GNOME 2.13.1 Development Release

Depends what you mean by "in Gnome". Abiword and Gnumeric are trying to create a GNOME Productivity release set. Inkscape and GIMP could be in a Graphics release set if they thought it would be useful to use our schedule for development. Meanwhile, nothing stops you or distros from using these wonderful applications.

I mean, when I look at the list of modules that is supplied with this article.

They are not on there.

The "GNOME Productivity release" you speak of. Is that the glibrary?

The "Gnome Productivity release" is a proposed release set to complement the desktop, platform, and bindings release sets. It doesn't yet exist.

I don't know, but I'm waiting for it too. I want Gaim transparency like *shudder* it can in Microsoft Windows.

Hey that is a realy brilliant coment, I nearly fell of my chair laughing....

Gdesklets are a cool idea and they are attractive, that is why they are popular. So they'll probably start getting included in gnome by default as soon as gdesklets quits being a buggy kludge.

It has already been decided that python programs will be accepted just fine into gnome.

Infact the proposed (extremly cool) deskbar applet IS written in python, and that is not an issue coming up on the discussion. (There is a problem that it needs a certian python binding that isn't part of the desktop, but I am getting offtopic.)

*weep* I just want to know the new features of everything. I guess I'll have to wait.

I was happy to see that not only is Fast User Switching (link in the introduction) finally being proposed for inclusion, but there are plans to get it integrated into the screensaver as well.

Imagine a typical multi-user standalone PC. First user leaves the desk for a while but prefers to keep his/her session private. S/he activates the password-protected screensaver and buggers off to wash the dog or something. A minute later another user sees the system vacant and decides to check his/her email...

This is great for families, dorms, schools, businesses...

That's an interesting objective. But it helps to be practical sometimes.

Given some of the limitations of the embedded and PDA hardware, this could be considered interesting and practical.

no.

you have to restrain on a pda and avoid to launch all components only useful for a DESKTOP based computer.

gnome on a pda (or part of gnome used in nokia 770) are not simply the "whole gnome stuff for desktop computer and pray it's nice"

yes they can now use better tools to profile gnome, NO there are too much needs and graphicals things in gnome to simply be all in 128mo ram

of course, you are not forced to put 35632 applets in your panels and a big huge wallpapers with gdesklet and Gimp 2.4 in the same time.

No, not really.

Sure, you might be able to make Gnome faster by using things like outline dragging and lower resolution. But the majority of the performance benefit is going to come from optimizing Gtk+, Pango, Nautilus, etc... which will be in any configuration regardless of platform. That will make it usable on things like PDAs AND much faster on normal desktop computers. The graphical needs of Gnome are not that great. Seriously...no 3D, no transparency, no drop shadows, no funky animations. There is no reason why Gnome shouldn't be snappy in 128 Mb RAM, and I hope it makes it there someday.

Trying to make gnome Faster and less memory hungry is not useful only to old PCs. It applies to many uses of today's machinery.

For example, imagine a P4 with gentoo emerging something on the background. Because of the compilation taking place, the high CPU load and lack of memory make gnome slow and unresponsive, however expensive your hardware might be.

In such cases the benefit of (speed and memory) optimization is huge.

If the binaries are smaller and the libraries are less, then the desktop loads faster. This affects all PCs as 7200RPM hard drives are still the standard and have been for quite some time.

Whenever I do a gentoo emerge, gnome stays very responsive to me. Linux is really great with its task schedualing.

If gnome and emerge together use more memory than you have, the system will be swapping. This is completely unrelated to task scheduling.

My hardware (bi AMD, 1 GHz memory, SCSI disks for 3 simultaneous desktops) never experience these problems even with high loads like that, so it's actually dependant on hardware.

Faster is of no use to me, but faster start times, less latency, less visual glitches and less mem leaks to X (granted, there are few now, only with Python/Mono/Java apps, so not really gnome core fault) are what I would want.

Well, 200 MHz may be pushing it, but 400 MHz and 128 Mb RAM is a good target. You can still find a lot of old K6-2s and PIIs around. I think it is reasonable to shoot for good performance on one of those. Partly because a well-optimized code base would make the desktop perform better in general everywhere, and partly because old computers still get used...a lot. Wouldn't it be nice if linux could actually live up to the myth that it can revive old hardware that the "Windows upgrade treadmill" forces you to throw away? And don't say Blackbox or XFCE. The usability of those sucks compared to Gnome, KDE, or even Win98.

I'd like to agree that a 400MHz 128RAM boxes should be a target for GNOME. GNOME aims to be usable on the whole world, there are huge translation efforts done, so it sould be usable in developing countries too. For example here in Poland one can see plenty of Pentium III 400-600MHz with 128MB memory in use running Win98 SE, even (especially?) in commercial environments. I assume that companies would be happy to reduce spending on licences for Windowses and Offices if they had chance to, unfortunately both KDE and (especially) GNOME isn't an option due to speed considerations.

If it will turn up that the speed of GNOME can't be pushed to such limits, perhaps an user should be able to disable some components?

Perhaps there should be a simplified/stripped-down version of Gnome, or people could simply settle for Xfce as a lightweight WM while still being able to run GTK/GNOME apps? Nautilus is still somewhat of a resource hog.

Other idea would be to use the low-end systems as thin clients. Ubuntu Breezy already includes LTSP (www.lstp.org).

It would also be interesting to see if someone (an OSS group?) could create a RAM collection scheme where people with "obsolete" (to them) RAM DIMMS and SIMMs could mail their unused memory to be used in non-profit OSS projects like schools.

Actually, old CPUs and graphics cards could be collected and categorized as well. Esp. developing countries could use any and all upgrades they can get.