Mark Shuttleworth Talks About What Ubuntu Contributes

[achaios]

Visions and marketing my a$$.
The question here is simple:
"does Canonical contribute code to open source"?
And the answer is a big NO. They are just taking credit for other people's work. The rest is fanboyism and tribalism.

For example read this:

https://www.linuxfoundation.org/sites/main/files/publications/whowriteslinux.pdf

Red Hat, IBM, Novell, Intel etc.
Canonical is nowhere to be found.

[allquixotic]

But Canonical does create code (do read the Mark's blog post, it's an interesting read). They may not spend as much energy on Mono/Gnome/[your favorite app] but they write code and share it with the rest of the ecosystem as they should.

Frankly, this "Canonical doesn't write code" mantra is bollocks.

I never claimed Canonical doesn't write code. It's the way they go about it that people question. Papercuts more or less stands on its own as the only example of Canonical putting their own staff to work on Free Software projects that are not strictly service-based.

We've got the Ubuntu One platform (including the music store), which is highly monetized, and AFAIK the backend is closed source. If users decide they don't like Canonical's pricing model, they can take it or leave it. The client code is open, but the backend is run as a tight ship, not unlike iTunes or any of a myriad of other walled gardens. Even opening up the cloud storage portion of Ubuntu One would be generally useful, as an alternative to something like FTP or Samba. I prefer the user interface and integrated nature of the Ubuntu One client to something like FileZilla, but it's simply not possible to set up your own Ubuntu One server without starting a reverse engineering project from scratch.

Then we've got Launchpad, whose open source code is so difficult to decipher and set up that most people give up and just use the official launchpad.net installation. I went from zero to having an FTPES server backed by an OpenLDAP directory server over SSL in about 12 hours. I set up a mail server with IMAP+SSL, SMTP, roundcubemail, and local mboxes in about 4 hours. I accomplished both of these tasks as a relative "newbie" to these environments. There were a ton of configuration files to edit, and a mind boggling array of settings and options. It was hell -- or so I thought, until I started playing with Launchpad.

I spent over 3 days trying to configure even a basic Launchpad service on my own box, and gave up without being the least bit successful. Canonical has an incentive to make Launchpad the open source project as sysadmin-unfriendly as possible, while making their own (monetized) Launchpad.net as user-friendly as possible. Are they intentionally making it difficult to install? Your guess is as good as mine.

My point is that "use and give back to upstream" has been a necessary mode of operation for the success of almost all free/open projects. Well, that's a bit black and white; I will qualify it further as "use, and give back, to the extent you are able". So we don't expect Grandma to be able to contribute anything back, unless she happens to be a technically savvy grandma with a career of state-of-the-art software engineering behind her! The more capable you are, and the bigger (in terms of manpower) your entity is, the more that is expected of you, relative to the amount of useful work you get out of the software as a user.

But look at Canonical. By your own insistence BlackStar (and I agree with you 100%), Canonical does write code. They have a staff of professional software engineers. But how do they invest their engineering resources? Do they work on generally useful projects that are fully open source and applicable to a wide array of Linux distros? On the whole, not really. Do the contribute back to the specific projects to which they owe the majority of their success? On the whole, not really. I will be interested to see whether Canonical ever steps up their upstream contribution as their revenue stream increases with their popularity and business acumen. If they become bigger than Red Hat but still contributing less than Gentoo, we'll know for sure that something is very wrong with their philosophy, and it will be abundantly clear that they have defied the tacit contract between distributions.

[BlackStar]

I have yet to see a Papercut that isn't upstreamed. Launchpad may be hard to setup but that's not saying anything.

Ubuntu One is kind of a peculiar thing, simply because noone is using it (guess what, most are using DropBox which is even more closed) - and while I don't like how Ubuntu One is being developed, I am willing to cut them some slack. It's not as if Novell et al(*) don't release closed-source applications either.

[Panix]

Maybe it's just me but it seemed from the article/blog, that the justification is (btw, simplified) we are responsible for recruiting a lot of people to Linux and open source software so we should be allowed to be exempt from significant code contribution upstream. Does anyone else wonder if this is a reasonable perception?

If you've watched any of those videos, the Novell employee who (yes, he uses OpenSUSE), seemed to take a few digs at Ubuntu for this in his presentation.

Ubuntu's marketing was exceptional and imho, a significant reason why they were able to build the project they did and of course, their devs and all the contributors as well but more than any other distro, the marketing was extensive.

I think Debian is left out of the praise and they do a lot that Ubuntu is able to benefit from. Yet, they don't seem to have hardly any marketing hype. I think the contrast is interesting. Anyone else think so?

[monraaf]

Mark Shuttleworth Talks About What Ubuntu Contributes

I was hoping to read something concrete, but yet again the usual blah blah about human generosity and the obligatory children reference. Always a good thing to throw children in your pitch, as any skilled politician can tell you. I can't believe people are actually buying this.

[mugginz]

Yet again I see no good deed goes unpunished

[Hoodlum]

I was hoping to read something concrete, but yet again the usual blah blah about human generosity and the obligatory children reference. Always a good thing to throw children in your pitch, as any skilled politician can tell you. I can't believe people are actually buying this.

Pretty much and unfortunately there has been no increase in measurable Linux use since the arrival of Ubuntu, either. Fortunately in the last few months alone Android has done a much better job in bringing people to Linux.

[crazycheese]

If they become bigger than Red Hat but still contributing less than Gentoo

Man, I agree with you 100%, up to this point RH, Debian and Gentoo are MAJOR contributors. You cannot say bout "Ubuntu contributing less than Gentoo". Its like comparing contribution rates of 1 to 140. You cannot compare (current) contribution amount of Gentoo to Ubuntu. Ubuntu vs Mandriva, maybe

[crazycheese]

I think Debian is left out of the praise and they do a lot that Ubuntu is able to benefit from. Yet, they don't seem to have hardly any marketing hype. I think the contrast is interesting. Anyone else think so?

Yes, I agree. It like serious grown-up vs schoolboy. All fancy stuff & talks, but zero output. Ehe.

[pingufunkybeat]

Man, I agree with you 100%, up to this point RH, Debian and Gentoo are MAJOR contributors. You cannot say bout "Ubuntu contributing less than Gentoo". Its like comparing contribution rates of 1 to 140. You cannot compare (current) contribution amount of Gentoo to Ubuntu. Ubuntu vs Mandriva, maybe

RedHat and Novell are major corporations, as is Nokia. They can afford paid developers, and it's right to expect them to.

But Debian and Gentoo are community projects. I don't think that they fund any developers at all.

I don't see why they are expected to contribute more than the most widely used distribution which makes money selling Linux.