Andy: The more I think about it, the more I think ‘where do we find capital’ is not going to be an approach that is successful for us as a community, despite what I’ve said about .gnome. Centralized capital means a centralized point of failure; it means reduced competition; it means that the little guys on the end points have a harder time getting involved. And it means problems scaling- the foundation is just never going to be able to support something like flickr or last.fm for all GNOME users, IMHO- the requirements are just too large. And that is even with our current user base.
Mike Linksvayer is probably right: what free software needs is a developer-friendly, user-friendly p2p platform, so that we can do all the things flickr and others do, but do it with shared bandwidth instead of centralized bandwidth. Hard, I know, but quite possibly necessary. Maybe we need to beg the Coral CDN guys for help :)
It is worth noting that integrating GNOME and the web isn’t just about innovation for end users, though that needs to drive everything we do. (Just being ‘usable’ and Free/free isn’t good enough, sadly.) We need to think about developers, too- in the future, Microsoft, Google, etc., will all be offering servers and services along with their desktop API offerings. Tim O’Reilly has a great bit on this here. Money quote: “Being a developer ‘on someone’s platform’ may ultimately mean running your app in their data center, not just using their APIs.” We need to think hard about that future- I’ve always thought that we spent too much time competing with the Windows 95 user experience, but it turns out we’re still competing with the Windows 95 developer experience too. More on this from Jon Udell and O’Reilly, again.
Finally, before I run off for the night, I don’t think the foundation’s problem is a conflict of interest, though I agree that Novell/RH/Sun/etc. and the community are not necessarily always on the same page with regards to brand strength. The big problem with the foundation and sponsors (and really in large part the community) is that in large part everything all three groups do is target established markets and needs. Apple, Yahoo, Google and Microsoft are most dangerous to us when they go out and create new markets (iPod, photo sharing, etc.) and then lock us out of them. Novell, RH, and Sun are all primarily in the business of targeting existing markets (mugshot, whether you think the software is good or not, is an attempt to break out of this rut), and in general the community is in the habit of targeting the last proprietary software many of us used- mostly Windows 98. There is (sadly) little conflict there- we’re all on the same, old, page, and we need to get out of it if we’re to advance. That’s bigger than any silliness about brand or competition between GNOME and the vendors.
Edit later: The excellent Kragen Sitaker also reaches the P2P conclusion, though for different reasons, in a really thought-provoking piece here.
But what if the GNOME applet had (optional) seamless access to Flickr, and I could browse that way? Then cache the pictures I like and were available in the right sizes and under the right license terms? I’d be a fan. Now I know that Luis and others believe that P2P is the way forward, not centralized services such as Flickr, but one is here now and the other is not. Would a Flickr back-end be an earth-shattering feature? The kind that would sell a new operating system? Certainly not. But the strength
But what if the GNOME applet had (optional) seamless access to Flickr, and I could browse that way? Then cache the pictures I like and were available in the right sizes and under the right license terms? I’d be a fan. Now I know that Luis and others believe that P2P is the way forward, not centralized services such as Flickr, but one is here now and the other is not. Would a Flickr back-end be an earth-shattering feature? The kind that would sell a new operating system? Certainly not. But the strength
[IMG RSS] Planet Gnome Thomas Thurman: Dentist NumbersLuis Villa: more ramblings on GNOME and the webبهداد اسفهبد: Cairo 1.2.2 and Pango 1.14.0 releasedBryan Forbes: Jogo da VidaAndy Wingo: the verb calificar and the dance sardanaJohn Palmieri: The easy way to help out smoketesting the GNOME release
I rather made a livejournal post about this. Here’s a copy and paste:
“maybe Open Source never seems innovative because we’re always trying to measure it against proprietary software. What the hell defines innovation in software though? I consider Beagle, Compiz and the evolving X servers, and GLSCube to all be innovative software. Is it because we say well Apple has spotlight and aqua and microsoft has WinFS? Beagle came before spotlight and microsoft cancelled WinFS. As for Compiz, plugin based compositing manager is almost an innovation by itself. Aqua certainly doesn’t offer that. Isn’t the greatest innovation the very openness of it though and the consequential freedoms it offers? In addition, I do think Open Source has identity and uniqueness. It’s not revolutionary but I don’t consider Mac OSX or Vista to be either. We’re still stuck in the desktop paradigm of the 90s with a little net fucntionality thrown in as a 21st century afterthought. Our problem isn’t so far from that of Vista or Leopard, either the technology for a revolutionary desktop is lacking or such an advance would go unadopted as it would be too radical a departure from the current state of things. Is our innovational limit our user base? What exactly is holding us back and if nothing was, just where the hell do we want to go? What is the goal of the innovation? Wouldn’t it be to get the machine to be as efficient towards accomplishing various tasks as possible? In this regard, isn’t this simply transparency of OS and Applications in general as an end goal? You either reinvent the computing paradigm or you don’t. If innovation is just successful experimentation though, then I think Open Source is doing damn fine.”
Regarding targetting current needs versus creating new markets: We are doing both, gradually, and we need to do both.
However, I too worry about all the new opportunities for lock-in. Sooner or later locked-in products become shitty (due to the lack of market pressure experienced by monopolies), creating incentives to switch despite the difficulties of doing so, but it would be nice if we could avoid that wrenching experience.
Hmmm…There is a very big difference between trying to create a distributed system on top of a few hundred (or thousand) highly reliable nodes on a fast network and trying to create a distributed system on top of tens of thousands of occationally connected, highly unreliable nodes on a very, very slow network. If you want numbers read this: http://www.usenix.org/events/hotos03/tech/blake.html The title “High Availability, Scalable Storage, Dynamic Peer Networks: Pick Two” says it all.
It might be worthwhile to talk to people who are already into P2P and also have an affinity towards free software. One thing I suggest to check out is the presenters list from CodeCon (http://www.codecon.org/). I’ve only ever been there once, back in 2002, but it was an absolutely amazing bunch of people.
One note of caution, though: Don’t let the perfectionist approach that is sometimes present in the lurker scene of P2P get to you.
Murray: example of us creating a new market, besides mugshot? (I suppose also Maemo, maybe.)
Jay: interesting link, thanks. Like I said, the p2p approach is clearly not easy, and I’ll admit potentially not possible, but I think it is worth trying because it is the only one that maps well to how our community developed and works.
Brit: I don’t think efficiency really counts as innovation in most cases. It certainly doesn’t help us get more users, either. I’ll add that Compiz and Beagle are definitely *not* innovative. Good and useful software? Absolutely. Doing something that had been talked about by programmers for years? Absolutely. The web guys are using their platform to do things no one else in computing has ever done- global availability, easy collaboration, easy sharing, etc. The competition is not currently Vista or OS/X- if you’re comparing us to what those two have released now, you’re looking at the wrong competition. Instead, we are competing with web apps and with the web-backed Windows and OS/X of the future- both of which are coming at us with collaboration and distributed data baked in, eventually. These are complex problems which will take years to work on, so we have to start now.
> Murray: example of us creating a new market, besides mugshot? (I suppose also Maemo, maybe.)
Well, I can conveniently define a new market as innovation and new user experiences, here are some things that we did or are doing. But we’ve mostly either failed to execute them (and hence been overtaken, hopefully not irretrievably) or are failing to execute them just yet. It’s not an exhaustive list.
– Standards-based telephony (Ekiga and/or the Telepathy)
– Standards-based instant messaging (Jabber with Gossp and/or Telepathy)
– Annotated self-created, self-mixed, hyperlinked, open licensed, democratic, community-forming video (Annodex)
– P2P file sharing (IFolder?)
– Music sharing (Various things in various music players, mostly Jabber based, I think)
– Blog feeds and aggregation (I think this is working – it hasn’t gotten away from us yet)
It would be nice to really execute our ideas. That might require money. One positive thing is that most of these services/experiences could be supported by (targeted) advertising. If I was a millionaire I’d see in them a way to make money by creating a new commodity, without the need to shaft my customers.
Luis, final thoughts on this. No more pestering, I promise.
Again with the cut and paste from the livejournal dealie:
“GNOME is getting ready for the Web app war. Yes, web apps are growing in importance and are often easier to develop than desktop apps as well as offering features (collaboration-wise) that desktop apps are hard pressed to match or compete with. There are a few reasons we don’t need to rush though. First, the desktop paradigm as it stands today will last at least another 5 years, probably 10. People seem to forget this in the Web 2.0, web apps = salvation, etc. rush. Second, we need to give it a bit of time to cook. Yes, the desktop’s days are numbered. They always have been. We’re not there yet. Yet. In the mean time, let other people experiment and observe what works. Open Source is good at this. We should do a few trial runs and implement our own collaboration-enhacing innovations (like Telepathy) where we can\see fit but we don’t need to storm ahead blindly wihout having a clue what we’re getting into. We talk about Topaz being a completely new desktop experience but just how new is something we should be very careful about. We want to be forward-thinking, not forward. When you’re forward you actually pass the user base at a certain point and this leads to dead tech. There are plenty of visionary techs that never make it for this very reason. We can get ahead of the pack without being a horse ahead of a carriage, right? Third and finally, we’re not done with the groundwork. Of course, we never will be. That’s one of the pros of Open Source. Refined ad infinitum. Sometimes it complicates but generally speaking, this is a good thing ™. There’s plenty to do to get the Linux Desktop Experience to compete on the DESKTOP level. Besides web apps will always need a browser and an OS to speak to the network card. We’ve got some time to prepare for the next war. We’re already taking first steps.”
[…] For those of you out there wondering why not simply use Writely, I think this approach offers some advantages. For example, you can edit your documents using a full blown word processor with all it’s capabilities. In my opinion, hacking a word processor in a browser is a bad idea. On the other hand, with Writely you only need a browser to access your documents, which is a massive plus. Oh, and finally, this approach might finally shut Luis up with all his we-should-integrate-the-web-with-our-desktop-whining :-P […]
[…] The document has moved here. […]