I started writing this as a comment on Andy’s post, but it ended up quite long so I decided to blog it instead 🙂
From a user’s perspective, I’m inclined to agree—to put it bluntly, I can’t really recall the last major innovation I saw on my GNOME desktop. After all these years, we’re still mostly working on things that Windows and, more recently, OS X, can do for us already if we’re prepared to pay Apple or Microsoft for the privilege.
Clutter is one such example—it’s a promising technology for developers, and maybe it will turn out to be a “way out” of the “decadence” Andy describes. But if we just use it to write clones of Front Row or CoverFlow or any of Apple’s other CoreAnimation-based apps, then Apple will have moved on again by the time we’re finished and we’ll just be playing catch-up again.
In any case, Clutter is just a toolkit, and toolkits don’t help us with the vague lack of coherence and integration on some parts of our desktop today, or the need for more clarity of purpose going forward. For that I applaud Havoc et al.’s vision of an online desktop—while it’s not one that desperately excites me personally, I don’t have anything better to propose, and having everybody focusing on something is surely better than nobody focusing on anything 🙂
As for the HIG—I really don’t believe that it stifles creativity any more than (say) Apple’s HIG stifles theirs, but it probably has been a victim of it’s own success to some extent. It was written six years ago, and as such, it describes the best practices for six-year-old GNOME apps. We do have plans to substantially revamp it, but it won’t happen overnight.
That said, the HIG shouldn’t drive major changes in GNOME’s UI, it should reflect them. But at the same time, the gtk guys (and others) need to know what changes to make for 3.0 that will support the sort of changes we might want to make going forward.
Everybody probably just needs to get talking and generate ideas to move that process along again, but I also wonder if maybe we don’t have enough rockstars in our community right at the moment—it’s much easier to reach a goal when two or three people are driving you there, than when a hundred people are all trying to make their own way.
Integration is a tricky problem (obviously as there are always at least two apps involved)… maybe starting some sort of integration wiki page would be useful, people could put up suggestions and link to relevant bug numbers for the involved projects, for each particular suggestion?
This is not a problem! I want my desktop to be boring. I want things to be obvious, “hrmmm, new type of network hardware, ok, let’s add a Wifi Icon” but other than that, no need for big changes.
I’m still advocating a Recent Files place that shows files that you actually recently used: Downloaded, copied off a flash drive, etc.
Gnome is not _SUPPOSED_ to save the world, just let the world have computers that don’t bug the living daylights out of them because other people have their fingers too deep down in them. I’m happy to see PolicyKit & PackageKit, and think they will do alot more for the user-experience than a revamp of the underlying nuts & bolts of the widgets and such.
@Rob: true, but innovation doesn’t have to be intrusive or flashy either– quite the reverse; it’s often said that the best user interface is the one you never notice. (And I’d say we have no shortage of things that aren’t “obvious” right now.)
exactly calum, integration allows you to spend more time doing what you were trying to do rather than fiddling trying to make your computer do what you want. Thus spending less time on it and more time living…
the “flash” that people want is generally about making the UI consistent with our interactions with the real world because those interactions are more intuitive to the human mind. (ex: minimizing windows in OSX visually takes you to the place in the bar as if it were a real object being stored temporarily)
Thus bringing us back to point #1, less time worrying about using something correctly and more time living.
Linux now has an excellent chance to reunite forces and promote the best desktop ever. Today the QT libraries is in better state that the GTK ones, maybe that sign that on the long path we should abandon multiples libraries, and stop doing duplicated jobs.
I rather think we are innovating with things like Soylent and Conduit; if we can get them nicely integrated throughout the desktop, we’ll be ahead of the game.
That said, I haven’t used Windows or OS X for a while, so I may be wrong and they may have had such functionality for years.
Telepathy could be that next great innovation. Not just because it allows to do the same things again (voip, irc, msn, aim, etc) but because it allows to be more transparent. There does not have to be distinct application, or the activities bound to just one. You can with some work for instance add file sharing telepathy plugin into all gnome applications, something akin to gvfs. The difference is not technical, it is major however in usability terms. Fluency, transparency.
I could see things like scheduling (free times from calendars, meeting rooms), IM chatting, email, collaborative document authoring, voip, document management systems integration, collaborative bookmarking, blogging, etc melting together into the point where you can point at any target anywhere and just tell via a set of pluggable ‘verbs’ what you want to do and things would just simply happen… That is something *NO* platform has nowadays, and as the network effects and #1 use of computers is communicating, it would enhance that beyond anything we have ever seen. It would be a major selling point, even forcing the others to adapt or die.
The sad thing is, most of the Gnome developers just are absolutely clueless about marketing, usability, and having any sort of visions.
One more example off the top of my head. Consider having a presentation made. Now, you’re in that big hall and it’s your turn. Desktop search for your document. Ask the presentation to start. No starting of OOo, nonononono!
Most of the places have wireless networked projectors nowadays already. Something that is entirely unsupported in Linux. But what if you could just say “I want to present this”, and as the one projector was networked, and uPNP and the other fancy technologies worked, it would be presented. Not via normal OOo, but via that new fancy presentation interface that we saw a while ago – showing the presenter the notes and lines to speak while sending the slide automatically to the projector without the rest.
Now, often there are teams where some people are remote. You’d be presenting to many people in the same room but some would be remote. How about adding feed to them into the picture. Voila, telepathy again unbeknownst to the user starts up some sort of broadcasting system and asks for people to join the channel (or some similar method, the logics can differ a lot) if they have not offered to join earlier. Voila, they get the slides and join the meeting via voip etc.
All in 2-3 extra clicks. Small things that make users really really love the platform.
GNOME is GTK+/C – very high learning curve, difficult to develop on it, you lose time for very easy tasks, while you could spend it for something better – hey, don’t say there are wrappers…they SUCK, and there’s quite NO documentation..
KDE is QT/C++ – professional development environment, easy learning curve – but the DE atm, sucks.. (too many options, and Ks :))
hello there, can we in 2008 use _technologies_?