hi Matthias,

I do like you a lot, you have lots of energy and zeal, and are not afraid to post ideas and questions, that said.

Right now tho, i think you're allowing yourself to get pessimistic about Arch and Linux in general. My advice to you is: DONT.

Can i ask what you do for a living?

I personally work at a software outsourcing house somewhere in Nigeria, Africa, and work takes sooooo much of my time i don't have time to do lots of other things. My work is not even linux related, so i don't even get to use linux during the afternoon, only at night like this, or when i do a 'candelstine' operation and do some personal work for other ppl, out of work hours ( i have a number of Arch Server installations running mainly as routers and proxies for ppl).

Last year, i started the Transaction Patch thingy. I really didn't intend it to go there, but i had a funny reverse engineering thing with makepkg that i posted to the mailing list. Judd himself suggested that I could make it into the Transaction Support thing that's in the eventually plan. I liked the idea and went ahead to build it. Right now, its more or less stable, with a little loose nuts to tie. Problem is, recently, i've been put incharge of Operations at my station, so i don't even have time to sleep at night, talk more of code... its really bad.

Imagine then... this is just me, with one meagre patch, and i can't even dedicate enuff time to it. Imagine what some of the devs have to do, and NO ONE is paying them to do it. And they still have to go about their normal lives, and make a living.

What i've come to realise in Open Source is that the software gets a life of its own. A couple of guys (like GT, who always lets me know he's just downloaded my lattest patch), are using my patch, hey... that makes me happy. I don't know if it will ever be included in the mainstream pacman. Ofcourse i hope it will be (i can add that to me CV :) ), but hey... i can't force Judd to do that. Same thing with pacman-i18n. I dare say you can't force Judd to add that. I'm also sure a couple of guys are using it already. What you should/could do, is to continously polish it.

Remember there's a big rewrite that may soon take place when libpacman is finished, and i seem to percieve that most plans are hinged on that now. I'll just say... be patient, and don't loose your fire. Judd may not personally tell you this, probably because he gets LOTS of mails as well, but i'm sure he appreciates the effort.

You said it yourself, you still use Arch, since there's no other REAL alternative after one has tasted Arch. don't loose vision my friend. YOu're valuable to the community, and are much appreciated.

Ok... that was waaaaaaay too long. Sorry for taking up soo much bandwidth :)

Matthias-Christian Ott wrote:
Daniel Zilli schrieb:

Dear friends,

Is a pleasure talk with you.

I am Daniel Zilli, an active member from Brazil Open Source community. I
am specialized in Free Software and Linux Operation System with tree
books released.

During the last two months I am being using and analise the Arch Linux
distro. To be honest with you, I am little amazing about the Arch.
During this two months of test the Arch showed be a stable, simple and
efficient distro. I am really do not see big different between Arch and
the big players of market.

So, with intention to help this distro become more and more mature, I am
will draw some points that can help the distro give a big step to the
hall of fame.


1) Arch need urgent a new installer! We know that the current works, but this today is not enough. We need an intuitive, easy and efficient installer. If you think a little, we can see that the new users make the opinion about a distro during the install process. So, if the install process is difficult, they will think that the distro is difficult too. Think about this! If you allow me, I can show to you guys a new installer in a fews days. (now you know about my syntax errors :-) ). Plus. I can add support to i18n in this installer too.


2) One thing that made me very happy in the Arch, was the package manager. Absolutally pacman is one of the best. But, has just one big defect. The extension. The extension "pkg.tar.gz" is too long! The most famous package manager using short extension like rpm, deb and tgz. Look the difference:


bash-3.0-3.pkg.tar.gz
bash-3.0-3.rpm
bash-3.0-3.deb
bash-3.0-3.tgz

Be more direct is essential. We know that pacman is a tar.gz based, but
we are not force to use tar.gz in the extension. And more. This extension can be confuse to the new user that see "tar.gz" in the end. Someones will not associate with a package manager. My suggestion is use a new extension like "pac".


bash-3.0-3.pac
bash-3.0-3.rpm
bash-3.0-3.deb
bash-3.0-3.tgz


I think "pac" is beauty and easy to remind.


3) Arch need more press. If the people don't know about the Arch, how will they use ? We need that Arch become famous. Why? To bring more people to use and contribute to the distro. For this happen here my suggestion: - Write articles about Arch Linux in each country involved here. Ex: I am leave in Brazil, so I will write articles to the Brazilian people. Now if Arch has an English or Chinese buddy, they can do the same thing. - To a distro be fames, depends the number (or quality ?) of packages that they have it. Look, almost any site that has a program in binary mode; this binary is rpm or deb. So we need do this too. And more. We need tell to each maintainer of some the famous packages (openoffice, postfix, gimp, kde..etc) that there is a package to Arch Linux and link to them. - Be more "commercial" in the right way. What do you think about Arch Linux Foundation ? Sell t-shirts, cap, cup..etc. this stuffs. - Use as default in the Arch Linux wallpapers in the Windows Manager. The others distro to this, why we not ? We have a lot of pretty wallpapers, so we can use them.


For now is that!

I really want to help. My suggestion is here. Think and tell me what you
decided!


PS: Judd recived this email too.

Bests,
Daniel Zilli


Yes, you're absolutely right and most of your ideas are good, but Arch Linux won't change. You can have that good ideas, but nobody will change something -- this mail will probably ignored by many developers.

I'm pissed of Arch Linux and Linux in general which are both consisting of unfriendly, uncooperative people, a bad structure, bad styled and structured code and a bad realization. But in this world there's nothing better -- regrettably.

Have look at Arch-i18n which was initialized by me: It consisted of about 20 people whos target it was to translate pacman and Arch Linux. Everything was well until we needed the help of Judd Vinet. I contacted him several times (the last mail I send yesterday to awake Arch-i18n) by mail. In this mails I just pleased him to add pacman-i18n to the Arch repo and mentioned that Pacman-i18n is ready to replace Pacman -- he didn't respond (I know that German People are not very favored in the world, but this is not my personal project -- it's a project of about 20 people). I had to talk about this mails on the mailinglist to get him responding. This mail never contained a response to my/our please. I contacted Tobias Powalowski who wasn't able to help me, because he hadn't the permission (he's the only developer I know who is cooperative), he said I should contact Jason Chu after having mentioned the problem to him (after 2 mails) he didn't respond.
Because of this Arch-i18n is now nearly dead. That's the plain truth, that's poor -- no cooperation, no structure -> no improvement.


This is just one example for the inferiority of Arch Linux and its development. Have a look for example at the pacman concept (storing files with some information files in compressed file and having a files based database without advanced search algorithms, a structure (and nobody of the Arch Linux functionaries wants to change this!)), the package submition system (it's horrible and the AUR is just a minimal betterment) or the split of Forum and Mailinglist -- everything is crummy in the (Arch) Linux world.

In my opinion the Arch Linux functionaries are crummy software ingeneurs and maintainers and and most work is done dowdy -- they (like famous developers like Linus Torvalds) only aim to get the software/their project working at the moment -- not in the future --, work unstructured, can't take criticism, aren't amenable for innovations and only think their work and their opinion is right.

This world is a real disaster (specially the Arch Linux world) -- no improvement ans structure caused by the facts I mentioned above. That's very very poor and crummy.

Now I want to comment your ideas:

1) Nobody will change it or it will be a debased one (created with XDialog or something like this). But anyway this would make Arch Linux popular and userfriendly and is designated by a lot of users.

2) I would change the extension too, but pacman absolutely sucks.

3)

   * Nobody wants to have Arch in the non-english (The Arch Linux
     functionaries press seem to dislike non-english presentations)
     press. For example last year someone asked to have Arch Linux at
     the Linuxtag (Linuxday) 2005 -- nobody except me respond (you can
     look this up in the Mailinglist archives).
   * As I mentioned above Arch Linux is unstructured and nobody will
     care about this
   * I think the maintainers are too lazy to customize the distro
     editing packages as a non maintainer is horrible


With this words I want finish my presumably last mail to the Arch Linux functionaries and Mailinglist (I won't leave Arch Linux because there's no real alternative, because Linux itself is similar to Arch Linux). Keep in mind that this is my personal opinion and I don't want to discuss it -- but maybe it forces you to think about everything and to change that (an uprising would be apt in this case).


Matthias-Christian Ott

_______________________________________________
arch mailing list
[email protected]
http://www.archlinux.org/mailman/listinfo/arch



_______________________________________________ arch mailing list [email protected] http://www.archlinux.org/mailman/listinfo/arch

Reply via email to