Google: "Android is the Linux desktop dream come true" (der Standard)
If you look at Android we have lots of partners. We have chipset partners, we have handset partners, we have carrier partners. They all want to use Android and they all want to have something special about themselves. So they want to use Android for that specialness. What that means is that one handset vendor probably doesn't want to interact too much with the other handset vendors because they are competitors. And Android gets caught in the middle of all of this. And the bigger question then becomes how you architect software that it's still useful around that kind of model."
Posted Jul 12, 2011 15:53 UTC (Tue)
by hitmark (guest, #34609)
[Link] (12 responses)
Posted Jul 12, 2011 16:36 UTC (Tue)
by dberlin (subscriber, #24694)
[Link] (2 responses)
Posted Jul 12, 2011 17:01 UTC (Tue)
by pabs (subscriber, #43278)
[Link]
Posted Jul 12, 2011 17:06 UTC (Tue)
by jospoortvliet (guest, #33164)
[Link]
(I agree it'd be awesome if Google works closer with the community but asking them to be more like Maemo is just... bad taste... :D)
Posted Jul 12, 2011 19:34 UTC (Tue)
by Cyberax (✭ supporter ✭, #52523)
[Link]
A lot of solutions in Android would have been a huge NO-NO if developed in community and could have delayed it for _years_ (wakelocks, for example).
Posted Jul 12, 2011 19:59 UTC (Tue)
by nhippi (subscriber, #34640)
[Link] (7 responses)
I mean there is a need to meet half-way instead of expecting the other participant to walk all the way to meet you.
Posted Jul 12, 2011 20:44 UTC (Tue)
by dlang (guest, #313)
[Link] (4 responses)
the kernel developers work with commercial product developers all the time, and do it successfully. but successful interaction also requires that the commercial product developers not do everything in secret and then present the final result as a 'take it or leave it' proposition
Posted Jul 13, 2011 8:02 UTC (Wed)
by nhippi (subscriber, #34640)
[Link] (3 responses)
At and Android and consumer electronics in general, things don't quite go that way. Getting new shiny products out faster than competitors is CRITICAL and nobody cares about long term support. There is no time to argue in open on howto to implement an interface (including bikeshedding and taking in account corner cases that might materialise in a parallel universe with different rules of physics). Likewise, there is almost non existing customer demand for long term support. Once the product is in the market, engineers are already working on the next one.
So, if people want more community interactions from google and other android players, the community collaboration model needs to take in account wider user cases than just servers. And that isn't doublespeak for "accept crap from vendors as-is". The only truly important thing is that community collaboration must not slow product development cycle.
What I want to say, is that if people, like original poster, want google to deal more with community, they need to make sure there is an incentive to do so. "Making community happy" is a surprisingly strong incentive, but if it comes with the price of slowing down product cycle, things are likely to continue like currently...
Posted Jul 13, 2011 19:42 UTC (Wed)
by Kluge (subscriber, #2881)
[Link]
"Likewise, there is almost non existing customer demand for long term support."
Not at all. I, for one, would like more long-term support, but I realize that it's not in the interest of the carrier or device manufacturer.
"The only truly important thing is that community collaboration must not slow product development cycle."
That's far from the only important thing. Android partners also want to be able to market new features when and how they wish. Sometimes that means keeping information from the public (or a public mailing list).
I don't have a solution to that problem, but I think it's clear that it exists.
Posted Jul 14, 2011 5:11 UTC (Thu)
by niner (subscriber, #26151)
[Link] (1 responses)
"At and Android and consumer electronics in general, things don't quite go that way. Getting new shiny products out faster than competitors is CRITICAL and nobody cares about long term support."
Couldn't be farer from my reality. We don't care at all for long term support on our servers, because there's new hardware that needs support anyway and our programmers don't want to wait five years till they can use those new features in the new software versions. Our software would simply not work without those recent PostgreSQL features and developing for Perl 5.8 would be a nightmare.
On the other hand I would so love to get those bugs in my phone fixed. I had no need for a new phone at all if it weren't for those annoyances. But chances of getting them fixed are zero because "once the product is in the market, engineers are already working on the next one."
So my next phone will run 100% free software, because I'm just too fed up with everything else.
Posted Jul 15, 2011 7:56 UTC (Fri)
by khim (subscriber, #9252)
[Link]
First, the question: how much kernel work is going in your organization or sponsored by your organization? If the answer, as I suspect, is "not much", then your experience it not all that relevant to the kernel development, don't you think? Well, this is typical situation for small organization (especially startups) - but these don't pay salaries to kernel developers directly or even indirectly (by buying RHEL subscriptions). In my case we are almost ready to switch... drumroll... from Python 2.4 to Python 2.6. We support both (of course), but most new development is with Python 2.6. And this situation is fairly typical for large companies (who can actually sponsor kernel work). Will you pay for that? How many other people will pay for that? If the answer, as I suspect, is "why should I pay? I already paid for the phone once!" then you've just gotten the reason not to fix bugs in old stuff: there are no money in that. Even radio firmware? That I'm yet to see. All the biggest problems I've had with my phones are radio firmware related (phone shows me as connected, but noone can reach me - this kind of things).
Posted Jul 12, 2011 20:47 UTC (Tue)
by lutchann (subscriber, #8872)
[Link] (1 responses)
Further advantages can be had, free of charge, by additionally observing the rules of the kernel development community.
Who is meeting whom half-way?
Or are you arguing that Google is different, because Android's market share means that Linux somehow "needs" Google?
Not being sarcastic; I'm honestly curious why you think Linux developers owe anything to ODMs/OEMs and commercial middleware providers.
Posted Jul 13, 2011 10:19 UTC (Wed)
by dgm (subscriber, #49227)
[Link]
In the case of contradictory needs, maybe we could find a "creative" solution, or agree on some balance.
Posted Jul 12, 2011 15:59 UTC (Tue)
by dcg (subscriber, #9198)
[Link] (4 responses)
Posted Jul 12, 2011 16:20 UTC (Tue)
by zlynx (guest, #2285)
[Link] (2 responses)
Or did you mean the rest of the Android userspace?
Posted Jul 12, 2011 16:34 UTC (Tue)
by dberlin (subscriber, #24694)
[Link]
So it's more realistic to say "they cannot afford to rewrite it several times when there is no clear decision maker with the authority"
Posted Jul 12, 2011 17:57 UTC (Tue)
by dcg (subscriber, #9198)
[Link]
That said, if Android was developed as a open project, there would be chances to solve design issues way before they are adopted by vendors and app developers. With the current closed development model, it's not surprising that they constantly find issues to merge features back into mainstream projects...
Posted Jul 13, 2011 4:51 UTC (Wed)
by gnu (guest, #65)
[Link]
And add to it the vendor locked propreitary hardware.
Posted Jul 13, 2011 2:58 UTC (Wed)
by gmaxwell (guest, #30048)
[Link] (8 responses)
Android is "Linux" only in the it uses the Linux kernel but the development methodology, the level of openness and freedom, the application APIs, the common software tools, are pretty much entirely unlike the software stack we associate with "Linux desktop".
I'm pretty sure that, assuming compatible hardware, I could swap out the Linux kernel for the Freebsd one on 99% of Linux desktop users and they wouldn't notice.
I doubt anyone had much of a dream of a Linux desktop which only included the Linux kernel, burred away under layers of non-linux-system-stuff with a mostly Java userspace.
Posted Jul 13, 2011 5:40 UTC (Wed)
by sramkrishna (subscriber, #72628)
[Link] (2 responses)
Of course after reading this article, if most of the google folks have linux desktops how are they enjoying Google Linux themselves without a Linux based importer?
Posted Jul 13, 2011 9:04 UTC (Wed)
by petur (guest, #73362)
[Link] (1 responses)
I observe that Google is good at *using* linux and other open/free tools to create their business, but then don't care about supporting that community.
Posted Jul 13, 2011 16:16 UTC (Wed)
by imitev (guest, #60045)
[Link]
I agree Google could do better, but you're not very fair - for instance they invested/invest quite a lot of money in GSoC, which many projects benefit from (maybe not the one Google is *using*, but at least there's some feedback to the "community" at large.)
Posted Jul 13, 2011 12:46 UTC (Wed)
by tialaramex (subscriber, #21167)
[Link]
I suspect it's as sadly mistaken as the idea that Ubuntu, Fedora or similar Linux distributions would be a seamless "upgrade" for many Windows users.
Posted Jul 13, 2011 16:32 UTC (Wed)
by tuna (guest, #44480)
[Link] (2 responses)
I thought Linux was a kernel. So what you are basically saying is that Android is not Fedora or Ubuntu?
Posted Jul 13, 2011 16:51 UTC (Wed)
by gmaxwell (guest, #30048)
[Link] (1 responses)
Or gentoo, or arch, or slackware, or anything even remotely like them.
Can you really say you've had dreams of a "Linux desktop" which would have been met with by copy of windows with the NT kernel swapped out for the Linux kernel but everything else kept the same?
I don't think it's unreasonable to say that kind of "linux desktop" that anyone would care about having would have more than just the Linux kernel in common with the Linux systems we use today especially since the kernel is basically invisible to desktop users so long as it works correctly.
Posted Jul 17, 2011 9:26 UTC (Sun)
by tuna (guest, #44480)
[Link]
It seems that you are confusing a "Linux desktop" with the current Linux/Xorg/Gnu/Gnome/KDE/More... based free OS distributions we have today.
Posted Jul 13, 2011 19:00 UTC (Wed)
by Pawlerson (guest, #74136)
[Link]
I'm pretty sure it won't work for 99% of Linux desktop users on random hardware. Other would notice slower response times.
Posted Jul 13, 2011 4:17 UTC (Wed)
by Hausvib6 (guest, #70606)
[Link] (2 responses)
Posted Jul 13, 2011 5:44 UTC (Wed)
by fujii (guest, #60630)
[Link] (1 responses)
Posted Jul 13, 2011 12:09 UTC (Wed)
by Cyberax (✭ supporter ✭, #52523)
[Link]
Epic fail.
Posted Jul 13, 2011 6:42 UTC (Wed)
by YangBaxter (guest, #25377)
[Link] (1 responses)
In my Linux destop dream, we have open software that just runs
Posted Jul 13, 2011 8:47 UTC (Wed)
by lkundrak (subscriber, #43452)
[Link]
Posted Jul 13, 2011 6:44 UTC (Wed)
by wertigon (guest, #42963)
[Link] (1 responses)
Posted Jul 14, 2011 22:51 UTC (Thu)
by jjs (guest, #10315)
[Link]
Posted Jul 13, 2011 8:15 UTC (Wed)
by Uraeus (guest, #33755)
[Link]
Posted Jul 13, 2011 13:16 UTC (Wed)
by xxiao (guest, #9631)
[Link]
Posted Jul 13, 2011 21:52 UTC (Wed)
by shmerl (guest, #65921)
[Link]
Posted Jul 14, 2011 0:45 UTC (Thu)
by jch (guest, #51929)
[Link] (4 responses)
Where I come from, a Linux desktop is free of anti-features designed to meet the desiderata of the telecom companies (or any other third party).
A case in point: support for ad-hoc wifi networks. The support for that exists in wpa_supplicant, but apparently Android's upper layers deliberately filters out ad-hoc networks. Support for ad-hoc is issue 82 in the Android bug tracker; at the time of writing, it has 1005 comments.
Android might be a popular Linux desktop, but it certainly doesn't represent what Free and Open Source software stands for. For us Free Software hackers and enthusiasts, it is at best irrelevant, at worst a costly diversion.
-- jch
Posted Jul 15, 2011 8:09 UTC (Fri)
by khim (subscriber, #9252)
[Link] (2 responses)
Got that: Linux desktop is by your definition is super-ultra-niche-product which few people ever care about. Because in real world there are always some third party which acts as mediator between developers and users (be it Canonical, RedHat or Verizon) and of course what the end users see is limited by desiredata of said entity. Nope. It's hugely relevant. Since your Linux desktop is amusing obscurity it needs someone else to push the compatible hardware platform: traditionally it was Windows, nowadays it's Android. Still much better then BlackBerry or Windows Phone 7, don't you think?
Posted Jul 15, 2011 13:33 UTC (Fri)
by jch (guest, #51929)
[Link]
The difference, of course, is that when I buy a licence for RHEL, I am RedHat's customer, and RedHat has no interest in throwing anti-features at me. Not so with Android -- it's the carriers that Google needs to cater to.
> your Linux desktop [...] needs someone else to push the compatible hardware platform
Indeed, that's a good point.
-- jch
Posted Jul 18, 2011 17:33 UTC (Mon)
by shmerl (guest, #65921)
[Link]
Posted Jul 15, 2011 11:59 UTC (Fri)
by sorpigal (guest, #36106)
[Link]
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
This is about money, as usual...
Couldn't be farer from my reality.
We don't care at all for long term support on our servers, because there's new hardware that needs support anyway and our programmers don't want to wait five years till they can use those new features in the new software versions. Our software would simply not work without those recent PostgreSQL features and developing for Perl 5.8 would be a nightmare.
On the other hand I would so love to get those bugs in my phone fixed.
So my next phone will run 100% free software, because I'm just too fed up with everything else.
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Android is not the Linux desktop dream come true
> interact too much with the other handset vendors because they are
> competitors.
and hardware vendors who are concentrating on their business:
competing for the best hardware possible.
Android is not the Linux desktop dream come true
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
by closing Android? in that sense it will be more like Window$ as far as desktop goes.
Google: "Android is the Linux desktop dream come true" (der Standard)
Google: "Android is the Linux desktop dream come true" (der Standard)
Well, Ok...
Where I come from, a Linux desktop is free of anti-features designed to meet the desiderata of the telecom companies (or any other third party).
Android might be a popular Linux desktop, but it certainly doesn't represent what Free and Open Source software stands for. For us Free Software hackers and enthusiasts, it is at best irrelevant, at worst a costly diversion.
Well, Ok...
Well, Ok...
I took the "Linux desktop dream" comment another way: It means that Android is, finally, a user-centric Linux-based system which is successful in the marketplace. It's not the Linux desktop we want, necessarily, but it's the massive end user adoption of Linux that desktop Linux advocates have been seeking, even if it's taken a form that we do not prefer.
Google: "Android is the Linux desktop dream come true" (der Standard)