LWN: Comments on "LC Brazil: Consumers, experts, or admins?"

CZFree.Net

wentasah — Thu, 09 Sep 2010 20:27:11 +0000

Maddog's project sounds like Brazilian variant of CZFree.Net. It is a network run by volunteers around the whole Czech Republic. It started at times when we had a single company which had de-facto monopoly on telecommunication services and the only available option for having internet at home was the classical 56k modem with connections charged by time spent on-line.

People started connected their houses together with Wi-Fi, home-made free-space optical links and various other technologies. Every part of the city had one or more small ISPs and people learnt a lot about the networking technology, Linux, FOSS etc.

Now we have a normally working telecommunication market and commercial companies typically provide better services than a bunch of volunteers, but many parts of the network professionalized and provide professional networking services now. I wish Project Cauã success because CZFree.Net was really a great experience for me.

Canonical

dgm — Thu, 09 Sep 2010 14:46:18 +0000

Yes and no. Releasing everything together has is benefits too. You only need to test one version of every app, for instance. Having to test many versions of every app for regressions before updating a library increases the testing burden quite a bit.

Canonical

kragil — Wed, 08 Sep 2010 20:56:54 +0000

The _vision_ (long term) has no more GTK. Older documents were clearer in that they clearly said that GTK is for lecacy/compatibility.
Other pages on meego.com make that fairly clear
http://meego.com/developers/meego-developer-story
http://meego.com/developers/meego-api

Anyways, the point I wanted to make is that Canonical wants Ubuntu to be its own thing and that is not only the due to the OEMs.

Canonical

jspaleta — Wed, 08 Sep 2010 18:26:24 +0000

Pardon. No gtk in it?
http://meego.com/developers/meego-architecture
I see gtk and friends listed in middleware. And the available meego 1.0 live images for netbooks still depend on gtk heavily do they not? It's only the other targets like the handset target that is qt based.

No matter. Even if it moves completely to Qt and uses the same infrastructure that underpins KDE's netbook interface its still a _differentiated_ UI purpose built for consumer device targets. So the point I'm trying to make is still valid. Meego like Moblin and Maemo before it are all purpose built differentiated UIs that leverage technologies that are shared with other interface concepts.

The fact that Meego envisions jumping toolkits is immaterial to the market forces driving differentiated UI in consumer devices. Though I will say its interesting that no one in the laypress..even here..has really picked up on the touch framework work that meego has been slugging away at quietly leveraging qt. My understanding is the meego handset and in-vehicle 1.0 releases both make use of the meego touch framework. It would be interesting to see a well written comparison between meego's touch framework and the framework that Canonical put together.

-jef

Canonical

ccurtis — Wed, 08 Sep 2010 15:38:04 +0000

Though it is also against my nature, I personally prefer "more thinking, less writing" over the alternative.

Canonical

Trelane — Wed, 08 Sep 2010 14:43:32 +0000

Last I knew, it was NOTABUG. I tend to agree with this assessment.

Canonical

kragil — Wed, 08 Sep 2010 11:30:59 +0000

When I say Gnome 3.0. I mean Gnome-shell.

Editor: How is the temporary edit button coming along. Thinking before writing is not my cup of tea.

Canonical

kragil — Wed, 08 Sep 2010 11:29:24 +0000

Well, Nokias(and Intels?) vision for Meego has no GTK in it. I wouldn't call that Gnome tbh. The GTK/Clutter stuff is Intel/Moblin legacy cruft.
And I wouldn't blame the OEMs as much. Canonical doesn't want to be your regular Linux distro vendor anymore. That is why I said that Ubuntu _Desktop_ (think even 11.10) will most likely not ship with Gnome 3.0. Canonical can't outdo RH/Novell on that one, so they will avoid that battle altogether and ship a pimped Unity (or something along those lines) with the default desktop. They picked that name for a reason, they want it on netbooks, desktops and touch devices.
They still profit from new Gnome technologies (Mutter, Clutter, etc), but what the user sees will be totally under their control and effectively create (albeit very weak) lock-in. Building your own UI ontop of Mutter/Clutter/JS is still hard, but not nearly as hard as it used to be.

Canonical

ummmwhat — Wed, 08 Sep 2010 08:08:01 +0000

And that: "the Canonical Design team regularly undertake formal user research. All our research is always shared under the Creative Commons Attribution license."

Canonical

dlang — Tue, 07 Sep 2010 22:51:41 +0000

I have no problem with kde apps being shipped and updated separatly from KDE (or gnome apps separatly from GNOME)

I think this would be a very good thing.

there is desktop infrastructure, and then there are desktop applications. Most applications can be used on either desktop nowdays, and I personally think that the attitude that you must use KDE to use any KDE apps (or GNOME to use any GNOME apps) is preventing competition between the apps developed for the different desktops.

Canonical

nicooo — Tue, 07 Sep 2010 22:42:13 +0000

My point is that most of the apps that kde uses could be released individually since there are already some that do so. Also, IMHO it's a lot harder to test when everything is released at the same time. If a program crashes after updating a single library it's easier to find the problem.

LC Brazil: Consumers, experts, or admins?

smoogen — Tue, 07 Sep 2010 21:54:28 +0000

I would guess that the answer is exactly what you fear. This business has a high risk level and not a high enough return on investment... and US businesses usually only go into that region if there is a government incentive or penalty not to. Add in virtual monopolies on telecommunication and your profit potential goes way south.

why such approaches are not tried in the US

smoogen — Tue, 07 Sep 2010 21:47:31 +0000

It depends on how you define Internet access... if you are talking greater than 1.5 Mb/s then there are large parts of the US without that. And you might actually get better coverage in some cities in Brazil because they have 4G networks (or some equivalent). My guess is that the maddog network would be about putting dishes on top of buildings and doing networking that way as its easier than digging up roads etc. [Though it depends on the city and area of city in Brazil.]

LC Brazil: Consumers, experts, or admins?

martinfick — Tue, 07 Sep 2010 20:29:03 +0000

This project does not sound like it is about simple ISP service, since in this case the buildings are already supplied with wholesale ISP service. I assume that a big piece of this project is the intended cost savings over the otherwise likely monopolistic ISP service. Do you think even wealthy US residents would not benefit from potential cost savings and potential improved service?

Not to mention that I suspect a healthy percentage of US buildings to not have ISP service. Unless you live in a metropolis, most buildings in the US likely house lower income families. But, even a small percentage would still be a very large number of individuals that could benefit.

Canonical

dlang — Tue, 07 Sep 2010 20:28:58 +0000

I would not consider any of those apps as ones that would be hard to update.

however a new version of QT, GNOME, KDE, glibc, etc or changing what compiler is used to compile the distro would require a LOT of testing to make sure that it works well with everything else. These are the types of things that I think work well on a six-month upgrade cycle where several of these things get udated at once as opposed to a full rolling-update distro like gentoo where these may appear at any time.

Canonical

drag — Tue, 07 Sep 2010 19:31:16 +0000

There are a large number of ways to contribute to 'Linux and open source'.

If Canonical continues with things like http://design.canonical.com/ then that is a huge boom for Linux desktop efforts in general. Even if they do not contribute a lot of code back upstream, simply taking the code they can get and making it perform correctly and be friendly will show Gnome developers in what directions they should proceed. It's one thing to just blindly write software that you think will provide benefits to people... it's quite another writing software you _know_ will provide benefits.

Even if Canonical does not produce a single line of code back to upstream their efforts have contributed far more for the Linux/Open source desktop efforts, specifically, then any other distribution I can think of. Just having something friendly for non-technical people is valuable in it's own right and after using a number of different distributions over the years I can say that Ubuntu has gone the farthest in user-friendliness then any other distribution despite it's technical deficiencies and numerous mistakes they've made.

> I still think RH has the best and most sustainable FOSS business model so far, hell

Yes. Proof is in the pudding, (when profits == pudding) Making money and finding a way to be valuable to many different organizations is a huge huge win for Linux. Redhat's contributions back to OSS is probably the most valuable any corporation has done to date (in general). Fantastic company.

> I might even switch back to Fedora if their update madness can be tamed (e.g. not include brand new stuff like SystemD etc.)

Funny. I was thinking of switching to Fedora specifically because of things like systemd. :)

Canonical

nicooo — Tue, 07 Sep 2010 19:28:17 +0000

> there are parts of a linux distro that tie in with lots of other things in the distro and are rather hard to upgrade independently (things like KDE, GNOME)

They should be able to upgrade independently. Right now I'm using k3b, ktorrent, kid3 and digikam. I've never had problems with any of them.

why such approaches are not tried in the US

clugstj — Tue, 07 Sep 2010 19:12:15 +0000

Where are you going to find large apartment buildings with no internet access in the US?

Canonical

jspaleta — Tue, 07 Sep 2010 18:59:05 +0000

I really don't think you can single Canonical out about wanting to offer a differentiated user interface with their Unity offering. They aren't the only ones building differentiated interfaces. Meego has a differentiated user interface which leveraged existing Gnome technologies. Litl's interface is essential a differentiated (and proprietary) user interface on top of GNOME technologies (and derived from Ubuntu as well, but doesn't actually use any of Canonical's in-house built technologies afaik).

Canonical's OEM servicing business strategy really puts them in a tough position because its ultimately OEM interests which are driving what Canonical is building in-house. If OEMs want a differentiated stack and want it by a specific delivery date... who other than Canonical is there expressing an interest in being paid to do the work? If there were not OEMs looking for differentiated stack, Canonical wouldn't be building them.

Take for an historic example the Mi interface created by Canonical and paid for by HP. Unity could end up just like the Mi interface, functional and utterly forgotten once its clear that the software doesn't actually help the sponsoring OEMs sell devices. We really won't know the fate of Unity until we start seeing how the Unity based OEM pre-installs from Canonical partners fare in the marketplace.

-jef

LC Brazil: Consumers, experts, or admins?

martinfick — Tue, 07 Sep 2010 18:30:56 +0000

"With funds obtained in this way, Maddog is able to say that the project is being launched with no government money at all."

He is able to say anything he wants, :) but that doesn't meant it isn't a lie or at best deceptive. Taxation is taxation, even if the money never passes through the governments coffers on the way to its destination.

It does sound like a great project though. I might ask why such approaches are not tried in the US (and I hope that the answer has nothing to do with the funding approach)?

Canonical

dlang — Tue, 07 Sep 2010 18:12:36 +0000

I can definitely see an advantage to their approach.

there are parts of a linux distro that tie in with lots of other things in the distro and are rather hard to upgrade independently (things like KDE, GNOME)

on the other hand, there are a lot of things that really can be upgraded independently (firefox, openoffice, games)

having a distro that split between the periodic update and the rolling update models, doing periodic updates for the things that have lots of interdependencies and rolling updates for the things that don't could be a very useful variation.

even better would be if the user could specify which packages get modified in which manner (with suitable defaults).

Canonical

kragil — Tue, 07 Sep 2010 17:59:06 +0000

From where I'm standing it looks like Canonical wants to create a real Windows/OSX competitor by doing things separate from rest of the community and emulating Apple and MS.

By now I think Ubuntu will _not_ get Gnome-shell as a default. More likely is a pimped up version of their Unity shell. I think Mark read "Differentiate Or Die" a few times too many(often?). By now he wants Ubuntu to be the platform for for-pay cloud/music/app/digital services/goods and be its own separate thing.

Problem is that Google has the exact same idea with Android and ChromeOS, but with much less for-pay stuff and enormous industry support.
I still think RH has the best and most sustainable FOSS business model so far, hell, I might even switch back to Fedora if their update madness can be tamed (e.g. not include brand new stuff like SystemD etc.)