Daily Archives: 2006/10/13

Are Mac lovers dumb?

If you read the comments on “when the bough breaks” one might come to this conclusion. The think that formats like PSD an H.264 are free open file formats, which they are definitely not. They do not get the point and they insist that their Macs “just work”.

This is the result of an oversimplification in Apples marketing strategy. Although they sure have some fine people inventing new hardware Apple continuously betrayed users, locked them in and mistrusted them. And if Apple does not trust you, how come you trust Apple?

Some people coming from Windows may think switching to a Mac might be good idea – ans also some Unix geeks think that a Mac saves them time. Well: it is not. This is true if you are a real simple user that just needs a computer to tank your iPod – and it is only true if you are willing to pay a price. The price is the loss of your freedom and you can also count it by your expenses. Better have a credit card right by your side if you use a Mac. The Mac is just selling “it just works”. Why else did I have to help a Mac user hours and hours to get things working that I could do with my Linux box within a few minutes? Ah yes it is easy if your Mac dealer lives around the corner and you pay him for every help you get.

The truth is that there is no such computer that “just works”! Said but true. I am using GNOME on Linux. GNOME also has Human Interface Guidelines like Apple has and it also tries to make things simpler. But I would not say that it “just works” – but it often works much better than a Mac. The only argument pro Mac I accept is if you have an application you need and that only runs on a Mac. Than you HAVE to use it, poor guy/girl. 🙂

The concept of Linux and Free Software in general advocates the freedom instead of simplicity. It includes the possibility to develop applications that are easy to use, while Macs philosophy does not include freedom. So it might be easier at some points – but if it is more complicated you do not have the freedom to make it easier. Even if you could.

You might now say that the Mac also is able to use free software. True, but really a Mac that tries to use free software is much more complicated to use than a Linux that is build upon free software. I could not get Gimp-Print to work on a Mac, even though I use it for years on Linux. On a Mac many things are hidden because Apple thinks you are stupid. I admit GNOME is also hiding more things than KDE because it also thinks users are stupid. But in the end it is free! What is hidden is up for discussion. You can send in bug reports and it even might get changed in the next release or you can apply a patch and have what you want in only hours.

So Apple is only as far interested in its users as they are paying for their products. Apple really does not care about you! Apple wants your money, therefore it makes nice looking and working products. They don’t do it because they want to help you. In GNOME and KDE users and developers are one family that help each other and try to understand each others needs, even if they do not agree every time.

If you like freedom throw away your Mac-Os and install some decent and trustworthy OS. If you are rich and have people that help you in handling your Mac, keep using it you do not want or need freedom.


Filed under Free Software

Inclusionism vs. .* ?

Inclusionism is a philosophy held by Wikipedians who favor keeping and amending problematic articles over deleting them.(read more at Mediawikis Meta Wiki) – thats just as a start for discussing some issues that Wikis andopen source projects do have in common. I think one could call me an inclusionist. But not in all aspects. I love the inclusion of everything so nothing will be lost.
I have just experienced the force of Wikipedia Deletionists in the german Wikipedia. I wrote an article about unconferences. That was deleted, but only weeks after that a similar article with the title BarCamp was written: The deletion was led by a highly respected Wikipedian who is a journalist . She search for the german word “Unkonferenz” and only got 64 results. So somehow this article got deleted, because she argumented it was not relevant. Indeed, if she had searches in english she would have gotten 1.2 million results. This is just one but a good example how “quality assurance” indeed leeds to less quality and redundant work. I would not say that wone should never delete a Wikipedia article but it should be the last choice for really stupid articles that do not make any sense at all. At least a REDIRECT should be possible.

Similar problems come up in open source. See the article “About leaving” from Russel Coker. Fedora decided not to support Xen for older CPUs like his (that he got from Red Hat as he left the company). There we see the problems with the classic WONTFIX approach. I found that distributions like FreeBSD have this WONTFIX attitude more often – and sure Fedora also does this more often than Debian. I think it is understandable if you have limited ressources but want to get a working release in time. I think the probklem is starting if people have less opportunities and are forced to switch (like from Fedora to Debian). There are other examples where people siwtch from Debian to Fedora for similar reasons. On my partI switched to Fedora because Debian never had uptodate software. Ubuntu really filled a gap here. Maybe I had switched to Ubuntu and not Fedora if it had existed at this point. I think the problem is that users often like to have a fork, something slightly different but that the efforts for switching or extending are often big. We still have different package formats. So as a user you often stuble accross a site where a developer builds only for his distribution (that might be Gentoo, Debian, or Fedora/Red Hat, etc.) and that just does not is what you are using right now. I think there is something wrkong in this development processes. People are starting to write for a distribution because they have limited ressources. At the same time that emans that many users will not be able to uses their packages and that some extra work will have to get in migrating the package.

I think the classic distribution development is outdated. Understandably but outdated. This approach only makes sense for tools that should only be run on one distribution and never on another platform. I am not a developer but I think it should be possible to automatically build for different distributions
while programming? Ok, sometimes you want to build on one library version that does not (yet) exist on other distributions, but this could be solveable. Maybe a tool could help you in deciding where to build on knowing what libraries all distributions use. Also I think development tools should enable users to write code that immediately is published online like with wikis or with Gobby. So certainly a devtool needs a jabber chat built in! So I am talking about live programming. And I am also astonished that translations of GNOME still happen via mailing lists and not while the code is written. This is ridiculous. This takes weeks instead of hours or minutes to fix some characters. The problem is that open source development is “traditionally progressive” – but in fact it often does not usenewest technologies to do the job better. hail to Launchpad that enabled very easy translating without people having to subscribe to mailing lists and so on. This social software stuff really is about enabling people to help each other more easily without much administrative hassle. But organisations like Fedora think that open source development needs strict organisation, while indeed it does not. Or better: development does not, creating a distribution DOES. Fedora really chose the opposit principles to Debian. While Debian chose to release when “it is ready”, Fedora chose to release on a regular basis. But a release often is nothing more as a working snapshot. I found the efforts of FedoraUnity interesting because they were able to build releases of their own without all the administrative overhead. I really think development and release building should be a complete seperate process. Distributions should not be proprietary. So it just does not makes much sense to waste time in trying to build a product as whole. I think distributions need to share much more ressources. I think distributions like Gentoo, Fedora and Debian should have a collective developer base. So that many packages get audited and worked on together and only after that the packaging happens in the manner that all developers try to automatically build packages for all possible distributions. And after that distributors fetch these packages and make installable ISOs out of it. in open source combining powers always will leed to better results. Do your forks like Inkscape did from SodiPodi but do not develop for a distribution. Development needs freedom. Freedom from policies like DFSG or Fedora meritocracy. Its just plain stupid to bind users and developers to a specific philosophy. Science and softwre development need as much freedom as can be. If the common basis would be bigger I could see much more choice for each distribution. The authors could enforce the use of specific licsenses if they want, but it should not be the distributions that bind the developers. And that you can see on many distributions nowadays. Take OpenBSD, take Debian, take OpenSUSE or FreeBSD. All do have their philosophy but mostly development and distribution comes in one. Let the developers discuss their philosophies and let the users decide what they want. The problem today is that you really would have to build Linux From Scratch and do all the work of you want to have your freedom. But doing all the work is also redundant and unneccessary!

Leave a comment

Filed under Uncategorized

Old hardware or new hardware?

I like the OLPC project. They build small, affordable laptops for developing countries. I think this is the right way to introduce Linux. Some people have stated that they think it is better to reuse old hardware. But I have come to the conclusion that that really is a trap for Linux. Why? Because full hardware support is always a problemm, not because of Linux, but because most hardware vendors do not care about Linux support. I have faced the problem dozend times, that you try to install Linux and some things do not work perfect. And it was mostly hardware issues. A project that takes many different kinds of hardware will have to deal with multiple obstacles PER COMPUTER. They would not be able to make one distribution that fits all. And on the other hand I think it is very interesting to begin to develop hardware for Linux or a specific distribution. I think that the efforts to support a wide set of hardware often result in slow process.

If we would have started Linux with one computer that would have been supported 100% Linux would be much more stable than today. I think it would not have been as successful. It was attractive because you can just buy a PC in any store and TRY to install Linux on it. But I think this is not really it. PCs are mostly built to work with Windows and nothing else. So we really need Linux Computers (Desktop and Notebook). And this could also leed to hardware development that fits to the software. Apple has often shown what power could be in this combination. On the Linux side this would be an open process. Linux could initialize hardware developments everybody can use. Unlike this new Intel-BIOS that is only supported for the new Intel-based MacOS X.

I think it is time to rethink Linux deployments. Today it is not the way to install Linux on old machines and to make them cooperate with Windows. It is much better if you try to make a system that replaces Windows soft- and hardware a 100%. Cities like Munich think that this is unrealistic, but I really think it is unrealistic if you think you can live with 80% Linux and 20% Windows. You will the problems in the areas where those systems have to mix, because Windows does not want to be mixed. It only likes to number one and every effort to include it in an Linux IT-concept leads to much more problems. It may not be possible to have 100% Linux from day one, but if this is not the goal, better do not migrate!

Linux must leave its underdog habits: “We are cheap, we can install on your old hardware. Ok we try to run your Windows software on top of Linux” , etc.!

Leave a comment

Filed under Linux