GNOME 2.12 Previewed

An anonymous reader writes "Davyd Madeley has completed his Prerelease Tour of GNOME 2.12. Scheduled for release on September 7th, 2005, GNOME 2.12 has picked up a new theme, some features popularised by Apple's System 7, some new multimedia tools and plenty of bug-fixes."

Comments

BSD ?

[mbyte]

Is this some subtile joke by the editors among the BSD is dying trolls ?

Re:BSD ?

[TummyX]

Or perhaps it's cause Gnome works on BSD?

Re:BSD ?

[trollzor]

I have seen other GNOME stuff listed under the BSD section a couple of times. It's all a bit of BS if you ask me. Because the G in GNOME stands for GNU. Perhaps timothy mistakenly believes GNOME is a BSD project, I dunno, anything is possible with these editors. I don't mind the BSD guys using LGPL stuff or GPL stuff, hell I use the OpenBSD derived ssh stuff which totally kicks ass (ubuntu lists it as OpenBSD derived in the bootup and shutdown I believe), but credit where credit is due.

Re:BSD ?

[Anonymous Coward]

Netcraft Confirms: GNOME is dying.

The current number of servers Netcraft reports as powered by GNOME now stands at zero. It's official. GNOME is dead. Upon hearing the news, creator Miguel de Icaza was seen working at Taco Bell, depressed, sweeping the floor, and muttering the words, "ay carajo..." The KDE dragon was unavailable for comment.

Or something. I don't quite have the knack for these.

Re:BSD ?

[ajuin]

Actually, there is a BSD-licensed compiler. It's called TenDRA. Unfortunately, it's not a drop-in gcc replacement.

Re:BSD ?

[The_Wilschon]

The real question is, however, how quickly and how well were these 53 vulnerabilities patched, and what proportion of openSS{L,H} installations were patched?

The number of vulnerabilities is not strictly correlated with a piece of software being insecure. It is also correlated with other things like "how hard are people looking for vulnerabilities?" and "how honest are the developers about admitting (and fixing) vulnerabilities?".

Re:BSD ?

[Homology]

Your compiled lists includes programs that uses OpenSSL, so the same vulnerability is listed several times
CAN-2005-1247 . It's like counting each and every vulnerable application that uses a zlib with the recent security holes.

Re:BSD ?

[Wdomburg]

Honest? Like De Raddt initially refusing to divulge details of an exploit in a bid to blackmail distributors to upgrade to 3.3 and enable privilege seperation, despite it being broken in numerous ways (e.g. privsep+compression broken on 2.2 kernels, broken or incomplete integration with PAM on numerous platforms).

The fact of the matter is that the two packages, individually and combined, have had many more vulnerability days than anything else we run in-house, including many that are not only at least as wide spread, but have a reputation (not always deserved with current versions) for insecurity (e.g. sendmail, BIND, wu-ftpd).

It's also worth noting that the window of vulnerability is not how long it takes for a patch to come out from the upstream vendor; you need to also account for packaging, testing and roll-out.

And regardless of how timely one closes the hole, each upgrade still has a cost. People who've never done large scale administration tend not to understand this, since as often as not they just compile a new version and slap it on, but when you deal with a large number of machines and have service level agreements to meet you need to be a lot more rigorous in the QA phase and often need to schedule a downtime window with your customers.

Re:BSD ?

[Wdomburg]

There's only three entries on the list that aren't with OpenSSL directly: CAN-2004-0607, CAN-2004-0975, and the one you cited. So it's fifty rather than fifty-three.

Re:BSD ?

[Homology]

To nitpick a bit more : The GP talked about OpenSSH, not OpenSSL that is not developed/maintained by the OpenBSD project. Though I'm pretty sure that OpenSSL cares about security as well :-)

Like in all software, there is security issues, but it's not quite as bleak as the list you presented (one reflect FreeBSD on Alpha, another about a script on Debian, if I remember correctly, some changes from default config (like disabling Priv Sep)).

Have a look at OpenSSH Security for the various security issues. The opensSL project has a similar page.

Re:BSD ?

[synthespian]

Well, to defend the honour of the OpenSSH, let's get on with this off-topic thread...

To sum it up, the OP is a TROLL; I counted at least
12 reports where OpenSSH was not even mentioned, let alone OpenBSD; and in fact, most of them refered to OpenSSL, like you said (and I'm not going throught the TROLL's whole list).

Re:BSD ?

[The_Wilschon]

Fair enough. I was merely pointing out that the raw number of vulnerabilities doesn't really tell us much of anything. Now you've told us a good bit more. Thank you.

What about Beagle?

[rekrutacja]

Is this mature enough to include it as standard? Desktop search is key missing feature in Linux...

Re:What about Beagle?

[ralinx]

as long as Red Hat is opposed to distributing Mono, Beagle (written in C#) will never be in the default Gnome Desktop... at least not the one Red Hat will ship. So who knows... there may be another high profile fork (red hat gnome vs novell gnome) coming up soon.

Re:What about Beagle?

[Anonymous Coward]

no beagle is only a trojan horse done to justify
mono as default platform into gnome.
i've heard that someone are working to produce a beagle replacement in python
http://img185.echo.cx/img185/2971/pybeagle47ya.png

Re:What about Beagle?

[KeyserDK]

To be honest, the widespread porting of the lucene engine(beagle backend) is the highest amount of forking i've ever seen.

There exists a port for every language, it just doesn't make sense. The basic algorithms for searching, and storing indexes hasn't changed for quite some time.

In the digital library space there even exists quite old (10 years) open source software such as zebra[1] which can handle large indexes fast. There are actually open standards[2] for information retrieval (IR), but nobody in the open source desktop space seem to know about it(?).
[1]http://www.indexdata.dk/zebra
[2]http://www.loc.gov/z3950/agency/zing/

Re:What about Beagle?

[dalutong]

I understand that Ubuntu will have it included by default in their next release (which will probably be on the same day as the release of GNOME.)

Re:What about Beagle?

[Korgan]

I hope not. I've been playing with Beagle 0.12 and its definitely not ready for prime time. Its a great service and the 'best' front end is very nice (although I personally prefer the web frontend as I usually already have a browser open) but its not anywhere near solid enough.

Its a mission to get going from source at the moment and even if you run a distro that already includes it, it doesn't take much to break it. Upgrading Firefox/Mozilla is enough in some cases (thanks to best's reliance on the Gecko libs).

Don't get me wrong. Personally I love the technology, but I really don't think its anywhere near ready for mainstream use. Great if you want to try out bleeding edge tech or help improve the software, but not if you just want a search tech that works.

The other downside is that beagled has to be run by the individual users when they log in. It refuses to run at boot as part of the init scripts. So its got to be included as either part of the xinit or shell rc scripts. Thus automation is going to be needed on the admins part at the moment. Sure, this can be done as a default part of a distro, but given its not ready yet... ;-)

Re:What about Beagle?

[irc.goatse.cx troll]

[offtopic]
"So its got to be included as either part of the xinit or shell rc scripts. Thus automation is going to be needed on the admins part at the moment. Sure, this can be done as a default part of a distro, but given its not ready yet... ;-)"

Just include it in the /etc/skel/ files?

Re:What about Beagle?

[mewphobia]

To be honest, I haven't used linux as a desktop for a couple of years, due to work, but what about locate?

Re:What about Beagle?

[Laurance]

and it is an ugly interface.

Re:What about Beagle?

[dalutong]

Hmm.. I disagree. First a disclaimer: I am working on a very poor memory -- beagle might not be included at all.

But I have been using beagle on my ubuntu machine using breezy backports. Ubuntu maintains its packages well -- I have not had beagle break at all over the past couple of months of using it. Setting it up to start when I start gnome has been the only kind of work I've had to do.

I'll admit, though, that the search is still fairly slow and not great at finding what I need. I am using version 0.11 though.

So I think that stability won't be the issue. The package in ubuntu will be well maintained. I just hope it speeds up a bit. And I wouldn't mind some search tips to help me find my data more easily -- especially an easy-to-find help button explaining precisely what beagle can and can't search.

Re:What about Beagle?

[ashayh]

Installation is not the only problem. mono/beagle frequently use 30-50% of CPU time and/or memory.

Re:What about Beagle?

[Korgan]

/etc/skel is one option if you intend to normalise it as the default search method on your platform across all users. But given how much of a hog beagled is when it comes to resources, on a busy server thats going to hurt.

In my office I run an Ubuntu termsrv set up. No one has anything on their own machines except for a basic OS with an X server on it. Cuts back on hardware costs. I'm sure I'm not the only one that has set things up this way. Now add in a whole heap of mono processes going and you're gonna hit trouble. Its hard enough having just a few people loading OpenOffice.org apps at the same time, adding in multiple Mono processes running for each user (beagled constantly in the background, best in the fore) and suddenly you need a much more powerful server to cope with that kind of load.

Sure, /etc/skel would work for desktop/workstation/laptop machines which are likely to have only one person using them at a time, but I'm not that rich. I run a small business and manage the IT side of things myself. I simply cannot afford the overhead required by multiple mono/beagled threads at once.

This is partly why I'm glad Gnome didn't let Ximian bully them in to coding core parts of the platform in C#/Mono. I've been a Gnome user since 1.2 and I would truly have had to change desktop if they went down that path. It just wouldn't be practical or useful for anything other than desktop/workstation/laptop installations.

One thing a lot of people seem to forget. Just because its Free doesn't mean it costs nothing. :-) In the case of Mono and C# apps on *nix, it costs a lot in memory/cpu. Sometimes beyond what I would consider reasonable.

Doesn't stop me trying them on my own machines at home, but at this point in time, I wouldn't consider it on any of my business related machines.

Re:What about Beagle?

[Korgan]

I fully agree. See my reply to 'irc.goatse.cx troll' to see my concurring arguments :-)

Re:What about Beagle?

[r_jensen11]

For the last time, it's not Beagle, it's Bagel! Have you actually opened up the virus and examined it? I have! It creates regestry entries under the name Bagel, not Beagle! Open it up with vim sometime, and search for Beagle. Then search for Bagel. Once you realize that vim went into hex mode, you'll be able to find Bagel, but not Beagle.

Re:What about Beagle?

[Chandon Seldon]

Although I wouldn't consider changing a working network architecture for a program like Beagle, your worry about RAM / CPU time may indicate that A.) your setup has scaling issues or B.) you haven't properly investigated the scaling problem.

In an optimal shared-server environment using commodity hardware and your own floor space, RAM and CPU time are pretty damn cheap. Like, if this program wants to use its own CPU and 512 megs of RAM 24/7, that should be a good deal if it will save 16 man-hours of time.

Re:What about Beagle?

[Feztaa]

Actually, Ubuntu's next release will be in October. They follow a 6-month release cycle that has them releasing a new distro every April and October.

Re:What about Beagle?

[dalutong]

I'm a little confused... my post didn't say when they were releasing their next version. (I checked my post -- just said i thought it'd be in the next version.)

Unless GNOME isn't going to be released with Ubuntu. I know the last releases matched, and I thought both were on 6-month schedules...

Re:What about Beagle?

[Korgan]

This is from my personal machine at home. Bit gives you an indication.

----8----
[steve@pyros ~]$ free
                          total used free shared buffers cached
Mem: 515340 449500 65840 0 27032 264536
-/+ buffers/cache: 157932 357408
Swap: 1048568 0 1048568
[steve@pyros ~]$ beagled
[steve@pyros ~]$ free
                          total used free shared buffers cached
Mem: 515340 468728 46612 0 27048 264560
-/+ buffers/cache: 177120 338220
Swap: 1048568 0 1048568
[steve@pyros ~]$
----8----

I'm sorry, but I'm not going to take a nearly 20meg hit on ram for each user that wants to run beagle so they can have flash/fancy search capabilities. Even on a machine with 1 or 2gig, that adds up quickly.

Now add in multiple instances OpenOffice.org writer and calc, multiple instances of the desktop environment (in my case, Gnome 2.10), a browser, Evolution, quite often Rhythmbox, plus whatever other tools the users are using at the time...

I'd rather not waste 20meg a user on something that should not in anyway use near that much. My issue stands. I will not use beagle (or Mono) in a production server environment until it is a lot cheaper to run. Why should I throw more ram and cpu's at it when the same (or similar) can be achieved for a much smaller cost?

For a small business like myself, with less than 20 employees, Mono and mono-based applications are simply too expensive right now. No matter how great or flash or whizbang or cool the application might be.

Re:What about Beagle?

[Hatta]

Desktop search is key missing feature in Linux...

Um what about locate and find? I know they don't have a cute puppy dog, but the functionality is there. And with programs like grep, sort, and especially xargs they're far more useful than the alternatives on other platforms.

Re:What about Beagle?

[Korgan]

Someone mod this one up. :-) +1 Funny

Re:What about Beagle?

[Chandon Seldon]

First, I wouldn't blame Mono for the problem. Having a little application level memory bloat in a version 0.11 application seems reasonable. I'm sure they'll get around to memory optimization when it's the right time.

On the other hand, 20 megs really isn't that much RAM. If this application had any notable value to your users, it would be nearly free to provide that RAM for the application, even with 20 users.

I mean, by your benchmarking system 7 megs of RAM is normal for a freshly started text editor. If search functionality is really important, I don't see "uses the RAM of three text editor instances" as terribly bad.

nat@icewing:~$ free
total used free
Mem: 1036316 795084 241232
nat@icewing:~$ gedit &
nat@icewing:~$ free
total used free
Mem: 1036316 801980 234336
nat@icewing:~$ emacs &
nat@icewing:~$ free
total used free
Mem: 1036316 808164 228152

Re:What about Beagle?

[Korgan]

Except that I can acheive the same thing with existing tools that don't stay resident. Whats so difficult about writing a tcl/tk (or even a PHP-GTK) wrapper to find/locate/grep/awk? Adding in smarts for the various formats is simple.

Sure it might be marginally slower because it doesn't have a resident daemon in the back end constantly crawling the $home directory, but over all it could provide exactly the same functionality without the 20meg hit for each user, plus the cost of the 'best' front end itself. It'd also only be resident when the user actually wants to use it, it wouldn't be slowing things down by crawling the directory tree in the background when that user is inactive (but other users might not be inactive at the same time) and the search results could be displayed in a way that was just as pretty. ;-)

I'm not saying that Beagle is bad. I use it at home for myself. But I'm the only one on the computer at the time I'm using it and I'm personally happy to take that performance hit.

My point is simple though. Mono/Beagle ARE expensive. Especially here in NZ where the cost of hardware is no where near as cheap as it is in say Australia (closest neighbour) or even the US.

I don't deny that Beagle is still a 0.12 release. In fact, thats part of my argument. I do not consider it ready at all for prime time use in a production/server environment. It is large, clunky and not optimised at all yet. The original poster wanted to know why Beagle isn't used as the search tech in Gnome 2.12... Quite simply, its still far to much of an alpha quality project in my opinion. Unless more people help out with it or more time is spent on the project by those already working on it, its going to be a while before I change my view on that. It doesn't stop me using it personally, but it does stop me even considering it in my office.

Re:What about Beagle?

[yozzman]

The fact that the over-hyped desktop search is perhaps missing in Gnome does not make it "not mature". Desktop search is a useful but highly uncritical feature that's been getting lots of press recently, but it hardly stands as a criterion on deciding whether a desktop is mature or not. By that criterion, few desktops were mature a few months ago, and few even are currently by default. Plus, I agree that anyone considering that desktop search is needed for a desktop to be mature should take a look at locate, find...

Re:What about Beagle?

[rekrutacja]

Oh, how wrong you are. I'm a journalist which writes one or two small news pieces daily, and one or two bigger articles a week. After several years of career my hardrive is full of texts i wrote in several different file formats (msword, openoffice, abiword, claris, rtf, txt - you name it).

Google desktop search i installed at work is a remedy to a lot of daily problems for me. Find what someone said three months ago, or six, i don't remember, maybe it's not in the text file, but rather in email, and i don't remember his name as well, and context, just that he said something like... oh, there you are, thank you google.

Before that i was starting serach, going for a walk, repeating search with different parameters, going for a walk...

Desktop serach is mission critical feature for me from day one it appeared.

Re:What about Beagle?

[yozzman]

Well, ok, let's say it's mission critical for you. That doesn't make it mission critical for everyone, or even for most people. GDB is mission critical for plenty of developers, but totally irrelevant for most people. It's availability has nothing to do with the maturity or relevance of a desktop.

Same goes for desktop search. Most people find their documents quite easily enough thanks to classifying their documents in relevant folders, and mails in relevant mailboxes. A quick search on those is easily performed.Then there are the geeks who master find, locate and all that. It leaves few people who actually *need* a desktop search by the likes of google desktop search or Beagle. Sure, it's a nice feature, but hardly critical.

And btw, last time I checked, Google desktop search wasn't a standard part of the windows desktop environment :-p

Re:What about Beagle?

[mixmasta]

Probably wouldn't be 20 mb per user, even before optimization, thanks to shared libs.

Re:What about Beagle?

[Chandon Seldon]

The more I think about the specific problem that's being solved here (speeding up a "Find" tool), the more I agree with you that from a software development standpont a 20 meg user-specific memory resident daemon is ludicrous. This is a task for 20 lines of Perl and a nightly cron job.

On the other hand, I still stand by the point I was trying to make to begin with: Heavy runtime environments like Java and Mono work well, and the couple megs of RAM that you pay by using programs written in these frameworks are usually a very good deal. Twenty megs of RAM per user for functionality that relevently and positively impacts productivity is an amazing bargain.

Re:What about Beagle?

[Chandon Seldon]

Hmm... while actually building an index file, 20 megs is small potatoes given a sufficiently large set of files to search.

Re:What about Beagle?

[Feztaa]

Oh, heh. I was going to say that I was hallucinating, but I see the source of the confusion now. You said:

(which will probably be on the same day as the release of GNOME.)

But the article states that GNOME will be released Sept 7th. So even though you don't mention a date, you are implying that Ubuntu will be released in September, given the context of the article.

There you have it.

(and I believe that the new ubuntu will have the new gnome).

Re:What about Beagle?

[dalutong]

You're absolutely right. I didn't see the date in the article. Oops.

nice headline...

[w4rl5ck]

for a second I was... "hey I have to install that imme... wait... I already did... I... *click* 2.10... [strange feeling]... ah, 2.12..."

can someone correct the headline or something? :)

Re:nice headline...

[etrnl]

Emailed the on-duty before it got posted, but... *sigh* Apparently noone was watching...

Totem

[astralbat]

I'll look forward to the day when Totem manages to play DVD's better than Xine.
Even changing the GStreamer backend for the Xine backend, Totem still never manages to play half the movies I seem to give it.

I do like the idea of a GStreamer based Mozilla plugin though. It will give users a great choice to drop the ugly Mplayer based plugin.

Re:Totem

[mixmasta]

I look forward to the day Totem can play for 5 minutes without crashing... Actually it can if I don't try complicated things like hitting the pause or FFWD buttons, etc. I'm using Gstreamer, maybe that has something to do with it.

Anyway, I don't watch media files much, so I can wait until it matures.

Re:Totem

[RossyB]

The good news is that Ronald has been working at this pretty much non-stop for 2.12, and GStreamer is *way* better at playing DVDs than 2.10.

It Just Works Philosophy

[vectorian798]

From the article: More software is taking advantage of the Hardware Abstraction Layer from Project Utopia. HAL-aware applications can display more information to the user, as well as benefit from "it just works" plug and play style hardware support. GNOME-VFS in GNOME 2.12 has improved integration with HAL, and now gives more visual cues about the types and names of media devices.

I am looking forward to this feature, especially - just another step towards making Linux more user-friendly.

In fact, this prerelease tour shows many exciting features for those who want to see a real desktop linux - improvements to Nautilus, a panel with Edit Menu option compliant with Freedesktop.org spec (how long have we been looking for something like this?), and more. Yay

Re:It Just Works Philosophy

[dbIII]

GNOME-VFS in GNOME 2.12 has improved integration with HAL, and now gives more visual cues about the types and names of media devices.

Open the iPod bay doors HAL.

Re:It Just Works Philosophy

[zsau]

Indeed. All those people who have ever said 'GNU/Linux will never be ready for the desktop' were wrong, and were always going to have been wrong. It just takes time. Not because free software development is an inefficient process, but because when a bunch of volunteers get together, they're going to create something they want to use first. It just so happens that the typical Unix user was happier with another system.

Now as that's changed and dollars've also become a driving force, GNU/Linux desktops are heading in a different direction, but it takes time. Windows wasn't written in a night, and neither can Gnome have been.

Re:It Just Works Philosophy

[cortana]

And MICROS~1 'stole' it from Apple.... etc :)

Re:It Just Works Philosophy

[zsau]

I was talking about free software development. I do not know how open-source development may relate to that kind of stuff, but there is nothing intrinsic in free software that causes it.

If you get a bunch of volunteers and watch them to make proprietry software (e.g. shareware) you see the same problems. If you pay a bunch of coders to write free software, you'll get similar results to proprietry software developed under similar circumstances.

Free software development is not the same as open-source software development (even if all free software may be open-source software and vice versa). That's why there's two terms!

(My experience as basically a non-coder is that if I have a problem and can provide adequate help, the developers try to give me adequate results. I do, however, remember that they're volunteers and have day-jobs and the like.)

Re:Gnome vs. KDE

[Anonymous Coward]

I don't want to troll, but I have always wondered... Why are there two major windows manager projects?

I don't want to flame, but I always wonder... how do people like you manage not to have seen this question discussed to death in every single previous Gnome or KDE-related discussion here on Slashdot since the dawn of time?

Re:KDE

[etrnl]

They do, from time to time. Take a look at the Extended Window Manager Hints Spec (that I was involved in administrating for a bit before it got too technical for me...).

They don't always work together very well, but given the basic design differences in architecture, that's to be expected technically. Personality wise... well, it's my experience that the more intelligent a geek is, the higher probability that they believe that anyone who disagrees with them is an idiot. (De Raat, Stallman, etc) That just breeds personality conflicts. (Linus seems to be an exception to that, for the most part.)

Efficiency

[Alain Williams]

The feature that I want is: efficiency.

Gnome is great at turning a fast computer into a sluggish one. Just because you have all of those CPU cycles doesn't mean that they have to use them, especially when lots of them seem to be wasted.

For instance: if you look (strace) at a typical gnome program when it starts up, it stats zillions of files; many of them more than once. This is why startup is so sloooooow.

Oh, I am trolling am I ? We all have fast computers so why am I making a fuss ? Think about: being able to save power (improve battery life) with a slower CPU laptop; people in the third world who cannot afford the super computers that we, in the 1st world, have on out desktops; think about sharing a server between many people (eg LTSP).

It would be nice to see a gnome release that just concentrated on making the code faster.

Re:Efficiency

[ssj_195]

I think both the major desktops are beginning to start the optimisation drive round about now; the thing is that software development usually proceeds in fits and starts, with different parts of the software development process going through the phases of Making It Work, Making It Work Well and Securely, and, finally, Making It Work Fast. I'm going to stick my neck out and say that even though both Desktop Environments have been around before the year 2000, both have undergone more intensive development, or at least had more features added, over the last year or so than ever before in their history (actually, this goes for Desktop Linux in general, from the kernel to X to the toolkits to the DE's to the distros themselves), so there are a lot of rough and unoptimised new additions in there.

Fortunately, unlike a certain other purveyor of Desktop OS's, the devs are actually fairly committed to making everything faster and less resource hungry (witness the GNOME optimisation bounties, and the efforts of the Ubuntu team). Robert Love gave a very interesting talk on optimisation of the desktop environments (I can't find a link right now, but the talk was called "Optimising GNOME", although some of the library-level changes could be conscripted by KDE and anyone else, really). KDE posted some resource-consumption figures for the (very rough and unoptimised) KDE4 port of Kate, and it already looks significantly better. Add in the upcoming xgl et al, and things should hopefully get to the absolutely perfect state of getting faster and faster while still adding features that every developer yearns for :)

Of course, it's pretty much impossible to continuously increase functionality without paying some price in terms of resource-consumption, so you might be better off going to less featureful DE's like, say, XFCE, if you prefer speed over functionality.

Re:Efficiency

[stilborne]

we (KDE) have been getting faster with each release in the KDE3 series. optimization isn't exactly a new thing on our plate, but you are right that KDE4 will likely show additional improvements.

Re:Efficiency

[dleifelohcs]

lol @ BOTH desktop environments.

We all know that there are only TWO out there. ...Gnome and Enlightenment DR17.

Re:Efficiency

[DeadMeat (TM)]

The feature that I want is: efficiency.

Gnome is great at turning a fast computer into a sluggish one. Just because you have all of those CPU cycles doesn't mean that they have to use them, especially when lots of them seem to be wasted.

For instance: if you look (strace) at a typical gnome program when it starts up, it stats zillions of files; many of them more than once. This is why startup is so sloooooow.

I've been following GNOME development on Planet GNOME and there's recently been a push to profile and optimize core GNOME libraries and apps to bring down startup time and memory usage. I doubt you'll see much talk about it in the GNOME preview release notes, though, because (a) it's still ongoing; and (b) the GNOME release notes are pretty high-level, so it doesn't include much beyond user-visible changes.

I agree that it's about time, though: I'm saddled with a 4-year-old Vaio notebook until my new Powerbook ships, and running GNOME 2.10 with 128 MB of RAM is just painful. With XFCE 4.2 at least I can load Firefox and Evolution simultaneously without the hard disk constantly thrashing; under GNOME I had to resort to elinks and mutt just so that the system was usuable.

Re:Efficiency

[Timbo]

For instance: if you look (strace) at a typical gnome program when it starts up, it stats zillions of files; many of them more than once. This is why startup is so sloooooow.

You kind of contradict your own point here. CPU cycles or less CPU intensive code is not what is needed, but less IO intensive code. It's all Amdahl's law see, and IO is the biggest f.

Re:Efficiency

[jejones]

Amen. When Firefox used its own file browsing dialog, it was kind of slow the first time I saved in a directory with lots of files, but on the second and following times came up immediately.

On my system at least (Ubuntu; maybe I should uninstall firefox-gnome-support...), Firefox now uses the GNOME file browsing dialog for saving files, and it takes forever for directories with many files, and doesn't seem to cache much, because the second through nth times around are just about as slow as the first.

Re:Efficiency

[Alain Williams]

No - after the first stat all the inode information will be in memory - there is no I/O, just a context switch to the Kernel to do the stat & back again ... ie CPU cycles.

Anyway - the point is that there seems to be much unneeded work done, they are much too slow, that is all that matters.

Re:Efficiency

[Timbo]

I wasn't refering to your example specifically, merely disagreeing with your assertion that Gnome is slow because of demands on the CPU. This is not the case. It is slow because it is heavily IO bound.

Robert Love - Optimizing Gnome

Re:Efficiency

[Chandon Seldon]

The slowest core you're going to get in a desktop or laptop today is a 550 mhz low power VIA processor with 128 megs of SDR RAM - and you'd need to try *really hard* to find it. From what I hear, the VIA processors are about the same as origional AMD Athlons clock for clock. Todays Gnome desktops could use some optimization for that class of system to run smoothly.

On the other hand, those processors are already obsolete, and the replacement is in the distribution pipeline. The next generation stuff from VIA is 800 mhz minimum and supports DDR RAM - about the same as my slowest test system. Current versions of Gnome should run fine on that as-is.

So... optimization for smooth performance is useful on: Rare low power systems that are already obsolete, and common desktops from five years ago. By the time an optimization cycle starting now were to finish, no new system will be slow enough to need it.

Re:Efficiency

[Olaf Underbridge]

I agree that efficiency WBN, but reliability is even more important. I switched to KDE when Gnome Panel 2.8 kept abending on my first day at a new job; because of Gnome, it actually wasn't until the third day that I had a usable workstation, and that was because on the evening of the second day I emerged (Gentoo) KDE. Gentoo is unreliable to build as well as to run. Emerges of Gnome 2.10 abort more often than not, whereas emerging KDE 2.4 Just Worked.

Re:Gnome vs. KDE

[Rapsey]

I dont think KDE has more money. I dont know of any company that puts money into KDE, but a few that put money into GNOME.
They do have more developers. Simply because its much easier to develop programs for KDE than it is for GNOME.

Re:Gnome vs. KDE

[bhalo05]

It seems that so much time is put in KDE and Gnome, that if the two teams worked together, they might make something superior to what they made on their own

Sigh. There's no way they'd be working together anyway. Gnome devs love C and GTK. KDE devs are C++ experts and like QT.

Besides that, Gnome users like Gnome. KDE users... well, like KDE. They can choose because both are different and there are many different kinds of users, you know...

There's not going to be a single desktop environment. Period.

Re:GNOME?

[Anonymous Coward]

Oh my! This is a poor attempt at a troll.

Gnome has had a consistent icon theme, mainly developed by jimmac, for a really long time. It is really quite different then the crystal icon theme in KDE and I actually think that it is the foremost icon theme in all operating systems today. I dont have a clue how you can see something that has been ripped from KDE.

And by the way, both KDE and Gnome are developed by international communities so if you want an "all american" desktop please use something from a large software vendor instead.

Re:Gnome vs. KDE

[Internet_Communist]

well to start gnome is written mostly in C and KDE is mostly in C++

so right there is a major difference in both coding style and what not...You couldn't exactly "integrate" them.

I also prefer the looks of gnome but I know just as many people who like the look of KDE better. It's very subjective.

My biggest concern is my programs not matching. Seeing as I like GTK themes better then most KDE themes, and nothing exists to match GTK themes on KDE (just the other way around) I'm stuck with just attempting to match my colors...Sure this is all apperance and doesn't say much about function but it's still pretty annoying...

Little annoying things like that are my main issue, and that's mostly just GTK/QT differences, not really kde/gnome....I don't actually use a DE, though I use a few gnome programs and thus have gnome installed, well partially anyway. I have konqueror installed so I can test my webpages with KHTML as well, plus I have a few apps which are QT only...etc..

so yes it's daunting but I don't see anything happening any time soon

and that's not to mention XFCE which is written in c++ but uses GTK libraries through it's own wrappers or something like that....

but in the end the question is, who do you really want using linux anyway? Do we really want your average joe on linux? Or trying to install/configure it? In the work place that's not so important, someone can set it up, put some big firefox/word processor icons on the desktop, and that's the end of it...

so what's really going on here? trying to dumb linux down enough to home users who don't want to take the effort to learn it?

I just don't see that happening.

and please note that gnome and kde are not window managers, they just include one. You can use any window manager you want with gnome or kde. Gnome uses metacity by default, and used to use sawfish before that. KDE uses kwin. There's a pretty big difference between toolkits like GTK or QT and window managers like *box,windowmaker,metacity,etc...Your comment makes me think you have no idea what a window manager does.

Still ugly fonts

[AvantLegion]

Neat features. Yes the font in every screenshot is still ugly and would be laughed out of the room if it were the default look for a GUI on an OS from Microsoft or Apple. They're thin, brittle, and chintzy. Though it's not necessarily the fonts themselves. Even when I copy over the fonts from Windows, they make those fonts look thin, brittle, and chintzy. Why can't, say, Times New Roman be rendered in Linux and match Times New Roman in Windows, without the crappy errors with italicized letters and such? All the tinkering with AA and subpixel rendering settings in the world still can't match what a few clicks to turn on and tune Cleartype does. It saddens me to leave my Linux desktop at home, go to work on a Windows PC, and marvel at how much better websites and such look.

Give me font rendering that doesn't suck.

Re:Still ugly fonts

[ocelotbob]

Write congress and tell them to reform patent legislation so that the Xorg folks can use the same techniques that Apple and MS does. It's not gnome's fault that the patent system is broken.

Re:Still ugly fonts

[Gleng]

If you're using a Debian based distro, you can run:

dpkg-reconfigure fontconfig

And then select the bytecode interpreter from the menu. Fixed. :)

Re:Still ugly fonts

[kelnos]

It's not like MS has had the best font renderer forever either. I'm forced to run Windows 2000 on my work machine, and the fonts there don't look quite as nice next to those on my Linux boxes.

I will admit, XP's Cleartype really does look nice, though it goes a bit too far into blurry-land for my taste, depending on the output device. At least the subpixel hinting is tweakable on X11 with fontconfig/Xft.

Re:Still ugly fonts

[nihilogos]

I really don't know why people still go on about fonts in linux. I haven't had a problem with them since XFree86 3.3.something.

If I've been working exclusively on linux for a few weeks I gag when I boot up windows and look at the fonts for the first time in ages. And vice-versa when I have been exclusively on windows for a while. It just depends what I'm used to.

Re:Still ugly fonts

[d99-sbr]

Microsoft release a ClearType calibration PowerToy a while back, and this made ClearType a whole lot better. Previously, I too thought it looked too blurred.

Re:Still ugly fonts

[Gleng]

Also, here's a copy of my /etc/fonts/local.conf which disables antialiasing on font sizes under 10 points, so you get nice crisp Windows style rendering on small fonts. I use Tahoma size 8 on my Gnome menus and it looks pretty sweet.

<?xml version="1.0"?>
<!DOCTYPE fontconfig SYSTEM "fonts.dtd">
<fontconfig>
<include ignore_missing="yes">/var/lib/defoma/fontconfig.d/ fonts.conf</include>

        <match target="font">
            <test name="size" compare="less">
                <double>10</double>
            </test>
            <edit name="antialias" mode="assign">
                <bool>false</bool>
            </edit>
        </match>
        <match target="font">
            <test name="pixelsize" compare="less" qual="any">
                <double>10</double>
            </test>
            <edit mode="assign" name="antialias">
                <bool>false</bool>
            </edit>
        </match>

</fontconfig>

Re:Still ugly fonts

[aussersterne]

"Thin, brittle, and chintzy?"

Um, I like "thin" and I don't find the Linux AA fonts to be "brittle and chintzy" which are certainly odd adjectives to use for a font (brittle? did you try to break your fonts by tapping them with a reflex hammer? chintzy? did you ask them for money?)

IMHO, the Cleartype fonts are just less readable than the subpixel AA fonts I see in, for example, KDE. I've spent tons of time trying to find a way to get Cleartype in XP to render Arial the same way that KDE does and as far as I can tell, it's not possible.

But in any case, I find the font rendering in Linux to be much more book-like. I find the unnecessary heavyness and color fringing (both adjectives, note, that actually say something about what I'm experiencing) of Windows fonts to be distracting and straining on the eyes.

Re:Still ugly fonts

[Linus Torvaalds]

Yes the font in every screenshot is still ugly

They don't look ugly to me at all. On the other hand, when I boot up Windows to test a website in Internet Explorer, I think the fonts are supremely crappy.

Re:Still ugly fonts

[ocelotbob]

No, I'd not accept broken system, and encourage others to complain loudly to their congresspeople and get things changed. Too often, software patents are so broad that anything even closely resembling the item in question requires expensive lawyers to be consulted. Remember SBC's claim that they invented the hyperlink? The system's broken, plain and simple.

Re:Still ugly fonts

[ocelotbob]

damn, slashdot needs editing. That should have been I'd rather not accept a broken system.

Re:Still ugly fonts

[Stevyn]

I see it the exact opposite way. Fonts on linux are so much nicer than on Windows. The anti-aliasing on two gentoo boxes, one a dell laptop and the other with a dell lcd looks a lot sharper than on windows. Whenever I reboot into windows xp, the fonts look fuzzy and I end up disabling Cleartype because it's hard to read.

I also just put windows vista beta 1 on the computer with the dell lcd and the fonts still look bad. Maybe it's just preference or maybe it's because I'm used to the fonts in xorg, but it looks a lot sharper to me.

Re:Still ugly fonts

[zsau]

Fonts are funny things. Everyone seems to like different ones. I dispise Cleartype on Windows (it's better without anything). And Mac fonts I find to be ugly, too. The fonts in my GTK+ 2 environment, though, I find to be absolutely supurb. If you gave me Windows-like rending of fonts on my GNU/Linux box, I'd punch you in the face.

(However, I consider Times New Roman to be godawful no matter how it's rendered, or even in print, so I almost always use only Bitstream Vera Sans/Serif/Sans Mono, the TeX Computer Modern series, and a handful others like Gentium for special characters. I even have my web browser configure to use my fonts and only my fonts, to the best of its ability.)

Re:Still ugly fonts

[mrogers]

Also, be sure to enable bitmap fonts - Debian disables them by default, but they're much sharper than antialiased outline fonts at small sizes.

If web pages still look blurry, add the following line to ~/.bash_profile to disable antialiasing in GTK2 and Gecko:

export GDK_USE_XFT=0

Re:Still ugly fonts

[SpinJaunt]

huper% cat ~/.fonts.conf
<?xml version="1.0"?>
<!DOCTYPE fontconfig SYSTEM "fonts.dtd">
<fontconfig>

  <match target="font">
    <edit mode="assign" name="hinting"><bool>true</bool></edit>
  </match>
  <match target="font">
    <edit mode="assign" name="hintstyle"><const>hintslight</const>
    </edit>
  </match>

  <match target="font">
<!--
    <test compare="less_eq" name="size" qual="any"><double>10</double></test>
    <test compare="more_eq" name="size" qual="any"><double>16</double></test>
-->
    <test compare="less_eq" name="pixelsize" qual="any"><double>7</double></test>
    <test compare="more_eq" name="pixelsize" qual="any"><double>15</double></test>
    <edit mode="assign" name="antialias"><bool>false</bool></edit>
    <edit mode="assign" name="hintstyle"><const>hintfull</const></edit>
  </match>

  <match target="font">
    <edit mode="assign" name="antialias"><bool>true</bool></edit>
  </match>
  <match target="font">
    <edit mode="assign" name="rgba"><const></const></edit>
  </match>

  <match target="pattern" name="family">
    <test qual="any" name="family"><string>fixed</string></test>
    <edit name="family" mode="assign"><string>mono</string></edit>
  </match>
  <match target="pattern" name="family">
    <test qual="any" name="family"><string>console</string></test>
    <edit name="family" mode="assign"><string>mono</string></edit>
  </match>
  <match target="pattern" name="family">
    <test qual="any" name="family"><string>mono</string></test>
    <edit name="spacing" mode="assign"><int>100</int></edit>
  </match>

  <match target="pattern" name="family">
    <test qual="any" name="family"><string>Helvetica</string></test>
    <edit name="family" mode="assign"><string>sans-serif</string></edit>
  </match>

  <match target="font">
    <test compare="more_eq" qual="any" name="weight">
      <string>bold</string><string>black</string>
    </test>
    <test compare="more_eq" qual="all" name="weight">
      <string>bold</string><string>italic</string>
    </test>
    <test compare="more_eq" qual="all" name="weight">
      <string>black</string><string>italic</string>
    </test>
    <test compare="more_eq" qual="all" name="weight">
      <string>bold</string><string>oblique</string>
    </test>
    <test compare="more_eq" qual="all" name="weight">
      <string>black</string><string>oblique</string></te st>
    <edit mode="assign" name="antialias"><bool>false</bool></edit>
    <edit mode="assign" name="hintstyle"><const>hintfull</const></edit>
  </match>
</fontconfig>

might want to change hintfull to hintslight.

Re:Still ugly fonts

[Breakfast Pants]

Cleartype can't even handle a non-LCD display. No matter what you get colored banding on a CRT at the edges of fonts. You can turn on regular font smoothing but then that doesn't even get applied to small fonts.

Re:Still ugly fonts

[LnxAddct]

Interesting, I run Fedora and think the complete opposite. Fonts under my linux box and Mac look great compared to Windows, I run at pretty high resolutions so that may have an effect. Maybe you should try a different distro.
Regards,
Steve

Re:Still ugly fonts

[Xyde]

I don't mean to sound rude or judgemental, but since when exactly did the OSS community give two shits about the ins and outs of the patent system?

LAME infringes, libPNG infringed all through the unisys debacle, faac, ffmpeg would etc - that's just off the top of my head (and I'm not generally familiar with most free stuff)

Why should something as fundamental as font rendering be the exception? Couldn't you add a --FUGLY-FONT-RENDERING=OFF option at compile time to enable the enhancements? I don't know if you linux users can't tell the difference or don't care, but compared to quartz or even windows, it's really quite vile. Don't direct me to the red herring which is your advanced settings box, the hacky, stilted output won't improve until it's gutted and fixed properly - trying to do it with patch after insipid little patch is futile.

I guess I'm quite opinionated on this subject but it's not until you've come from a system with legitimate font rendering and try to use linux that you realize just how much of a shambles the current system is. When the metrics are wrong in the first place all the antialiasing methods in the world won't make it acceptable.

This isn't a whiny complaint about the current state of it as I don't even use linux - just an observation. I don't design fonts, do DTP or anything else that would make me particularly anal in this area, just general usage.

FWIW quartz isn't perfect either - I still don't think you can define specific antialiasing methods on a per screen basis when using a multi monitor system (ie - subpixel for internal LCD but normal smoothing for external CRT - global only) and anything drawn using quickdraw (rare, though quicktime is still littered with it) looks like garbage...

Re:Still ugly fonts

[RAMMS EIN]

Funny enough, the fonts I get from X have been way better than what I have seen on Windows for the last few years. Many fonts that come with a typical X installation are crappy, but the Bitstream Vera fonts are absolutely top quality. FreeFont and Nimbus don't suck, either.

Re:Still ugly fonts

[RAMMS EIN]

``At least the subpixel hinting is tweakable on X11 with fontconfig/Xft.''

Which is a nice illustration of "the good thing about Windows is that it works the way many people want it to. The good thing about Unix is that you can make it work the way you want it to."

Re:Still ugly fonts

[Xyde]

I'm an OS X user with a pretty good grasp on unix. I tend to use the best software for the job as opposed to favoring certain projects because of their ideological principles.

This means yes photoshop instead of GIMP, mail.app instead of kmail, LAME instead of itunes (if I need to use mp3), apache instead of , quicktime instead of faac, and x.264 instead of quicktime. No favourites, just what's best.

I find the howling and bleating when MS bundles IE with Windows hilariously entertaining, especially in the face of the more...fanatical linux supporters. You get the One True GPL and all of it's inherited morals forcibly inserted into the oriface of your choice, with the DVD set being express posted the next day.

Your perceived value of "free" (as in beer) declines sharply when community-induced monthly sacrifices of "any competing commercial software" become the norm. You'd stop but the consequences (facing exclusion by a significant portion of whatever linux irc channel you happen to be on that week) aren't worth it, considering sound still doesn't work just right. An updated kernel module seems to be in CVS but there's some uncertainty about license compatibility so wait because You-Do-Want-To-Be-A-Good-OSS-Citizen-Don't-You?

Feeling a little restricted?

Fortunately for us there's always the Church of the Brethren, or even Christian fundamentalism for those who require rules which are significantly more flexible. Additionally, recent polls indicate the rituals and associated order are a good 53% more socially acceptable when compared to your average GPL fanatic.

If you found the above offensive in any way the recommend fix is to apt-get install sense-of-humour-0.3. This is very much a work in progress and may not catch all - subtle sarcasm seems to slip under the radar quite frequently. Questions regarding the quality of the code will be deleted immediately. Ungrateful attitudes are not appreciated - if you don't like it or think you can do better then bloody well code/fix it yourself. This software is absolutely free. ;)

Re:Still ugly fonts

[say]

Uhm... I repeatedly get compliments on my Linux desktop having better font rendering than Windows. And it's just a vanilla Ubuntu install with font packages like msttcorefonts.

Re:Still ugly fonts

[FauxPasIII]

> Give me font rendering that doesn't suck.

Okay, this is ridiculous. I mean, is somebody paying these people to troll on /. ? It's been a couple years
since I've seen anything on ANY platform that compares with Bitstream Vera Sans rendered by freetype2.
TNR on Windows XP tuned to the hilt doesn't hold a candle to it in my opinion. Lucida Grande on OS X
looks alright, but starts to look smeared at tiny font sizes, again IMO. Bitstream Vera Sans looks gorgeous
on any gtk 2.x app and on FireFox, and so far is the benchmark for best looking fonts on the planet.

Look, there are still plenty of valid criticisms of Linux for the desktop. I'll even spot you some; wireless
drivers are still very inconsistant in their behavior, and power management support is dodgy and hit-or-
miss. But you really need to drop the fonts argument, it just doesn't fly anymore.

Re:Still ugly fonts

[IamTheRealMike]

Or ... you could use the "Fonts" control panel applet and disable AA text there!

Re:Still ugly fonts

[ookaze]

Well, I'm same as you.
I tried several Windows fonts, and finally, reverted everything to Biststream Vera fonts on my Linux desktops.
I looked at the article from Firefox on Windows 2000, with Arial for Sans-Serif and Times New Roman for Serif : it looks ugly compared to the fonts pictured in the article, and I must say Arial must be the font I prefer on Windows (and on video, I subtitle with it).
Where it's even better, is the fact that it contains nearly all japanese characters (kanji) with which I work a lot, as I mix japanese and french in lots of documents and for some file names. To do that on Windows (in the case it's possible) I would have only one big Arial-Unicode font that comes with MS Office.
Plugging the japanese input method into Gnome and natively draw kanjis in a box, or enter the kana to see them transformed into a glyph in your editor or in Nautilus is amazing really.
I mean, while some people discuss on fonts they fing ugly (notice they don't say it's unreadable anymore, all they can say is it's ugly), people like me actually are productive with the great Gnome desktop and its fonts, doing things yet impossible to do in every Windows actually released.

Re:Still ugly fonts

[leobh]

Try Luxi Sans with grayscale antialiasing (or sub-pixel if you're on an LCD screen) with hinting off. See my my desktop. It's an incredibly nice looking font, nicer than MS' or Apple's fonts IMHO, at large and small sizes. I don't know what there is to complain about really, however most screenshots of GNOME have the not-so-nice Bitstream Vera Sans, so maybe people are not aware of the Luxi set of font's existence/niceness

Re:Still ugly fonts

[Breakfast Pants]

You still can't really turn off sub-pixel hinting in cleartype. That powertoy lets you change the font contrast and change between two common LCD sub-pixel layouts. You still get strange coloration on the edges of fonts on a CRT with either choice.

Re:Still ugly fonts

[d99-sbr]

I have never tried ClearType on a CRT, I guess it is designed to look good on LCD.

But if you don't like sub-pixel hinting, can't you just use the old font smoothing?

Re:Gnome 2.2 ey?

[nijk]

How can such a beautiful person be so stupid?

What I don't understand

[dnaumov]

about GNOME is: why can't they fix and polish existing applications, instead of COMPLETELY REPLACING them with new applications like they do now. People do not want to change 1/3 of their desktop applications every year, I know I don't.

P.S: Why is this in the BSD section anyway?

Re:What I don't understand

[ciroknight]

Well, the only completely replaced feature I see is PDF viewing, which, from my point of view, was in heavy need of being replaced. The applications which used to exist in its place were rough at best and terrible in my opinion.

Meanwhile, I think this release of GNOME is going to be leaps and bounds better than 2.10. I really like the fact they've done their best to get Cairo up and running (that OpenGL rendering feature is something they've been needing for some time, especially where the rest of their desktop system is so slow). Hopefully with some more work they can keep improving the usability.

Last note: I really like the browser now that they've added a relative URL button bar, but it's so.. well.. less powerful than having a text bar where you can simply type the location and instantly go there. But I guess it's better from a GUI point of view if you don't have that ability at all. Also, I wonder if they've ever fixed the slowness of loading folders.. my thoughts would be to cache the thumbnails and lists of the directory contents.. I mean how hard is it to make a graphical view of the "ls" command!?

Re:What I don't understand

[ReinoutS]

I really like the browser now that they've added a relative URL button bar, but it's so.. well.. less powerful than having a text bar where you can simply type the location and instantly go there.

I guess you mean the file manager, not the browser :-)

Anyhow, try pressing Ctrl+L, you'll be able to enter a location like you used to. That key combo also works in the filepicker, the web browser and maybe even more places in GNOME.

Re:What I don't understand

[aztracker1]

Because it's a common desktop on BSD as well as Linux... So, the same reason it's in the Linux section.

Re:Gnome vs. KDE

[Anonymous Coward]

I remember once there was a war, but since then there has never been another war because as a species humans learnt their lesson.

Re:KDE

[ocelotbob]

Why do I have to choose between the great features of Windows and OSX? Why can't they shut the fuck up and work together? The two groups have vastly different philosophies as to what a desktop should be. As you stated yourself, there are strengths and weaknesses for both. The one true anything simply won't happen.

Re:KDE

[Anonymous Coward]

Er. You don't have to "choose", for the most part. GNOME applications typically run just fine on a mostly-KDE desktop and vice-versa.

It's mainly the fanboys who don't STFU. Under the auspices of the http://freedesktop.org/ organisation, KDE and GNOME (and other minority desktop!) developers regularly work together, standardising interaction protocols and whatnot. KDE and GNOME have different design philosophies, and I happen to prefer KDE (though I wish it wasn't written in Qt-extended-C++). I don't WANT to see one or the other go away, though, because friendly competition drives innovation in the linux desktop.

Re:KDE

[Phleg]

The same reason that you have to choose between Microsoft Windows, Mac OS, Linux, and *BSD. The developers all have a different perspective on what defines good software, different project goals, different target audiences, and these differences are irreconcilable for the purposes of a single project.

I have to admit, I fail to see what is so utterly difficult about this concept that causes people to be so blind to the answer, despite the fact that they accept it on faith for everything else: why we have competing cars, fast food restaurants, colas, and so on.

new features, new shmeatures

[KZigurs]

ok, but when will I finally see a list of REMOVED features.

You know - those features that was recognized to be shitty and unusable. Removed default applications that simply don't work(r). Sourcebase size shrinking by megabytes. Abstraction and unification instead of the Linux Way(tm).

Yes, I'm flaming. But honestly - what's new? Desktop theme? Cool rendering approach? And why desktop envorement should ever mention HAL?

(yes, but I really like the fact that now Gnome is copying System7. Actually it's really a progress - all the usability quirks from Microsoft Windows have been copied already, yes?)

Re:new features, new shmeatures

[sploxx]

ok, but when will I finally see a list of REMOVED features.

You know - those features that was recognized to be shitty and unusable. Removed default applications that simply don't work(r). Sourcebase size shrinking by megabytes. Abstraction and unification instead of the Linux Way(tm).

I don't know about you, but lack of features in gnome is exactly why I switched back from Gnome to KDE.

The gnome devs. - IMHO - pulled out their chainsaws to remove every Gnome feature 'that is not in use by everyone'. (In the gnome 1.x -> 2.x transition). And they tried to enforce certain 'standard' ways of workflow. Both are things which I do not like. It may be perfect for others, but not for me.

I rather have a feature I need than a 'non bloated' application which is crippled. That is what I like about KDE.

I really do not understand all the fear about application bloat. Most of the time it is way easier to make features optional (with compiler flags, shared libraries etc.) than it is to implement new features. As a developer, I would rather spend time making new features than to think about what I could cut away from my application. If it is not easy to remove or disable the features, then it is IMHO not a problem of bloat but a problem of spaghetti code, i.e. a broken architecture.

Well, a few other reasons why I don't care so much about code bloat:
- my /usr is 2.5GB. That is only 1% HDD usage!
- the dynamic linker in linux only loads what is needed from HDD -> I don't care if my application or libraries contains code I do not use.

One last think that I'd like to add is that I still care about 'cycle bloat', i.e. applications that waste time just because it feels better to run it on top of a VM (GUIs implemented down to the pixel level in JAVA...). This annoys because it directly slows down my work. On the other hand, a few more MB/GB on HDD do not harm in any way.

Re:new features, new shmeatures

[strider44]

(yes, but I really like the fact that now Gnome is copying System7. Actually it's really a progress - all the usability quirks from Microsoft Windows have been copied already, yes?)

I laughed when I saw that. From what I understand the Gnome dialogue boxes have OK on the different side to Windows purely because they didn't want to do anything windows does. Correct me if that rumour's wrong, but it is funny anyway!

What is this crap about Linux copying everything Windows does anyway? People seem to see KDE or Gnome and say
"OMG you have a taskbar, Linux must be the same as Windows!"
"OMFG you have an internet browser! SO DOES WINDOWS!"
"OMG, there's a place to configure your system in Linux. You can configure your system IN WINDOWS TOO YOU FUCKING PLAGERIST!!!

Re:new features, new shmeatures

[Pecisk]

Hmmmm, trolling and flaming is nothing bad these days, hey, you can get even modded Interesting.

"Default applications" works. It works on mine Gentoo with GNOME 2.10, It have worked on Debian/Ubuntu with GNOME 2.6 and 2.8. And what a heck 'Sourcebase size' have to do with usability of GNOME?
There are only few features which have been removed in GNOME 2 series - I even don't remember, if any.

It seems that you haven't used GNOME not at all. "New desktop theme" Cleanlooks is the best in the market for GNOME/GTK and for that reason it will be included in GNOME 2.12. It really pushes Windows XP default theme out of the window.

Cool rendering approach - when someone hypes Windows Vista, they mentions cool new looks, transparent windows, etc. Cairo, which is base for all such features in GNOME/KDE, will be supported in GTK 2.8 and it IS BIG thing.

You are flaming. And honestly, I don't see any point to have you Interesting modding.

Re:new features, new shmeatures

[ciroknight]

Yeah, but.. Quick question: How big's your system memory?

I contest that at lower than 256MB of ram, both current desktop environments become almost unusable. Currently my desktop beats the requirement 4 fold, but when I was getting my computer up and running, I through in the first stick of ram I could find (a 128MB stick) and GNOME was slower than Grandma's molasses in the dead of winter.

KDE's no better, as it actually FROZE on the machine.

Reasons more users should care about codebloat even if you don't:
1) System memory is a priviledge, not a right.
2) Dependency hell is worse than Windows 98 on a bad day. (four versions of the same library in RAM at the same time? HOLY HELL BATMAN!) [this is actually translatable to: "Using the latest and greatest library is *not* nessicarily a good thing"].
3) Code bloat tends to make maintainance hell.
4) Code bloat contributes to cycle bloat (something you said yourself that you cared about, almost contradictory to your previous point).
5) Most people do not have emmense hard drives (you may have a 250GB hd, but my mom's computer has a 40GB, and my sister's laptop, only a 20GB. Hell, my work desktop machine's hard drive is only 6GB, though I often will bring in an external) [but I guess this point doesn't really matter, because Linux isn't your mom, your work, or your sister's Operating System now is it ;)].
6) Because only you can prevent forest fires.

I'm not a big fan of either major desktop environment right now, but GNOME's easier for me to use coming from Mac OS, plus it doesn't look like Windows, and those two things together make it my choice (though it saddens me, because I love C++ and GTK's C++ support, in my opinion, is absymal, thus forcing me to use WxWidgets for all of my GUI work (which is like using Microsoft Foundation Crap, only it goes on every machine!). *sigh*

Re:new features, new shmeatures

[sploxx]

Yeah, but.. Quick question: How big's your system memory?

1GB. Ok, that's still a bit above average, I suppose :)


2) Dependency hell is worse than Windows 98 on a bad day. (four versions of the same library in RAM at the same time? HOLY HELL BATMAN!) [this is actually translatable to: "Using the latest and greatest library is *not* nessicarily a good thing"].


As far as I have seen this, it is more a problem of distributions offering libraries in a lot of versions (such as debian - which I am using). If the API is clean and the distribution sets a standard by including just a certain version of a library (SuSE/former Redhat), I see no problem here.
But in most cases, even on debian you can pin all programs to a single version of a library. With recompilation from source, of course...

3) Code bloat tends to make maintainance hell.
4) Code bloat contributes to cycle bloat (something you said yourself that you cared about, almost contradictory to your previous point).

Well, I tried to express this, but english is not my mother tongue, so I may have been unclear about this:
If you imply code bloat with new features, then I disagree. What must also be fulfilled is that the spaghettiness/entropy of the code increases with the new features (for example a lot of patches from a lot of different people who do not hesitate to look at your current code).

5) Most people do not have emmense hard drives (you may have a 250GB hd, but my mom's computer has a 40GB, and my sister's laptop, only a 20GB. Hell, my work desktop machine's hard drive is only 6GB, though I often will bring in an external) [but I guess this point doesn't really matter, because Linux isn't your mom, your work, or your sister's Operating System now is it ;)].

Yes. But if less-used features of the application can be removed by the means of compiler switches or plugins, they do not need to be removed completely.

To sum it up: I just don't think throwing features *completely away* to get the software running on older hardware is the right way to go.

Re:new features, new shmeatures

[John Nowak]

Gnome has the OK buttons always on the right so that you don't have to go hunting for them. They're always at the bottom right. Even if there are 6 six buttons, it is always in the same spot. This is unlike KDE, where I have to look for the OK button in every dialog. No it isn't a major thing, but it is this attention to detail that makes Gnome (and OS X) great in my opinion.

Re:new features, new shmeatures

[10Ghz]

I contest that at lower than 256MB of ram, both current desktop environments become almost unusable

maybe, so what? What is it with this "This computer ran Windows95 just fine, and it must run KDE4/Gnome 2.12 just fine as well!". Instead of trying to run today's software on yesterdays machine, why not invest few bucks and buy more RAM? 128MB is ancient by now. 256MB is very low-end these days. And from what I have seen, at least KDE works just fine on 256MB.

5) Most people do not have emmense hard drives (you may have a 250GB hd, but my mom's computer has a 40GB, and my sister's laptop, only a 20GB. Hell, my work desktop machine's hard drive is only 6GB

my personal laptop has 6.4GB HD. And it has KDE, with free space to spare. I fail to see the problem here. Even the smallest HD's that are available today are downright huge for regural work. No, P2P is not "regural work".

plus it doesn't look like Windows

How does KDE look like Windows? And is it automatically a bad thing to look vaguely like Windows? Do you think that things should look different from Windows for the sole reason of lookind different than Windows? Hmmm, Gnome does have taskbar in the bottom of the screen, so it looks like Windows as well...

Maybe if KDE moved K-menu to top right corner and taskbar to left side of the screen, then you would be happy, since it would look "different than Windows"?

True Transparencies?

[xjerky]

TFA didn't seem to mention anything about it. I would hope that 2.12 can utilize X.org's native transparencies that have been present for months now.

Re:True Transparencies?

[kelnos]

I would hope it doesn't, actually, at least not by default. The X Composite extension is still buggy as hell, and requires a recent video card with good drivers (pretty much only nvidia, and then only using the binary-only driver, and to some extent ATI), otherwise it's painfully slow. Hopefully that should improve with X.0rg 6.9.0/7.0.0 onwards, but I don't think that'll be timely enough for GNOME 2.12.

Re:Gnome vs. KDE

[stilborne]

> I dont know of any company that puts money into
> KDE

SUSE/Novell, Trolltech, Mandriva and Linspire all pay people to work on KDE directly, to name just four companies you probably know by name. i could also name a bunch of small you companies you don't know who each fund part of a developer to several developers, ranging from co's like kitty hooch who funds quanta developers to KDAB who does a ton of work with KDE and groupware..

Re:Hidden sinker

[bhalo05]

I've never heard a newbie complaining about the variety of linux desktops. They may have problems installing software, and of course headaches with hardware drivers and kernel compilation. They also say things such as "does do not this thing run Half-life 2?". But I don't remamber any user moaning about the fact that there are two different desktops. In fact, they usually just use the one their distribution uses by default, and don't try the other until they are not newbies anymore.

Re:Gnome vs. KDE

[lasindi]

Why are there two major windows manager projects? Not like lots of other smaller projects like IceWM. It seems that so much time is put in KDE and Gnome, that if the two teams worked together, they might make something superior to what they made on their own. Does KDE and Gnome have the same goals, or are they very different?

I'm no expert by any means on either KDE or GNOME; this is all from what I've gathered as a KDE user, so don't quote me on any of this. I personally wouldn't want the two to become one because they do seem to go into different directions. A perfect example is their file browsers. I've always loved Konqueror, especially since it means I get to use tabbed file browsing. Nautilus, on the other hand, decided to use a "spatial browsing" interface, which opens a new window for each folder you open. Personally I can't stand this, but it was decided on after much deliberation by the GNOME people, so apparently some people like it. KDE also behaves a lot more like Windows than GNOME does. Some people dislike the Windows interface, but for newcomers to the Unix world it is useful to have this to ease the transition. So long as you can use KDE apps in GNOME and GNOME apps in KDE, I think there's no problem keeping the two projects separate.

This is my guess, correct me if I am wrong. KDE has more developers and money. Gnome has fewer people, but more creative people. KDE will give you everything and the kitchen sink. Gnome will find ways put a twist into things, to make it fun.

I'm not sure if GNOME or KDE has more people; I've always been under the impression that they have about the same number. As far as corporate sponsorship goes, though, companies like Novell are going for KDE, whereas Red Hat has poured a whole bunch of resources into GNOME. As far as putting a new "twist" into software, yeah I'd say that's true of GNOME. The difference is that, IMHO, the twists just make the software harder to use. But again, this is all in the eye of the beholder. Different people like different features, and that's why I'm fine with two different desktops.

For me, the difference boils down to this. GNOME does what it's supposed to do very well, and it's lighter-weight and cleaner. But what GNOME is supposed to do isn't what I want (like spatial browsing). KDE is supposed to do what I want, but it feels slower and there are weird bugs that can be annoying (example: my desktop icons magically rearrange themselves sometimes).

Wow!

[David Horn]

My boss saw this over my shoulder and is almost (but not quite) possibly thinking about maybe trialling Linux on his home machine...

Re:Wow!

[GeffDE]

Yes, right after he fires you for being on slashdot all day...;-)

Re:Wow!

[Anonymous Coward]

It's alright - it's the weekend. He probably meant 'girlfriend'

Re:Wow!

[InfiniteWisdom]

On Slashdot? Think again.

Re:Wow!

[RogerWilco]

And people are recommending hacking your X.org config scripts in the same thread. I think your boss should probably wait with his trial, until you can actually use Linux without hacking config scripts.*

*Note I use several machines that do not even have X installed, just CLI, VI rulez!

Re:Gnome vs. KDE

[Mad Merlin]

...that if the two teams worked together, they might make something superior to what they made on their own.

I doubt it. To make a comparison, if you have two trains facing opposite directions and you stick them together, what happens? No, they don't end up getting twice as far, they end up hindering eachother's progress, and the final combined result is less than what they could have accomplished on their own. While they may seem to do similar things on a very superficial basis, they have very different goals and very different ways of getting there.

Re:Gnome vs. KDE

[twener]

> I dont know of any company that puts money into KDE

You then can learn something at 1, 2, 3 and 4 or in incomplete summary: HP, Intel, Novell, Trolltech, Linspire, Mandriva and countless medium-sized businesses.

Spatial Tree file view and huge directories

[Travoltus]

How much time and cpu power/RAM will be needed for this feature to process a tree with a hundred directories, one directory having a total of, say, 5000 pictures/thumbnails?

Re:Spatial Tree file view and huge directories

[ScriptGuru]

The tree expander only loads the subdirectory when expanded. So it takes roughly the same amount of CPU power/RAM as viewing the folder and double-clicking the subdirectory. More because of the extended architecture, and less because you won't be creating a new window each time.

Re:Spatial Tree file view and huge directories

[DeathAndTaxes]

Any view of a folder with 5000 files will take a long time to load...Even if you have the thumbnails off. If it's all images and the thumbnails aren't generated yet, take a nap or come back tomorrow...It takes a while for this to happen.

Yes!

[FrostedWheat]

New clipboard management, based off the Freedesktop.org specification and tightly integrated with GNOME, allows for objects to persist in the clipboard longer than the lifetime of an application

About time! Closing the application and losing the clipboard contents always annoying me and was a real embarrasment for Gnome. I'm glad it's been fixed but I wonder why it took so long.

Re:Gnome vs. KDE

[twener]

Add IBM to above, not really medium-sized - no, just missed to copy it.

Re:Gnome vs. KDE

[kelnos]

and that's not to mention XFCE which is written in c++ but uses GTK libraries through it's own wrappers or something like that....

Bzzt! Xfce is written in straight C using GTK+ directly. We have a couple support libraries with utility functions and custom widgets, but there are no "wrappers or something like that".

There are currently C++ and Python bindings, but the desktop itself is all written in C.

Did they fix the Gnome Settings Daemon?

[k8to]

It used to be (as of Gnome 2.10), that if I would launch Natilus from a shell, or if another application (such as totem) launched Nautilus withous asking me, that I would get a window where every icon was a grey piece of paper. Apparently Nautilus is incapable of showing the correct icons for the files unless gnome-settings-daemon was launched beforehand. Further, Nautilus would open a window the size of the screen and draw the desktop onto this, as well as draw a static image of the desktop to the root window.

It is possible to manually ask nautilus to please not draw to the root window. This request will also ask it not to open the useless desktop sized annoyance window that some window managers aren't aware it intends to _be_the desktop. It is _not_ seemingly possible to ask nautilus, such as in a settings file, to please launch gnome-settings-daemon if it is not running. Since I do not run the gnome panel, and other gnome desktop tools, it is never already running, and the behavior of the window and the look of the icons is so unhelpful to make me upset that the window is on the screen.

It would be a pleasure if the Gnome developers were to consider improving the behavior of Nautilus and/or the gnome-settings-daemon to handle users of gnome applications who are not necessarily interested in running the entirety of the gnome desktop experience.

Re:Did they fix the Gnome Settings Daemon?

[moranar]

nautilus --no-desktop
As it says if you do
nautilus --help
But I don't really know about the correct icon for file types. Nautilus has done this for at least a year, and quite possibly more.

Re:Did they fix the Gnome Settings Daemon?

[nihilogos]

I just start the gnome-settings-daemon in xinitrc.

Re:Did they fix the Gnome Settings Daemon?

[MichaelSmith]

It would be a pleasure if the Gnome developers were to consider improving the behavior of Nautilus and/or the gnome-settings-daemon to handle users of gnome applications who are not necessarily interested in running the entirety of the gnome desktop experience.

Gnome is a desktop environment. I don't think the developers care what happens outside Gnome. It's a shame because some of their applications are pretty good.

I had the same problem with gnome-settings-daemon under RH 9. FC3 seems to handle this situation with more grace, but given the complexity of Gnome, etc; it is impossible to say what has changed to improve the situation.

Re:Did they fix the Gnome Settings Daemon?

[k8to]

Filed long ago. I'm just bringing it up in hopes of discussion.

Re:Did they fix the Gnome Settings Daemon?

[k8to]

Yeah, this is exactly what I was referring to when I said:

It is possible to manually ask nautilus to please not draw to the root window. This request will also ask it not to open the useless desktop sized annoyance window that some window managers aren't aware it intends to _be_the desktop.

The obnoxious thing is, however, that if some other app, for example totem when chosing something to do with plugins, launches nautilus, it does not know that it needs to pass --no-desktop. And so by choosing an item in totem, suddenly my whole screen is covered with a big dumb window, and my root window is painted with useless trash, which stays around even after I find the nautilus process id and kill it, since the desktop "background window" doesn't respond to the close box correctly.

Re:Did they fix the Gnome Settings Daemon?

[k8to]

Yeah, I'm unwilling to do this since it uses more memory than is reasonable when I use gnome apps pretty infrequently. I'm kind of baffled why they can't just fall back to configfiles.

Re:Did they fix the Gnome Settings Daemon?

[nandhp]

There is an option to disable the desktop. http://gtweakui.sourceforge.net/ or gconf-editor can set it for you.

Re:Did they fix the Gnome Settings Daemon?

[potHead42]

You can turn it off permanently with gconf-editor, just change /apps/nautilus/preferences/show_desktop to false. But yeah, this really should be in the preferences somewhere...

Re:Did they fix the Gnome Settings Daemon?

[k8to]

I disagree that it should be in the preferences. I think if programs are going to be able to launch nautilus, it should launch with or without the desktop intelligently.

If i actually launched gnome with like gnome-session & , I'd want a desktop, but if totem opens a single nautilus window to allow me to drag plugins into it, it shouldn't create a desktop. And I don't think this is totem's job. Totem just says "show me this directory". Nautilus should be aware of the context of the request, because it's the one starting up as a result of the request or knowing it's already in charge of the whole desktop.

Basically I think the problem of how to launch is quite solvable, and could be documented fully to avoid surprises. And it frustrates me that the nautilus developers have not done so.

I do thank you for mentioning this setting and it mostly does what I want, although I continue to be sad about gconf. .nautlusrc would be preferred. A hive of xml is not more discoverable or accessible, it is less in both ways.

Re:Did they fix the Gnome Settings Daemon?

[juhaz]

No. They couldn't possibly have fixed it, because it wasn't broken

You can set icon theme from config file without running g-s-d, just like you can set widget theme. And like others already mentioned, you can tell nautilus to not become desktop with a gconf setting. So neither of these problems actually exist.

I am inclined to agree that documentation for these and some other semi-common issues is less than optimal, however first google result for

nautilus without gnome

would've gotten you this http://mail.gnome.org/archives/nautilus-list/2004- September/msg00095.html thread, and the instructions contained therein:

You should be able to just put
 
  gtk-icon-theme-name = "<themename>"
 
into your ~/.gtkrc-2.0.

Re:Did they fix the Gnome Settings Daemon?

[k8to]

It doesn't add extra complexity when the data is coming from the same files anyway. The program could

1 - parse the files itself using the same exact code dlopened from a library and hidden entirely from the applications asking.
2 - autofire the settings daemon again entirely hidden from the the host app
3 - do something entirely stupid that doesn't work reliably

let's go with #3.

Re:Gnome vs. KDE

[twener]

> Why are there two major windows manager projects?

Because GNOME was started as reaction to Qt not being GPL licensed in ancient times.

KDE Looky-Likey

[SpinJaunt]

Yep, Gnome starting to look more like KDE, they might be better of doing a merger *hint*.

One pet peeve of most *Desktop Enviroments is the size of the icons by default -- yes I know they can be changed quite eaily.

  * I jest. I jest.

Re:Gnome vs. KDE

[grumbel]

In the beginning it was a license issue, KDE was based on the non-free QT while Gnome was purly based on LGPL stuff. However in the meantime stuff has changed, QT is now free and the goals have shifted. From what I can tell KDE tries to go the feature-bloat and customization aproach, implement whatever seems usable care about sorting stuff later and give the user the freedom to configure pretty much anything. Gnome on the other side tries to develeop the desktop environment for your grandma, if a configuration option might be not totally necesarily it gets moved into gconf (aka Gnomes Registry clone) or even removed completly. Gnome people also try to get it right, instead of basing to much of their stuff on existing UI principles, so you get for example the different button order in dialogs in Gnome.

Both of the aproaches are pretty much incompatible, you have a hard time getting stuff extremly simple and at the same time full of features and configurablity. So in the end I am very happy that we have both. Beside from that they don't work against each other, they just try different routes. When it comes to important core concepts they actually follow the same standards, see freedesktop.org.

Explanation of the basics?

[rduke15]

Your comment makes me think you have no idea what a window manager does.

Neither do I, really.

I always wondered, and would very much like to find a good and simple explanation of these things.

What are exactly all these pieces, and what are the relations between the OS, X (are there alternatives to X on Linux?), GTK/QT/other?, Gnome/KDE/others?, Window managers, "Desktop environments", etc.

If someone knows a good page giving an overview of all this, that would be nice. And how this architecture is comparable to or different from the ways Windows and Mac manage the GUI, would be interesting too.

I'm really missing a broad understanding of this subject.

Re:Explanation of the basics?

[johnnyb]

The OS:

Some call the OS just the kernel, and others put the OS as the kernel + libraries + basic utilities, and others put the OS as "everything that ships with your distribution". All of these are more or less correct, and just a matter of perspective. I personally prefer the latter two because it is the measure of compatibility of downloaded tools.

X is a graphical system. It is actually rather barebones, as it doesn't specify anything about how the controls work. It doesn't include a text entry box, buttons, graphics, or anything. It is simply a basic toolkit for network-driven graphics. It doesn't even hook up to sound.

A "toolkit" is a set of widgets -- text entry boxes, standard icons, buttons, etc. GTK and QT are toolkits.

A widnow manager is just what it says -- it manages windows. The top bar of windows, with their close, minimize, and maximize buttons, are actually drawn by the window manager, not the application. The Window manager often, but not always, draws to the background (sometimes the file manager does this). The window manager is involved if you have multiple desktops. It handles minimizing and maximizing. It handles laying out the windows, and telling them where they can and can't go.

A "desktop environment" is a complete collection of tools and specifications for a desktop. For example, the GNOME desktop environment consists of (a) the GTK toolkit, (b) a set of libraries for making applications work together in a consistent manner, (c) a panel and a set of applets (most desktop environments include a panel), (d) a set of specifications for interaction. These specifications include specifications for the function of window managers, specifications for human interface interactions, specifications for the handling of clipboard data, specifications of standard application interfaces for doing various things like printing, configuration, etc. Many desktop environments specify a default window manager as well.

Windows and Mac roll all of these parts into one. All you have is the desktop environment, which contains everything else. There is _some_ separation, but it is not as clear-cut as it is on UNIX. Whether this is a benefit or a flaw depends on your perspective.

Re:Explanation of the basics?

[pantherace]

A window manager simply manages where X11 windows are displayed.

The simplest window manager is probably twm. (In that it's nearly the minimum, even if it doesn't do it the now conventional way.) If you want to see what it does, login to a box with a session setting of 'failsafe' (Available on most linux boxes.) Then run 'twm &' (ampersand to put it in the background, so you have a command line free) then type in any favored application's name.

Now, other window managers add upon the simple positioning of twm, toolbars with more useful things, close, maximize, etc. Then you have things like virtual desktops. (Implemented in any GNOME-compatible window manager, such as metacity, KDE's kwin, icewm, fvwm, and most others on X11.) And some options to allow windows to say things like I don't want a border, or request a size. (Run kicker in twm to see the annoyance of borders on apps which should have them.) This all belongs to the window manager.

Now, given that most people don't like just a plain screen, you add a panel. KDE's is Kicker, I think GNOME's is gpanel. This gives you the panel at the bottom, a windows like toolbar. It's not limited to that, I personally have 4 panels per screen, each at the center which auto hides. Something that both GNOME and KDE can do.

Now, add in something to control the appearance. In GNOME's case it used to be a browser, recently replaced. KDE having kdesktop. Both containing the icons on the desktop, and background, etc.

GTK, Qt, FLTK, Motif, etc. are all toolkits, and manage what is drawn into the windows. Essentially you say: Please draw X widget here, and they provide the appropriate messages to the X server. You don't have to use them, as you can call xlib directly, or alternatively OpenGL. However, that's not recommended at all.

X11 works by passing messages to the sever: lots and lots of messages. It's handled in a way that it doesn't matter where the client is provided it's authorized. Thus providing the easy way for display across a network. Windows and Mac OS X use message passing as well, but in a local only method. It might be possible to abstract them similarly, but I personally doubt it will happen. X11 isn't really any slower than Windows or Mac OS X, contrary to what many people wishing to ditch X11 say. Drivers are generally better optimized on Windows (Manuf's drivers) or Mac OS X (very very few cards supported). Windows and Mac OS do not have a distinct window manager, or desktop from the OS, and I don't remember exactly how those are managed internally.

KDE and GNOME are called Desktop Environments, because they provide all the above, as well as other useful programs. KDE pretty much providing anything you'd want to use from one place, GNOME more leveraging already existing GTK programs. However, for the most part you can use programs that only come from the DE. (Example, at the moment, I have all KDE apps open except firefox. and I've got browser, IM, and anything else you'd expect to find.)

There exist a few alternatives, Qt/Embedded (What the Zaurus uses from Sharp) or Opie (opie being a fork of Qt/E, periodically semi-synced see openzaurus.org), GTK's embedded (name escapes me, see openzaurus.org), DirectFB (GTK was/is being ported to it), and svgalib (requires root permission, links is one that support svgalib output). While most of the examples are embedded, X11 is not necessarily heavy. Keith Packard had a server for ipaq which was 600K, kdrive being the name. The Agenda, a nice little linux handheld used X11 on a 66MHz processor.

Hope this helps to give you the brief overview, as well as a starting point for where to look further. More or less all the current systems are fairly similar, the differnce being how they seperate out the different layers, or in a DE/OS's case mash them together.

Re:Explanation of the basics?

[FullCircle]

I'll get as close as I know how...

Linux is an OS. It can give access to hardware, memory, etc. and run applications. It can't draw, but it can show text. Think of DOS on steroids.

X is really just a big application that can draw on video cards that it knows about and use pointing devices such as mice, but X knows squat about windowing. If you run an app on X without a window manager, it just sticks in the upper left corner of the screen and has no title bar, close button, can't be resized, etc. It's just up there.

DirectFB is one semi-popular alternative to X, XDirectFB is X on DirrectFB so that there are actually apps available. There are others, but X has all the apps due to popularity and completeness.

The window manager (twm, metacity, kwin, kahaki, etc) keeps the apps separate and allows you to place them. Beyond that, its just more features.

A widget toolkit knows how to draw widgets for buttons, window borders, sliders, etc. Those are like GTK, QT, wxWindows or motif.

A Desktop is just a bundle of related apps that someone picked or wrote specificly to work well together, the toolkit they are written with and usually, but not always, a window manager. You can use other window managers with an integrated desktop, but normally people use the default one (metacity for Gnome, kwin for KDE). XFCE is another really good desktop based on GTK.

If you pick a window manager, toolkit and filemanager that you like, you have made your own desktop. It may not be a very well integrated one, but by definition, it is.

On Mac and Windows, there aren't really any choices, but these parts are there under the hood. They don't really make a distinction in names because these aren't usually something that sells an OS to the masses. It's a geek detail for low-level programmers and you don't get to choose a replacement. Well, you CAN use wxWindows, GTK and QT on other OS's in apps you write, but by default every application uses the default widgets.

On Mac, Aqua is a well known word and roughly equates to a desktop and X. Remember hearing that OSX is based on BSD? It is running a Mach kernel (similar to the Linux kernel) and FreeBSD spinoff (but similar to Gnu) utilities. They call this Darwin and is equivalent to what we know as Linux or Gnu/Linux. Aqua gets loaded in on top of all that to make a GUI like X and a desktop.

Correct me if I'm wrong, but I think Quartz is similar to the Cairo drawing library. It makes for easy programming of drawing operations and speeds up the process of actually drawing anything to the screen.

On Windows I've heard that GDI is similar to a Window manager and also a brain dead Cairo or Quartz. Remember that as far back as Windows ME, Windows sat on top of DOS just like X and a desktop are on Linux. On windows 3.1 this was very clear as you bought them separately. Windows was like X and a desktop, nothing more. Even now on 2000 and XP, if you boot into a recovery mode you can run NT without the Windows GUI. They just don't give you the option so easily anymore.

I hope this helps and that I'm not too far off in my analogies.

Re:Hidden sinker

[kelnos]

I've always felt that the fact that there are two major free *nix desktops is a real detriment to normal people adopting free OSes.

Why does that have to be the goal? I'd much prefer a wide variety of choices over being friendly to the average-joe user. Am I being selfish? Not really. The average-joes can have Windows and Mac OS X. I'll stick with Linux.

Note that there's nothing stopping a company from taking a snapshot of GNOME or KDE (or whatever), and spending a year or two turning it into an average-joe-perfect distribution. IMHO, selling to the teeming masses is more the job of a commercial distro vendor than hackers working on a desktop environment. Let the hackers have their fun (I know I do), and let the businessmen make their money by appealing to the largest customer base.

Looks fantastic!

[a.different.perspect]

I'm terribly impressed - 2.12 looks like it will fix basically every gripe I have with out-of-the-box GNOME. The implementation of a clipboard service and menu editor is long-overdue, and most users have had to find third-party programs like gnome-clipboard-daemon and Smeg; the switch to Clearlooks saves us all a download; the extra configurability of Sound Juicer means I'll be able to switch from Grip; DVD support in GStreamer means I won't have to change the Totem backend to XINE, and a Firefox plugin for it means I don't have to have other players installed. The GNOME project have done really great work here: 2.12 seems to be a big step towards making GNOME a self-contained and complete desktop environment.

Re:Gnome vs. KDE

[Elranzer]

My biggest concern is my programs not matching. Seeing as I like GTK themes better then most KDE themes, and nothing exists to match GTK themes on KDE (just the other way around) I'm stuck with just attempting to match my colors...Sure this is all apperance and doesn't say much about function but it's still pretty annoying...

Isn't that what the whole Bluecurve project was all about? Making Gnome and KDE apps look almost like they were the same interface? The final version of it (in Fedora Core 3 and RHEL 4) seemed to pull it off nicely.

Too bad that starting with Fedora Core 4, Redhat is switching the default theme to that ugly-ass Clearlooks and it's light-blue/light-brown color scheme. I guess since Fedora is suppossedly no longer a fork of Redhat, they have to pull the plug on redhat-artwork which means R.I.P. Bluecurve. :(

Nobody? I do

[ThreeDayMonk]

Don't make the mistake of assuming that because you don't want something, nobody does.

I like big icons, because they are easier to click. I have plenty of screen space, so wasting it isn't a problem.

I like stripes, because it's easier to track the information across.

I even like the spacial browsing mode of Nautilus.

Now, you can say that I'm an idiot for liking these things, but if nothing else, it proves that some people are happy with the direction Gnome's taking. We just don't tend to make a loud noise about it.

Re:Gnome vs. KDE

[bnitsua]

Obviously, he's a time traveller from the year 1996, so the whole concept of "Slashdot" confuses him. He seems to understand Karma Whoring pretty well, though.

Re:Gnome vs. KDE

[MichaelSmith]

Why are there two major windows manager projects?

The communists asked themselves the same question about any and all products.

The answer is that Humans are naturally competitive. To do their best they need to think the other mob are nipping at their heels. It's just human nature and it is the reason why products which achive a near monopoly find it so hard to keep quality high.

This post prepared on a GTK web browser under fvwm.

OS X-alike font rendering in Gnome

[ThreeDayMonk]

Personally, I despise Windows' hinted rendering. I like the heavily anti-aliased look on OS X, especially when dealing with Japanese text.

If you want to emulate OS X's font rendering, that's easy to achieve in Gnome. Just go to Font Preferences, Details..., and set Smoothing to Subpixel (or Greyscale for a TFT) and Hinting to None. Then walk away from the computer for a few minutes, because it looks weird in direct comparison. When you come back, enjoy the smooth text!

Re:KDE

[kelnos]

(Linus seems to be an exception to that, for the most part.)

Really? By all accounts, he seems just as egotistical and arrogant as those you've mentioned. He just tends to be quieter about it (doesn't speak outside of lkml quite so much), and (dare I say it) more sensible and practical in his ideas than true zealots like RMS, ESR, de Raadt, etc. I guess that's the difference: he may be an ass from time to time, but at least OSS isn't a religion to him.

What about MIME types/file associations?

[WWWWolf]

This isn't intended as a troll. I'm a GNOME fan. Just a frustrated one, regarding one particular feature of the system.

I don't know, GNOME's file associations seem really, really tricky to deal with. A few revisions back (2.6), I was marginally aware of how to manage file associations through control panel. I could not add my own icons to file types at all, but I at least managed to say which apps I wanted to be shown on Nautilus menu, which were available at all, and which was the default application.

I have no freaking idea where this thing is actually stored. In GNOME 1.x, they used some kind of really broken text file format. In early 2.x, they seemed to just keep using it. Nowadays, I have absolutely no idea how it stores the associations. Is it somewhere in gconf database, finally? I also have no idea how to really manage these file associations in 2.10: Nautilus isn't particularly helpful and I couldn't find the knob in the control center.

So is the file association stuff getting better at all? How do I manage the file associations in 2.12? And do they finally have some working way of adding icons to file types, or an actually understandable way of making icon themes?

Oh, and about article: Seems interesting. The HAL looks particularly droolworthy, based on a random and uninformed glance it seems to finally beat KDE =) And the Cairoification is always a good thing, it's definitely going to drag X11 kicking and screaming to the future, the direction where Microsoft and Apple are running blindfolded =)

Re:What about MIME types/file associations?

[smallfries]

The simplist way seems to be just use the tools. So use the custom icon button in a nautalis properties dialog, and choose an icon.

Obviously this isn't always desirable, I wanted custom jpg icons for each of my album folders and it would have been a bitch to do using a gui. So to get at the config files, do one by hand, and then look in ~/.nautilus/metafiles/

The format is quite straightforward xml and its easy to tweak by script.

Re:What about MIME types/file associations?

[Korgan]

I have no freaking idea where this thing is actually stored. In GNOME 1.x, they used some kind of really broken text file format. In early 2.x, they seemed to just keep using it. Nowadays, I have absolutely no idea how it stores the associations. Is it somewhere in gconf database, finally? I also have no idea how to really manage these file associations in 2.10: Nautilus isn't particularly helpful and I couldn't find the knob in the control center.

You can change the file type associations by going to the properties of a file in Nautilus, then go to the "Open With" tab and make your changes there. :-)

This has been the case since at least 2.8 (I'm typing this on an Ubuntu 4 machine) and is definitely the case in current 2.10 machines (Ubuntu 5, FC4 tried and tested.)

Re:What about MIME types/file associations?

[cortana]

You are looking for info about the freedesktop.org shared mime specification.

Re:What about MIME types/file associations?

[Dolda2000]

I have no freaking idea where this thing is actually stored.

I also had trouble finding that, until I tried Autopackage. As it turns out, both Gnome and KDE use Freedesktop's MIME specification and store the files in ~/.local.

The ~/.local directory is quite a wise addition, and is a user-specific directory equivalent to /usr (with bin, lib, share, etc.). Autopackage uses it to install per-user programs, and Freedesktop uses it to install per-user MIME-settings (in ~/.local/share/mime and ~/.local/share/applications).

Apple System 7 ??

[rduke15]

I admit I don'tknow anything about Gnome, but seeing a reference to Apple System 7 doesn't really sound very exciting, and doesn't quite make me want to give it a closer look.

Maybe some details could coinvince me?

Re:Apple System 7 ??

[peterprior]

It's referring to nautilus's spatial tree file view.

Re:Apple System 7 ??

[DeathAndTaxes]

I have fond memories of running MacOS 7, and I remember who slow 8 felt when I "upgraded". I seem to recall checking my "About this Mac" box and seeing my system memory running at about 6MB of RAM with Virtual Memory enabled. As for the actual folder tree view thingie, I do like it, and I miss it. It's nice to have that *option* in list view. One other thing I'd like to see implemented (as an option, off by default) are interactivity sounds. I remember 8.0 or maybe it was 8.1 that introduced sounds. You grab a window and a very light, unobtrusive sound would follow where you dragged that window around on the desktop...IN STEREO. It was annoying for about 3 seconds until you got used to it, and when you don't have it any more (*cough* MacOS X *cough* Windows *cough* linux-anything), it takes a long time to get used to it not being there any more...It's part of the feedback that you got used to. And of course, while a bit off-reply, there's my ever-present astonishment with the lack of individual background pics for each workspace. At this point I prefer to just refer to the gnome devs who refuse to accept it as RAVING LUNATICS. Ever seen a thread where someone proposes it? It's not pretty. Claiming to have a focus on "spatiality" while not allowing each workspace to appear "spatially different" is retarded, IMO. I'm sure I'll think of more, but not for now.

Re:Apple System 7 ??

[DeathAndTaxes]

Man, I guess I should have tried the "preview" button on that one...My default pref changed from "Plain Old Text" to "HTML Formatted"...WTF?

Re:KDE

[kelnos]

Ok, so we merge KDE and GNOME into KNOME. What next? Well, why should we choose between the great features of KNOME and Mac OS X? Well, ok, let's merge them too. So now we have MaKNOME X. Well, there's Windows out there too, and it does have a few nice features, so why not do another merge? And thus WinMaKNOME VistaX is born. And a thousand marketing gurus' heads explode.

Seriously, though, why should unification be the ultimate goal? Different people have different ideas about what makes a good, productive, usable desktop environment. Trying to make a one-size-fits-all monstrosity would be just that: a monstrosity.

And remember, we're talking about OSS here. Put up or shut up. If you're a GNOME user and like a certain feature KDE has, bust out some coding skills and write it yourself. If you can't do that for whatever reason, find someone with the ability and pay them to do it. In the end, no one is beholden to you, and there's no such thing as a free lunch.

s/TFT/CRT/

[ThreeDayMonk]

In case it wasn't obvious, that should read "for a CRT" above.

Re:Gnome vs. KDE

[Almond Tree]

Gnome is better ... and vi. There, it's settled

Because

[Mr. Underbridge]

speaking *very* generally, KDE apes windows and Gnome apes pre-OSX Mac. Diametrically opposed viewpoints. KDE goes for more options/control, Gnome, generally, less.

You can't combine those. And hell, what's wrong with choice? You've got 27 linux distros, you can't handle two major window managers? You're lucky it's just 2!

Nice work... shame about those icons

[delire]

... s/icons/wrinkles

Bring on the spat of posts telling me I can change the icon theme, as of course I do, but I'll say it again: Gnome needs a new default icon set.
The icons in most of those screens are sadly still as dull, muddied, venerable and depressing as they were 6 years ago, when I first tried Gnome.

The forward and back arrows in Nautilus seem to have absoutely no graphical correllation with the rest of Gnome's visual landscape (except the Refresh icon). The ~/ icon still looks like a little squashy mushroom house from a childrens novel and the icons in the menu editor (for menu groups) have no internal correlation other than they exhibit a tongue-and-cheek dig at futurism. Who actually thinks of a typewriter when looking for 'office', let alone a bricklayers tool when thinking about development?. Is this theme targeting a 50+ demographic? For icons so small, that aliasing really eats into their form and lastly the colour space of the icons seems all over the place, as though to solve the lack of a common palette they have simply mixed Khaki greeen into everything. This one thing KDE has really sorted out.

From what I have seen of Gnome desktops over the years, these default icons have a life expectancy of about 2 weeks (especially that home icon). Why not finally lay them to rest - or just move them down the theme list, far away from 'Default'?

Re:Nice work... shame about those icons

[fruey]

You seem to talk as if you're reasonably knowledgeable about UI elements and colours. You also realise that different icon sets exist and use them. But my question is how involved you have got with the development process in Gnome? Whining on Slashdot is easy; making educated argument to the Gnome dev team is harder, but you sound like you have the mettle, so is there money where your mouth is?

Re:Nice work... shame about those icons

[moranar]

The question is not about "who actually thinks of a typewriter when looking for 'office'?" but "who doesn't think of wordprocessing or text editing when he sees a typewriter icon?".

I don't think Gnome needs a new icon set, I think you do. I like the current one.

Re:Nice work... shame about those icons

[delire]

I would like to get involved in developing an icon set for Gnome, but sadly my commitments to improving and/or working on other projects take precedence for the time being. This crosses my mind often enough to justify making time for it at a later date, perhaps.

Re:Nice work... shame about those icons

[Mprx]

I'd always thought the bricklayer's tool icon was supposed to be a wire coathanger. The correct explanation makes about the same amount of sense.

Re:Nice work... shame about those icons

[ibentmywookie]

At the risk of sounding like an AOL using "Me too!" poster, I just want to say that I agree with you over the GNOME icons being dull, dull, dull.

I really wish they would fix them, I hate the default icon theme in GNOME. It just pulls the whole desktop down.

Re:Nice work... shame about those icons

[Nicholas Evans]

The question is not about "who actually thinks of a typewriter when looking for 'office'?" but "who doesn't think of wordprocessing or text editing when he sees a typewriter icon?".

Oh, THAT'S what that thing is. Now my menu has but one mystery left: why is the programming icon a clothes hanger?

And no, I'm not joking. I just lack artistic vision.

Re:Nice work... shame about those icons

[moranar]

I think that's a small (I don't really remember the English term for it) shovel used to mix cement. It's indeed a stretched metaphor. I might be wrong, since my menu has the mandriva icons, which do not feature that. I'm quoting from memory.

Re:Nice work... shame about those icons

[Nicholas Evans]

Ahh, I totally see it now. It's not a shovel for mixing cement, its what is used to flatten the cement once it is poured.

Many thanks.

Re:Nice work... shame about those icons

[bartok]

What anoys me about the Gnome icons is that they are 24x24 pixels which make the toolbars look bloated. I wish there was a set of 16x16 that we could switch to.

Re:Nice work... shame about those icons

[dodongo]

I thought it was the ugly duckling or something -- it always registered as a swan to my (admittedly just glancing) eye.

Chalk this one up in the "you learn something every day" column!

Re:Nice work... shame about those icons

[jvital]

Thank God you're not making Gnome icons.

I love Gnome icons, the way they are. They have a unique style, i just love it. And, i hate KDE childish/shiny icons. See, it's all a matter of point of view.

Re:Nice work... shame about those icons

[vga_init]

I agree with you about that home icon. I imagine that there are many distros that DO change the GNOME default icon theme. As a Fedora Core user of late, I'm quite pleased with their default GNOME theme in FC4. I believe the icons are the same from bluecurve, but the window decorations are now less blinding.

GNOME's default look and feel have been graciously taken care of for me by Red Hat, of which I am very thankful, but as a BSD user I'm still left to deal with GNOME as it is.

Re:Nice work... shame about those icons

[delire]

I love Gnome icons, the way they are. They have a unique style, i just love it.

I agree.. around 17 completely unrelated, unique styles rolled into one. Like a bonus pack but lacking that elusive added value factor.

Re:Nice work... shame about those icons

[WWWWolf]

I agree a new icon theme is needed. The icons don't need to be made more interesting though. They need to be duller.

Originally, Nautilus had actual, genuine Susan Kare stuff. Not just style imitation - real stuff. I thought it was spartan, but it worked, unlike most of the modern icon stuff. I liked Mac and Win3.0 UI designs a lot. I liked Nautilus a lot.

Nowadays, it seems icons have to be fluffy, ephemeral, and cute. I can live with cute, but the fluffiness and complicated looks of the icons have to go. We need simpler icons. We need icons that indicate what they represent and not a damn thing more.

Plus, simpler icons are faster to make so the icon sets can be made really comprehensive, too. It's frustrating to have an icon theme you like and it lacks icons for majority of cool filetypes...

Re:Nice work... shame about those icons

[rcbarnes]

*Shudders at the image of a catatonic AOL user who somehow wandered onto slashdot from the fluffy, nonthreatening kittycat pages...*

Not that I don't appreciate all those AOL users, I work in a tech shop and they are my bread and butter (virus cleanups for $70 a pop 'cause AOL antivirus sucks, and the NAV "Scan for viruses" button is too scary...) Seriously, I had a woman panicking on the phone about pushing a button (perhaps shut down?) for fear of losing all her data, despite all the reassurance I could muster).

Re:Hidden sinker

[Haeleth]

No, you see, this is why distros have defaults that newbies can't change without knowing what they've done.

You don't say "which desktop are you using", you say "what kind of Linux are you using". If they say "Red Hat" or "Ubuntu", you know they've got Gnome. And if they say "SuSE" or "Linspire", you know they've got KDE. Simple, see?

Re:Gnome vs. KDE

[dbIII]

Why are there two major windows manager projects?

As I see it - politics, misunderstanding a software licence and unwillingness to negotiate over a licence. In the end zealots got bored when they realised the first group really wanted the same thing when the first group gave in to the demands, and in gnome developers took over from politics, and we ended up with two decent collections of software, of which the actual window managers are only a minor part.

KDE was influenced by CDE, a desktop environment on Solaris which showed that not everyone wants to have the same desktop environment, but has some nice features. Gnome was originally a backlash against a software licence used by KDE, and originally was some sort of odd mixed KDE (ie. CDE once removed) and MS Windows based on some code taken from the drawing program "the gimp". The project became more popular and less politically driven, breifly included Enlightenment as it's window manager (until the Enlightenment people ran screaming for the hills a few weeks later because gnome broke all of their cross-platform code and didn't care) and eventually became cross-platform and the useful thing you would have seen over the past few years. Now about the only vestige of it's beginnings is stuff like the windows registry style gconf which really is aimed for single user stand-alone systems and not for anything with aspirations beyond being a personal computer (ie. like something on a network!). There is a tool developed this year that allows gconf settings to be exported to other users on the same machine, so it's getting somewhere.

As for the actual window manager, you can use plenty of different ones and still use KDE apps or gnome apps - including the taskbar and menu style things.

Still slow and buggy on my nvidia

[ThreeDayMonk]

Even with nvidia and the binary drivers, X Composite is flaky and slow on my computer (2.6 GHz P4, not bleeding edge, but not slow either).

With Composite enabled, I frequently suffer from complete lockups of Xorg (the mouse moves, but nothing else works).

So yeah, I'm with you on the "hope it doesn't".

Re:Still slow and buggy on my nvidia

[strider44]

Have you included

Option "RenderAccel" "true"

in your xorg.conf?

Lies! The true unified desktop is GDE!

[Glytch]

Good sir, I disagree with your design philosophy and will be forking the KNOME project.

Re:Lies! The true unified desktop is GDE!

[kelnos]

Hmm, but then later on, everyone would complain about GDE and KNOME having two different feature sets, and we'd have to unify again. GKNOME? KGDE? GDEK?

No, I've got it: GDK.

Wait... Crap.

Re:Hidden sinker

[MrHanky]

I just think it's useful to look at it from the perspective of general users and worth considering.

OK, tell me: How is this useful? Will KDE developers one day lay down their C++ compilers and start programming with GTK? No. Not going to happen. And users obviously disagree on which desktop is better, so none of them will go extinct by natural selection. So your comment is just wishful thinking, and not useful at all.

Here's a different wish: I wish people would stop posting comments like yours to every fucking story about Linux desktops. I wish you would consider that perspective.

God. Now THAT's a mean trick.

[Stauf]

It's like all my fonts snapped out of focus. Maybe we should tack on an appropriate disclaimer: not for LCDs, your eyes will bleed .

Re:Hidden sinker

[dbIII]

Tech: Well, which desktop are you using?

So it's MS Windows - is that 95, 98, ME, NT3.51, NT4, win2k, XP, 2003? All have their quirks, advantages, disadvantes, and users always hide their start menu entries in different spots if they've had the thing for a while.

Live with it, it isn't a monoculture anywhere, not even with Macs.

On linux at least you have the advantage that the tech support person could ask you to run switchdesk - or more likely they can get you to put something in a shell window (bash, csh, ksh whatever they are used to) or get you to let them ssh in to solve the problem.

It's not for everyone

[ThreeDayMonk]

Aye, it's not for everyone. Some people seem to really hate "blurry fonts". I, on the other hand, like it - and it is the way that OS X does it.

It might help that I sit a good metre away from my TFT (yes! LCD!) monitors, so I can't discern the individual pixels anyway.

Software patents and innovation

[Richard W.M. Jones]

But on the other hand, not having access to those techniques forces the X.org people to come up with innovative solutions to the same problems.

That's what patents are supposed to do. In practice there are two problems with this. Firstly what happens if the provably best algorithm is patented? (think: compression - eventually someone will come up with an algorithm which is provably optimal, and patent it). Secondly what happens if you need to implement the algorithm to interoperate? That's the case with these fonts: the fonts include hinting programs, so in order to display the fonts as intended you simply have to be able to run those programs. Unfortunately there is a patent on running those programs. No amount of "innovation" is going to help you here.

Rich.

Re:Software patents and innovation

[HuguesT]

Hello,

Provably optimal is not as impressive as it sounds. For example Huffman coding is provably optimal for some definition of optimal. Yet arithmetic coding is often in practice more efficient, and Huffman coding is seldom used in practice (except with JPEG) because of the dictionnary problem.

To prove something optimal you have to spell out the conditions. In Huffman's case on of the assumptions is you can't compress any datum with anything less than a single bit. In fact you can ; and arithmetic coding can achieve that (arithmetic coding can compress whole groups of data with very few bytes, hence each datum can is on average compressed to less than a it).

Moreover, many problem don't have an optimal solution that is computable in reasonable time. Think NP-complete problems. Unless P=NP this is not going to change soon.

In the font hinting case I thought I read somewhere that there was no point in enabling the patented technique in Xft because the default technique works as well or better anyway.

The problem might be somewhere else altogether, like in the font design themselves.

Re:Software patents and innovation

[Tim C]

Firstly what happens if the provably best algorithm is patented?

Well, in my opinion algorithms shouldn't be patentable, as the implementation is covered by copyright. In this situation, however, I suggest that you find another algorithm that is good enough and use that instead. It's not like there's any shortage of them...

I do take your point ("if everything's patented, you're shafted"), but I simply don't think that's a likely problem.

Re:Software patents and innovation

[DeadMeat (TM)]

In this situation, however, I suggest that you find another algorithm that is good enough and use that instead. It's not like there's any shortage of them...

Which is exactly what the FreeType people did. Normally TrueType typefaces contain a miniature program to help with hinting on a per-typeface basis. Apple's patents prevent FreeType from using these programs by default. So, FreeType 2 contains generic auto-hinting code.

The result is fonts are hinted differently than Windows and OS X. Whether the hinting is actually worse is mostly a matter of opinion -- I prefer FreeType's hinting over Windows's for the most part, although there are specific typefaces (like Verdana) that are hinted pretty badly with FreeType if you don't use anti-aliasing.

At any rate, if you live in a country that isn't covered by Apple patents (or if you just don't care about violating them), you can unofficially enable the patent-protected hinting in FreeType by getting a copy with said code enabled. Instructions are available here.

Re:Software patents and innovation

[cecom]

Unfortunately small fonts with antialiasing enabled actually render _a lot worse_ with the byte code interpreter than with auto hinting. Go figure.

Re:Gnome vs. KDE

[johnnyb]

"Why are there two major windows manager projects?"

The idea that more people working on it brings better results is simply untrue. In fact, they are both probably a lot further along because they are separated. This allows them each to look at the problem from a different perspective, and build out those ideas, and thus provide new ideas to the other group. With a single project, everyone would have to agree about the project direction before moving forward. With multiple projects, multiple directions can be explored fully, and then each project can see which parts worked well and which parts did not.

HAL is desktop-agnostic

[21mhz]

HAL is not bound to GNOME in any way. KDE can use it for its benefit, too.

Re:HAL is desktop-agnostic

[twener]

> I reckon KDE have their thumbs up their asses if they want to stick with DCOP for their desktop IPC interoperability/integration layer.

Sorry, but ever heard of "backward compatibility"? KDE has a working IPC mechanism for ages which it isn't going to dump within the KDE 3.x series. Which doesn't shall say that KDE isn't supporting D-BUS for some purposes already (HAL, media:/ KIO slave).

> But, I also think that they're waiting for D-BUS to hit stable

A stable/compatible staying interface of D-BUS would be nice, yes.

> Enough politics.

You mean trolling?

Re:Gnome vs. KDE

[twener]

> You only need to look where they are putting the real money and engineers.

Several Novell engineers are working on KDE itself or integrating applications with it. Not even talking about other stuff like conferences, sponsored server hosting etc.

> It makes me laugh when KDE devotees

It makes me sad seeing GNOME trolls like you...

> The fact is, the Novell money is going into GNOME and NLD

And NLD, for the case you didn't notice, is not a "GNOME only" show.

> the SUSE name and KDE are legacy

Both wrong.

Re:Gnome vs. KDE

[Internet_Communist]

well thats good to know, I somehow got the idea it was in C++ reading something on the xfce site about C++, maybe I just was reading about the bindings...

Re:Gnome vs. KDE

[bhalo05]

if linux wants to get anywhere with mainstream users, there's going to have to be one superior desktop environment, rather than two slightly inferior ones.

I think there are far, far more important issues preventing Linux sucess on the desktop.

if developers are split between two desktop environments that essentially do the same thing, it gets linux nowhere.

As a I said, even if developers really wanted to work on a single desktop and agreed on using the same technology, they're probably not prepared to do that. Do you think a Gnome developer has the skills required to start hacking in C++ and QT technologies right now? I'm almost sure the answer is no. The same is true for KDE developers and C/GTK. If Gnome had never been started, I'm pretty sure Gnome devs would be hacking on any other thing, but not in KDE anyway.

The fact that matters is that this discussion is always pointless because as long as both have active developers working on them and users willing to use them, none of them is going to go away, it's simply not going to happen. It's time to finally accept this fact and live with it.

Re:Gnome vs. KDE

[Internet_Communist]

sure, bluecurve does that, if you like the bluecurve theme.

the GTK-QT engine makes QT themes work on GTK. Something that did the opposite (qt-gtk engine anyone?) would fix the other end of this situation. Well, for GTK and QT anyway.

Of course the other option which you mention, making a version of the theme for all the different toolkits, works too...but that seems like a lot of work, and so only is done in the rare instances such as the one you mention. I know there's also a "keramik" and "geramik" theme, similarly that match on both toolkits...but again, I don't use/like those themes and thus we're back to step one...

Re:Gnome offers nothing for efficiency-freaks

[moranar]

Tried setting "Default zoom level" to 25 for list view? It makes for tiny icons, you see, hardly higher than a 12pt font. Font I need to actually see and read what's on my screen.

If you want efficiency drilling for files, use find or the console. If you want something in the middle, then understand that your particular favorite shade of gray is not the same for everyone. I find my gnome desktop incredibly useful as is.

Re:Still ugly fonts - this works too!

[bogaboga]

1: Install Microsoft true-type fonts.

2: You could install them via this script: http://vigna.dsi.unimi.it/webFonts4Linux/webFonts. sh

Then do the following:

Configure X and Gnome to 96 dpi sudo cp /etc/X11/xorg.conf /etc/X11/xorg.conf.bak sudo gedit /etc/X11/xorg.conf Locate Section "Monitor" and add the following lines before EndSection: # DisplaySize 270 203 # 1024x768 96dpi

# DisplaySize 338 254 # 1280x960 96dpi

# DisplaySize 338 270 # 1280x1024 96dpi

# DisplaySize 370 277 # 1400x1050 96dpi

# DisplaySize 423 370 # 1600x1400 96dpi

Uncomment the line corresponding to your current resolution.

To get other values, use the following formula:

displaysize = {pixelsize}/96*25.4

Remember:

The display size must be "right" so adjust those values till you get your size right.

Already on Ubuntu Breezy Badger...

[vhogemann]

I'm typing this from a Gnome 2.11.90 desktop, and I must say that it works perfectly.

Gnome is almost complete now, most of the interface work is done, and now all that's left are those little details that make all the difference.

Right now it dellivers a very nice experience, and my wife, a History teacher with serious problems with modern technology, really prefers it over Windows or KDE. I really like the Gnome approach of make things simple, and then add the needed funcionalities one step at the time, instead of the KDE "everything-plus-the-kitchen-sync" aproach.

The rougth edges lie on things like better hardware discovery, and better integration with the underlying OS. But these will go away as the HAL matures, and more and more scripts are added to it's library.

Re:Hidden sinker

[e133tc1pher]

Mod parent ^^^^^^^^ : ) Its like /. just talks in circles. Hey btw, gentoo owns your gay ass fedora and debian blow!!!

Re:Gnome vs. KDE

[FidelCatsro]

Obviously, he's a time traveller from the year 1996, so the whole concept of "Slashdot" confuses him. He seems to understand Karma Whoring pretty well, though.

I think he may need a few more lessons

nome vs. KDE (Score:-1, Troll)
by John Seminal (698722) on 31-07-05 12:42 (#13206787)
(Last Journal: 22-02-04 3:07)

Re:Hidden sinker

[caino59]

You got it.

I provide support for Windows 98SE - XP and Mac OS 9 and X.

Each have their different quirks, menus and commands.

From a support standpoint - if I need someone's IP address - Windows 98-ME needs to run winipcfg. In Win2k-XP, ipconfig needs to be run the command line.

At least if someone calls with Linux I know all I have to do is have them open a terminal and run ifconfig. No matter the distro.

Block middle click too, please

[Kristoffer Lunden]

If they could only add an option to block "middle click pastes" too, it would all be perfect in clipboard land. But browsing the web, evertime someone even thinks this thought it is immideately flamed through the ground by all the people who knows how superior this way of doing things is, and that also knows that there is no chance in hell that anyone could do this by mistake.

Heads up: I'm not proposing to remove it, or even turn it off by default. I just need a way to turn it off manually. It is extremely annoying, and I (and other with me) *do* click middle by mistake - often - and that is a hell when scrolling around code in text editors... Yep, a lot of it probably owes to the mouse I have, it has a tendency to get stuck slightly on scrolling, which results in a click. But really, do I need to buy a new mouse for something as simple?

I don't use, want or need it, and it hinders me in my work. I would really like to see it go. (Maybe it really is a X.org issue in the end, though. Not sure where it would be best to implement it).

Re:Block middle click too, please

[Anonymous Coward]

Implement it? You'd have to un-implement it instead. You're talking about the X11 primary selection. Which is a different thing to the highlight - explicit copy - explicit paste of the clipboard. Its pretty orthogonal really, if you do the "usual" copy/paste you won't be using it, you won't even see it. In your case, its biting you because of your mouse's errant clicks after having something highlighted in another application.

Remember that you're still using X and its a very standard way of interoperability between X apps.

So in summary, I see where you're coming from, but the problem really sounds like your middle mouse button going off when scrolling. That would also be a problem in other applications that use that mouse button for other things. Or in another OS if you dual boot for example.

Now, assuming you don't mind that naughty mouse - theres a trick you could do. X, that thing whose primary selection you dislike and a lot of us love, could ironically save your butt. Its an odd fact that mouse buttons 4 and 5 are used for up/down scrolling. Odd because when those side buttons came along they ended up generating 4 and 5 and the wheel gets 6 and 7. Have a look at /etc/X11/xorg.conf for ZAxisMapping.
Now to make this sane, theres a configuration tweak using xmodmap swapping 4,5 and 6,7 so the wheel movements are back to what apps expect and any side buttons work again. Find that, change it so your middle button (3) gets moved to some other button - 1, maybe, or an unused one, and you should be saved. You could even emulate middle button using left and right together to still have it when you need it.

Where the mapping is done is distro specific, and you might need to add a section to xorg.conf to say you have more buttons than you do to get a spare one to map it to. Google is your friend.

Re:Block middle click too, please

[mrogers]

As far as I can tell from the wiki the new clipboard manager won't affect middle button pasting (which uses the "primary" selection rather than the "clipboard" selection). Perhaps you could change the mapping of your middle mouse button from 2 to 6?

Re:Block middle click too, please

[cortana]

If you use the actual GNOME browser, then you won't get this braindead behaviour. If you insist on using Firefox, etc, then you can disable it in about:config, search for middle and play with the options there.

Re:Block middle click too, please

[DeathAndTaxes]

You could try running imwheel and gaining even more functionality out of your mouse. I mention it in my blog, and I'm not afraid to whore it out! ;-)
http://www.delayofgame.com/?q=node/45

Re:Block middle click too, please

[arose]

Get a better mouse (or use the keyboard when editing text). Would you be complaining about context menus if you would right click a lot by mistake? If you would I suggest an Apple branded mouse.

Re:Block middle click too, please

[mixmasta]

try ^Z in the meantime.

Re:Block middle click too, please

[Kristoffer Lunden]

Nah, actually I am talking about implementing - a conditional around it. I'm not talking about removing it, just the option to disable.

And I do know about several different tricks involving xmodmap and imwheel and whatnot, but sometimes it just doesn't work. I've been doing this dance before, but right now I can't get it to work, for some reason.

Ok so it was a few months since I tried last, I think the main point I'm trying to make is this: I've been running Linux for years now and I'm getting very tired of editing arcane config files, all of which have their own format, at the same time I've seen this particular request raised time and again on mailing lists, in forums and other discussions, by lot's of people who actually have a problem with this (see Google). And it's always been dismissed with the usual "we know what's best" attitude.

My humble, not-knowing-enough-to-do-it-myself-attitude is that it really can't be that hard to have a simple boolean (probably in X) that says "don't use primary selection", or even just blocking middle mouse via simple Gnome checkbox. I can't imagine this is hard or that it should violate any "best practices". To me, this is both better usability and more choice. And yes, I'm fine with leaving the default as-is, and have a hidden little checkbox somewhere in the system settings or whatever.

As can be seen in this thread, the general attitude is the general "just edit config XYZ", which I think it's time Linux got away from (mainly because I've done that for years and it gets tedious hehe ;-) or "you don't understand". As I expected. Sad but true. Not that I don't appreciate tips, and those tips *are* useful. Just that they don't solve the problem, just the symptom, and pretty awkvardly at that. Never been fond of sweeping under the rug. :)

I will willingly admit that it is not a *huge* issue, but just because it isn't, why couldn't it just be fixed? =)

Anyhow, while we are at it, xmodmap/imwheel-like functionality would not be boring to have in a nice interface... if someone is looking for something to do, I'm sure that would be pretty popular, because that is one of the common issues still being dealt with often in various forums and the like.

Re:Block middle click too, please

[Kristoffer Lunden]

Funny guy =)

Well I have to, don't I? That is tedious as it is, but the worst thing is actually when things has gotten pasted with me noticing it. That can be a bit painful to clean afterwards, if I've been writing good stuff after it and can't ^Z.

Re:Block middle click too, please

[Slack3r78]

Well, think of it like this. If your mouse wasn't wacky you could totally ignore primary selection.

Wrong. I have a perfectly working mouse and find primary selection constantly getting in my way because many apps misbehave to send everything to the same clipboard buffer. The most common place I have this problem is web browsing - highlight the address bar to clear it, only to find you've lost the URL you thought you had copied which is now replaced by the one you were trying to clear. Disabling the middle mouse button entirely doesn't work as some apps (again, namely web browsers) actually make use of it these days.

It's just so big

[threaded]

Why do things just keep getting bigger? I liked it much better when it was nice and small and cute.

Re:Gnome vs. KDE

[Andre]

Qt still is not as free as GTK+, because it is a library and licensed under GPL. GTK+ is licensed under LGPL for a reason. If you use Qt, the license of your application cannot be chosen freely. Either you use the GPL, or you pay for the freedom to choose another license.

A GUI toolkit is part of the critical infrastructure of a software component on the desktop. Every application needs such a component (apart from fullscreen applications like games).
And a GUI toolkit is commodity, nothing special anymore.

Many developments begin at home, and these developments are the programmers' own crown jewels. I want to secure my investment in time and energy, and want to be able to deploy my ideas anywhere I see fit. Of course, I want to take my developments to the workplace and go on without interruption. This is freedom, and highly productive.

Re:Gnome vs. KDE

[Gnulix]

KDE has more developers and money. Gnome has fewer people, but more creative people.
I think you switched KDE and Gnome in your sentence. Gnome is supported by companies such as Sun and Novell. Gnome has a huge PR/propaganda machinery. And so on...
KDE is more of an community effort and seem to try to compete much more on merits and features, rather than PR/propaganda.

Is it actually "spatial"?

[argent]

Is it actually "spatial". It looks like the Apple tree view, but the appearance isn't why the Classic Finder is described as "spatial". The Finder has the annoying property (annoying to me anyway, the "spatial Finder" fans love it) that the location on screen of any Finder window is fixed, and there is a single view of any object in the file system. In the classic Finder, if you open an icon-view window of an expanded directory in the tree-view, it actually collapses, and if you expand it again, the icon-view closes. Every time you open a folder, it shows up in the same place, and even if you use cmd or opt (I don't recall which) to force it to keep the same window when you open a subdirectory it closes and re-opens in the "right place".

Is all that actually happening here? If not, they shouldn't call it "spatial".

Re:Is it actually "spatial"?

[Knuckles]

Where have you been these last 2 years? Yes, it's spatial, and it's been flamed to hell and back for it.

Re:Is it actually "spatial"?

[argent]

Where have you been these last 2 years?

Using a real Mac instead of a fake one?

Yes, it's spatial, and it's been flamed to hell and back for it.

That's just wrong.

Re:Is it actually "spatial"?

[Knuckles]

Huh? What's wrong? That it's spatial or that it's been flamed?

Mind you, I'm not saying that the flames were
deserved :)

Re:Is it actually "spatial"?

[argent]

Huh? What's wrong? That it's spatial

Yes, particularly if there isn't a non-spatial version.

BTW, that link just says "Too many connections".

Re:Is it actually "spatial"?

[Knuckles]

there isn't a non-spatial version

Excuse the question, but do you have any idea at all what you are talking about?

First, there is a non-spatial version, just go to Edit -> Preferences and check "Always open in Browser Mode". This will turn off spatial nautilus and give you browser (i.e., Win File Explorer) mode.
Or you can rightclick a folder and select "Open in Browser Mode", which will give you a non-spatial browser mode window.

Second, since they switched to spatial mode in Gnome 2.6 (see link, it works now), there obviously was a non-spatial mode before the switch

Re:Is it actually "spatial"?

[Knuckles]

I forgot: in addition to (1) and (2) in my previous reply, the main error in your post is a logical one:

If there isn't a non-spatial version, my statement that it is spatial is not particularly wrong. Quite the opposite, if there isn't a non-spatial version, it follows that it is spatial, thus I am right :)

Re:Is it actually "spatial"?

[argent]

Excuse the question, but do you have any idea at all what you are talking about?

Of course I don't: I don't use Gnome... if I did I wouldn't have been surprised at the idea that anyone would have actually thought resurrecting the spatial Finder was a good idea. That's why I wrote "if there isn't a non-spatial version".

Excuse the question, but do you understand what the word if means?

Re:Is it actually "spatial"?

[Knuckles]

We both managed to misunderstand each other completely. Sorry for my part :)

Re:nautilus default behavior

[ebuck]

Spatial navigation can be turned on or off via a configuration setting. AFAIK Gnome doesn't have anyone providing a "just this distro" feature set.

That said, after all the hype about how annoying / frustrating spatial navigation could be. I was very disappointed to find out that many were making a huge fuss about almost nothing.

I suggest you try it for a couple of weeks, odds are you will not be infuriated, but then again, odds are you will not be awestruck. If you don't like it, get familiar with the config setting to toggle it and off.

Yes

[ThreeDayMonk]

Yes, I have it. Does it conflict with Composite?

Re:Yes

[strider44]

no but the lack of it is one of the reasons why a lot of newbies looking to try Composite get bad performance.

Re:Yes

[shywolf9982]

Ok, that's gonna be terribly OT. But it seems like any SDL application HIGHLY dislikes the Composite extensions, killing the window manager and other nasty things when i try to fire them up with the composite extensions enabled. Other than that, I noticed it runs better on my old athlon 900 with a geforce2 than on my sempron 2600+ with an ati.

Re:Yes

[DashEvil]

Do you want me to post a screenshot of me using XComposite to play back a PrBoom demo (partially transparent, with drop shadows enabled)?

PrBoom is an SDL application.

Re:Yes

[ookaze]

That's because the GP is clueless.
If you activate the Composite extension, with NVidia binary drivers, you lose the ability to use GL "areas" (don't remember the correct name).
So I guess the SDL apps he is trying to use are OpenGL apps (like Privateer Remake for example).
What happens then, is that the game crashes immediately, but that's not SDL fault actually.
Now, there is an option, "AllowGLXWithComposite", which allows you to use the two together, but then you have a highly unstable desktop. That means each time you use a OpenGL app (like many screensavers), you take the risk of locking the XServer, and ultimately, lockup the machine entirely.
And some Composite apps are then sure to lockup your desktop. For me, it happens anytime I use the KDE shadow or translucency options, for example.
I never had a problem with PRBoom though.

Some old NVidia drivers (in the 6XXX versions) actually work well with both, but they are not compatible with latest kernels, so it's out for me.

Re:Yes

[shywolf9982]

LGeneral uses OpenGL? And btw, I tried the "AllowGLXWithComposite" and it does not work. By the way, enabling "AllowGLXWithComposite" lets the GL screensavers run just fine, and any app containing a gtkglarea control works fine as well. Also, cube works (even if not in fullscreen mode). Or the problem might reside in LGeneral (even if I doubt this).

Re:Hidden sinker

[ebuck]

Ahhh... The joys of passing the "Mom" test.

Look, you Mom isn't going to know that there's another desktop unless you tell her. She will just know her desktop.

My mother-in-law wanted a computer to surf the internet. I built a cheap linux box, indicating that she was getting "cutting edge technology" far superior to Windows. She asked me (proof of how far Microsoft's advertising budget has reached) if it was harder to use than Windows. My response was, "No, but if you learn something and then change, a lot of people focus on the steps being different instead of focusing on the underlying thing they want to do."

Within a few weeks she knew more about KDE than I did. I guess I skimped on actually reading most of the end user documentation.

Re:Hidden sinker

[mattyrobinson69]

no, in linux you have to say "open a terminal and type 'su -c ifconfig' - then theyll say "its asking for a password, what is it?"

people are retarded, there's no two ways about it.

What about column view

[SassyDave]

I like the new look of Nautilus in browser mode:

        http://www.gnome.org/~davyd/gnome-2-12/images/naut ilus-browse.png

But what about a column view like the OS X Finder has?

        http://arstechnica.com/reviews/4q00/macosx-pb1/ima ges/column-view-big.gif

Re:What about column view

[Lispy]

Can I have a pony? :)

Re:What about column view

[Sark666]

How about can I have a god damn invert selection.

I can't believe nautilus doesn't have this. Or doesn't as of 2.10.

I really liked gnome for a long time but more and more I find it too simplified. Although, I also agree that kde can throw everything but the sink at you.

Gnome needs to put the options in tiers, the simplified ones are tier 1, and more advanced options are tier 2. Not disregard tier 2 entirely. Maybe not the idea solution, but I really like the look n feel of gnome in general, but sometimes I do need those 'advanced' features like invert selection, or how about in the gnome find, there is no way to stop it from searching sub directories! That's just ridiculous.

And for the limited options nautilus really does feel like a pig. And some things could be obvious speed ups. Like in details view don't bother showing icon previews (which konq does btw).

Re:What about column view

[arose]

I'm not at a computer with Gnome right now, but I believe rubberbanding all icons should inverse the selection, this is of course no replacement for a real 'invert selection', why don't you file a feature request?

System Seven Savvy

[Red_Winestain]

GNOME 2.12 has picked up ... some features popularised by Apple's System 7

It's good to see GNOME catching up with a state of the art (circa 1991) operating systems. However, I've heard there were some nifty new developments within the last dozen years or so that might be worth looking at...

Re:System Seven Savvy

[fontkick]

I wouldn't ignore older ways of file management just because they are "old". There are a lot of people out there would like Windows Explorer to have dual pane file nav, like several popular DOS treeview apps.

The feature in System7 (refined in OS8 and OS9) is the triangle icon next to the folder that lets you expand the contents of the folder and collapse it. I'm not sure how the gnome implementation will work, but it works well on OS9. The usefulness of this comes when you need to create new folders, rename folders, rename files and do a lot of manual work to organize a file/folder structure exactly the way you want it. I think Mac OS9 is the best OS for this. Window scroll speed, rename delays, and ease of cursor insertion into the filename are all critical elements for a pleasant experience with this type of organizing, and so far the only OS that does it perfectly is OS9.

In addition, in OS9 you have access to a folder tree view in the Open/Save dialogs, which lets you easily navigate through a folder structure to put a file where you want it. OSX uses this exact way of doing it, so to say that the older OS is somehow irrelevant, well, OSX's GUI builds on everything that came before. It's really just a prettier version of OS9, OS8 and System7, as much as OSX fanatics don't want to admit it.

They could have easily compared the new Gnome feature to the same exact thing in OSX, which is still implemented as a carry over from OS9.

Re:System Seven Savvy

[be-fan]

Not really. MacOS Classic was pretty much the height of UI design. OS X is catching back up to where they used to be, but they aren't there yet.

Evolution

[jvagner]

The only thing I'm really dying for in GNOME is that Evolution be compiled with the plug-in that allows for the creation of tasks from email messsages dragged to the icon. That would improve my life tremendously. As it is, I'll stick with Thunderbird (I don't want to re-compile Evolution every-time I upgrade my GNOME).

Re:Evolution

[arose]

Shouldn't it be possible to compile plug-ins separately?

A coveat

[coolGuyZak]

Unless, of course, you are using it to change (rather than view) your network connection :)

Re:Gnome 2.2 ey?

[Poltras]

http://www.google.com/search?q=gnome+%22three+poin t+zero%22

Few things:

[coolGuyZak]

1) System memory is a priviledge, not a right.

For the most part, I agree. This is one of the reasons why I advocate making pluggable components, and why the GP was proposing compile-time removal of code :)

3) Code bloat tends to make maintainance hell.
4) Code bloat contributes to cycle bloat (something you said yourself that you cared about, almost contradictory to your previous point).

Yes and no. It depends on how well the underlying system is written. For instance, the KDE codebase is considered "bloated", and yet it's incredably easy to follow. (Once you know the base functionality in kdelibs, of course).

Of course, you will always have some small errors popping up in the code, but the devs are only human. One of the funniest ones I can think of from KDE was the flickering of new tooltips on rendering... It turned out that someone had added an "update" call in a function where it shouldn't have been.

5) Most people do not have emmense hard drives (you may have a 250GB hd, but my mom's computer has a 40GB, and my sister's laptop, only a 20GB. Hell, my work desktop machine's hard drive is only 6GB, though I often will bring in an external) [but I guess this point doesn't really matter, because Linux isn't your mom, your work, or your sister's Operating System now is it ;)].

Mostly true. What should be emphasized is both Gnome & KDE are trying to compete with "those other OS's" these days. So, anything which is not "mom and pop" friendly could be concieved as a bad thing.

Concerning HDD space, I have a laptop with a 20 GB drive, and the entire install takes up only about 6 GB total. That includes a full KDE install (along with several versions of the multimedia that comes with it), all of linux, and whatever else gentoo saw fit to put on it... And this is compared to my base Winows install, which is about 2Gigs in the "C:\Windows" directory alone. Installing the plethora of applications that you get within a full Gnome or KDE install would push that number much higher. than the paltry 6 GB...

In other words, while it is an issue, I think it's a bit overrated. ;)

When will they modernize Epiphany?

[mnemonic_]

http://www.gnome.org/~davyd/gnome-2-12/images/epip hany-search.png

So Epiphany uses Gecko. That's great. But when will they get rid of the silly separate button bar? They should trim the buttons, make them compact move them to the same bar as the address box. Why have they not kept pace with modern browsers like Firefox and Safari GUI-wise, and instead have taken to the original layout of Internet Explorer 3.0? It doesn't make any sense. It's not any easier to use when the button area is cluttered with functions better relegated to the menus. And its less usable for everyone when there's no google search box.

I don't get it; why did the GNOME team decide to use a space-wasting, cluttered UI in Epiphany? Why not a simpler layout in the vein of Safari?

Re:When will they modernize Epiphany?

[arose]

You may not like it, but that does not make it cluttred. Either way you can edit the toolbars in Epiphany and Icon/Text preferences for all of Gnome.

It's not about the features

[Atomic Frog]

The people making the core components of Linux and associated machinery still don't "get it".

It's not necessarily about the features. There's enough already. It's about presenting and organizing them in some consistent manner that is understandable to the average user.

Pick any standard distro that uses GNOME. Now take a look at Sun's Java Desktop (which is essentially the same thing). Don't look at screenshots, donwload it, install it and use it.

The guys have Sun have clearly done a far better job organizing things and cleaning up items (i.e. duplicate items in various menu places, configs that exist in places that didn't make sense, etc) that the average user would think it was something totally new.

Even farther up the road, look at what Apple's tiny band of engineers have done vs. a worldwide pool of GNOME or KDE dudes. Come on, it is basically *NIX underneath, as you all know, but the difference in user interface is night and day.

Re:It's not about the features

[Jussi K. Kojootti]

If you have evidence backing your statement that people working on OS X user interface are a "a tiny band of engineers" compared to Gnome, please lay it out. Otherwise I have a hard time believing it.

Robert Love's talk at guadec

[Zeroth_darkos]

Here's a link to the mentioned talk: Optimizing GNOME

Re:Still ugly fonts - this works too!

[DeadMeat (TM)]

If all you want to do is adjust the font DPI, you can do that from the GNOME Font Preferences. KDE's Control Center has a similar option. No manual configuration-file editing is necessary.

The configuration-file editing is only necessary if fonts are the wrong size because X guessed your monitor size incorrectly (which is very rare in my experience, since it just fetches that information straight from the monitor, but it does happen). At any rate, Windows doesn't have the ability -- GUIfied or otherwise -- to override monitor geometry, at least as far as I can tell. I'd be a little surprised if OS X does, although it might since Macs are used in graphics work so often.

Re:Gnome vs. KDE

[PhucYuew]

"Competition is a by-product of productive work, not its goal. A creative man is motivated by the desire to achieve, not by the desire to beat others." --Ayn Rand

Color Me Amazed

[Doc Ruby]

Antialiasing in the color picker? The edges are smoother, but the colors are wrong. It's bad enough they used that feature in that app, but to use that terrible example to show off the feature?

Re:Color Me Amazed

[Breakfast Pants]

I thought the same exact thing. Their circular selection mechanism is so large that it is probably extremely imprecise anyway.

Re:Color Me Amazed

[baquiano]

The color wheel is a inherently imprecise selection mechanism, meant to pick a color in a fast and simple way, e.g, "I want a light-greenish tone". If you want a precise color, you should use the RGB or HSV spinners.

Complaining about lack of pixel-perfection in the edges of the color wheel is like complaining about lack of pixel-precise proportion of scrollbar thumbs.

Re:Color Me Amazed

[Doc Ruby]

Yeah, it's imprecise. But antialiasing makes it inaccurate. I can pick a color that's been antialiased with the neighboring white background. Which therefore looks like the color I want, but includes color data from the neighbor which isn't related to the actual color data at all.

So antialiasing makes it prettier, if that's your style, but it makes an already inexact tool worse. There are other GUI widgets that start out better, I suppose. But showing an app that is made worse (even if expectations are low) by antialiasing was a poor choice for demonstrating antialiasing as a desirable feature. They should have shown an antialiased GUI that is either not degraded, or improved, by the visual improvement - like type, or composite images, etc.

Multimedia

[Silkejr]

It's great to see more multimedia stuff being worked on. Ease-of-use in multimedia is a big selling point for new users of Linux, from what I've seen when I show different distros to friends.

VFS / FreeDesktop.org Status?

[Doc Ruby]

How about the URL and MIME integration of GNOME-VFS? It's supposed to let apps hand off data, as per the FreeDesktop.org specs, with exactly the same API calls as in KDE (and other systems subscribing to the FDo spec). It kinda worked before, but didn't really interoperate with the KDE implementation. Is that working in GNOME v2.12? How about the KDE work to comply?

This interop really would go a long way towards a useable Linux desktop for everyone. Right now, users must choose which of GNOME or KDE they use. Or install both, and either switch, or put up with mismatched widgets. But developers must pick which environment to code into their apps. The FDo specs offer any app the critical ability to toss data to another app, across a "three-tier architecture" boundary. The protocol lookup lets an app that gets a pointer to an object (URL) call another app to retrieve the data, without knowing which app - it just asks the OS to look it up. The app can then do the same for processing or presentation of the object, according to the MIME type of the data (or decoded from the URL, at the app's choice). This extremely powerful IPC lets any apps integrate, even using custom protocols and x-<whatever> MIME types for custom integration.

But if it only works on GNOME or KDE, stuck inside its ghetto, it's not going to get used by developers. It's not going to be expected by users. It'll remain a half-assed solution to problems internal to a given desktop choice. When instead it could make Linux apps integrate easily, across all kinds of divides, like an automated backdend clipboard. Instead of a fragmented desktop market with limited integration, we'd have a unified market, with choices available for ways of doing things, and seamless app integration. Without excluding any apps - rather, including every app that calls the unified API.

This tour mentions GNOME-VFS only in its HAL/GUI feature. What's the status of the real guts? And is KDE working yet?

Re:Works For me (TM)

[astralbat]

Absolutely. The problem is not with the xine backend or the decryption libraries etc.. but with Totem itself. It often refuses to auto play my dvd-rom drive (as xine or mplayer does).
If you do ever manage to get a dvd-rom to play, the navigation slider is innacurate, and skipping forwards or backwards often causes the audio to go horribly out of sync, although I'm not sure if that is more to do with GStreamer.

For most media files though, Totem is fine and I'm looking forward to the improved DVD capabilities coming with Gnome 2.12.

Braindead

[vthokie69]

Now if only they would revert their braindead file chooser to something more usable by adding the text entry back.

Gnomdows

[celephaix]

Gah! It looks like.. Windows! I mean, I thought that was a screenshot of the Win 98 file browser for a second there. Which is too bad.. I use Gnome to take an occasional break from ugly Windows interfaces.

I don't like the current trend of trying to win over Windows users by emulating Windows. OS X can win over Windows users by just having a better interface. Instead of copying Windows, why not innovate?

Re:Gnomdows

[Slack3r78]

Actually, if it looks like anything by my eye, it's classic MacOS. This would make sense given that the Gnome devs are following many of the same interface guidelines as MacOS did prior to OS X.

Re:Gnomdows

[celephaix]

This is one of the screenshots I was refering to: http://www.gnome.org/~davyd/gnome-2-12/images/naut ilus-browse.png Compare that with this: http://www.baycongroup.com/win98/w98explo.gif The web browser looks pretty similar too. Although, I just noticed they let you use Google's cache for pages that can't be found. Now that's a really good idea.

Read again, please.

[Kristoffer Lunden]

If you'd actually read my post, you'd see that I was complaining about X pasting text when I was editing code (this is the biggest issue). No mention of a browser.

But I see that you are only on a mission to promote your favourite product no matter the subject or context, something that almost always backfires. And you absolutely did not make me more positive to Epiphany, a browser that I like but don't think is good enough yet. Calling other browsers names like that does not help one bit. It just shows uncertainty and weakness. Think about that, and you'll make a lot better job promoting it.

Thank you.

Re:Read again, please.

[ReinoutS]

Epiphany, a browser that I like but don't think is good enough yet.

If there are any specific things you would like to see changed so that it becomes "good enough" for you, you're free to file a bug, post to the mailing list or visit us on the #epiphany channel on GNOME irc.

Re:Read again, please.

[Kristoffer Lunden]

Actually, I've been participating a bit with wishes on the wiki, and I'm trying it out now and then just to see how progress goes.

I think I like the way it is heading pretty well, just that it isn't there yet, but I see work is being done. The idea of a "Firefox" or at least Gecko based browser that integrates completely into Gnome *is* a good one.

The only thing I think might be a mistake is to not support Firefox extensions and instead go with the C/Python-based ones. On one hand, it's not such a bad idea for some extensions, especially when it comes to performance - on the other hand, you miss out on a huge wealth of very useful plugins, some of which are to services where there probably(?) never will come a Epiphany alternative, say like a real Google bar or the Stumble upon-plugin.

Even not counting those, people are making tons and tons of great extensions to Firefox and that is one of its major strengths. That is sad to see getting thrown away. Most of the engine for supporting it should be still there, shouldn't it? Or does Gecko not render the complete interface? I don't remember...

Of course, ideally, I suppose all widgets should be non-XUL for consistency... which might not be really doable. Or maybe it might, XUL could quite possibly use any toolkit to actually render the components if they contain the same set.

Well anyway, please keep the C/Python ones and use it for core extensions, but if at all possible, please support installing extensions for Firefox as well. If the trouble is with installing, architecture etc and not with XUL/Javascript/DOM, than I'm sure it would be possible to work with the Mozilla team and come up with something all could use. If I remember correctly, installers nowadays are "just" RDF configs that says which files do what, more or less, so it ought to be possible to use even if installed in a completely different place.

Then again, I don't actually know anything about this, but hey it is Slashdot, isn't it? =)

Anyhow, keep up the great work. I for one will at least keep trying Epiphany now and then.

Re:Read again, please.

[cortana]

I'm sorry, I hastily read your post and understood that you were complaining about the irritating middleclick = go to url in primary selection behaviour that Mozilla defaults to.

I'm not sure what you mean by "calling other browsers names". It is annoying to see the good work of the Gnome developers being overlooked when people use Firefox under Gnome, and assume its poor platform integration on Gnome itself, which is what I thought you were doing.

As for your pasting issue: I found it annoying when I moved to the platform at first... then I got used to it. I even had a shitty old mouse with a weak spring underneath the scroll wheel that made trigger a middle click often. I simply replaced the mouse. If you don't want to do that, I believe others have stated that you can remap the middle button to some other high-numbered button that doesn't do anything.

Re:Not so simple!

[kbielefe]

I favour object-oriented (OO) over procedural, ... OO code has been verified to be better.

That's probably why Gnome/GTK+ is one of the best examples of object-oriented C code I have seen. Just like procedural code can be written in Java, even though the language requires everything to be part of a class, object-oriented code likewise can be written in most every language. OO is a design methodology, not a programming language feature, although certain languages are unarguably designed to be better suited to OO programming than others.

Personally, I learned the basics of OO programming from the camel book. Larry Wall's witty writing style and the pragmatic way classes are defined and used in Perl helped OO theory to "click" in a way that any C++ text I had read previously failed to do for me.

I'm not trying to say that Perl or C are better than C++ for object-oriented programming. I'm just affirming that OO programming is not limited to C++ and Java.

Re:Hidden sinker

[Anonymous Coward]

Two reasons:

(1) The free-software movement (of which GNOME claims to be a part) has the goal of giving users freedom to do with software what they want or need. Mac and Windows, while being fine systems in many respects, do not. Why do "freedom" and "don't have to be an uebergeek to use it" have to be mutually exclusive?

So yes, I think your position is a bit selfish: freedom should not simply be for the technological elite!

(2) Making things easy for the toughest group to do this for, tends to make things easier for *everybody*.

• Those big handles in hotel bathtubs were designed to make it easier for handicapped people, but they make it easier for everybody.
  
• Electric starters in automobiles were originally built for women, but I don't know any man alive today who wishes his car had a hand crank.
  
• And I know plenty of geeks who said "We don't need to improve search...", until they actually used find-as-you-type in Firefox and then immediately want that feature in *every* program.
  

Even if we never do get a significant share of Joes to use GNOME, we'll have made things much better for people who are not average Joes.

Make it work, make it right, make it fast.

[Dante]

Make it work, make it right, make it fast.

They aren't really present

[Andy Dodd]

They've been available for months, but have been considered heavily experimental and will still be considered heavily experimental for quite some time.

Even with a recent version of xorg and the latest NVidia binary drivers, generally considered the most robust and stable implementation of the composite extension available at the moment, using the composite extension is CRASH CITY. The composite extension still causes massive problems with the GLX and XV extensions, so much so that using GLX and Composite simultaneously is disabled by default. I've had so many problems with that extension that I outright disable it in my X configuration.

All I want to know is...

[Feztaa]

Will GNOME 2.12 make it into Ubuntu Breezy?

Re:All I want to know is...

[arose]

The Ubuntu release cycle is based on the Gnome release cycle: when a new gnome is released the Ubuntu team packages it up, irons out some bugs and a new Ubuntu is released. So yes, Breezy will be released shortly after Gnome 2.12 and will contain it.

Re:Gnome vs. KDE

[Breakfast Pants]

As I understand it you cannot develop a non-GPL application for KDE because QT has released their stuff GPL, whereas with GNOME you can because everything is LGPL. Is this correct?

GNUstep vs. GNOME vs. KDE

[O2dude]

GNOME and KDE are application development frameworks as much as they are 'Desktops'. It's is therefore ironic that the GNU version of the highly respected Openstep/NeXTSTEP application framework, which by the way is now of course the foundation of Apple's Cocoa/Mac OS X stuff, receives so little coverage or interest.

It seems to me that GNUstep (http://www.gnustep.org/) offers the cleanest framework for application design in Open Source land with a totally Kick-arse development environment etc. etc. etc. Am I the only person that finds it rather odd that so few people use it?

Strangy strangy...

Re:GNUstep vs. GNOME vs. KDE

[RAMMS EIN]

Well, let's be fair...it's butt ugly. I know work started on some themeability, but I haven't seen any way to properly skin it yet, which I consider a huge weakness. Also, it had some stability issues last time I tried it. I guess that goes a long way in explaining why people don't use it and it doesn't receive a lot of press.

It's a shame, really. OpenStep was far ahead of its time, OS X shows how beautiful and powerful it really is, and the GNUStep folk are working hard to not only implement OpenStep, but also Apple's extensions. With Renaissance, you can even create apps on one that will compile and work on the other, and this very easily.

Re:GNUstep vs. GNOME vs. KDE

[Daverz]

GNUStep is damn difficult to install. While I like a lot of things about Objective-C, I don't really care for manual memory management. If and when it starts working with PyObjC, I'll be there.

Re:GNUstep vs. GNOME vs. KDE

[Somnus]

I use GNOME, but my heart is in GNUstep. Now that camaelon is working, they just need some apps.

Right now my must-have apps are Evolution, Epiphany, Gaim, Gnumeric and Rhythmox. If GNUstep gets suitable equivalents, I'll be all over it.

I'd also be interested to know if GNUstep has any equivalent to HAL, gnome-volume-manager, gnome-vfs, etc.

Prerelease Tour of GNOME 2.12

[kiddie]

GNOME 2.12 will be released to the world on September 7th, 2005, culminating 6 months of very exciting work by members of the project. A number of exciting technologies come together in GNOME 2.12 that will set the standard for free software desktops to come. you can read much more here

Re:Gnome vs. KDE

[Klivian]

No, it's not correct. You can develop with any sort of open source license on KDE. The KDE library's are LGPL or BSD (While most applications are GPL, but that does not affect other application developers) and the Qt libs are GPL or QPL. The QPL gives you right to develop using the Qt libs using any open source license, the only restriction are that the source has to be freely available(But again that's more or less what the open in open source means:-). If you want to develop closed source programs, you have to buy a developer license for Qt from TrollTech. So the correct answer is, you can develop non-GPL application for KDE, as the Qt licenses makes it possible.

Re:Gnome vs. KDE

[seguso]

Either you use the GPL, or you pay for the freedom to choose another license.

You pay for the freedom to remove freedom, yes. Can't see anything wrong with it.

Re:Gnome vs. KDE

[Andre]

You pay for the freedom to secure developers' jobs in many software companies.

People have to pay for the product in some way. Normally, if people get something for free, they won't pay for it. Only if there is added value in doing this, they pay. For example, buying an audio CD often gets you a beautifully designed cover, lyrics, good audio quality, and a very robust medium. Many people like this and pay for it, even if the music can be downloaded somewhere for free (paying only for bandwidth). They know that the artists get their part as well, so the added value can be small. Of course, if the audio CD is not conforming to Red Book (CDDA), the "added value" might be negative ;-)

For an application, using a free license and offering service is another possibility, but it depends on the product and the market. This is at the discretion of the company to decide, and customers can then decide if the offer is acceptable or develop a free alternative (but everyone has to earn money somehow in the meantime).

Re:KDE

[Eslyjah]

You make a good point, and choice is good, yadda yadda, but I wish the open source world could pick one toolkit. It makes sense to me that some people like Xfce and some people like GNOME. I love that Xfce gains from GNOME's improvements to the GTK+ toolkit. GNOME and KDE using different toolkits is unfortunate, though I don't know what could be done about it.

Re:Hidden sinker

[Omega Blue]

I mean, if you're teaching the system or providing help desk support, you're dealing with two different beasts when walking a user through KDE vs. through Gnome.

If you are teaching how to use the Linux Desktop or working in technical support, chances are you know which windowing manager you are dealing with. I am quite certain that, within any organization, the desktop is standardized, so there is no doubt at any moment on this issue.

Re:KDE

[kelnos]

Still, you really just don't get it. Qt and GTK+ represent two different ways of thinking with regards to how a GUI toolkit should be written and used. Telling the open source world to "pick one toolkit" is tantamount to telling half the OSS GUI community that their way of thinking -- something rather subjective and variable -- is wrong. Is that something you really want to do? I didn't think so...

*BSD is dead

[weaselcho]

Now that I've seen this topic, I'm sure about it. BSD is dead as D.E.A.D... Next thing you see: Xtux's new version reviewed in the BSD section. This whole thread is a troll

Re:Replies to Sarojin & kbielefe

[juhaz]

Performance and code size probably do play a part in the decision of language to use for core libraries, but another important thing to consider is that C libraries are easy to wrap - they can be bound to be usable from any other language. Thus GTK, and to lesser extent GNOME libraries, are perfectly usable from Python, C++, PHP, Perl, Java, C#, probably Objective-C and lots of others too.

It might not be exactly impossible with Obj-C libraries, but it would certainly be much harder, and so there would be not nearly as many language bindings as there are now, and restricting programmers language choices is definitely NOT a way to increase productivity, OO or not.

Re:nautilus default behavior

[juhaz]

Ubuntu, probably. Hoary had modified nautilus behaviour too. Otherwise spatial, but it always closed the previous window when opening new (same as opening with middle click in vanilla gnome).

Re:nautilus default behavior

[juhaz]

AFAIK Gnome doesn't have anyone providing a "just this distro" feature set.

GNOME doesn't, but they can't prevent a distro from distributing a patched version. And Ubuntu has done so before.

Re:Hidden sinker

[cant_get_a_good_nick]

So it's MS Windows - is that 95, 98, ME, NT3.51, NT4, win2k, XP, 2003? All have their quirks, advantages, disadvantes, and users always hide their start menu entries in different spots if they've had the thing for a while.

Was always fun to have to install a newer version of IE to get some kind of common control. Back in the day I needed to install a new version of IE to get a (16bit) properties dialog to work the way I needed. The "single windows UI" is a myth.

Re:Looks like an interesting combination

[aztracker1]

My personal take, is the window managers have been ready for several years now, for the general public and add features sometimes ahead of, sometimes behind commercial OSes(mac/windows) ...

That said, I think a *LOT* needs to be done to make commercial applications viable across multiple distros, with minimal effort, which is where linux comes up short a lot of the time... I honestly feel that progress is being made, and offering frameworks such as Python, Java, and Mono with the more popular distros can only serve to help this effort... managed environments are enough for *most* applications (not all, but most)... with a common library setup, differing versions under the covers become less important.

an earlier thread in this article makes points about beagle, and the framework overheads in play, but for a "user" desktop it's more important to have some comonality with the distros for commercial apps to develop on, and be widely distributable without a compile under each distro. Not as good an idea for the server space, but drastically needed for the desktop space...

On a side note, the PC-BSD distribution (bsd not linux) is a great base, though java, mono, and python install packages would be a wonderfull addition.

Try Galeon

[Ignominious]

I agree. I'd use Galeon in the meantime (Epiphany forked from it). It has more configurability and comes with really nice tab features. You can edit the toolbars (Edit->Toolbar) much the same as Firefox. It's basically Firefox with much better GNOME integration.

Rubbish

[Ignominious]

WTF? They don't look at all similar considering they're both file managers.

Thumbnails definitely weren't in Windows 98.
Windows 98 didn't have Places (common useful network/filesystem shortcuts that can be fully customised by the user).
Windows 98 didn't have a directory extent toolbar which is simple but very useful.

You're complaining about the fact that there are two large panels similarly placed even though the usability is fundamentally different. Don't you think you should perhaps actually have tried using GNOME recently before criticising it?

Re:Rubbish

[celephaix]

I think we're talking about a bit more than similarly placed panels, although that's another problem. I'm not saying they look exactly alike, but Windows 98 was the first thing that came to mind when I saw those giant buttons that say back, foward, up, stop, etc. with icons that are obviously Windows inspired.

I use GNOME on a regular basis and it's of course functionally different, but why try to look like Windows? What's the point?