The State of GNOME

[Albert Strasheim <fullung ilink nis za>]

Hello,

According to a Slashdot article posted yesterday, GNOME will soon or has
entered the 1.0pre stage. Whilst I cannot confirm this bit of news on the
GNOME Website as of this writing, it seems quite likely, taking your 
recent releases into account.

Excuse me for being blunt, BUT WHAT ARE YOU THINKING? 

GNOME's version number is greatly over-inflated, and IMHO, this piece of
software isn't anywhere near being near being ready for release.

I'm running the most recent version of GNOME [RPM-wise, at least], on a
Red Hat 5.2 system with Linux 2.2.2.

The panel dies off every now and then. Sometimes it doesn't even save its
own configuration, leaving me to redo all my launchers, menus and drawers.
There are numerous little buglets in the panel configuration code, for
example if you've removed the directory from where a launcher got its
icon, edit the launcher's options, and try to change the icon by clicking
"No Icon" button, nothing happens.

The GNOME Control Center cores every now and then. Despite the pathetic
state of Windows 95, at least Microsoft got it stable enough to prevent
its Control Panel from blowing up in their users' faces at random
intervals. [Don't get me wrong, I'm a huge Linux/Gtk+/GNOME fan.]

Half the games in 0.99.7 release didn't even work. 0.99.8 is a little
better, but not near production quality. Whilst this may seem as something
very trivial, the first thing a newbie user looks for are some Games to
keep him entertained, after he's broken the rest of his system. When the
games blow up in his face [literally], he/she is going to go look
elsewhere for some shiny, new stuff.

Another classic example is gtop. A simple utility, you might think. Allows
me to check out the CPU usage of the processes running on my machine. Yet
when I fire up gtop, it sucks up 11% of my CPU. There's no such thing as
immature code to destroy performance.

And where is Enlightenment? Methinks someone needs to tell Rasterman to
get it done so that we can have a new stable release to look at. GNOME
isn't going to pull it off if the only Window Manager that works properly
with it [but not always] is icewm.

I'd expect GNOME 1.0 to be at least as stable as KDE 1.0, which at this
"late" stage it most certainly is not. I find the fact that you feel it is
almost ready to be released at the least distressing. This is the product
that was selected by Red Hat Software to make up an integral part of the
future Linux desktop, yet when I click a few buttons, it dies. Compared to
KDE [which, unfortunately, is based on the seemingly unpopular FreeQt],
it's miles behind, and not catching up fast.

I don't know how much of this instability is to be blamed on Gtk+/GLib
[which has really turned into a mess with the 1.0/1.1 issues], but it's
still most horrible.

Stop moaning and do something, I hear you say. At this late stage, bug
reports seem to be going unnoticed, and pleas for fixes and new features
are being ignored.

I decided to send this message (in a slightly older form) to Michael
Fulbright [aka Dr Mike]:

Date: Thu, 25 Feb 1999 15:29:42 -0500
From: Michael Fulbright <msf@redhat.com>
To: Albert Strasheim <fullung@ilink.nis.za>
Subject: Re: The State of GNOME 

I understand you concerns, but the 1.0 release does not indicate that 
anyone feels the job is done. Everyone will continue to work hard to 
improve GNOME.

Work has been going on for over a year and it is important to show
progress. The 1.0 version has some special stigma attached to it
historically so its good to have things working well.

I encourage you to share your concerns with the gnome-list mailing list
where I think many people will be able to answer your specific questions.

Cheers
Dr Mike

-

Indeed, the 1.0 release does not indicate the job is done. I think
Rasterman will tell you you can never have enough features [or I think
he'll tell you that, according to the interviews with him]. But in my mind
the 1.0 release says "Hello, users. We've made this shiny, new thing.
Please play with it. Learn to love it. Use it. Advocate its use." At this
stage a 1.0 release of GNOME will be saying "Hello, uh, user(s). We've
made this thing that doesn't always dump core. If you promise not to click
on any buttons, it'll make your desktop look pretty. Tell your closest 
friends, but be sure to warn them not to click anything."

If the Linux kernel had been in it's 0.99 stage for a while, and Linus had
decided that he [and others] had put in enough work, and that people
wanted to see progress, and mindlessly bumped up Linux to 1.0, he probably
would've killed it. "Here's 1.0. But the ext2 driver wipes your disks
every now and then, but we'll fix that later." That's progress. Indeed. 

I'm sure Red Hat is feeling some commercial pressure to push out a mature
desktop product with Red Hat 6.0, but at this time GNOME 1.0 isn't it. Red
Hat could end up doing serious damage to their own name [think 5.1 with
it's 20+ errata entries, just worse], and the name of the GNOME project.
The worst thing about receiving flames from KDE'ers would be that they
have a valid point. :)

I'll leave you with a simple analogy.

GNOME is like a game of Gnometris [one of the games that work, and one
with an author that replies to my mail :)]. You start out on level 1, and
everything is easy and fun. You can place the blocks exactly where you
want, and steadily build up your line count. The level goes up to 2.
You've accumalated quite a few points, but not enough to make it into the
high scores. Here and there you have to drop blocks in ways you'd rather
not prefer, but don't have an adverse effect. Yet. Level 3. All is fun and
nice. Still, a few odd shaped blocks don't quite fit, and you accidentally
drop a few "good" blocks into the wrong slots. Suddenly, for some or
other reason, you decide you're ready to take on the world. You up the
level to 9. Suddenly blocks are falling thick and fast, and your once
ordered little stack turns into a mess. You only add a few blocks to your
line count. You do however score at a rapid pace too. You hit level 10,
and by this stage your game has totally deteriorated. You manage to pop a
few more blocks into place before GAME OVER flashes on your screen. You've
made number 2 on the high scores. But number 1 was SO close. Perhaps some
patience would have allowed you to set a new record.

The only way to find out is to start afresh. And after a whole year's
effort, who wants to?

The state of GNOME 1.0 is a troubling affair. At least, in my mind. I want
GNOME to krush KDE and wipe Windows, not get crushed and wiped by
aforementioned products.

I'm considering the writing of a Slashdot editorial about this topic in
the coming week [and perhaps a review of KDE 1.1 to go with it], and
anyone's [constructive] input would be greatly appreciated.

Regards,

Albert
Stable GNOME Fan