| |||||||||||||
Quotes of the weekQuotes of the weekPosted Feb 3, 2011 2:43 UTC (Thu) by yokem_55 (subscriber, #10498)Parent article: Quotes of the week
In GNOME 3.0, we're defaulting to suspending the computer when the user shuts the lid, and not providing any preferences combobox to change this. This is what the UI designers for GNOME 3.0 want, and is probably a step in the right direction. We really can't keep working around bugs in the kernel with extra UI controls.If this cannot be disabled easily, this is a horrible misfeature. When at my laptop I often leave a streaming music station running, while closing the lid. If the machine were then automatically suspended, said music could no longer play. Inane.
(Log in to post comments)
Quotes of the week Posted Feb 3, 2011 3:30 UTC (Thu) by Cyberax (✭ supporter ✭, #52523) [Link]
Seconded. I often close the lid of my notebook as a quick-and-easy alternative to locking screen. It's fast and intuitive.
Quotes of the week Posted Feb 3, 2011 12:42 UTC (Thu) by foom (subscriber, #14868) [Link]
Huh. Maybe I just don't know what I'm missing, but I've never missed this feature running OSX (which afaik doesn't have the option to not sleep on lid-close).
It does disable auto-sleep when closing the lid while having an external display/keyboard/mouse attached (or maybe even just some subset of those, I haven't tested). That seems to cover the normal usecase for me.
Quotes of the week Posted Feb 3, 2011 12:49 UTC (Thu) by Cyberax (✭ supporter ✭, #52523) [Link]
>Huh. Maybe I just don't know what I'm missing, but I've never missed this feature running OSX (which afaik doesn't have the option to not sleep on lid-close).
And I was using Insomniax on Mac OS to get the same behavior, while cursing the stupidity of OS X designers.
Also, I quite often close the lid of my notebook to move it to another place. I certainly don't want it to hibernate every time.
Quotes of the week Posted Feb 3, 2011 13:41 UTC (Thu) by fb (subscriber, #53265) [Link]
> Also, I quite often close the lid of my notebook to move it to another place. I certainly don't want it to hibernate every time.
I do the same, specially when I need to carry something else.
I always wished there was a configuration setting such as "suspend if closed for more than 60 secs".
Quotes of the week Posted Feb 3, 2011 16:27 UTC (Thu) by sbohrer (subscriber, #61058) [Link]
At my office we all have Mac laptops most running OS X. It is very common to see people wandering the halls with their laptop deliberately half open to prevent it from sleeping.
Since I run Linux on mine I've always been able to happily close my laptop as I walk back to my office.
Quotes of the week Posted Feb 3, 2011 18:14 UTC (Thu) by ikm (subscriber, #493) [Link]
> Maybe I just don't know what I'm missing, but I've never missed this feature running OSX (which afaik doesn't have the option to not sleep on lid-close).
The difference is that OSX does this flawlessly, while with Linux chances are high something is screwed after you resume (if you manage to resume at all). Hence the original GNOME posting about not providing an option despite the need.
Quotes of the week Posted Feb 3, 2011 19:09 UTC (Thu) by mjg59 (subscriber, #23239) [Link]
I strongly disagree. The majority of machines now have reliable suspend/resume under Linux. If yours doesn't, that's a bug in the kernel. Should userspace add workarounds for users to avoid kernel bugs?
Quotes of the week Posted Feb 3, 2011 19:19 UTC (Thu) by ikm (subscriber, #493) [Link]
I don't know which majority you are referring too - it certainly does not include any of my laptops. Answering your question: yes, userspace should add workarounds if it is supposed to be used by an end-user (as opposed to early adopters and alpha testers).
Quotes of the week Posted Feb 3, 2011 19:24 UTC (Thu) by mjg59 (subscriber, #23239) [Link]
Every laptop I've tested in the past two years. Have you got a bug open for the ones that don't work?
Quotes of the week Posted Feb 3, 2011 19:28 UTC (Thu) by ikm (subscriber, #493) [Link]
No, I have other stuff to do in my life than tracking obscure issues. And I have had enough of this - the next laptop of mine will be a Mac running OSX.
Quotes of the week Posted Feb 3, 2011 19:32 UTC (Thu) by mjg59 (subscriber, #23239) [Link]
Expecting software designers to cater for bugs that haven't even been reported is pretty unreasonable.
Quotes of the week Posted Feb 3, 2011 23:44 UTC (Thu) by Cyberax (✭ supporter ✭, #52523) [Link]
My laptop can't hibernate - it just doesn't resume. Is it a userspace issue or kernel space? How to debug it? Frankly, I don't care.
My another notebook with crappy Poulsbo video card fails to resume from suspend-to-RAM in about 10% of cases. Oh, and it uses a proprietary driver. You do accept kernel bugs for proprietary drivers, don't you?
This change is nuts. There are good use-cases for doing just nothing instead of inventing 'solutions' aimed to work around the bugs in certain hardware.
Quotes of the week Posted Feb 4, 2011 12:41 UTC (Fri) by rahulsundaram (subscriber, #21946) [Link]
If you don't care, ignore it. If you care, file it with the hardware details, kernel version and let the developer ask for more information if necessary.
Quotes of the week Posted Feb 4, 2011 12:44 UTC (Fri) by Cyberax (✭ supporter ✭, #52523) [Link]
I've had experience with an attempt to post a bug for a tainted kernel. I don't want to repeat this experience.
Quotes of the week Posted Feb 4, 2011 14:39 UTC (Fri) by rahulsundaram (subscriber, #21946) [Link]
Easy enough to see whether any proprietary drivers are resulting in the breakage by disabling them temporarily. If the problem is still there, report to LKML. If not, consult with the proprietary driver vendor.
Quotes of the week Posted Feb 4, 2011 14:45 UTC (Fri) by Cyberax (✭ supporter ✭, #52523) [Link]
No, it doesn't happen. However, there's no Compiz and OpenGL running and I'm fairly sure it's caused by Compiz.
Quotes of the week Posted Feb 4, 2011 15:51 UTC (Fri) by rahulsundaram (subscriber, #21946) [Link]
Then file it against Compiz. I don't see the reason to resist doing that if you have already figured out the source.
Quotes of the week Posted Feb 5, 2011 0:58 UTC (Sat) by giraffedata (subscriber, #1954) [Link] Expecting software designers to cater for bugs that haven't even been reported is pretty unreasonable. True, but then ikm never said he expects software designers to address these bugs. I don't report bugs. The probably of the report resulting in a fix is too small to justify my time. I just accept my loss and look for the next best alternative. I'm cognizant of the fact that if the bug affects only me, there's a vanishing chance that some developer working for free will be interested in fixing it, and if it affects a lot of people, someone else will have reported it or, better yet, some developer with the means to fix it will have experienced it himself.
Quotes of the week Posted Feb 5, 2011 1:03 UTC (Sat) by mjg59 (subscriber, #23239) [Link]
At this stage, if you have a machine with an in-kernel graphics driver and it fails to suspend or resume then there aren't huge numbers of people reporting the bug. Really. Suspend/resume is part of the certification process for RHEL now and I get all the related bugs. There simply aren't that many, and in general it's something like hibernate failing because someone's swap partition vanished. I really would encourage you to report any suspend/resume issues to bugzilla.kernel.org and the massive probability is that it'll be tracked down quickly.
Quotes of the week Posted Feb 5, 2011 1:05 UTC (Sat) by ikm (subscriber, #493) [Link]
Even if there's an NVIDIA blob involved?
Quotes of the week Posted Feb 5, 2011 1:09 UTC (Sat) by mjg59 (subscriber, #23239) [Link]
If you can repeat the bug with nouveau, then file it. If not, file with nvidia. I'm afraid that I work for a free software company, so we tend to make decisions based on the code that we actually ship.
Quotes of the week Posted Feb 3, 2011 19:38 UTC (Thu) by boudewijn (subscriber, #14185) [Link] Have fun then... I had one for a year, in 2010, and I had my share of having to track obscure issues. When I got it there was a hair wedged between backlight and pixel panel. When I gave it back a year later, there were random vertical blue and purple stripes over the screen, unless I pressed the enclosure in exactly the right way. It wouldn't come out of sleep about once a week (meaning about 10% of the times I put it to sleep). I wouldn't go to sleep, too, and then get really hot in my backpack. And one day sound broke. Then there was an Apple OSX upgrade that forced me to reinstall. Poof! went my development environment. And I'm fine with Safari wanting a holiday on the beach, but I got so regularly beach-balled that I felt a proper outsider who had applied for a posh London club. Watching a dvd was extremely annoying, because unlike dragon or xine, I couldn't skip the mandatory warnings at the beginning of the dvd. And in that year I've seen the gray kernel panic screen in a veritable babel of languages at least five times. I don't think there's any platform, any computer system, that works as advertised, without grief, without bugs, without hardware problems, without obscure little issues you either learn to live with or have to track down.
Quotes of the week Posted Feb 3, 2011 20:32 UTC (Thu) by jmitchel (guest, #11611) [Link]
I missed it. The combination of cats running around and an open laptop is a bad one.
Quotes of the week Posted Feb 3, 2011 3:32 UTC (Thu) by djm (subscriber, #11651) [Link]
I thought the "developers know best" idiocy peaked with GNOME 2.0, but wow was I wrong. This certainly clarifies which desktop environment I won't be using in a couple of years.
Quotes of the week Posted Feb 3, 2011 4:46 UTC (Thu) by linuxjacques (subscriber, #45768) [Link]
I agree. I have no idea who Gnome is targeting these days.
Quotes of the week Posted Feb 3, 2011 5:25 UTC (Thu) by jmspeex (subscriber, #51639) [Link]
Gnome is targeting grandmas and they'll keep doing that until the only user left is the developer's grandma.
Quotes of the week Posted Feb 4, 2011 4:50 UTC (Fri) by djm (subscriber, #11651) [Link]
So grandmas are the key to the coming decade of the Linux desktop? I never figured...
Quotes of the week Posted Feb 6, 2011 3:44 UTC (Sun) by nicooo (guest, #69134) [Link]
Maybe not the key, but still a significant role:
http://linuxgrandma.blogspot.com
She doesn't use Gnome ;)
Quotes of the week Posted Feb 4, 2011 17:28 UTC (Fri) by MisterIO (guest, #36192) [Link]
I guess they thought the spatial mode for nautilus wasn't enough, so now they've come up with this. (I mean removing the option from the ui, not the choice of the default, which I think is reasonable)
Quotes of the week Posted Feb 3, 2011 6:26 UTC (Thu) by amaranth (subscriber, #57456) [Link]
Well in GNOME 2.x your music player can inhibit suspend so I see no reason why it wouldn't be able to in 3.0, you just need a player that supports this.
Quotes of the week Posted Feb 3, 2011 11:43 UTC (Thu) by jospoortvliet (subscriber, #33164) [Link]
That is equally stupid, I often want my laptop to suspend when I close the lid. Why not let me choose the conditions so it is predicable and works how I want it? My laptop has no multimedia keys so turning off the music first takes some effort - and when I close the lid of my laptop, I often do it for a good reason - to QUICKLY suspend-to-ram because I'm interrupted by someone talking to me or something like that! This would make me have to tell them to shut up so I can figure out how to stop my music player THEN close the laptop lid...
Quotes of the week Posted Feb 3, 2011 14:08 UTC (Thu) by __alex (subscriber, #38036) [Link]
This is only a problem if your music player doesn't allow you to inhibit suspend.
Quotes of the week Posted Feb 3, 2011 14:55 UTC (Thu) by jospoortvliet (subscriber, #33164) [Link]
No, it's only a problem if your musicplayer DOES inhibit suspend. I don't want it to - I want my computer to act predictable and how I want it... So when I configured it to suspend when I close the lid, it should suspend. If I don't want it to do that I'll put it in a power management scheme which won't suspend the lid if I close it. I don't want apps meddling with that. Sure, a app should suspend screensaver or suspend when I am watching a video - but when I close the lid, the darn thing should OBVIOUSLY go into suspend...
The whole idea of building heuristics (which will always fail in many situations) for this is stupid, sorry. Good defaults - of course. Trying to out-smart the user in a way that will never fully work and then not allowing the user to fix it - braindead.
Quotes of the week Posted Feb 6, 2011 23:15 UTC (Sun) by gmaxwell (subscriber, #30048) [Link]
Yea, freeking fantastic there. So you've turned the volume (almost) all the way down. A half hour later you toss your laptop into your bag, expecting it to suspend because thats what it always does, latter you open your bag to find that your laptop has become a wad of melted plastic.
Gnome needs to fire all its usability people because they don't seem to understand this simple relationship:
"Always exactly the behavior I wanted" is better than "deterministic but not always what I want" which is better than "sometimes not the behavior I wanted and not deterministic".
One beautiful thing about computers is that they can be pretty deterministic. If something is not always right, but at least consistent we can work around its idiocy. If it's not deterministic then we have to constantly check that it's not screwing us over.
It is usually not a usability improvement to make the machine right a little more often but far less deterministic. A laptop that always suspends on lid closed or never suspends on lid close is not a fire-hazard. Making that depend on an external monitor being connected is probably safe as long as the detection is _very_ reliable (and consistently wrong when its not reliable). Checking to see if a music player is running? I think thats going to ruin a lot of hardware.
Quotes of the week Posted Feb 10, 2011 10:46 UTC (Thu) by __alex (subscriber, #38036) [Link]
I didn't say "if your music player inhibited suspend" I said "if your music allowed you to inhibit suspend." The implication is of course that there would be a checkbox in preferences that would let you manage the behavior.
Generally laptops have thermal cutouts so I'm not sure what this firehazard stuff is a bit silly.
I would prefer to have customizable (yes hahaha this is Gnome, they don't do customizable) automatic suspend behavior over having to manually suspend it myself or having it automatically suspend whenever I close the lid.
Perhaps some sort of global "inhibit suspend" option in the power management applet would accomplish this in a more obvious way than making it a per-app configurable.
Quotes of the week Posted Feb 3, 2011 18:25 UTC (Thu) by iabervon (subscriber, #722) [Link]
You know, the GUI necessary to make it possible to create both your desired configuration and the default policy is way beyond GTK's dynamic form capabilities. Since you've got specific desires that you can explain in detail, you'll surely do better with gconf than any possible dialog box. The best configuration system, in my opinion, is to give the user the choice of a few provided policies (for people who don't want to think about configuring this particular aspect) plus the ability to use a user-generated policy (for people who want to come up with one). Madness lies in trying to provide users who think through their policies with a simple choice of the policy they'd come up with.
Your laptop doesn't have any multimedia keys? All of the computers I've bought have at least had a Pause/Break key. Of course, it took a certain amount of cleverness to get it to actually control anything.
Quotes of the week Posted Feb 3, 2011 20:41 UTC (Thu) by jmitchel (guest, #11611) [Link]
Sure... gconf. When I was young and never left the house, I could be bothered to figure out what random incantations I needed to make things work, or just ignore the problems. Back then I had Linux on everything. Now I just want my computers to work, for features I counted on last year to stay where I found them, etc...
Quotes of the week Posted Feb 3, 2011 21:27 UTC (Thu) by foom (subscriber, #14868) [Link]
If you want nothing to ever change, then just never upgrade anything. I know someone who is still running RH6 on a 10-year-old machine for just that reason. :)
On MacOSX there is almost a whole industry devoted to making fancy GUIs to expose hidden preferences only available from command-line options. That seems like a pretty sensible way to do things: keep the default GUI simple, and if someone wants some advanced feature, let them either use the commandline or download an advanced-gui to do it for them.
Quotes of the week Posted Feb 4, 2011 5:32 UTC (Fri) by jospoortvliet (subscriber, #33164) [Link]
Well I still prefer if developers try to build a GUI that can do both. Sure, it's more work, sure it is harder - but if you succeed you got really a superior product. Some applications manage to do this - a great example is gwenview 4.x, which is 1000 times simpler than its 3.x version but has actually MORE features. It's possible ;-)
Something like GConf is just an abomination, imho. KDE has something like that kconfigeditor I believe - blegh. I must say I like recent ideas to build in search in configuration interfaces, let them be build dynamically etc - that might be very helpful.
Finally, why install a separate app or tool if you could have an advanced button in a normal gui hiding the complexity but having it available for those who need it?
Quotes of the week Posted Feb 4, 2011 11:49 UTC (Fri) by Felix.Braun (subscriber, #3032) [Link] Something like GConf is just an abomination, imho. KDE has something like that kconfigeditor I believe - blegh. I must say I like recent ideas to build in search in configuration interfaces, let them be build dynamically etc - that might be very helpful. Have you used gconf-editor recently? I find it's a decently discoverable GUI for so-called "hidden" preferences. Unlike Windows RegEdit it even explains what every switch does and which possible values it has. IMHO this is equivalent of having an "Advanced" button in every GUI. They are just all kept in a central place. Of course I don't know whether the old gconf-editor will be ported to the new dconf world. But I suppose it will get written if there is a demand.
Quotes of the week Posted Feb 4, 2011 12:39 UTC (Fri) by rahulsundaram (subscriber, #21946) [Link]
dconf-editor already exists.
Quotes of the week Posted Feb 4, 2011 13:21 UTC (Fri) by Lennie (subscriber, #49641) [Link]
I wonder why anyone would create a KDE application which starts with a g. ;-)
Quotes of the week Posted Feb 4, 2011 11:47 UTC (Fri) by Cyberax (✭ supporter ✭, #52523) [Link]
So map the "close lid" action to suspend. It's certainly possible right now (and AFAIR it's the default).
Personally, I've mapped the "Power" key to suspend.
Quotes of the week Posted Feb 3, 2011 7:49 UTC (Thu) by jmorris42 (subscriber, #2203) [Link]
> If this cannot be disabled easily, this is a horrible misfeature.
Amen! I leave terminals open to remote machines so I set mine to only suspend if the lid closes while on battery power. An idling laptop with the power management working right doesn't draw that much electricity so I don't really care if it is still running if I close the lid and sit on on a table beside my chair in the living room for an hour or two.
And yes I know all about screen but reconnecting is a lot more work than just opening the lid and having a quick at progress on a long running job or checking something. And screen disables scrollback.
I really don't know who GNOME's target audience is, but it clearly isn't me. And do they really think they have a shot at general audiences in this age of Android and Meego? On the off chance an ODM were to allow a machine to escape into the wild with Linux, they will stuff one of those on as teh desktop. Just hope their rot doesn't go all the way up the stack to the desktop plumbing in freedesktop.org's stuff because about the only way to avoid their alterations to the *NIX Way is *BSD and I'm not too sure how long they will hold out.
Quotes of the week Posted Feb 3, 2011 9:41 UTC (Thu) by Klavs (subscriber, #10563) [Link]
>And screen disables scrollback.
Nope. It just maintains it's own for obvious reasons.
ctrl+A esc - and you can use the arrow keys and pageup+down to go through your screen history (and the length of it is ofcourse configurable).
Quotes of the week Posted Feb 3, 2011 9:43 UTC (Thu) by Klavs (subscriber, #10563) [Link]
and to comment on the actual article :) - I've been using KDE for a looong time for exactly its lack of configurability. Gnome is definetely not targetting me :)
"screen disables scrollback" Posted Feb 3, 2011 9:59 UTC (Thu) by marble (guest, #2719) [Link]
Edit your /etc/screenrc (or .screenrc) and add:
termcapinfo xterm|xterms|xs|rxvt ti@:te@
If you're using Ubuntu, that line is there by default, but commented out.
Random GNOME joke Posted Feb 3, 2011 12:48 UTC (Thu) by ballombe (subscriber, #9523) [Link]
At least, it states clearly that GNOME 3.0 is targeted only at UI designers.
Quotes of the week Posted Feb 3, 2011 13:10 UTC (Thu) by jengelh (subscriber, #33263) [Link]
This is all very reminiscient of http://tech.slashdot.org/story/07/02/17/0219225/Godwins-L... .
Quotes of the week Posted Feb 3, 2011 18:12 UTC (Thu) by Frej (subscriber, #4165) [Link]
I think rhythmbox has (had?) an option/plugin to disable suspend when closing the lid.
suspend blockers? Posted Feb 5, 2011 17:02 UTC (Sat) by dmk (subscriber, #50141) [Link]
We need suspend blockers!
|
|||||||||||||