HP Drops Gnome 2 Efforts
nauta writes "Now is official, HP will not make further investments in Gnome. They will stick with the old (and crappy) CDE. Here is the announcement This is the official statement if they are pressed for an explanation:
'The open source development of GNOME v2.0 was still on-going at the end of 2002, and did not stabilize in the timeframe that HP had earlier anticipated. This and other business and industry factors required us to re-assess our plans.'"
While W2k is an improvement over NT in terms of reliability, it still bluescreens occasionally. I note that the oldest IIS webserver finally managed to rack up 2 years, just in time for Slammer - but that every Unix and it's dog routinely exceeds that. And XP is a reliability unimprovement. And Foghorn Leghorn - er, I mean, Longhorn, or BlackComb, or whatever it's called today is gonna be all shiny new and with a fabulous and innovative range of unforeseen bugs too.
Meantime, I get plenty done and there are no Windows machines in the house at all to "do stuff" with. I may not have the latest frilly border on my documents, and each screen I face may have more than three things to click on, but my documents and programs do come out hot and on time.
If you ever come to visit Western Australia, call ahead. I can show you a bunch of kids doing video editing on their Linux boxes and a highly productive office kitted out with nothing but Linux. No Windows, no bluescreens, yes productivity.
Got time? Spend some of it coding or testing
What does "stabilize" mean, anyway?
Are you kidding? That has to be one of the top complaints regarding alot of OSS development, including Gnome.
I do alot of testing and bug stomping for some Gnome packages, and I've frequently heard Gnome developers describe many Gnome and Linux libraries such as GTK as "moving targets". By the time you finish developing for version a.b.c, version a.e.f was released, and it breaks compatability with version a.b.c.
As a Gnome user, I've tried to compile everything from Source on a number of occasions. The dependancies drive me up the wall.
I use prepackaged products such as Gargnome, but it only solves some of the dependancy hell. If I want that new version of software X, I need to go and find and compile the newest version of several other packages.
"Can of worms? The can is open... the worms are everywhere."
HP wants to write commercial proprietary applications for GNOME. They cannot do that when the development has not stabilized. HP does not want to develop for and support a moving target, and their customers won't want to install a patch every week just because someone at GNOME changed the API. Geez, even Windows managed to keep a stable API through three different desktops, nine major release versions, and one complete decade!
A Government Is a Body of People, Usually Notably Ungoverned
There's more to a successful product than quality engineering. Every product has a finite window of opportunity. If you miss that window, all your potential users have gone on without you, using some other product to satisfy their needs.
Look at Mozilla. That project has been wandering in the wilderness since 1998. If they had produced a useful, stable product back in 1999, when Internet Explorer still only had half the market, people might have resisted the pressure to switch.
In 2003, IE has ninety-six percent of the market. That's a huge mass of people who have every motivation not to switch back. So what if Mozilla is now technically superior? There are a zillion web apps that are designed around IE's quirks and "innovations". Users of these apps will never switch back -- and Mister Bill gets to dictate how web browsers "should" work. Depresssing thought.
gnome/gtk libs has been ABI/APIstable since 2.0. It seems they are pretty comitted to do just what you want.
still reading?
The GNOME API remainded backwards compatible (IE, no functions taken away, only functions added) throughout the 1.x series.
The transition to 2.x allowed the API to change, meaning that applications that were written to 1.x APIs would not always compile with 2.x libraries. This is common, and KDE and QT do it as well. The 1.x and the 2.x libraries are parallel installable, so that you can have both installed on your system.
The GNOME development platform is now backwards compatibable in the 2.x series and will remain so until 3.x
At least, this is how I think it works.
KDE 3.0 to 2.0 was not a big step and the main reason for it to go from 2.x to 3.x was the major change in Qt. KDE 1.x to 2.0 was however a huge step and changed things pretty dramatically. GNOME just went through the same step for GNOME 2.0 and the API is not expected to change much for quite some time. GNOME 2.0 came out last june. GNOME 2.2 came out 2.5 months ago, and GNOME 2.4 will be out in september, all with the same basic API.
GNOME is not even meant to change that much from 2.x to 3.x, so the API should now be reasonably stable for quite some time. It was perhaps not ready for HP in time, but it is there now.