GNOME 2.10 Beta 1 Public Testing Release

I happy gst is coming along, BUT FOR THE LOVE OF GOD can we PLEASE have an option to turn off automatic thumbnailing?

because no one has written support for mplayer. gstreamer is of course advancing at a good pace, wmv 9 support (without win32 codec usage) is upcoming as well unless I imagined the blog entry on planet.gnome.org. So better thumbnailing is coming.

KDE 3.4 implements the freedesktop.org trash specification (~/.local/share/Trash). And I guess so does GNOME 2.10.

From what I understand In KDE 3.4 we will have the ability to trash many files with the same name but located in different folders, and still be able to restore them. This means that there need to be some logic involved it is not just about a where the trash is located in the file system. Is this logic included in the spec? If so will gnome have it.

That's part of the spec.

As I have said for other apps, apps such as Anjuta which have a specialist audience should not be part of GNOME.

Then FFS, go install one. But quit complaining about it here, mine works great without any "enhancements" leave it alone TYVM.

Go install one? A working clipboard manager??? Uh, how about telling people to go install working right click? Or working file open dialogs! Your response is beyond idiotic.

The parent poster is 100% correct that this is an embarassing hole in gnome. If it works great for you then you are not really doing much copying and pasting or you have just gotten too used to the idea of working with a clipboard that is cripled:

a) copying anything but plain text is flakey and at best only works between apps that made agreements about copying such things. A generic system needs to move from discussion phase to implementation and standard.
b) in gnome stuff is only held on the clipboard as long as the app you copied from is still open. Try copying text from gedit, close gedit, then paste that text into another document. Doesn't work.

Continuing to bring up this glaring hole in *basic* desktop functionality is not pointless complaining. Your dismissal of the parent post is so insanely stupid. Wow you are a dumbass.

"b) in gnome stuff is only held on the clipboard as long as the app you copied from is still open. Try copying text from gedit, close gedit, then paste that text into another document. Doesn't work."

Not much consolation to those who insist on running G everything, but klipper seems to work and docks fine in the tray area if you have that applet running.

that make Gnome unfit for the corporate desktop. Much more is missing.

How about some way to manage users and groups through LDAP, perhaps something like the user management tool in MacOS-X or KDE, that not only handles users and groups but also samba accounts and e-mail aliases.

How about being able to reliably connect to windows NT shares on windows networks. KDE have managed to do this for a long time now why not Gnome as well? This is a serious showstopper since it is highly unlikely that a company will replace all their windows stuff over night.

For the same reason, another important factor to land Gnome on the corporate desktop would be a windows port of Evolution. Companies are not going to stop communicate just because part of their desktops run another OS.

Evolution needs major improvements. E.g. it is impossible to move mail from one IMAP folder to the other by drag & drop. It is impossible to set permissions on imap folders so that other users can access them as shared folders. It also needs some more scalable way of storing calendar events and tasks, perhaps they should do like KDE and store it in imap folders. Support for Sieve server side filtering for IMAP would be nice too, if we wan't evolution to be an outlook/exchange killer (and that's probably necessary to conquer the corporate DE)

One other thing that need to happen is to simplify the view of the file system for ordinary users. I'm thinking of something like what MacOS-X have done. In OS-X normal users doesn't see folders like /etc, /bin, /usr/bin, /lib/, /usr/lib/... The ordianry user doesn't need to see these places as he normally can't write anything to them, and whatever applications or info available here could be pressented in the application menu, or some some settings panel. The sysadmin should of course still be able to see all files and folders (this would be an improverment over OS-X where they are hidden even to the admin account). This would speed up navigation to folders that the user actually use.

Evolution needs major improvements. E.g. it is impossible to move mail from one IMAP folder to the other by drag & drop.

The only problems I have moving mail between IMAP folders is when moving mail from a public folder to a private folder (or vice versa) on a Microsoft Exchange server. All IMAP clients I've tried have the same problem, but the fault is in MS Exchange, not the clients, see IMAP4 and public folders. As a workaround, I first move the message to a local folder .

The ordianry user doesn't need to see these places as he normally can't write anything to them, and whatever applications or info available here could be pressented in the application menu, or some some settings panel.

I hate this kind of behavior. An ordinary user may not be able to write to those directories, but they may want to be able to read from them. There is no reason to make the filesystem inaccessible. At the moment, the default Gnome desktop makes it so users have ready access to their files and settings. They shouldn't "accidently" start roaming the entire filesystem looking for something. However, if they want to get to the rest of the filesystem they can with a few clicks. I it might be possible to make the Home folder a bit more obvious, but don't remove access to the root filesystem.

Yes, you may need the information in the hidden directoris, but being able to see it through the file system is not necessarily the right way to do solve the problem.

Let me give you an example: Assume that you have a problem with name lookup of IP addresses. You may then have to look at /etc/nsswitch.conf, /etc/resolv.conf, /etc/named.conf and perhaps some /etc/rc.d/ file.

Ordinary users like your secretary, your lawyer, or the person at your local library will have no idea how to interpret this information they need some kind of user interface that can guide them better than what nautilus can do.

Another example: George calls you on the phone, saying that he have made the sales report from last month available to you, and that you can get it from his home directory. Now you have to remember that his home directory is /home/sales/gsmith and you will have to browse the file system to get there. On your way you will see a lot of directories you don't need. If you are Georges boss you would probably want to find a directory called "George Smith" not gsmith and you don't want to be bothered with folders you never enter.

So, whats needed is not only to hide files, but to make it possible to map user related concepts to file system objects. E.g. Your users would probably talk about projects rather than /usr/local/shared/projects. To be successfull we need to talk the language of the user, he is usually not paid to be a computer expert. It is bad enough that windows users have to learn about C: etc. Lets not repeat that mistake on the free desktop.

I think the problem is that today many people that use Gnome are home users and all the information on their computer belongs to them or some familly member that may share it with them. In this situation seeing all of the files system makes perfect sense. If you have installed Gnome at home you are also likely to have some technical skills or at least have access tosome familly member possessing such skills and you are using the system of your own free will.

But if you work at IBM or some other big company you, use whatever system your employer puts in front of you. This means that you may not have as big incentive to learn what varios diretoris is used for as if you had installed it yourself on your computer at home. You just want to get the work done as quickly as possible so that you can collect your paycheck.

From a strategic point of view, I think it is better to develop for the corporate users than for the home user. They are the ones with money to support development. They are the ones that can put economic clout behind demand driver compatibility for hardware etc. They are also the ones that have most to gain from stream lined multi user systems such as Unix. When Unix and gnome is in place at work it the home users will follow, and with them the game and entertainment industry.

OK... to hide files; like in OSX use a .hidden file (like OSX uses I think)

Evolution windows port; Novell is going to do this (woopee!! - hopefully more gnome libs will come over too hence more apps)...

ACLs - Yes this would be good and is probably needed in the meantime something like this
http://freshmeat.net/projects/guiaclmanager/
can help...

Windows connections - I find these get better each release maybe it'll be good enough for you in a couple of releases...

Yes; it is frustrating to wait for these things but as people not contributing to it it's the way it is; it's fine for home use at the moment and I eagerly download each new release to check out the new features.

But y good luck with the waiting; if you submit bugs hopefully someone will take them up :)

>OK... to hide files; like in OSX use a .hidden file (like OSX uses I think)

You mean this is possible in the current version of Gnome???!!
If that is the case, the we just need to alter the defaults.
Any links to documentation on this?

Most normal desktop users don't use ACLs.

In corporate environments where there is sensitve data (either because of secrecy/privacy or just because you don't want the wrong person to change/delete something) people use ACLs every single day. And yes, that users cannot do this in unix/linux *is* a deal breaker.

There are ACLs in unix/linux. The problem is the file manager nautilus!

a) There are acls in 2.6 distros only which are all very recent. Most corporations won't go beyond what is offered in RHEL (at my work) which right now is still 2.4.
b) Without good file manager integration there is no way even your average power desktop user is going to know how to use ACLs. I need to be able to do like windows (seriously) where I can just righ click->permissions and then put a little check box next to the name of the user I want to give a permission to. Most users don't even have a way to find out who the other users are without being root!

So while you are technically correct ACLs are not really here for mass consuption. Anyway the post was about ACL support in gnome, not acls in general.

Solaris and many other Unix flavors have had ACLs for years.
When this functionality finally gets integrated into Gnome/Nautilus
there will need to be some kind of fallback so that older systems
can be handled as well. The fact that some Linuxes don't support it
is no reason not to implement it in systems that do.

Besides, most corporate desktops run Windows, so where ACL use is common so saying that not many people use it just because RHEL doesn't have it is bending the truth somewhat. Besides RHEL 4 will be out soon, and I would be very surprised if that didn't have it.

If there is no GUI most users will not even know that they are there.

Its a bit annoying that a windows user connected too your server through samba have more power over the Unix filesystem permissions than the Unix admin.

It also gives you a feeling of unsecurity, how can you as a user be sure that other people doesn't have unknown permissions to your files, as Gnome not even let you see ACLs, let alone set them. Yes, you could use the command shell to find out but that requres a lot more knowledge on the users part.

This really is a serious bug, in most unix GUIs. I would hope that Gnome could be the first unix DE to fix it.

What "recent Evolution security problem"? It hasn't been mentioned in evo's blog, or on any of the Linux security mailing lists I watch...

> What "recent Evolution security problem"?

See the problem? No security overview page, no advertized contact address for security problems, unsigned source checksums on FTP. Doesn't it look like an anti-security systematic?

The one for which Novell/SUSE issued an online update today: http://cve.mitre.org/cgi-bin/cvename.cgi?name=CAN-2005-0102

Who cares ? It's not like you can't install more software just because someone says something is not "part of GNOME".

Why not replace epiphany with Firefox with a suitable skin. That way users from other platform would feel right at home.

Users of gnome would then have an app that sticks out as non gnome in its behaviour, e.g. slower ui response, submenus not behaving the same. In addition an app that didn't follow the rest of the gnome release schedule.

Have you actually used Firefox build for gnome recently? I does not, in any way, "stick out like a sore thumb". The gnome integration is really good.

1. Gnome theme is 100% perfect
2. Gnome file open/save dialogs are used
3. Gnome application preferences are used

There are two small things that make it possible to tell that Firefox is not 100% native:

1. Still need a theme to get the icons right (this is being worked on)
2. Still uses its own print dialog. However this one is sort of ok because a) the dialog is themed correctly, b) print dialogs in gnome are all over the map anyway!

So really, except for native gnome icon themes being used, what you have said is just complete bs and epiphany fan boy crap (which is all you ever get from epiphany uses since everyone who can think clearly on such things know that epiphany is a total piece of crap)

the firefox build in Ubuntu Hoary. no icons in the menus at them moment (admitedly that is small), slow menu display. submenu opening does not exhibit the same behaviour as gtk submenus. You can make firefox look like a gnome app, but it just doesn't behave like one.

Settle down, please. While I prefer to use Mozilla/Firefox (mostly due to inertia), your statements are not accurate. See http://tinyurl.com/4tgqa . Also, while it's possible the parent poster was an "epiphany fan boy" (which I can see nothing wrong with; after all, you appear to be a "firefox fan boy"), it is clear that he was civil. The least you could do if you are going to call someone wrong is to play nice (and try to double check your own facts).

why don't you make a specific reference to where I got my facts wrong? although your point about being polite is taken, I'm very sure that my facts are correct.

I read your original post as "There are only two things in Firefox which make it not 100% integrated with Gnome". That's a mistake, you didn't actually say that. You said that Firefox does not stick out like a sore thumb (which I agree with, and is one of the reasons I don't bother switching to a "more integrated" Gnome browser, besides the fact that I like some small things it does differently much better) and that "There are two small things that make it possible to tell that Firefox is not 100% native" (with the word "only" not present in the sentence). So, I apologize and take back what I said before. You were accurate.

...I did provide a link that would have proved the other statement incorrect had you made it (http://tinyurl.com/4tgqa), but it's irrelevant considering what you really said.

I would certainly agree with you if Epiphany was built on pure Gnome technology but it is mostly a wrapper around Gecko from Mozilla.
This means that it largely depend on Mozilla release scheduel for bug fixes etc. Whenever there is a new Gecko there needs to be a new
Ephipany.

As for look and feel that can be fixed with skins. If something can't be fixed it would be a better idea to work together with the Mozilla Firefox team to make the problem go away than make an entire product of their own.

By doing so, new gnome users would have access to a cross platform browser and that is something that would make a transition to gnome from windows easier. Webbrowsers, Mail and Calendar clients are part of corporate IT infrastructure, and if they are not cross platform they hinders migration to Gnome from e.g. windows.

is there a skin that will make firefox follow my toolbar settings if I change them, so I will have text/icons etc. like all my other apps?

Epiphany doesn't need a new release when there is a new gecko, as it is built against a specific release of mozilla or firefox. epiphany itself can develop along with the gnome schedule as the app isn't just dependant on gecko.

Look can be handled with skins. Show me a skin that duplicates GTK behaviour with submenus, or a skin that can make a XUL menu display as quick as a GTK one.

In terms of a web browser, there is nothing a cross platform app should give that a non cross platform one can. The web browser UI has stayed pretty much standard in all browsers. As for hindering migration, should we then also drop spatial nautilus, because it is different to how windows behaves? Should we drop the panel? Should we change the HIG to match windows, all to aid migration?

Yeah, sure. We are going to add another 200 applications until the end of the week. Would that be enough or do we need to add more? Just tell us.

That's funny. Well said, Anonymous George.

which is not part of gnome yet ...

Beagle is for 3.0

Beagle needs Mono and it's not certain that Mono will be included.

Actually the fd.org "crap" allows an app to specify if it should only appear under a specific desktop. There is also nothing preventing the "crap" from being edited. Gnome just doesn't currently have a UI for doing it. XFCE does IIRC.

and true :-)

Good one!!

There should be 5-10 items that cannot ever be removed and must appear (item for the system config and preferences, default browser and email, browers nautilus, favorites, history, recent (all ajustable) but the rest should all be open to the user to modify.