Could Trisquel 8 switch to MATE?
- Anmelden oder Registrieren um Kommentare zu schreiben
As of Ubuntu 15.04 onward, Ubuntu MATE is an official spin meaning MATE packages are available by default and will get feature and security updates from here on. Things can only get better with 16.04, for which Trusquel 8 will be based on.
While I applaud Ruben's work in getting Gnome 3 classic almost usuable, maybe its best to switch to MATE to get the pre version 6 look and features that made the Gnome 2 versions so great. Low in resources and works well without 3D accelleration.
Thoughts?
I apologize for the spelling mistakes by the way. I wish I could edit the original post. :-(
its run orca??
https://wiki.archlinux.org/index.php/MATE#Accessibility
"MATE is well suited for use by individuals with sight or mobility impairment. Install orca, espeak (Screen reader for individuals who are blind or visually impaired) and onboard (On-screen keyboard useful for mobility impaired users)
Before starting MATE for the first time, enter the following command as the user who needs accessibility features:gsettings set org.mate.interface accessibility trueOnce you start MATE, you can configure the accessibility applications via System -> Preferences -> Assistive Techologies, although if you need Orca, you will need to run it from the Alt-F2 run window in order to start getting speech."
Had Ubuntu MATE not become official spin, though, would MATE still have made it to the repos anyway (since I'm pretty sure it's present upstream at Debian)?
I was thinking about this too, even with the great work of Rúben and the others on the appearance of the distro. We have lost features like window tiling, and we have bugs like bad drawing when scrolling on lists. And, as far as I know, is all Gnome's fault. But then, we have MATE which could make Rúben's life easier since he is not fighting with the DE on how should look like, because MATE already looks and feels just the way we are used to.
Window tiling's gone??
If it is I don't think it's GNOME's fault since it was basically absent in GNOME 2 and is present in GNOME 3.
Well, window tiling wasn't a GNOME's feature, but we have it in Trisquel 6 thanks to Compiz. As far as I know, we don't in Trisquel 7 because it's impossible to use any window manager other than Metacity without changing GNOME's source code. And Metacity don't seems to have window tiling, at least no in Trisquel 7.
GNOME Shell allows you to maximize windows on the left or right side of the screen. Of course that's got little to do with what GNOME Flashback is capable of, but GNOME Shell is the normal GNOME experience.
"GNOME Shell is the normal GNOME experience."
Yeah, but as far as I know Trisquel aim to have a classic and lightweight DE for it's default image. And GNOME Shell needs hardware acceleration to run decently. That's why we use GNOME Flashback, but since it lacks features that MATE have without loosing anything (as far as I know), then we have a reason to make the change. Wouldn't you agree?
Perhaps I should've used "GNOME Flashback" instead of just "GNOME" in my previous posts.
It's disingenuous to suggest that GNOME has removed a feature it hasn't. What's happened is Metacity never had this feature in the first place (GNOME first got it with GNOME Shell), and supposedly (I've only heard it in this thread) it's difficult to use an alternative window manager with GNOME Flashback now, meaning you can't (easily) get the feature from Compiz anymore. The solution for GNOME Flashback would be to add the feature to Metacity.
That's not to say I disagree with the position that MATE is better. Honestly, I don't have a strong opinion either way; for my needs, both are about as disappointing (I much prefer GNOME Shell), though I think MATE does have a variant of mintMenu, which is a neat thing if that's true. I'd say what's most important is to look at the accessibility problems MATE has, and consider whether the benefits of MATE outweigh those problems. After all, one of the major reasons Trisquel uses GNOME is GNOME's great accessibility.
Yes, there is a version Mintmenu for MATE. It seems to prevent dconf-editor from running correctly when I install the version from Dagda, though (except I'm probably the only one who cares one wee bit about that).
Agree with you regarding GNOME Shell. But without marketing nearly as good as done by the chaps in Cupertino a switch to that would be at least a bit disasterous for Trisquel.
> "It's disingenuous to suggest that GNOME has removed a feature it hasn't"
Sorry for my imprecise use of the language, I'm still learning English. What I wanted say is that we lost a feature because of GNOME putting in the way, not that it was GNOME itself who removed it. Sorry about that.
> "and supposedly (I've only heard it in this thread) it's difficult to use an alternative window manager with GNOME Flashback now"
I read this in an article about Metacity on Wikipedia some time ago, though it's in Spanish http://es.wikipedia.org/wiki/Metacity#Controversia
My translation:
However, since Gnome 3 it will not use Metacity but Mutter, which does not allow the use of any window manager without rewriting the source code, this has generated even more controversy to the point that the Proyect GNU itself has greatly criticized the project and suggested solutions but apparently have not been heard by the GNOME project.
> "I'd say what's most important is to look at the accessibility problems MATE has, and consider whether the benefits of MATE outweigh those problems."
I agree, perhaps @Dave_Hunt can tell us more about those few minor accessibility gaps with MATE.
You're cunfused, then. That's talking about GNOME Shell, which is indeed tied to Mutter. Mutter is different from Metacity, and GNOME Panel should be perfectly usable with Compiz et al.
By the way, that sentence about the GNU Project criticizing GNOME for this should get a "citation needed" tag. I've never seen the GNU Project or FSF making any such criticism in English, though it could be that they've done so only in Spanish.
I thought that was the case since Trisquel 7 didn't include Compiz by default, but it seems that was not the case. I'm still wondering why Compiz didn't make it into Trisquel 7 though.
It's still available, just not used by default (as was the case for most distros that used GNOME 2).
But why is not installed by default? We lost the window tiling that we had in Trisquel 6 because of that. Was because of something related with GNOME?
Honestly, I think the question of interest is why it was enabled by default in Trisquel 6. That wasn't the case in Trisquel 5.5 (the first version of Trisquel I used), and I'd wager to guess that wasn't the case in any of the earlier versions either. So Trisquel 7 is just continuing the previous status quo.
One guess I have is that using Compiz by default (and falling back to Metacity, because remember, Compiz requires hardware acceleration) was an experiment, and deemed a failure because of problems it caused, or something along those lines. Compiz, after all, is quite an advanced window manager, so if you're going to use it, you should probably be using it on purpose.
> Honestly, I think the question of interest is why it was enabled by default in Trisquel 6.
I don't know. It was? I remember that was installed by default, but not that was enabled by default too. If I remember correctly, Trisquel 4 also had Compiz installed by default.
> One guess I have is that using Compiz by default (and falling back to Metacity, because remember, Compiz requires hardware acceleration)
Using Compiz by default and falling back on something else if fails would be a bad idea anyway.
Well, if we want to keep using Metacity we're never getting window tiling - and it's not GNOME's fault. Metacity never had it and Compiz didn't have drag-to-edge tiling until 0.9 anyway. It's got nothing to do with GNOME 3.
But yeah, if you want tiling with a stacking window manager that doesn't ask for 3D acceleration, I wouldn't call MATE an unwise option.
No, you can use Compiz with the default Trisquel Desktop, the reason why Rubén changed to metacity is because compiz does not work well with Orca. You can create your own Trisquel with compiz session. When I have time I will create a script that will do so.
I am a member!
This is first I've heard of compiz not working well with orca; I'll ask. Does either metacity or marco do screen magnification? Whatever we decide to do for Trisquel's window manager in version 8, we should make sure that accessibility feature works.
> This is first I've heard of compiz not working well with orca
Me too. Do you know what kind of issues exists between Compiz and Orca?
Yes, Ruben said that Orca couldn't read the name of the window so if someone cannot see well, he/she would not know in which window/container he/she will be in, and as you know, accessibility is one of the key points of Trisquel so he didn't pick compiz as default.
Regarding of what Dave_hunt asked: I don't know about a magnifying effect on metacity or marco, only the option to make the text larger. Sorry.
I am a member!
Oh, I forgot:
With LXDE and XFCE, user is unaware of what window would have focus, when she releases the 'alt+tab' (window switching) keys; Orca doesn't speak this. In orca releases before 3.12, the same information was missing when Mate was the desktop. In the Mate case, it was a matter of sym-linking the orca scripts for metacity and marco. Now, it' works pretty well.
I am a member!
I have Mate and orca working in another distro. There are a few minor accessibility gaps, but it is usable. Since Ubuntu Mate is an official spin, and Mate packages are also maintained in Debian (Jessie) main repos, I think Mate in Trisquel 8 is a fine idea!
BTW, can run orca in lxde
I am a member!
Perhaps there's been some improvement. Last time I tried orca in LXDE, the panels, menus, desktop, and notification area were not accessible, and the file manager was only accessible in list mode. I found, however, that mousepad (the default text editor) is accessible. The Mozilla programs and libreoffice are accessible, independent of desktop.
I'm pretty sure LXDE's default text editor is Leafpad.
It is Leafpad. Mousepad is Xfce's text editor. Xfce's logo is a nice mouse (not an evil one like Mickey who refuses to enter the public domain).
"not an evil one like Mickey who refuses to enter the public domain"
U mad bro? Who cares if Mickey Mouse is public domain or not?
You should! Why? See e.g.
Not because it refuses to enter the public domain, but because of the Mickey Mouse Protection Act:
After the United States' accession to the Berne convention, a number of copyright owners successfully lobbied the U.S. Congress for another extension of the term of copyright, to provide for the same term of protection that exists in Europe.[1]
I agree with "Who cares if Mickey Mouse is public domain or not". After all is just a tool for brainwash. But I do care when some people with money can manipulate the government and it's law.
[1] https://en.wikipedia.org/wiki/Copyright_Term_Extension_Act
Same. Great idea. MATE for days. I love MATE.
its nede systemd?
"It's systemd needed?"
According to Gentoo's Wiki, no. Isn't that great?
Does MATE rely on a specific service manager or init system? No, MATE has been tested to work with both OpenRC and systemd and might work on other service managers and init systems too (untested, but no known reason for it to break); systemd support was added in release 1.6.
Isn't that great?
It isn't. It means MATE still uses the unmaintained ConsoleKit framework that GNOME recently ditched in favor of logind, thereby solving several longstanding energy management issues.
+1 for GNOME Shell.
> "solving several longstanding energy management issues."
Can you give examples of those energy management issues?
I heard that from a GNOME developer at a conference. I do not remember the details. I guess it has to do with things like closing the lid of your laptop (what suspends it by default) while the system is shutting down. If you have ever tried that with Trisquel (which currently uses ConsoleKit), you have probably discovered that the system may be stuck and need a hard reboot.
logind brings cool new features too. Multi-seat (many screens, keyboards, mouses, etc. for one single machine) for instance. According to https://wiki.freedesktop.org/www/Software/systemd/writing-display-managers/ :
systemd's logind service obsoletes ConsoleKit which was previously widely used on Linux distributions. For X11 display managers the switch to logind requires a minimal amount of porting, however brings a couple of new features: true automatic multi-seat support, proper tracking of session processes, (optional) automatic killing of user processes on logout, a synchronous low-level C API and much simplification.
I always close the screen while it's shutting down, and I've had no issues.
As long as Trisquel aims to offer a "classic" desktop that looks good and doesn't require 3d acceleration by default, the options are pretty limited right now.
Gnome fallback/classic tries to mimic many Gnome 2 features, but it really is barebones and a chore to customize. Like I said before, MATE is now supported fully upstream and by switching the default DE to it, we get more options.
The Gnome 2 versions of Trisquel were some of the best looking out there and we already have stable themes that can be used and improved upon where needed for MATE. The current implementation is hacky and really isn't working to be honest.
As for Magic's issue with ConsoleKit, I'm sure it can be resolved in a future MATE release if its important.
"The Gnome 2 versions of Trisquel were some of the best looking out there"
You can say that again! Trisquel 4 was my favorite, and we can achieve the same look again with MATE easily, while with Gnome fallback it just keep getting trickier.
"As for Magic's issue with ConsoleKit, I'm sure it can be resolved in a future MATE release if its important."
I hope they come up with a solution that satisfied people who don't like systemd. Then everyone is happy.
Version 4 was my favorite as well. Excellent theme, icons, and an overall level of polish. I loved the level of opacity on the bottom panel and the menu.
It was so pretty that I instantly installed it on my brother's laptop which thankfully worked out of the box (video, audio, wireless) and he could surf the web, listen to music, and write documents for school in LibreOffice with no hassle.
I'm a very visual person and I can honestly say that the Gnome 2 based Trisquels were on par with the non-free operating systems in look and feel.
The magic was lost with Trisquel 6 onwards as we lost most of the visual flair and I couldn't customize my desktop as easily.
Now that MATE is supported and ready to shine, its time to seriously consider it default or at least offer it as another ISO alongside the regular and mini versions.
As for Magic's issue with ConsoleKit, I'm sure it can be resolved in a future MATE release if its important.
It is really important since ConsoleKit is not maintained anymore. And it turns out systemd-logind should be used in the next version (1.12): http://wiki.mate-desktop.org/systemd-logind
Let me point out (it does not seem clear to everybody) that the actual init system does not matter to GNOME or to the future MATE. systemd is an umbrella project that is more than an init system. systemd-logind is a component that manages user logins and seats and works with any init system.
Magic Banana, you can say that again.
According to Pottering and Kay Sievers (systemd developers), their end game is to have systemd be a complete OS. Even Linus had to block all code contributions from Kay because they were trying to hardwire the kernel to require systemd. You can search the net for all of this. Latest news, is they are working on systemd also becoming a bootloader (Gummiboot), to make the system more secure of course.. Again, search it.
Magic Banana, you can say that again.
Sure: "The actual init system does not matter to GNOME or to the future MATE. systemd is an umbrella project that is more than an init system. systemd-logind is a component that manages user logins and seats and works with any init system."
In fact, it is installed by default in Trisquel (most of GNOME depends on it) even if Trisquel still has Upstart as an init system.
See by yourself systemd-logind's dependencies: http://packages.trisquel.info/belenos-updates/libsystemd-login0
According to Pottering and Kay Sievers (systemd developers), their end game is to have systemd be a complete OS.
No. They have an opinion on the design of a good operating system. Many people do. What would that even mean for systemd to "be a complete OS"? Will it soon include a desktop environment? A Web browser? Games?
Even Linus had to block all code contributions from Kay because they were trying to hardwire the kernel to require systemd.
No. The (digital) fight relates to the keyword "debug" that can be passed in argument of the kernel boot line. That keyword was supposed to be generic, i.e., it could be used outside the kernel. Kay Sievers used to it for systemd but the debug mode of systemd was logging so many lines that it could even prevent the kernel from booting. That is all (and it is bad enough!). There is no conspiracy theory "to hardwire the kernel to require systemd". The fight does not even have anything to do with Linux's code.
Linus Torvalds "[doesn't] personally mind systemd, and in fact [his] main desktop and laptop both run it": http://www.itwire.com/business-it-news/open-source/65402-torvalds-says-he-has-no-strong-opinions-on-systemd (interview given little after the fight with Kay Sievers).
Latest news, is they are working on systemd also becoming a bootloader (Gummiboot)
No. Gummiboot has absolutely nothing to do with systemd (besides sharing developers). Guess what? It can even boot Windows! The bootloader loads a kernel. The kernel loads the init (pid 1). In other words, the bootloader never interacts with the init system. Ever.
Oh Magic Banana...The research is all on the internet from people working on the projects.
Fact, Linus has blocked Kay from submitting any/all code for the kernel until further notice.
Fact, soon systemd will default to gummbiboot as the loader, it has already merged with it. When Lennart and Kay were caught trying to make it require gummi, they changed it to default but not required.
Fact, Pottering has stated on his own blog and various mailing lists that he detests Linus and the entire kernel team and has mentioned that systemd could use its own RedHat kernel.
Fact, Pottering has stated that he does not care if systemd breaks other distros, or breaks other desktop environments, he only cares that it works with Gnome and the programs they view as being key for further development.
Gentoo, Slackware and Crux are continuing to pursue non-systemd development, for how long only time will tell. Manjaro is also offering Gentoo's OpenRC with mixed results.
the BSD's are even helping them, informally, as they are of course antisystemd, since any programs hardwired to systemd will break or not be included in the BSD repos/ports.
So yes, their end goal as stated by Pottering is to have a complete CoreOS, that includes everything from the kernel (Linus' or not) up to the DE (Gnome).
It is not true that the big distros chose systemd, it was forced on them, by making many key components require it. If it is so widely accepted, then why is it ripping Debian's team apart...their board is being conveniently replaced with RedHat employees...but as you say no conspiracy, nothing to see here folks...just smile and carry on...
The BSD guys aren't against systemd. Many of them oppose copyleft, but the main reason they don't use systemd is because they can't. systemd uses Linux-specific features, so it won't work with the BSD kernel. In fact, the OpenBSD project is working to come up with kernel-agnostic replacements for certain systemd programs that they want to include in OpenBSD (e.g. initd. Unfortunately, they are also trying to make these replacements available under a pushover license, so I suppose they're reverse-engineering the programs rather than simply going in and replacing any Linux-specific stuff with generic stuff, which would be easier; and this also means that their effort unfortunately benefits proprietary software developers. In any case, this is not "antisystemd" behavior. If OpenBSD wanted the latest GNOME without systemd, for example, it would be perfectly possible to remove the dependency on systemd from GNOME (and add back in support for that obsolete program logind replaced).
onpon...not true, they are completely against it. What they are doing is making their version which would be limited to an init system and not the entire monolithic software suite that is the ever growing mess of systemd.
BSD will never go for systemd as it violates POSIX, which the BSD's (and prior to systemd GNU/Linux did as well), adhere to.
You might enjoy these threads whether you are for or against it:
https://www.linuxquestions.org/questions/linux-from-scratch-13/what-is-so-bad-with-systemd-4175500300/
https://www.linuxquestions.org/questions/slackware-14/slackware-and-systemd-885228/#post4380112
https://forums.freebsd.org/threads/systemd-is-obsolete.49372/
https://forums.freebsd.org/threads/how-is-freebsd-coping-with-a-systemd-future.46667/
http://without-systemd.org/wiki/index.php/Main_Page
http://systemd-free.org/index.php
Interesting to note that you can make Parabola systemd free as well by following the tips from the above link. Dragora also will not use it.
Debate is good, and debate is the one thing RedHat and Pottering did not allow. But then again this is what happens when people right out of college think they know better than people with 30+ years of Unix/GNU/Linux experience.
Fact, Big Foot exists. It is all over the Internet as well.
More seriously, how do you explain that Linus Tovalds "[doesn't] personally mind systemd, and in fact [his] main desktop and laptop both run it" if systemd's developers "were trying to hardwire the kernel to require systemd" (the interview was little after the fight with Kay Sievers).
I have better things to do than discussing all your fantasizing. I will just take one example: "Debian's board is being conveniently replaced with RedHat employees". Really? Let us take a look at https://www.debian.org/intro/organization then. The ten officers (including the eight member of the technical committee, which chose systemd as the default init):
- Neil McGovern works for Collabora;
- Don Armstrong is an academic;
- Andreas Barth apparently works at argh.org (at least that is the domain of its email address);
- Bdale Garbee works for HP;
- Sam Hartman works for a small consulting company, called Painless Security;
- Tollef Fog Heen resigned as a Debian systemd maintainer last year ("because the amount of crap thrown [his] way just becomes too much"), hence is out of doubt;
- Steve Langasek works at Canonnical and compares Red Hat to a purgatory (https://wiki.ubuntu.com/SteveLangasek);
- Keith Packard works for Intel;
- Didier Raboud works for LIIP (a web design agency);
- Kurt Roeckx works for E-webshops (that sells e-books).
None of them works for Red Hat... what is pretty normal if you think a little about it! So, if Red Hat is trying to take over Debian, it is doing a terrible job.
It is exhausting to talk with conspiracy theorists. In the case of systemd, they are so numerous and talkative that some users can believe their unfounded statements. Moreover, it looks like 90% of them have no idea what they are talking about. In https://trisquel.info/forum/systemd-i-old-sysv-i-hope-some-tech-distros-stay-sysv-or-bsd-init-grsecurity-vs-security-updat chaosesqueteam was writing that "[systemd] forces us all to learn upstart garbage" (good one!). The current discussion started with HuangLao trying to mock me because I wrote that systemd-logind can run on top of another init. It really does. HuangLao probably discovered at that occasion that logind was on his Trisquel system, which uses Upstart. And, yet, in a subsequent post, he describes systemd as an "entire monolithic software".
It just seems there is no way to rationally convince a conspiracy theorist. Yet, big foot does not exist.
Magic Banana,
no reason to resort to this level of communication. I did not mock you, nor make fun of you...
Merely stating, facts about systemd, how that led you to conspiracy theories and big foot is beyond me.
Regarding Linus, he said "that he does not have have a problem with systemd as an init system". He does have a problem with binary logging, systemd hiding many of its processes, and the monolithic nature of the software suite, which is no longer just concerned about being a modern init system....
He has also mentioned in interviews and workshops, that he does not care for or support any one distro. His only concern is the kernel and he uses Fedora because it is what he always used, and he does not like to change or constantly maintain systems etc... He hardly ever uses GUI anyway, and works mainly from TTY and CLI. So the distro. does not really matter much to him.
Personally, I and many other long time Unix and GNU/Linux users think that systemd was rushed, has grown very large in a short period of time. The process should have been much slower, let some of the bugs work themselves out, and it should have remained a "modern" init system, of which many people have no complaint. Also, perhaps let people choose for themselves during installation etc... do you want systemd to control everything or a combination of systemd and openRC, or retain the old SysV etc...
But...like another person said, the OP was asking about MATE DE so...perhaps this will need to become a separate post/discussion.
I did not mean to offend you! :)
- Anmelden oder Registrieren um Kommentare zu schreiben