The GNOME Commentary

Sometimes I get upset over copyright, most recently Lily Allens public statements and requests that copyright be enforced with increasing strictness and demands that violators be persecuted harshly managed to trigger me. This is not about Ms. Allen’s unconsidered statements which artist Dan Bull and others already addressed. This is about a specific example of creativity that is very likely not legal but which benefiting everyone.

An artist going by the name Colorpulse recently released a piece called A Glorious Dawn which mixes samples of late astronomer and science popularizer Carl Sagan from his fantastic mini-series Cosmos to produce a truly fantastic song. A song that in all likelihood would cost him his house and livelihood should the current rights holder elect to sue him. Here is why this would be a bad idea.

Cosmos ran in the 80’s and most of todays young people have not seen it at all, despite the fact that it is a wonderful mini-series that while that old is still very relevant in todays world. Carl Sagan more than any scientist in recent history has managed to convey why science is both beautiful and fascinating and why looking up at the stars should be a source of inspiration and wonder as well as understanding. Sagan inspired millions to pursue a life of seeking knowledge, he explained science in a way that got people excited. All that is sadly being lost now and Carl was robbed from this world by cancer more than a decade ago.

Without belittling the efforts of such people as Stephen Hawking, Neil deGrasse Tyson and others I hope, I would state that there isn’t any single scientist who has managed to take his place and continue furthering the public understanding of science with the same vitality and excitement as Sagan did. Most in the younger generation have not had a chance to get to know his work, his books or his bubbling enthusiasm for understanding nor anyones popular works of science for that matter. We are seeing bits here and there in popular entertainment but gone are the days when Carl Sagan would wow audiences with tales of recent discoveries from the cosmos we all inhabit on The Tonight Show with Johnny Carson. Science isn’t for the public anymore and it suffers because of it.

This song is a tribute to Sagan and by providing a catchy way for the younger generation to learn of his work, a fraction of them might go and watch Cosmos. Some might even buy it or read one of his excellent books (personally I recommend The Demon-Haunted World and Pale Blue Dot). This would generate revenue and benefit society as well.

The young people for whom this experience proves to be the spark that lights a lifelong love of knowledge might grow up to be engineers, doctors, astronomers, biologists, teachers. For those that do not manage to get involve with science or do not wish to, one might hope that such an experience instilled at least a respect science and those who use it to let the universe reveal it’s workings to us and an understanding of the importance it plays. One can still look at the sky and wonder even if one does not do it through a telescope, I catch myself doing that on night walks and the experience and wonder is overwhelming even when I acknowledge that I alone will never understand it all. I am happy knowing that a multitude of people around me like me understand little bits and together we and generations to come might piece those bits together to form the whole picture.. still it remains beautiful, awe inspiring, terrifying and yet simply wonderful.

A Glorious Dawn is also a fantastic display of creativity, clearly Colorpulse takes something that was Sagans and makes it his own. This kind of experimentation and playful exploration of culture could lead anywhere. The people who discover their way of expressing themselves this way might end up enriching society in ways we could never have imagined. I thank Colorpulse for enriching society, for bringing joy to my heart and for making me think.

Killing that before it happens deprives us all.

Finally, it only took 3 months, 2 of those with a fix in LP

My faith in humanity is restored.

Yesterday as I was typing up yet another article explaining why people should use and contribute to Fedora, I noticed that my template list of important improvements to Linux done in Fedora was getting several pages long. Fedora does a lot of important work, much of it fueled by Red Hat but I often feel like there is ample room for other companies to take Fedora other places, interesting places we didn’t think of. To do with Fedora what we might not dream, things like building a nationally certified OS for schools and the interesting demands that would put on the distribution. Porting and developing interfaces for netbooks, and many other things which might ultimately blow our minds and take us new places.

But are we being hampered by Red Hat as our sole sponsor I wonder, while they unquestionably are doing a lot of good, there is much they cannot do or have no interest in doing. Fedora also does not seem to have been able to attract much additional corporate interest, which I hear from talking to such people is due to them not wanting another company to have the ultimate deciding vote over their work. So today when I read that Intel were handing over Moblin to The Linux Foundation while remaining key contributors I started thinking what if Fedora did something similar. Opening up Fedora to multiple contributors with a foundation at the helm rather than a single company. It would be bold move, I doubt Red Hat would be entirely opposed to doing so but also considering their historically very generous inventments in Fedora it would be unfair not to give them credit and a say in Fedora and it’s future direction. Something they might rightfully fear they would not get in that new situation.

I do believe that in a foundation of equal partners we could attract more big investors to take Fedora and scale it beyond its current limitations. Be those architecture ports or deployment scenerios. I also believe that it would fundamental be good for Red Hat as well as Fedora to help make such a move happen, they would tab directly into important markets and while their adjustments for their RHEL products might be bigger this way the gains in support and a wider community would amply make up for it I think.

As a Fedora contributor I always cringed when I heard Fedora mentioned as Red Hat’ community offering or simply as Red Hat Fedora Linux. I always felt like deep down Fedora was it’s own and attributing it all to Red Hat with the implication that it was their unsupported public beta tree was rather untrue and unfair towards non-RH contributors (which now make up more than 50% the last time I heard numbers – corrections welcome), I contributed because it was a best possible place to do so, I believe if we make it so it can be the best possible place to contribute for everyone. Be they private individuals, governments or companies I think Fedora has a lot of offer and I think that moving it beyond a one company sphere of dominance it could become _the_ reference Linux platform. I think that would make good business sense to strive to be that for Red Hat, I think it would be easy to get there if the freedom and will exists to seek that position. It would help create a real world standard, one that would work unlike LSB and other some what failed attempts at standardizing Linux as a whole.

Following a lot of soul searching, mostly based in the ton of friendly mail I have gotten lately I have decided to renounce my evil ways and remove all of Mono from my machines. I also have applied to be part of the Rhythmbox development team so to bring computing into the 21th century.

I plan to rewrite as much code in the only true language there is, ASM. thank you very much for showing me, through repeated use of sound reasoning and flawless logic, the err of my ways. I will forever be in your gratitude oh great ones.

Don’t forget to check upcoming blog posts which will soon be aggregated via BoycottNovell.om

For me personally only really the x86_64 architecture matters, I do have an ix86 machine sitting under my desktop but it is humming along nicely. Everyone really supports these just fine, but here is a thing I don’t get. Why do so relatively few distros support ARM and MIPS. Firstly let’s examine why these archs are important, ARM is widely predicted to take a huge bite out of the netbook market, a segment it is extremely well suited for. MIPS naturally is important because it is the foundation for the Chinese YeeLoong chips which we are likely to see the Chinese market gravitate towards. Being a market of a billion potential users with additonal opening to the west via netbooks this is definitely a place Linux distros should be putting some effort, we have literally no competition here, the first distro to integrate really good support will likely be able to pick up millions of new users over a short span of years. This is also a good target because this specific segment has shown signs of being the first truly open platform we might get, from the BIOS and up.

Having a significant chunk of the netbook market is likely going to be important. It is a good market for Linux as the use cases for a properly promoted netbook do not call for “regular” desktop tasks but more likely a specially designed interface such as is see with the EeePC or the Aspire One.

Thus in my search for a new distro to replace Fedora, one requirement I will set up will be support or planned support for ARM and MIPS. Not because I personally need them right now but because it is my belief that an important piece of the future for Linux lies in this direction.

A man I once was proud to call a friend, with whom I have shared many good moments over the years, Jesse Keating recently gave me some blunt advice on the advisablity of continuing my work with Fedora. Draw your own conclusions.

Along with the recent onslaught of mail I have gotten, focusing not on the technology or with aims of providing constructive criticism. Instead focusing on calling me names or implying that I am deluded and dishonest. Like the kind mail I found in my inbox this morning from one Matthew Woehlke:

David Nielsen wrote:

I have instead gotten disrespectful personal mail
[snip]
Looking back at the IRC log I am left with a foul taste in my mouth,
mentions of what RHEL supports and how it’s not a bad thing to
discriminate based on language.
[snip]
I am frankly disgusted by this whole affair, especially since my
honesty and motives were brought into question.

I’m going to guess a lot of that “disrespectful personal mail” revolves around the use of mono? And why shouldn’t it? Lots of people (myself included) have a special hatred of Microsoft’s Trojan Horse, and good reason to question the honesty and motives of people that push it. (Which is not to say I don’t believe there are honest people that are either deluded or simply don’t care.)

If you’re going to promote the technology of a Linux-hostile, GPL-hating, monopolistic bully of a company that regularly engages in racketeering, encourages people to violate the GPL, and is currently suing against Linux… well, some people aren’t going to like that .

Personally, the only thing I would want to do with mono code would be to port it to !mono. YMMV.

My conclusion is that any further work with Fedora is not something I have a desire to do, it will not be welcomed and I have no interest in fighting to contribute my time and work for free. I do owe the users of the products I maintain or do work on in Fedora an apology, I wish I could continue to say Fedora will provide a pleasant experience for your applications of choice. However as they do not for me, I cannot do so for you. For that I am sorry, I hope things will change and that capable people will pick up the work on Mono I outlined with such joy over the last week or so, I cannot in good conscience do so anymore.

With that I will delete my FAS account, remove myself from Planet Fedora and cease being involved with the project in any way. Not without a tear in my eye though, I have gotten to know a lot of great people in Fedora, some I have had the privillage of calling friends. I will miss working with them.

This is a follow up to the post I did a while back on why RPM and especially Fedora does everything in their power to make maintainers jobs time consuming, error prone and confusing. I will stress Fedora (and probably others) do this with the best of intentions, namely increasing the overall quality of the distro packages. However the implementation is all wrong. Read it here.

It is no so much about striving for perfection, it’s that a lot of the work is pointless. E.g. why doesn’t RPM have one policy that is easy to override that does the right thing in 90% of cases. You can see this with libtool linking files, .la. These are required to be removed, in all cases but a few. So one would suppose the correct thing is to have the default be nuking them and then for the cases where they need to be there, have there be work. It is however the exact opposite that is the case in current implementation. Every single time there is such an issue, the default is picked for the most amount of work for the most amount of people and the highest chance of causing an error.. every d..m time.

This problem is most elegantly explained in this talk from FOSDEM 2008 on the Conary packaging system. This is how all packaging should work from a maintainers point of view.

It is all about reducing the amount of work you need to do, the amount of things that can go wrong and the burden of maintaining the package – by picking good defaults and policy. RPM and especially Fedora has horrible policy (well it’s good in the sense that it is comprehensive and well thought out but horrible in the way that the guidelines are designed towards the exact opposite of good defaults, instead of applying them to the default they are applied to every single maintainer- namely making the maintainer do the most amount of work possible with the lowest amount of automated help).

This is only getting worse and worse as more packaging policies are written, RPM is not making the right decisions here and allowing powerful policy making. To you a Fedora user it will mean software will be updated less frequently, as it gets harder and harder to master all the policy manually fewer people will become maintainers. Reviews of new packages will either stall for a long time, require a masterful packager for the review (these are rare, looking at our review stats I would wager we have maybe 5-10 who truly master every aspect of packaging well enough to ensure every review is done by the book) or let packages in with lots of mistakes. Overall a poorer and poorer product ensues over time.

To me the maintainer it means more dull work, time which I could have spend interacting with users on bugs is spend in the jungle we call the packaging guidelines. Unless this changes and fast, Fedora is doomed. The sad thing is that openSUSE, Debian and Ubuntu are pretty much on the same train. Their defaults aren’t selected with this in mind either there. The existing package managers are just not designed with making complex packaging easy, and keeping work maintainable over time, in mind.

The only two ways to fix this is keeping the guidelines and keep throwing increasing amount of maintainers at your package database hoping their skill level is equal to or above that required to follow all the guidelines, even as they change. Or you can do what Conary does and just do the right thing by default.

I would like to stress that it is not that RPM (and I am sure this goes for dpkg as well but I have no intimate knowledge of how it is to work with) is inherently a bad product, it is in fact a very powerful tool. RPM and yum get far more flak than they deserve, their developers are very nice people who care deeply about making good tools, for the most part the manage to do this. This aspect of it though needs serious consideration, I personally doubt that what we are currently doing is in anyones best interest.

Again I urge that you watch the video linked above.

As I currently suffer a bit from the flu, I also have plenty of time to do work. So today I have been preparing a version bump for taglib-sharp which is needed for the upcoming Banshee 1.6.x series. I am still not in ACL for taglib-sharp so spot will either have to do that or add me. I also did some work on Banshee in Rawhide which will trickle down to the stable repos when 1.4.3 is released. This work is mainly little cleanups but also includes a request from upstream to help them identify bugs based on which distros they come from, this will hopefully help weed out bugs found in the Mono stacks around the various distros and now Fedora is participating – living true to the upstream mantra.

The feature I proposed yesterday seems to have been received fairly well so far, I am hoping to get some help from upstream presenting this and answering questions when FESco feels it is ready for evalutation. I would like to give as full a presentation of the advantages switching to Banshee would give us as possible, for this Gabriel and Aaron are definitely more qualified than me. Best surprise so far though, Paul Lange, who definitely gets great guy of the month award for his ability to get me energized and helping out all over the place, turns out to be the very guy who did the Rhythmbox db migration work upstream in Banshee. This pleases me as he is a nice guy who I am sure will respond nicely and swiftly to any bugs shaken out of this code.

The ongoing mass rebuild seems to have hit all the packages I care for, and luckily so far no fallout. Which is really rocking good luck. Aside for the changes that made my touchpad feel like horrible death and disabled all it’s cool functionality, Fedora 11 is really presenting some excellent changes and I look forward to the final product.

Also did some bug triaging work on a few Fedora components, overall I feel very productive today. It might be the fever speaking but life is good.

The wonderful Paul Lange has recently been having back and forths with me about the Mono SIG and already days after annoucing his interest in helping to improve Mono for Fedora users he is hard at reviewing and packaging up applications and libraries.  We are trying to get a Mono SIG mailing list setup though figuring out who to talk to in the beaucracy that is Fedora is a little hard. However the immediate effort to get gnome-do-plugins and all it’s dependencies is progressing very nicely.

Just this weekend FlickrNet and Mono.Nat were approved by review. I have Monsoon, WebKit-Sharp and Beagle-Xesam sitting in the queue for review and I have promised to do all the reviews to get gnome-do-plugins into Fedora as quickly as possible. I have also applied to be the co-maintainer for the following components: Banshee, Taglib-Sharp, Podsleuth and MonoTorrent to ensure that these components are up to date and the specs are as clean as possible ensuring the highest quality Mono for our users.

Also poked Paul F. Johnson, Mono maintainer extraordinare, on a number of reviews he has pending with me, hoping to get them finished or closed so other interested parties might take up that work.

Moonlight is still being difficult, and work has stalled a bit due to all this new ongoing Mono work with Paul.

Finally, I have proposed a feature for Fedora.

Interesting times are upon us.

The nice people over at Novell sent me a free openSUSE 11.1 boxset and an “iContribute” t-shirt, all for telling them that their wifi configuration made me cry. Never again will I discount the value of an honest 5 minutes of work.