[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [linux] Arte ... ou comment perdre une soirée



Le Lundi 18 Mars 2002 09:45, vous avez écrit :
> ...
>
> > Après tout, si c'était vraiment si simple, cette liste n'existerait pas.
>
> Il faut quand même dire qu'on remarque très bien au "niveau" de cette liste
> que Linux est devenu plus mature et plus facile à utiliser (je peux en
> juger puisque je suis dessus depuis le début ;-)): les questions se
> rapportent de plus en plus à des thèmes plus complexes comme le
> firewalling, la configuration d'Apache, de la programmation, etc...
>
> Et puis rien n'est parfait, il y a tjs des bugs et moyen de faire mieux ;-)
>
> ...
>
> > Bien entendu qu'il faut en parler autour de soi mais cela reste du
> > prosélytisme à petite échelle. Tant que les administrations, les PME et
> > surtout les écoles n'auront pas le courage de franchir le pas, cela
> > risque de prendre du temps. Heureusement, il semble que la Chine a
> > compris. Espérons que l'Afrique prendra le même chemin.
> > Malheureusement, nous, pays riches, pouvons nous passer
> > de réfléchir et simplement acheter les windown successifs.
>
> C'est en train de changer: l'Allemagne.
>
> Le Bundestag allemand (= le parlement) va passer tous ses systèmes
> informatiques (et il y en a bcp ;-)) sur Linux et OpenSource pour les
> serveurs.
> Pour les stations de travail, ce sera encore XP dans un premier temps.
> Plus intéressant: ils n'utiliseront que des protocoles et formats ouverts
> quand c'est possible.
>
> Mais c'est déjà un succès énorme pour Linux !!
> 1) on sait bien que si Linux a déjà un "grand" pied dans la porte, il va
> aussi s'infiltrer sur les stations de travail
> 2) ça montre l'exemple et ça va montrer que c'est possible, d'autres
> administrations européennes vont emboîter le pas (il y a pas mal d'activité
> dans ce sens, notamment en France et en Italie)
>
> La Belgique est évidemment et malheureusement un pays de lobbyistes M$.
> Le pire est en fait que les patrons/décideurs sont excessivement frigides,
> prennent peu d'initiatives, bref, très peu de dynamisme.
>
>
> ...
>
> > Personnellement, j'ai eu plusieurs fois la tentation de le balancer par
> > la fenêtre (et je suis sûr de n'être pas le seul) mais je m'accroche à
> > essayer de résoudre des conneries qui ne devraient plus poser de
> > problèmes, comme l'horloge qui n'est plus jamais à l'heure, comme
> > l'envoi d'un fax aussi "simplement" que d'imprimer (ça, j'y suis arrivé)
> > ou l'installation d'un graveur IDE. Et pour être tout-à-fait juste, à
> > part StarOffice, je n'ai pas trouvé dans Linux de logiciels aussi
> > performants et surtout aussi conviviaux que Photoshop (ou PhotoPaint),
> > Corel Draw, Autocad, OmniPage, Winfax, Easy CD Creator, et je
> > passe 3DMax & Maya (animation 3D), Edit & Première (montage vidéo),
> > et sûrement quelques autres que j'oublie et qui m'empêche de jetter
> > simplement Windown.
>
> Oui. Tu as raison.
>
> Mais tu ne trouveras pas sous windoze qqe chose d'aussi performant
> qu'Apache, MySQL, BIND, sendmail, proftpd ou pure-ftpd, etc... etc...
> (oui, je sais, certain d'entre eux tournent aussi sous windoze, mais moins
> stable et moins vite)

Bref, le troll continue ...

Quand j'ai bazardé Windows pour Linux il y a deux ans (je ne suis qu'un 
pauvre utilisateur) je sentais bien que c'était un O.S. en pleine expansion: 
passionnant à suivre.

La réaction est néanmoins très dure.
J'ai fait la semaine un petit exposé sur Linux à mon boulot (j'y suis 
malheureusement le seul Linux-Geek) et les réactions furent les suivantes:

1) oui mais Apache c'est difficile, il faut être connaisseur. WebStar est 
tellement plus facile, il suffit de faire clic-clic. J'ai ensuite installé en 
vitesse Apache sur Linux et MacOSX pour essai en interne mais la personne n'a 
même pas voulu voir combien c'était "difficile".
2) Autre réaction: c'est un cliché de dire que Linux ne plante pas et Windows 
bien. Au XXX (centre XXX à Uccle) le serveur DNS sous Linux plante souvent. 
Ceci dit j'ai eu deux gros plantages avec Cyrus-Imap (clavier tout-à-fait 
bloqué mais était-ce Cyrus-Imap, aucuns problèmes avec Postfix très stable).
3) On m'envoie un fichier.pdf de WebStar qui prouve que WebStar est 2 fois ou 
plus plus rapide que Apache, essais sous MacOSX uniquement. J'aurais voulu 
vous envoyer ce fichier.pdf pour commentaires mais 300k cela prend trop de 
place et c'est pourquoi j'avais demandé sur la liste s'il était possible de 
transformer un fichier.pdf en fichier.txt ou fichier.html.
4) Content d'apprendre que le parlement allemand va s'essayer à Linux. Il y a 
quelques jours, ce n'était apparemment pas encore décidé d'après ce que j'ai 
pu lire dans un article.
5) mon chef de service, qui est loin d'être pro-Linux, m'a transmis il y a un 
mois un document d'origine inconnue qui est apparemment une étude et une 
évaluation de l'OpenSource commandée pour le gouvernement anglais. Cela fait 
170k donc trop gros pour cette liste mais voici un extrait de l'introduction: 

Analysis of the Impact of Open Source Software"

October 2001

0.1Background

Open Source Software (OSS) is software whose source code is openly published, 
which is usually available at no charge, and which is often developed by 
voluntary efforts. This paper examines how the OSS movement could impact the 
UK Government's policy towards the worldwide market in software 
infrastructure1 and software applications2.

OSS has leapt to prominence by starting to take a significant market share in 
some specific parts of the software infrastructure market. For example, since 
1998 Linux has been one of the fastest growing server operating systems3. 
Indeed, in a few important market segments, such as storing Web Pages, OSS 
software is far and away the market leader4.

The software industry is very fast moving, and frequently throws up promising 
new developments that initially promise to make great changes in the 
marketplace, but which ultimately fail to live up to their initial press 
hype. Our first key conclusion is that OSS is indeed the start of a 
fundamental change in the software infrastructure marketplace, and is not a 
hype bubble that will burst. This is perhaps surprising because OSS does at 
first sight appear to be a bit of a paradox.

Given that OSS software is often developed by largely volunteer efforts, how 
can software, such as the Linux operating system, compete with software such 
as Microsoft Windows, which has had, and will continue to have, billions of 
dollars invested in it? In particular, how is it that the best OSS is 
perceived by many to be at least as reliable as market leading proprietary 
software? The body of this paper explains how this has happened. OSS's 
credibility has been established as market giants such as HP, IBM and Sun 
have thrown their weight behind it.





0.1Future Trends




Where will it all end? We now have to move into the area of predictions5.

Within five years, 50% of the volume of the software infrastructure market 
could be taken by OSS6. We expect that OSS's position in the small server 
market (file and print servers and Web servers) will grow fastest. 

OSS's position in large servers (e.g. those managing massive multi-user 
databases), such as those that underpin many large Government procurements, 
will grow from its current position of near zero penetration, to a position 
where OSS is a viable option, within 2 - 3 years.

Within the developed world, we as yet see no sign that OSS will become a 
viable alternative to Microsoft Windows, for users' (general purpose) desktop 
machines in the corporate or home PC markets. However, OSS on the desktop may 
soon become a significant player on the desktop in the developing world. For 
these reasons the study recommends against any preference for OSS on the 
desktop, but also recommends that this issue be reassessed by the end of 
2002, by which time early trials of the use of OSS desktops may have 
generated sufficient evidence to warrant a reassessment.

OSS is already suited to restricted functionality desktops, such as those 
used in industry for point-of-sales and point-of-service8 terminals; and in 
these areas OSS's market share is likely to grow significantly.

We expect OSS to rapidly become the market leader in consumer computing 
devices10.

We expect the market for new portable and consumer computing devices (such as 
set-top boxes and smart mobile phones) to remain very dynamic, with no 
dominant market leader emerging. OSS is however likely to be a significant11 
player in this market.

We expect that the software infrastructure that is implemented on top of 
operating systems (so-called middleware) will move gradually from proprietary 
products towards OSS.

All of the above predictions relate to the software infrastructure market. As 
yet, it is not possible to predict that, within the developed world, OSS will 
make such a major contribution to the software applications market. There are 
a few OSS applications that are becoming significant, but it is too early to 
say if a trend is developing.

Etc ... Conclusion: idem que parlement allemand Linux => serveurs, MS => 
Desktop pour l'instant à revoir d'ici deux ans.


-- 
Alain Barthélemy

cassandre@gminformatique.com
bartydeux@gminformatique.com
http://bartydeux.gminformatique.com
_______________________________________________
Linux Mailing List
Archives: http://unixtech.be/mailman/listinfo/linux