is the complete lack of basic ‘all-applications-use-it’ API for this stuff.
Currently, no matter what the GNOME or KDE devs do, Firefox, Thunderbird, OpenOffice and Eclipse ( and this is just a short list of the major productivity apps i use) won’t play nicely in this ‘internet desktop’ scenario.
Please, don’t work on ‘GOD’, work on (and what a nifty acronym this is – ) ‘LOSE’ – The Linux Online System Environment’.
Edited 2007-07-29 21:19
There is a lot of work going on in the online-desktop area in KDE, it’s just not mentioned as such. See OpenID integration, GHNS2, Decibel and realtime collaboration, stuff like that. Many of the games are going network-multiplayer thanks to a framework in the games lib for that.
The aproach KDE is using totally diferent from the one proposed by Havoc, you should know that.
Edited 2007-07-30 23:21
Yes, they don’t call it GOD, and they don’t have a grand vision for it but are just coding away. That’s the biggest difference. In terms of infrastructure and code, they’re ahead…
The aproach KDE is using totally diferent from the one proposed by Havoc, you should know that.
The trouble is, no one has any idea what on Earth Havoc is proposing. At least KDE has code written for this, they have KHotNewStuff and they’ve had IOSlaves for years to connect to many different network protocols. All this GOD stuff just sounds like some pretty extreme vapourware in search of the next big thing.
Is he proposing that a desktop be run remotely, like Terminal Services? Is he proposing that only documents are stored online? If that is so, how do you handle an unreliable network connection being cut off in the middle of saving a document? How will things like social networking be integrated, because there are no applications and no infrastructure for it? How do you handle multiple web services, since many might require their own authentication and might duplicate an awful lot of your own information?
your ordinary GNOME desktop will not vanish, it’s still under heavy development and evolving in beautiful way, you will not be force to use Online GNOME and your application will not stop working being offline.
Online GNOME is something very interesting, trying to take advantage of web, think about it, wouldn’t be cool if you can see Google calender events and to do list from tadalist.com when you click clock on your GNOME panel? Google notes and tomboy… contacts, all synced across you GNOME desktops at home, work and Mobile devices… there’s some hard problems Online GNOME trying to fix… trying is better than doing nothing and watching competitors getting ahead of us, even if this project fail, at least GNOME did try
A friend of mine, from long ago, used to say: “Better to have it and not need it than to need it and not have it”.
Good advice for a wide range of situations, I’ve always thought.
While nobody will be forced to use the online functionalities, it will inevitably have an impact on their traditional desktop.
As I mentioned in my initial reaction about GOD, resources (developers, time, money) will inevitably be drained to focus on the new online vision. Offline applications won’t go away, but they might evolve slower and have more bugs. While the idea of synchronizing your data across multiple devices is neat, concerns about privacy are genuine. While being “online” doesn’t seem as bad as I thought at the beginning, it should be planned carefully. At first, private corporations can be cooperative, but they might change their tone once their advertising revenues goes down. Not sure that I want ads within my documents… neither many users who stay away from the commercial world for similar reasons.
In a nutshell, the new vision might alienate a significant part of the current user base. You mention even if they failed, they would have at least tried… Actually, I don’t believe it’s just a mere gamble. If they fail, they are pretty much toasted. Hope they thought about it.
Why does the world need to know I’m going to see my girlfriend and have sex with her in about an hour? Why does the world need to know I’ll be visiting the dentist tomorrow? Why does the world need to know what kind of toothpaste I’m using? Why does the world need to know my personal notes and random thoughts, unless I am consciously writing them down for a comment section like this one? Why does the world need to know how much money I make?
Why does the world need to know who I am, while they do not have any real interests in knowing that?
Seriously, where exactly is the “interesting” part of it?
It’s wonderful to have all that data synced to the devices one has, but there are a lot of better ways to take care of that.
I’d vote for making the computer personal again. It’s my shit, not another’s.
Please Apple, hire Havoc already and get him off Linux for good.
Well, in my opinion Gnome 1.4 was the best, and then Havoc started cutting features. And a decade later there is less features in Gnome than there was in 1.4
What an idiot.
Of course nobody is installing Gnome from source because it’s a nightmare only a few geeks can handle without a nervous breakdown.
Even great pioneers of Linux like PV from Slackware dropped Gnome because he had enough of Havoc’s bullshit.
Yeah, what a bright idea, make everything to be on-line by default, and have it all on one central world server.
So then the government can nicely supenea that one server and get everyone’s personal data in one easy swoop. Game over.
… I think they should work on improving the gnome-vfs support for universal access to network oriented file storage. By that I mean, yeah right now it seems in ‘some cases’ you can access a wide variety of Network Oriented ‘files systems’ but it is not consistently or widely made available. And as mentioned earlier, simply including slightly different Default setting of already capable applications can go along way to making this concept apparent.
I think it’s time the userland desktop environment VFS was scrapped entirely. We have FUSE in the kernel now and libfuse is being (has been?) ported to the BSD and Mac and probably easily portable to others.
I would say most of Gnome’s VFS capabilities are already available for FUSE, why worry about porting apps to a desktop environment specific VFS API when just about everything already uses POSIX compatible API’s?
Of course there may be issues with latencies with some FUSE filesystems (like access over the network) that some apps expecting local file access simply weren’t tested for, but surely heading into a multi-threaded multi-core world, shouldn’t we be thinking about asynchronous I/O for new apps anyway?
Overall I was glad to hear from GUADEC that the Gnome VFS is being overhauled
Agreed. If the network transparency (and anything else file related you can think of) is done by FUSE services then VFS and friends are completely deprecated.
What I would want is good integration with cashing file systems, Coda is an excellent example [http://en.wikipedia.org/wiki/Coda_%28file_system%29]. Also there are file meta-data APIs already in place on Linux and most other modern OSes. So why not expose any synchronization specific information as attributes that can be managed as part of the save/open dialogs for any file versions issues.
Beyond that at the application level there should be standard libraries like the Evolution address book server that can accommodate plug-ins that carry higher level data synchronization using Web services, mobile devices and what not. Again for any operations that integrate with the web move the function out of any specific application and create libraries and services that they can plug into, isn’t that the sort of thing dbus was written for?
None of what is useful that I describe is so big that any fundamental changes are required in GNOME or any other DE.
What I simply was speaking to was the need to be be unified and availible. Whether it is it is VFS or FUSE, cause after all there seem to be quite a few ‘frameworks’ being developed but a great deal without extensive implementation or adoption before another one comes in to replace it.
Well, there seems someone to be working to ensure GVFS can be used as backend for Kio in the future, so that’s good news. If GVFS really turns out that good, why not? But until then, the KIO slaves are superior, and with KIO-FUSE, any app can use them anyway. I don’t understand why nobody tries to get that working properly… Don’t want to depend on KDE stuff, I guess?
Havoc Pennington:But to me when I use a desktop today I feel like there are major negatives to it – there are all kinds of maintenance and backups to worry about. Just cleaning up the desktop background, cause it’s cluttered full of junk to not having calendar notifications, having recent documents be the wrong documents and not the online documents that I’m actually using.
Hmm maybe because you ripped half the functionality out of gnome and now refuse to include new useful features users have been asking of you for years. And what will happen when this new online desktop gets going and users ask for features in that, will it take you four years to incorporate those features just like now.
Call me a cynic, but I can envisage an online gnome desktop where users will have to wade through online-gconf editor to change proxy settings because changing proxy settings is just too confusing for mere mortal users.
Could you name any of the features that have been “ripped out”? There are a lot of complainers who keep perpetuating this myth. Yes: it’s a myth! Gnome 2.18 is far more featureful than 1.4 ever was. Sure, there’s the obvious exception of panel configuration:
http://www.gnome.org/~jdub/2003/sequelsyndrome/mgp00015.html
Bullshit!
The easy access to modifying menues was ripped out. It took several releases before a THIRD party project was included in Gnome. And of course it only has half the functionality it ought to have.
The Mime Type Editor was ripped out in Gnome 2.8 and hasn’t been replaced yet. There is a third party project (assogiate) but it’s not quite there yet – but look out for assogiate 0.4.x
When people complain about things being ripped out of Gnome they are NOT referring to missing functionality – but to missing EASY ACCESS.
One can still edit menues by messing with xml-files – and the same goes for mime types. But it just ain’t easy access! So this misunderstood “purity”-madness has made the desktop HARDER to use – and not EASIER.
Gnome 2.18 has of course more functionality than Gnome 1.4 – it is just a lot more unaccessible.
The two things you mentioned were ripped out during the 2.x series.
IMHO the original poster is right, 2.x was slower, but it did a hell of alot more then 1.x did.
Wow, how you got modded up so much is beyond me, What do you need a mime editor for. It’s totally useless to avarage joe and it’s been replaced with a more sane way to default apps to a mime type.
I think people like you rant because your missing the “Power user” features, are you seriously telling me that average joe will use be bothered about moving toolbars and editing mime types. I think you need a reality check when it comes to features that ARE really use to the user.
Yeah… +9 seems impossible to me…. I don’t get it. Weird.
But apart from that:
You are talking nonsens!
Take this scenario: Install Gimp – Gimp steals ALL your “Open with…” settings from eog, and now you have to right-click on 45 different imagefiles in order to restore your settings *sigh*
And every now and then you trip over an image format you haven’t restored, and when you expect eog to open it opens gimp instead… ARRGGHHH!!!
Just because it is a power user tool doesn’t mean it shouldn’t be there. A Power User Tool does NOT rule out a Newbie Tool as well – there is room for BOTH solutions!
You don’t have to be a “Power User” to move toolbars or have a need for a fast and easy way to change many mime types in one go.
Whether or not a feature is available has nothing to do with whether or not a newbie can use them or understands the issues. What matters is SOLELY: Does this make it easier to use? And yes, a centralized Mime Type Editor makes reverting changes for many Mime Types at one go MUCH easier.
A centralized editor can easily coexist with the existing decentralized way of 5 clicks to change one setting.
The existing decentralized way of changing settings for a mime type is ONLY sensible when you only need to change ONE mime type. Try many and you’ll see the flaws in ONLY having ONE way to do things.
Most persons don’t need columns in their text document – let’s rip it out of OpenOffice. Most users don’t need a fill option in Gimp – let’s rip it out. Let’s aim for the lowest denominator!
There’s a difference in restricting the user and making things easy.
What if I want GIMP to edit all my images, after all it’s what it for.
It’s quiet simple, right click, open with EOG(which is a option anyway, jpeg and png are the most common types, what 45 types do you mean?
The fact that there is 45 image types is the problem, so your missing the major thing gnome devs have to deal with.
yeah yeah, the world isn’t perfect but we choose to ignore that. OK, there maybe shouldn’t be 45 image filetypes, and maybe nobody should use anything but the gimp for editing (screw em if they want multiple colorspaces, good tablet support, proper 16 bit and RAW support or other stuff gimp doesn’t have).
Real life says: give the users what they want. GNOME says: they want this. If they don’t they’re stupid and should go away.
That solution is good for changing one or two mime types. But when you have images of many different formats (like when you are dealing with professional graphics) then you are in deep shit.
It’s okay Gimp is taking over the associations. But it’s not okay that it is so difficult to change them back. It is very cumbersome. And besides png and jpg there is also gif, tiff, bmp, xpm, pcx among the more commonly used formats.
What matters here is not the amount of standards but the lack of an easy way to change many mime type settings.
But in the end you still have to change all the mimetypes back regardless, so we are right back to square one again. Why people insist on having arguments about gnome features being taken out since 2.6 is beyond me, Live with or don’t use it. People will keep banging on about the feature THEY need, other DE’s have it so we must, otherwise it’s uselss?
People talk about these missing features like average Joe would use them everyday, when really it’s the power users ego thats really the problem.
Edited 2007-07-30 22:24
Why would I have to change them back? I wouldn’t have to change them unless the stupid packages kept changing them
What we need is not the option to change mime type settings – what we need is an easy way to change many in one go. It’s easy to change one mime type – but changing many mie types is less fun than playing with the Windows registry.
People talk about these missing features like average Joe would use them everyday, when really it’s the power users ego thats really the problem.
There is no such thing as a power user. You have a user, who at one time or another, may need to use a feature or an option that they may not use again for weeks or months. If that feature or option isn’t there then it simply makes life even more difficult. These may be ordinary users, sys admins, developers etc.
The people who keep regurgitating the above rubbish simply have no idea, and have no idea why KDE, Windows, and even the cornerstone of wannabeism for Gnome, Mac OS X include features and options that Gnome doesn’t.
You have a user, who at one time or another, may need to use a feature or an option that they may not use again for weeks or months.
The only solution to that is to have all imaginable features implemented in a program, but you seem to ignore the problems that would cause. For each feature you add there’s a cost, and some environments are willing to pay less than others.
I use GNOME and I don’t want all the features you love about KDE.
GNOME, KDE, Windows and Mac OS X have each drawn their line somewhere. Get over it.
These may be ordinary users, sys admins, developers etc.
Yes, and nothings stops these people from using GNOME. In fact, many of them do.
even the cornerstone of wannabeism for Gnome, Mac OS X include features and options that Gnome doesn’t.
That’s not quite right. The thing is that people accept Mac OS X for what it is, but many people (who don’t even use GNOME) feel the need to complain about it very often. For example, the way to set default applications for file types is exactly the same on both Mac OS X and GNOME, but which one do you read people complaining about?
Edited 2007-07-31 16:09
The only solution to that is to have all imaginable features implemented in a program, but you seem to ignore the problems that would cause.
No. Every desktop environment tends to have a way of managing your mimetypes reasonably for a reason, as well as a way of managing what you, not some developer you’ve never heard of, wants to have in your menus. It is functionality that never gets in the way per se (I’m not constantly running into the menu editor and thinking “Damn, not again”), but when you need it it’s there to hand.
You give me the name of any popular usability book and I’ll give you the exact chapter where it tells you that a piece of software should grow with your users and attempt to help them out in even the 5% of cases where they need a time and effort saving feature and where they want to do things in a different way. It doesn’t have to cost the Earth, and it doesn’t mean bundling the kitchen sink. Other desktops do it.
I’m sorry, but you don’t get away with the simplistic “Oh, we have to have every imaginable feature!” argument, because it’s a well travelled road in the usability world. It takes things called time, effort, research and debate as to what is needed. It’s not about culling functionality.
GNOME, KDE, Windows and Mac OS X have each drawn their line somewhere. Get over it.
Lets face facts. Gnome is simply inadequate for many Windows, OS X and KDE users out there. If Gnome doesn’t have the resources or means to implement features included in every other desktop environment there is then just say so, but please stop using usability as an excuse. The other desktops aren’t just simply wrong.
I use GNOME and I don’t want all the features you love about KDE.
Getting your/my mimetypes in a workable order so you/I can open files with the application that you/I choose using a decent tool, and using a decent tool to get your/my menus organised in the way that you/me want are not special features in any way, and there’s simply not a single reason why you shouldn’t have the option to do that. It doesn’t get in anybody’s way, and don’t even try to use the ‘clutter’ word because there just isn’t any.
At some point or another, it’s something that quite a few people end up needing to do.
Yes, and nothings stops these people from using GNOME.
So you think.
Sys admins:
1. There is no simple way to run an application as root or another user as there is on KDE, or as an administrator, as on Windows and even OS X. This gets useful for sys admins when they need to do certain file operations as root, for example, or run certain administrative applications as root.
2. There is no tool to configure Gnome’s own GDM login manager. If you want to install a workstation, or server, and simply lock down the login screen in many ways, such as limiting shutdown and reboot to the root user, there’s no way to do it simply. Get that. A graphical desktop environment that cannot be configured graphically!
The list can go on a pretty long way if you want. I suggest some people have a long hard look at what actual types of users are trying to use Gnome, rather than a simplistic, idealistic and narrow view of the only type of user some people think are using Gnome.
For example, the way to set default applications for file types is exactly the same on both Mac OS X and GNOME, but which one do you read people complaining about?
Both of them actually. It’s yet another piece of functionality lifted from OS X because certain people think it’s the way to do things. Apple largely gets away with it because OS X tends to come with a bunch of set apps that you’ll use, most of the same type. Windows and Linux tend to have many more different types of applications that people install, some of them that can open the same file types, that necessitates a convenient way of managing it all. However, it is still a bugbear with OS X and people still complain about it:
http://apcmag.com/6746/setting_your_default_application_in_mac_os_x
Edited 2007-07-31 18:30
When a feature that should be included is missing or got buried deep in the gconf database, it doesn’t hurt my ego. It hurts the usability of the software. I don’t mind if the average Joe can do the same tasks as me; I do mind if I am unable to cater my needs because of a lowest common denominator politic.
What is an average Joe, anyway? An uncle in his 50s? An elderly person? A computer neophyte? I would say it’s a young person (late 20s/early 30s) familiar with technology and computer paradigms. It doesn’t matter, though: whoever it is, there are other people using the desktop environment. Dismissing them is a huge mistake.
I am not pleading that GNOME should become a feature mess. However, simplicity should be make the software simple to use, not simplistic.
I know, “live with or don’t use it”. However, it only reinforce my impression that the GNOME community is becoming a clique.
Both the mime type and the menu systems have been replaced whole-sale during the v2 series with new cross-desktop solutions. Yes, there might have been regressions, but in both cases the current solution is superior to the old.
The current mime-type system allows for much easier configuration of mime-types than the previous one ever did. Just right-click the file, and select properties. There’s no need for a separate application. As a bonus: the new system “just works” when installing and uninstalling applications (something which can’t be said of the old one).
The menu editor is such a non-issue. This isn’t MS Windows where apps install entire folder hierarchies for themselves. I’ve yet to find a single Linux app which doesn’t properly integrate into the main menu.
When people complain about things being ripped out of Gnome, what they usually mean is: “I used to have to jump through hoops to do this. Where are the hoops? I WANT to do this the hard way.”
Cross-Desktop has nothing to do with this. It is completely unrelated to a centralized Mime Type Editor. I agree the new decentralized way (introduced in Gnome 2.8) is way better than the old crappy mime type editor.
But that DOES NOT mean that this solution is the only one which should be available. And being superior in terms of changing MANY mime types is incorrect. For such tasks the new way is very insufficient!
The new way does NOT allow for easier configuration. It doesn’t allow for ANY configuration EXCEPT favourite application. And it pretty much the same way the old decentralized way worked. In Gnome 2.6 we had a centralized and a decentralized way of doing it. Since Gnome 2.8 we have only had a decentralized way of making changes. True, better than the old decentralized way – but today you have to manually edit xml-files in order to make changes apart from the favourite application. And making changes to MANY mime types in extremely cumbersome.
We need a good way to change many mime types WITHOUT replacing the existing decentralized way – which is GREAT for changing one or two mime types.
Oh well, Thunderbird 2.0.0.5 is one. Scribus is another one (when using Gnome), esvn, rapidsvn, lincvs and the Gnome User Editor and a gazillion other apps which are misplaced in the Gnome Menu – or not installed at all. Especially games fail to be installed in the menues.
You are obviously not using any applications apart from Gnome Games and Firefox.
That’s pure arrogance. “Somebody needs a different solution than me. They are morons. Boohooo!” <– nice answer kelvin (and yes, it’s sarcasm!).
When people complain about missing options it is because they have DIFFERENT needs. Try reverting stupid changes (made by installing Gimp) for many mime types and you’ll see how insufficient the decentralized way is.
Get over it. People don’t have different needs, Gnome has all they need.
Yes, sarcasm, but hey, it is what it is.
We are drifting from the main topic, but:
Unless you are installing those packages from the original source, these associations should be a job for the package maintainer of your favourite distribution. On that aspect, I didn’t had such issue with menus in any major distro for a long time. Now, if you are installing those applications from the sources, I suppose it should be your job, since authors don’t necessarily know your environment that is going to be used for their application. Flexibility comes with a price.
However, I do concede that the GNOME folks are ridiculously anal about features. For instance, I shouldn’t have to use gconf-editor to set preferences. Yet, so many applications are burying “advanced” preferences in it, like power users are inexistent. Oh well, it’s not like the community have something to say on the directions taken by the GNOME team, the new online vision being a prime example.
Edited 2007-07-30 21:57
Still OT but is >>Unless you are installing those packages from the original source, these associations should be a job for the package maintainer of your favourite distribution.<< really true?
There is now the portland xdg desktop specification which aims at standardising menus installation (among other thing), so applications should be seen in the menus after installation (as for being in the “correct” location, sure that’s possible only for packages provided by the distributions).
WTF!?
How the **** did this get a +9 score?
I suspect like him/her they didn’t know about the menu spec change during the time people moaned about the menu editor, but oh look we have a menu editor now. I also suspect they either not folowed gnome for a while or dont even use it.
Edited 2007-07-30 19:06
Like some people say these days: being online makes a person less productive.
I think that goes for this “Gnome Online Desktop”-thing as well. While applications may not stop working when “the net” is offline, it *is* a huge source of frustration when that happens and personal data is not available, and for what? Well, so we can share our personal stuff with the rest of the world.
To be honest, I really feel the urge to share my point of view on how I think about people that want to drag my desktop into the web and add yet another reason for pulling out hears when it doesn’t work.
Great idea mr. Havoc. Great idea. Let’s all enjoy the new frustrations coming up, shoveled down to us as a new feature, while the real features that users are actually waiting for will be pushed aside, once again.
I don’t see why this is represented as such a radical reworking of Gnome. It seems to me that it really amounts to creating a bunch of add-ons, none of which requires any changes that are much deeper than cosmetic. For example:
*Some VFSs for flickr://, myspace://, etc.
*Some right-click menu options such as “Post photo to Flickr” and “Send file to my Myspace contacts” etc.
*A directory server maintained by the gnome.org people that would hold people’s identities on various services in a centralized location, allowing them to log on to all of this stuff simultaneously if they wanted.
People who like the above can use all of it, people who don’t are forced to use none of it. It’s mostly just some spit and polish to make the desktop easier to use. It could even be packaged separately from the rest of the Gnome system.
I’m sure there are ways to go about this project that would involve ripping out and replacing the object model, etc., but — why bother?
Well, that’s what I thought at first. And – KDE is doing this, they just don’t call it Online Desktop.
But on the other hand, if you really focus on this, you can invent new stuff, new things. So it might become intersting.
For me social networking sites are the new root of all evil, so I’ll do my best to avoid GNOME ships with Facebook or Mugshot or whatever social networking crap enabled by default. There is no point in developing free software, when your goal just is to replace Microsoft by some other Company owning your data.
I’m more confused by what Havoc is driving at than I ever was. Gnome’s developers also seem to have some conflicting corporate interests that are muddying the waters of what an open source desktop is about:
derStandard.at: As Novell seemed to agree on your vision, does that mean, that both companies together just can go forward anyway? Do you need the community to join at all?
Well I would imagine they do agree. This wouldn’t be a way of tying in something like Red Hat Network or Novell Network to the Gnome desktop, would it?
But to me when I use a desktop today I feel like there are major negatives to it – there are all kinds of maintenance and backups to worry about.
Yer, because it’s a desktop and the desktop is supposed to make all that easier……. At least you know what is going on, as opposed to having a bunch of proprietary crap sitting on a server where you don’t know what is happening.
One way to think about it is, say you’re a small business and buy Google Apps for your domain it keeps you from having to maintain your own servers, you know things like an IMAP-server.
Yer. You don’t need an online desktop to do that.
I think so, yeah. Potentially we will have to include Windows in that. So for instance: Firefox on Windows, why shouldn’t it know about the same things than Firefox on Linux. If you have services like social networks, you always have to support multiple platforms…
I find this very confused. How will you tie all those online services together, and what will be included in the desktop by default?
Mostly cause we won’t have any kind of fixed connection to one one service in our desktop, like they will have to “Windows Live”. We could have all the best-in-class services and basically everything you can find on the Internet integrated with the local desktop, while Microsoft will restrict it to their own stuff.
Yer, but how will you tie that together Havoc? A set of default services is going to have to be built into the desktop, as well as an easy way of adding new ones. How will they all tie together?
Some people just have the objection, that they never want there data on any server in any scenario, I think that’s just a very small minority kind of view.
Errrrrr, Havoc. This is a good part of the reason why open source software exists in the first place. It’s not just a shell for some online, proprietary and hidden software.
But in the end it’s all about the defaults and that’s something for the distributions to decide.
Yes, and the distributions will be tying their desktops into as many online, proprietary web services and online storage servers as much as possible.
And in my opinion the entire GNOME 1.0, 1.2, 1.4 series was pretty bad
Well that’s a ringing endorsement.
I mean if you look at the new features in every Mac OS X release, they could have never come up with this feature list, if their way to think about it would be “we are making a desktop”….
Goodness me, I wish some Gnome developers would just stop it with this “OS X is God” talk. If OS X is so good either just tell people to use it or outright clone the whole thing, and possibly get sued by Apple.
I’m not entirely sure whether an awful lot of these developers have actually used OS X as a day-to-day desktop. They just seem to have got it into their heads that OS X is the bastion of design, usability and innovation and they regurgitate it at opportune moments when they have nothing else to say.
The reality is not that many legitimate opportunities for things to do just like come along every day.
Well, you could have a bash at creating a desktop that has a good toolkit and some lovely looking and consistent APIs so that open source developers can create all these online applications reliably, and not leave users hanging when something goes wrong.
It’s all a bit sad really.
Edited 2007-07-30 09:51
Sniff, sniff. Is that the desire for a new revenue stream I am smelling?
Yes, let’s tie the desktop to online servers for the storage of personal user content. Access to those servers will be free of charge initially of course, but once everybody is dependent on an online server, why not charge subscription fees?
Who needs vendor format lock in, when one can just lock away whatever format a user uses lock, stock and barrel on your own for pay servers.
I would preffer gnome to migrate it’s base to a modern language. We have nice apps with modern languages, and it sure would bring more nice apps to linux.
And it could be another plataform, not just Mono. I know a few people that would work with linux, same people that returned to code for windows with the advent of .NET platform.
C++ WIN32 just sucks as Gnome’s C GObject (just one example) IMHO.
Just my 2c.
GTK already have nice C++ bindings, pretty good in my opinion.
I know, but the base GNOME is still using C with OO.
And OO doesn’t fit C by any means. You lose time that could be spent in coding the application itself instead of treating with GObject C and such.
But that’s just my opinion. This move in my opinion would bring linux to all-time-said “next level” and maybe we could finally have a “this year will be the linux year”.
Sorry for my crappy english
You will be glad to know that C++ is now a base languaje for GNOME, no more C only anymore, and some base GNOME applications are writen with C++ already.
Hmmm, now let’s see…There’s several things that come to my mind that would benefit from online storage, like f.ex. when you log in to GNOME for the first time it could ask for your login and password on the GNOME Online Server and then fill in all your details, bookmarks etc from there. Perhaps even the login picture would just be downloaded from the server if you have saved one such for yourself before, like for example on another machine. But then, I wouldn’t want ANY of my personal stuff accessible by others, and if the service became very popular and default on all distros I’m sure it’d attract a whole lot of hackers too..
Wow, that’s like an entirely new level of hubris.
Where’s HAL and SHODAN?
Pfff…they succumbed to the wrath of GOD. 🙂
How is this going to effect on whoms having slower speed internet accesses