Default desktop in 7 still looks horrible
I tried out the recent build of Trisquel 7 today, and the default desktop environment looks very amateur. The panels look outdated and the default fonts are ugly to look at.
You figured that things could get better than 6 considering the new libraries, but Ruben either doesn't care or lacks the artistic flair to do better.
Please don't be like other free software projects that are content with ugly just because its free. Any sane person runs the netinstall with the Cinnamon PPA if they want a classic desktop: http://www.omgubuntu.co.uk/2014/07/new-cinnamon-ubuntu-14-04-ppa-stable
Care to post a few before and after screenshots? I believe Cinnamon requires hardware acceleration, right?
I think it looks the same as Trisquel 6.
You can do software. I have an Nvidia card that works with free drivers and has 3D acceleration, so this isn't a problem: http://h-node.org/videocards/view/en/232/NVIDIA-Corporation-GF108--GeForce-GT-430---rev-a1-
With Belenos still being in alpha I expect the theme has yet to be completed.
Ruben has had way over a year to fix the theme with 6. It seems like he is going to leave it as is without any effort. What makes you think he is suddenly going to revamp the interface when the OS is already lagging behind in its release.
It sucks, considering there are improvements to the fallback session in Ubuntu 14.04 according to http://www.omgubuntu.co.uk/2014/04/ubuntu-14-04-classic-gnome-flashback-session.
This is why I don't recommend using the stock ISO images for Trisquel to anyone. Just start with a clean netinstall and add either Cinnamon, MATE, Gnome Shell, or XFCE. Heck, the Elementary OS desktop can be installed on Trisquel 6 if you have the PPAs.
Netinstall? I prefer debootstrapping :)
Having said that I think it'd be great if Cinnamion is the default DE in Trisquel, just open up a new repository for trisquel (other than edition, edition-updates, and edition-security) and title it belenos-Cinnamon!
As for hardware acceleration I think the question becomes pointless with the existance of Trisquel LXDE.
Or another solution would be MATE+Compiz, but just like in this shot:
Considering MATE's and Trisquel's GNOME 2 past, I doubt that'd be a problem.
Cinnamon, on the other hand, has got a GNOME 3 past, which is what Trisquel currently has got so I don't expect it to be a problem either......I'm not an expert, though.
Their new function (applets, panels, docks etc) would have to have it enabled/coded in. Gnome 3 had a special fundraiser and year of accessibility.
Historically Accessibilty has been one of the key factors in Rubén's decisions. Support for ATI/AMD users with no hardware 3D is another. However, as I've said elsewhere on the forum when the community repos arrives then logically it might be possible to have remixes. Only we'd need more devs and one to be interested for that to happen.
The GTK+ theme there was supposed to show "Trisquel" ('s current default theme) - Unfortunately it got covered in the screenshot by a menu.
Oh, and the shot actually shows Trisquel Dagda, in all of GNOME 2's glory.
And I meant MATE+Compiz+Docky/AWN/Cairo-Dock(GLX-Dock)
I think the beauty is important too, because it can show to new people that a truly free system can be wonderful in a glance.
Although is important to worry about desktop environments that require too much memory: many people outside Europe and North America still have computers from five to ten years ago. that's why there is a LXDE version.
I know that this topic come and goes once in while, but my vote would be to using MATE or XFCE as the default, and leaving LXDE behind.
Both MATE and XFCE are lightweight enough without the ugliness and inconsistence of LXDE. Both of them are based in GTK+2, so they have all the legacy of hundreds (maybe thousands) of beautiful themes easy to change. And they are consistent because they don't rely much on applications from other desktop environments.
Beyond that, the new Xubuntu 14.04 is outstandingly beautiful with his elemenary-os feeling. http://xubuntu.org/screenshots/
Even if is not choose, it would be nice to have an official way to implement the Xubuntu configurations at netinstall.
I still think Trisquel should use GNOME Shell by default and offer LXDE (Trisquel Mini) for those who want a lightweight system. The big reason GNOME is being used is accessibility, right? Wouldn't GNOME Shell be better for that? I guess it might be just knowing how to use it, but it seems to me that GNOME Shell has better keyboard support than GNOME Flashback.
The worst case with using GNOME Shell I can imagine is that people notice that the system is slow; it's really not that big of a deal to tell those people that Trisquel Mini is a better choice for them. Even if you need GNOME's accessibility and GNOME Shell doesn't run very well on your system, couldn't you just install GNOME Flashback yourself, after installing the GNOME Shell version and dealing with a little slowness for a short while?
I am a member!
@onpon4 I think the main reason GNOME is still used is the accessibility. GNOME Shell's accessibility is much improved since version 3.4 (as in Trisquel6), so, Shell with llvm pipe, or something, for accelleration should, perhaps, be the new default. Shell is slow, especially with accessibility enabled. Since I feel I don't benefit from Shell's overhead, I have installed Mate on a machine running Arch. Though Mate's accessibility isn't yet as mature as that of GNOME, it's better than that of XFCE and LXDE. I have no experience with Cinnamon, though, have heard, for a long time, that it is not accessible. Elsewhere in this forum, I've mentioned accessibility regressions that have cropped up in Trisquel 7 Alpha. They don't have to do with the desktop, rather, with a few of what I'd consider "core" applications. Maybe these are up-stream? Though, they haven't been reported in Ubuntu 14.04.
Maybe there should only be one Trisquel iso. The installer would automatically check the hardware for certain criteria such as RAM, availabilty of 3D acceleration, etc..., and automatically install the 'right' desktop: a basic one for machines which are either too old, or not free-software friendly (AMD/ATI cards), and a full-fledged one for the users who have a modern machine and the right kind of graphic card. The first option could be LXDE, Xfce or GNOME Fallback/Flashback (whatever it's called). The second option could be GNOME Shell or KDE.
I prefer GNOME Classic Mode from 3.8+ than Cinammon. RHEL 7 uses GNOME 3.8 Classic by default:
I like the default desktop, at least after I've changed the background image to one matching my screen resolution. It's clear, clean and simple.
I think that the current look, the gtk theme is really beautiful. The icons in the indicator panel applet changed and now they have a modern look. Probably with an update of some other icons, of the default wallpaper and perhaps the panel's default color, Trisquel 7 will be really perfect. Furthermore the greeter has a sleek look.
I am a member!
My preference would be for Gnome 3 Shell to be the default. Remember there is such a thing as software rendering (llvmpipe).
Users who either don't like or can't run Gnome 3 shell (low ram, etc.) can use Gnome Flashback.
If we must also provide a "mini" edition (why?) then something like LXDE or XFCE would seem appropriate. However, I can't see why this is different to or better than Gnome Flashback, tbh.
Because it uses LXDE, Trisquel-Mini requires very little RAM. 256 MB is sufficient. I doubt GNOME Flashback (with some heavy applications running, say Abrowser + LibreOffice) would properly run with less than 768 MB.
I am a member!
Ok, I guess that is a benefit.
But really, does anyone actually run a system with <1 GB of RAM nowadays? RAM is currently cheaper than it's ever been.
I understand the desire to cater for everyone. However, the way this discussion is going we're at risk of stagnating indefinitely due to the fact that some people continue to run legacy hardware. We can't let that fact hold back progress forever.
RAM is cheaper, but if you've got an old PC lying at home and don't know how to fit a new one or just have got more of a desire to look for a low-resource OS than to actually go to a computer store.....
And then there's also the fans of HP TC1100 computers which has a limit on RAM upgrades as far as I know. Not to mention the original Compaq TC1000 with an even lower limit plus an effingly slow Transmeta Crusoe processor.....
And I think a spin of Trisquel isn't as hard a job as, say, maintaining the repos......
Of course, I'm commenting about the Mini edition. As for the main edition, I do think Trisquel holds back on that front a bit too much, especially with the existance of Trisquel Mini.
I agree: I would like the "normal" edition to use GNOME Shell. I know how to get it (two packages to install). I am more thinking of new users for whom the defaults matter.
Trisquel Mini has its users. I have more doubts regarding Trisquel Sugar or even Triskel (although I believe a separate developer takes care of it what means no additional burden on quidam).
I think Cinnamon would be better. Or GNOME Classic but perhaps with the panel on top, Cairo-dock on the bottom setup.
Vanilla GNOME Shell would take new users for too much of a spin, I think. Though if Trisquel puts a welcome screen with a manual on how to use the desktop in it put in a detailed and attractive way I guess it'd be OK.
http://www.omgubuntu.co.uk/2014/07/install-budgie-evolve-os-desktop-ubuntu-14-04
There's this new DE from the guy that created (and abandoned) SolusOS and the Consort desktop. It looks promising and it clearly draws inspiration from ChromeOS.
Of course its still very early and the author is known to abandon projects in the past. Will he stick with this one or get bored?
I think Triskel is basically abandoned. It's just a metapackage with no custom settings. If you install the triskel package, the package manager doesn't even work because polkit-kde-1 isn't installed. Also, until I submitted a patch, Abrowser looked really ugly in KDE with the default Oxygen because the package that enables GTK+ programs to use Oxygen handles a list of XUL programs specially, and "abrowser" wasn't in that list. (Luckily, "icecat" is.)
Kubuntu has different settings (default, netbook, "low-fat", etc.) and has a few other customizations that make it more usable (such as "Notification Helper"). I really think the kubuntu-* packages (which are in the repos) should be Triskelized.
The polkit-kde-1 thing was fixed in 6.0.1.
I test installed Triskel 7 today. Some rough edges but it seems to
have inherited pretty directly from Kubuntu this time e.g. Amarok is
the default audio player and there's an empty cruft Firefox short cut
in launcher favourites. IMO it's still very usable though.
Leny
>The polkit-kde-1 thing was fixed in 6.0.1.
Cool. Then who can close the ticket I opened on that very subject?
>>The polkit-kde-1 thing was fixed in 6.0.1.
>
> Cool. Then who can close the ticket I opened on that very subject?
>
> https://trisquel.info/fr/issues/10293
I didn't know about that one. I've emailed SirGrant who is able to
close Issues.
Actually, the 6.0.1 package doesn't depend on polkit-kde-1, but the 7.0 one does. So I don't think it's fixed in 6.0.1 yet.
> Actually, the 6.0.1 package doesn't depend on polkit-kde-1, but the
> 7.0 one does. So I don't think it's fixed in 6.0.1 yet.
I did a Triskel install from the 6.0.1 netinst and didn't need to put
polkit-kde-1 on. It might be that the actual package dependencies
haven't been fixed. However, as the netinst is the official media
from which one would install Triskel I think it counts as fixed.
Leny.
It should still be added as a dependency. I installed from the regular Trisquel iso and installed the triskel package. (I have both GNOME and KDE)
> It should still be added as a dependency. I installed from the regular
> Trisquel iso and installed the triskel package. (I have both GNOME and
> KDE)
Rubén said in the developer meeting where Triskel was dropped that
defect support was now limited to the packages on Trisquel install
media, all others (apart from GNU FSDG bugs obviously) were to go
upstream. So it's a reasonable inferrence from known policies that
fixing subsequent Trisquel install media will have to do.
While I agree that all bugs /should/ be fixed it's a question of
resources limiting what can be, Kubuntu who have more resources dealt
with the absence of polkit-kde-1 in a subsequent release AFAICT. No
distro has the resources to fix all the bugs in any given release in
that release, so a fix in a later release is the norm for lower
priority bugs. The bug is only an inconvenience and therefore low
priority as the bypass/resolution is written all over the web on a
search of the symptoms.
To personalise the above - what is a better use of the time you
personally put into writing patches - fixing this and other Trisquel 6
bugs, or installing and testing the alpha by making it your daily
system then fixing any bugs you find in that? Your R6 fixes would
have to be forward ported so the answer is surely as we want each
release to be better than the last the more testing and fixing the new
release gets in development the better. In a case of severely limited
resources as with Trisquel there's a strong argument for 'all able
hands to the pumps' is there not?
Leny
... As in fix R6 bugs in the R7 alpha. we can always tell a user to
move to the latest release.
Leny
BTW, most patches have very little to change for the next release (or are unneeded), especially ones that just import from upstream, and most of my patches just import from upstream.
I don't know what you're talking about. The triskel package isn't from Kubuntu, so there's no upstream. It's one of Trisquel's own packages, and therefore it's Trisquel's responsibility to fix it.
I don't know if you're familiar with Debian packaging, but triskel is a metapackage, which means that it just depends on other packages. It takes literally one second to edit the debian/control file and add polkit-kde-1 as a dependency.
EDIT: The reason that bugs in Trisquel never get fixed is that very few people ever submit patches. You can check the trisquel-devel mailing list. This month, I submitted the only patch. Last month, mejiko and I submitted patches. I don't really blame people for not submitting patches, because nobody looks at them or the issue tracker (which is pretty unorganized), which probably discourages people. Of the 2 patches that I submitted that were ever committed to the package-helpers repo, both of them took a really long time to get committed.
And even if a bug gets fixed upstream, it's never imported from upstream. Just take a look at the issue tracker. There's a lot of bugs with status "patch (ready)" that were fixed in Debian from months or even years ago, and that still haven't been fixed in Trisquel (importing from upstream doesn't take a lot of effort, by the way).
The mention of Kubuntu was to indicate that a distro with a lot more
developer resources took ages to fix the same problem with their
kubuntu-desktop metapackage (and haven't backported it either). If
they are 'lack lustre' in fixing this sort of thing in a timely
fashion what is a 'one [part time] man and his dog' distro like
Trisquel going to be like?
So to restate: Yes, agreed 'should be fixed' all bugs 'should be
fixed.' But this is a case of the inelasticity of finite resources.
We got fix attempted in the next ISO release and then because there
was limited testing of it (mia culpa) not the 'correct' fix until the
release after that. The latter just like the equivalent upstream bug
seems to indicate. As it stands the bug is confirmed fixed
'properly' in a subsequent release of Triskel, which is 'good
enough.' Matching Kubuntu's performance on a similar bug in their
distro seems to be good work for a one part time dev distro to me.
I'll point out one line bugs still need diagnosing and the fix
building and QA/testing neither of which are necessarily in
proportion to the number of lines changed. If it was as simple as a
small source change, which many bug fixes are, then there'd be a lot
more bug reports with patches attached.
Leny
It was not fixed in the iso release. The actual problem is that, if you have Trisquel GNOME installed, you have policykit-gnome-1, which apper/muon are supposed to work with (but don't). Apper/Muon depend on either policykit-gnome-1 or polkit-kde-1, so if you already have policykit-gnome-1, it won't install the required polkit-kde-1 package. Since you did a netinstall, it didn't have either package and preferred to install polkit-kde-1 so it worked. The bug was never fixed in Toutatis.
I get the impression that most patches I submit to trisquel-devel are ignored. I understand that testing is important and takes time, but nobody even replies in the mailing list.
This is also a side effect of Trisquel being based on Ubuntu, because Ubuntu rarely updates their universe packages. I think Trisquel should be based on Debian.
EDIT: And let's end this discussion. It's going too far to the right :P
Anyway, if Trisquel is so concerned about Accessibility I think it's a bit of a waste that the main site doesn't market it as such.
Which is to say from Belenos onwards I think Ruben should make a bit more of an effort to market Trisquel as such.
This is a good point. The word accessability doesn't currently appear on the front page.
We don't have much about accessibility in the documentation either
Actually, to be fair, nothing on the front page really promotes Trisquel, I think. The closest thing to it is the description that Trisquel is a "fully free operating system" which I reckon I'd find uninteresting were it not for my interest in going fully free.
I reckon it'd be more attractive if the front page greeted visitors with big letters reading:
"Computing today is an integral part of civilisation. People use it everyday, promised that they get their freedom and privacy.
HOWEVER, we get neither. We are restricted when it comes what we can do with software. We can't study it, or even use it on more computers than its author allows us (i.e. Microsoft Office).
WE DON'T EVEN OWN IT. Under the letters of the law we're only licensed to use it. We are fooled into thinking that we are buying software, when in reality we're only allowed to, in many ways, HIRE them.
All the while companies do so while spying on us - and at the same time giving a misleading "privacy statement" (the reality being the statement of A LACK THEREOF), and then demanding their rights without fulfilling OURS.
Thus, we in the community of Trisquel users use a different kind of operating system. An operating system that solves those problems. An operating system where the user OWNS, not hire, the programmes. Where the user can study and modify it, or hire someone to do so. Where schools can customise it to suit their needs.
Or supermarkets, or auto dealers, or the White House, or the computer you use.
We are users of Trisquel, a Distribution of GNU + Linux. Join us, our fantastic community, and most importantly, our fantastic operating system.
One that offers not only freedom, or privacy. We also offer ease of use, and accessability-helping programmes for those that need them, and a simply beautiful working space.
And choice, and freedom from registration, or from Trisquel spying on you.
Yes, all those is yours. You OWN it, not HIRE it. We enjoyed making it. We hope you enjoy using (and owning) it as well"
Take a look at a video about the Golden Circle http://www.youtube.com/watch?v=l5Tw0PGcyN0 (basically ads should start with why the thing being advertised exists, and end with what the thing actually is) I reckon someone can manage a better thing than what I wrote up there.
And an ad I saw about (Google's?) Project Ara was also pretty attractive.
You are wrong. We don't own Trisquel as it is the copyrighted work of Ruben. Sure, we have the freedom to run, modify, and redistribute, but Trisquel uses packages from Ubuntu which in return is based off of Debian.
Well yes, the own v. hire thingy needs to be replaces with something a bit more legally true, if that's possible (thanks for noting that!), but at least you get where I was trying to go with my text....
I am a member!
@davidnotcoulthard This is a good start on the new front page text. I like how you distinguish Trisquel from the proprietary alternatives with which folks are more familiar. It may need some work on distinguishing Trisquel from the other GNU/Linux distributions. Every distro's front page I've read, so far, says basically the same nice things about libre software. Regarding marketing Trisquel's accessibility, it should be a little easier to do in 7 than it was in 6. Within about a year of its release, when I tried to promote Trisquel among my fellow GNU/Linux accessibility users, I was met with "this is a throw-back" and "why no accessibility in net install and mini", to name a couple of typical responses. Just curious, has anyone tested the screen magnification, on-screen keyboard, and high-contrast options? What of braille support? I can only test orca screen reader's synthetic speech output. Incidentally, the Debian net install has had speech and braille support available since Squeeze, by means of the espeakup screen review package for the console and the brltty daemon. There are two GNU/Linux distros pushing accessibility, they are Sonar GNU/Linux asnd Vinux. Both have non-libre parts, and won't be linked here. Since Sonar is based on Manjaro, wqhich is based on Arch, it's nearly bleeding-edge. The blind users I know best, prefer up-to-date packages to full freedom. .
t3g said:
I tried out the recent build of Trisquel 7 today, and the default desktop environment looks very amateur. The panels look outdated and the default fonts are ugly to look at.
You figured that things could get better than 6 considering the new libraries, but Ruben either doesn't care or lacks the artistic flair to do better.
No disrespect to Ruben and all the trisquel users, T3G has a strong valid point.
Lets consider an important factor, that always has been a burden to many, that without a proper hardware acceleration (video card) it looks very amateur. Even any good Intel CPU's with enough cached Memory, the GL drivers will do a pretty good job BUT THAT IS NOT ENOUGH.
when it comes to the DEFAULT page, yes t3g HAS A VALID POINT..IT LOOKS SHABBY, A BETTER PROFESSIONAL LOOK SHOULD BE IMPLEMENTED AND NOT SOME QUICK JOB FOR SUCH ILLUSTRIOUS DISTRO.
respectfully