The Linux Questions Thread: a bunch of pitfalls, but technically it's possible

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > The Linux Questions Thread: a bunch of pitfalls, but technically it's possible

«‹›1028 »

KozmoNaut: Apr 23, 2008; Happiness is a warm
Turbo Plasma Rifle

peepsalot posted:

How do I get google drive to sync my files because google hates linux.

Also why does google hate linux?

Celexi posted:

I have been using insinc as i have a bunch of space in google drive and not much in dropbox. it does work fine, unfortunately its not free

I've been using Insync for a while, and it simply works great. It seamlessly converts Google Docs files to and from OpenDocument formats when you sync, and it integrates well with Dolphin, Nautilus, Nemo, Thunar etc.

It's a one-time $15 payment for the normal license, it's not that expensive in the grand scheme of things.

# ? Jan 25, 2015 21:15

Adbot: ADBOT LOVES YOU

# ? May 10, 2024 07:29

Longinus00: Dec 29, 2005; Ur-Quan

Wheany posted:

I'm not sure if this is a Linux question or an Ubuntu question, but here goes: I'm making a continuous integration server for our company using Jenkins. One of the jobs I'm doing is supposed to migrate our old MySQL database to a new structure. One of the steps is dumping the current contents to text files.

At first I tried using the jenkins job workspace for the dumps: /var/lib/jenkins/jobs/migrate_production_db/workspace but I got this error:
"mysqldump: Got error: 1: Can't create/write to file <path to file> (Errcode: 13) when executing 'SELECT INTO OUTFILE'"

Okay, I googled that and it's probably AppArmor's fault.

I managed to get the job working by using
sudo aa-complain /usr/sbin/mysqld
and by having the jenkins job create a directory /tmp/migration_work and then using chmod 777 on the directory (because the directory is owned by jenkins:jenkins and mysql runs under mysql:mysql). This is not a huge problem as long as I'm messing around at home my own server, but I'd like to put Jenkins and the jobs onto an internet-facing server and I feel like disabling AppArmor would be pretty dumb in that case. Also, I'd like to fix the configuration on my own machine as well, since this was a "just loving work, goddammit" fix.

Anyway, what would be a less dumb way of doing this? I think using /tmp/migration_work is not a huge problem, but I think I'd prefer to use the jenkins workspace instead.

The less dumb way is to modify the appamor profile to allow whatever it is you're trying to do, look in syslog/audit when it complains to see what needs to be done. Anecdotally apparmor syntax is easier to pick up than selinux so fixing it should be straightforward. Put your rules in apparor.d/local.

As for your tmp directory, chmod 777 is dumb. Make the directory jenkins:mysql or whatever instead.

# ? Jan 25, 2015 22:33

Vulture Culture: Jul 14, 2003; I was never enjoying it. I only eat it for the nutrients.

Longinus00 posted:

As for your tmp directory, chmod 777 is dumb. Make the directory jenkins:mysql or whatever instead.

Or finally learn that newfangled circa-1997 POSIX ACL thing

# ? Jan 25, 2015 23:19

Cidrick: Jun 10, 2001; Praise the siamese

Misogynist posted:

Or finally learn that newfangled circa-1997 POSIX ACL thing

I'm not trying to be a smartass - do real production environments ever use ACLs? I've never actually seen one in the wild before, only ever in lab environments. Most people are content with standard unix permissions coupled with audited sudo access. I'm curious what places decided to use ACLs and why they chose to go that route.

# ? Jan 25, 2015 23:28

evol262: Nov 30, 2010; #!/usr/bin/perl

Longinus00 posted:

The less dumb way is to modify the appamor profile to allow whatever it is you're trying to do, look in syslog/audit when it complains to see what needs to be done. Anecdotally apparmor syntax is easier to pick up than selinux so fixing it should be straightforward. Put your rules in apparor.d/local.

As for your tmp directory, chmod 777 is dumb. Make the directory jenkins:mysql or whatever instead.

SElinux rules syntax is easy. Learning how to write policies which apply on older distros (optional includes) and keeping up with all the different types is hard/impossible.

Cidrick posted:

I'm not trying to be a smartass - do real production environments ever use ACLs? I've never actually seen one in the wild before, only ever in lab environments. Most people are content with standard unix permissions coupled with audited sudo access. I'm curious what places decided to use ACLs and why they chose to go that route.

Some, often for better alignment with NTFS ACLs on fileservers. They're not hard, just that it's really weird to get used to ls not showing you everything you might need to know

# ? Jan 25, 2015 23:39

jaegerx: Sep 10, 2012; Maybe this post will get me on your ignore list!

Cidrick posted:

I'm not trying to be a smartass - do real production environments ever use ACLs? I've never actually seen one in the wild before, only ever in lab environments. Most people are content with standard unix permissions coupled with audited sudo access. I'm curious what places decided to use ACLs and why they chose to go that route.

Redhat has a serious boner for facls. You'd be surprised when and where you can actually use them to your benefit.

# ? Jan 26, 2015 00:17

evol262: Nov 30, 2010; #!/usr/bin/perl

jaegerx posted:

Redhat has a serious boner for facls. You'd be surprised when and where you can actually use them to your benefit.

[citation needed]

ACLs are on the RHCE exam and everything, but ACLs and caps aren't even supported in RPM to my knowledge. You can do it in %post, but I don't think we ship a single piece of software with an ACL set by default, and we may not even have one that can manage ACLs from a frontend of any kind.

We have a hard-on for SElinux, in the sense that we aim to have every product usable with SElinux in enforcing, but...

# ? Jan 26, 2015 01:59

Vulture Culture: Jul 14, 2003; I was never enjoying it. I only eat it for the nutrients.

Cidrick posted:

I'm not trying to be a smartass - do real production environments ever use ACLs? I've never actually seen one in the wild before, only ever in lab environments. Most people are content with standard unix permissions coupled with audited sudo access. I'm curious what places decided to use ACLs and why they chose to go that route.

Extended ACLs are very common in file server environments, especially where you have full NFSv4 support (i.e. ZFS) since the ACL model makes a lot more sense than the POSIX one. (NFSv4 ACLs are basically a superset of CIFS ACLs.)

I've run them in other weird cases where they've been needed, but like evol262 said, many people avoid them for ease-of-use because ls likes to show a stupid little + on top of the standard owner/group/world permissions instead of doing something obvious that says hey person, check the extended ACL.

# ? Jan 26, 2015 02:16

ToxicFrog: Apr 26, 2008

On the subject of ACLs...

I run ZNC. It creates log files with permissions znc:znc 660. I would like to instead create them with ownership toxicfrog:znc 660.

Based on my reading, it should be possible to do this with ACLs by configuring the existing log files and directories with this ownership and setting ACLs that cause new files and directories to inherit the parent ownership, permissions, and ACLs. But every time I thought I'd figured out how it's ended up not working. And I'm not even sure where to start looking anymore.

Help?

peepsalot posted:

How do I get google drive to sync my files because google hates linux.

I've been using this. It has to be explicitly run every time you want to sync, rather than automatically syncing in the background, but it's pretty great.

quote:

Also why does google hate linux?

Google is like 90% Linux internally. Possibly more. :iiam:

I know that inotify and stuff works differently than whatever APIs they used for the windows client, but this is something that Dropbox, BTSync, and Insync have had solved for years, and the Drive for Linux client has been coming Real Soon Now since, what, 2012?

# ? Jan 26, 2015 02:38

Longinus00: Dec 29, 2005; Ur-Quan

ToxicFrog posted:

Google is like 90% Linux internally. Possibly more.

While the backend stuff is mostly linux what you'll see people using are macbooks.

# ? Jan 26, 2015 03:09

evol262: Nov 30, 2010; #!/usr/bin/perl

Longinus00 posted:

While the backend stuff is mostly linux what you'll see people using are macbooks.

Because you can get a terminal and SSH with reasonable power management and suspend/resume that works 100% of the time. That's pretty much the only reason. If Lenovo could get their poo poo together and make a Carbon X1 people actually wanted to buy, they'd eat the "development on Macbooks" market in 6 months for people who actually need to do serious work on Linux, which Macs are totally inappropriate for since Mach is the worst kernel and OSX is one of the worst imaginable platforms for systems programming or testing

# ? Jan 26, 2015 03:49

Gucci Loafers: May 20, 2006; Ask yourself, do you really want to talk to pair of really nice gaudy shoes?

evol262 posted:

Because you can get a terminal and SSH with reasonable power management and suspend/resume that works 100% of the time. That's pretty much the only reason. If Lenovo could get their poo poo together and make a Carbon X1 people actually wanted to buy, they'd eat the "development on Macbooks" market in 6 months for people who actually need to do serious work on Linux, which Macs are totally inappropriate for since Mach is the worst kernel and OSX is one of the worst imaginable platforms for systems programming or testing

Isn't the Suspend/Resume more of a Windows OS issue than it is with Lenovo's X1 Carbon?

If I had my pick of hardware, I'd be happy as can be with a Macbook Air, Dual-Boot to Windows if needed and remote into everything else.

# ? Jan 26, 2015 03:55

jaegerx: Sep 10, 2012; Maybe this post will get me on your ignore list!

evol262 posted:

[citation needed]

ACLs are on the RHCE exam and everything, but ACLs and caps aren't even supported in RPM to my knowledge. You can do it in %post, but I don't think we ship a single piece of software with an ACL set by default, and we may not even have one that can manage ACLs from a frontend of any kind.

We have a hard-on for SElinux, in the sense that we aim to have every product usable with SElinux in enforcing, but...

Was referring to the cert tests.

E: but speaking of rpm. What the hell is going on there? Is Redhat ever gonna take back development and improve it?

jaegerx fucked around with this message at 04:10 on Jan 26, 2015

# ? Jan 26, 2015 03:56

evol262: Nov 30, 2010; #!/usr/bin/perl

Tab8715 posted:

Isn't the Suspend/Resume more of a Windows OS issue than it is with Lenovo's X1 Carbon?

If I had my pick of hardware, I'd be happy as can be with a Macbook Air, Dual-Boot to Windows if needed and remote into everything else.

Windows does it fine, but cygwin is terrible, and you pretty much need to use a VM to get a Dev env, since a lot of python/etc packages behave differently on Windows, even if there were native ssh.

But macefi sucks for running Linux on Apple. And suspend/resume is still pretty bad on Linux. Even Intel's drivers have weird backlight issues on resume and so forth. I said the Carbon because Thinkpads have good Linux support and build quality, so if there were an alternative to Macs...

The Carbon is not great, though.

jaegerx posted:

Was referring to the cert tests.

E: but speaking of rpm. What the hell is going on there? Is Redhat ever gonna take back development and improve it?

The certs also include autofs and a bunch of other stuff you probably should know to be an "engineer" but that we're not big on.

libresolv and interop with suse is big, but rpm doesn't need a hell of a lot. We still push updates, but what does it really need? Especially with debates about what to even do about packaging?

# ? Jan 26, 2015 04:27

Vulture Culture: Jul 14, 2003; I was never enjoying it. I only eat it for the nutrients.

evol262 posted:

Because you can get a terminal and SSH with reasonable power management and suspend/resume that works 100% of the time. That's pretty much the only reason. If Lenovo could get their poo poo together and make a Carbon X1 people actually wanted to buy, they'd eat the "development on Macbooks" market in 6 months for people who actually need to do serious work on Linux, which Macs are totally inappropriate for since Mach is the worst kernel and OSX is one of the worst imaginable platforms for systems programming or testing

If you're doing your distributed systems testing on a single host you're doing it wrong to begin with

# ? Jan 26, 2015 04:34

Hollow Talk: Feb 2, 2014

evol262 posted:

Windows does it fine, but cygwin is terrible, and you pretty much need to use a VM to get a Dev env, since a lot of python/etc packages behave differently on Windows, even if there were native ssh.

But macefi sucks for running Linux on Apple. And suspend/resume is still pretty bad on Linux. Even Intel's drivers have weird backlight issues on resume and so forth. I said the Carbon because Thinkpads have good Linux support and build quality, so if there were an alternative to Macs...

The Carbon is not great, though.

-- snip --

The backlight is a problem, and so are the broadcom-wl drivers, which are a pain to get automatically depending on distribution. The backlight problem is solvable by using mba6x_bl. It's still somewhat annoying. rEFInd at least makes dealing with EFI bearable. Oh, and tripple booting is one of the worst things to set up, ever, because apparently Windows 7 doesn't recognise the keyboard/touchpad unless you install it via Bootcamp, which means you cannot change the partition size from inside MacOS X.

tl;dr: next time my computer breaks and needs urgend replacement, I'll go buy something else. But the decision was between a 2014 MBA and a Dell XPS13, after the old laptop was a Dell. :suicide:

# ? Jan 26, 2015 04:51

ToxicFrog: Apr 26, 2008

Longinus00 posted:

While the backend stuff is mostly linux what you'll see people using are macbooks.

...except the workstations are all Linux, and what people are doing with those macbooks (if they're SWEs, anyways) is SSHing or Chromoting into those workstations.

# ? Jan 26, 2015 05:09

evol262: Nov 30, 2010; #!/usr/bin/perl

Misogynist posted:

If you're doing your distributed systems testing on a single host you're doing it wrong to begin with

Well, I think the argument is less about distributed systems testing, and more about basic day-to-day stuff. I can do a lot of python/go work locally. It's obviously not ideal for everything, but there are cases where it's really useful to be able to rapidly prototype some console app. And I can spin up an image to do it or whatever, but there's a convenience "I want to get something done on a flight with no wifi" or whatever case that isn't well-satisfied by Windows without a VM, or OSX without one.

There's nothing wrong with that, necessarily, but the preponderance of macs for developers owes at least a little to just how much you can get done locally, even if you can't do real systems work.

Hollow Talk posted:

The backlight is a problem, and so are the broadcom-wl drivers, which are a pain to get automatically depending on distribution. The backlight problem is solvable by using mba6x_bl. It's still somewhat annoying. rEFInd at least makes dealing with EFI bearable. Oh, and tripple booting is one of the worst things to set up, ever, because apparently Windows 7 doesn't recognise the keyboard/touchpad unless you install it via Bootcamp, which means you cannot change the partition size from inside MacOS X.

tl;dr: next time my computer breaks and needs urgend replacement, I'll go buy something else. But the decision was between a 2014 MBA and a Dell XPS13, after the old laptop was a Dell.

rEFIt/rEFInd are mostly ok, but OSX has a bad habit of loving with the EFI table on major upgrades. Newer (2013 or 2014) macs support actual UEFI, so you can just use it natively, but it's terrible before that. The fact that OSX firmware updates are almost impossible to apply without OSX installed (even though they're plain EFI blobs, the EFI partition on hfs+ is not fun) still makes that kind of nasty.

But the backlight problems, and broadcom problems, and touchscreen problems and whatever else are problems on non-Mac platforms, too. Linux is way better on a laptop than it used to be, but it's still nowhere as good as OSX unless you're on a Thinkpad

# ? Jan 26, 2015 05:10

Hollow Talk: Feb 2, 2014

evol262 posted:

rEFIt/rEFInd are mostly ok, but OSX has a bad habit of loving with the EFI table on major upgrades. Newer (2013 or 2014) macs support actual UEFI, so you can just use it natively, but it's terrible before that. The fact that OSX firmware updates are almost impossible to apply without OSX installed (even though they're plain EFI blobs, the EFI partition on hfs+ is not fun) still makes that kind of nasty.

But the backlight problems, and broadcom problems, and touchscreen problems and whatever else are problems on non-Mac platforms, too. Linux is way better on a laptop than it used to be, but it's still nowhere as good as OSX unless you're on a Thinkpad

Firmware updates are the main reason I kept OS X on it at all. And I agree, there a quite a number of annoyances with it running Linux. Funnily enough, however, it was pretty much the best price wise for what I wanted. I wouldn't have thought Apple would ever be a sensible price choice, but there you go...

# ? Jan 26, 2015 05:14

jaegerx: Sep 10, 2012; Maybe this post will get me on your ignore list!

Been on apple since iBook G4 for all my Linux sys admin jobs. The only desktop I could handle with Linux was openbox. Now with vagrant and cheap vms I don't think I'll ever go back to a Linux desktop.

# ? Jan 26, 2015 06:13

Gucci Loafers: May 20, 2006; Ask yourself, do you really want to talk to pair of really nice gaudy shoes?

jaegerx posted:

Been on apple since iBook G4 for all my Linux sys admin jobs. The only desktop I could handle with Linux was openbox. Now with vagrant and cheap vms I don't think I'll ever go back to a Linux desktop.

Yea, I'm the same boat here too unless Steam continues to port more games to linux and maybe one day I'd completely switch over.

# ? Jan 26, 2015 06:41

SurgicalOntologist: Jun 17, 2004

Weird problem. Starting sometime recently I lost the ability connect from one particular machine (Ubuntu 14.10) using ssh with pka. I'm pretty sure it's on the client side since I can connect to my server fine from other clients, and from this particular desktop I can't connect to anything with private keys. Specifically it hangs after the debug1 message SSH2_MSG_SERVICE_ACCEPT received. I even checked against bitbucket.org using hg --ssh='ssh -vv' and it hangs at the same place. It seems to hang indefinitely and I need to Ctrl-C it.

# ? Jan 26, 2015 08:02

Gucci Loafers: May 20, 2006; Ask yourself, do you really want to talk to pair of really nice gaudy shoes?

evol262 posted:

Windows does it fine, but cygwin is terrible, and you pretty much need to use a VM to get a Dev env, since a lot of python/etc packages behave differently on Windows, even if there were native ssh.

But macefi sucks for running Linux on Apple. And suspend/resume is still pretty bad on Linux. Even Intel's drivers have weird backlight issues on resume and so forth. I said the Carbon because Thinkpads have good Linux support and build quality, so if there were an alternative to Macs...

The Carbon is not great, though.

What does cygwin have to do with this?

# ? Jan 26, 2015 09:15

evol262: Nov 30, 2010; #!/usr/bin/perl

SurgicalOntologist posted:

Weird problem. Starting sometime recently I lost the ability connect from one particular machine (Ubuntu 14.10) using ssh with pka. I'm pretty sure it's on the client side since I can connect to my server fine from other clients, and from this particular desktop I can't connect to anything with private keys. Specifically it hangs after the debug1 message SSH2_MSG_SERVICE_ACCEPT received. I even checked against bitbucket.org using hg --ssh='ssh -vv' and it hangs at the same place. It seems to hang indefinitely and I need to Ctrl-C it.

How long have you waited? This sounds like rdns. Disable lookups?

Tab8715 posted:

What does cygwin have to do with this?

That suspend/resume isn't a Windows issue, it's a Linux one, but Windows is an inconvenient work environment for many of us because native tools aren't there

# ? Jan 26, 2015 14:51

KozmoNaut: Apr 23, 2008; Happiness is a warm
Turbo Plasma Rifle

evol262 posted:

The Carbon is not great, though.

The new 3rd-generation Carbon X1 moves back to physical function keys and physical buttons for the trackpad. It's pretty rad now.

# ? Jan 26, 2015 15:07

Wheany: Mar 17, 2006; Spinya^{ha^{ha^haha}ha}ha_{ha_{ha_haha}ha}ha!; Doctor Rope

Longinus00 posted:

The less dumb way is to modify the appamor profile to allow whatever it is you're trying to do, look in syslog/audit when it complains to see what needs to be done. Anecdotally apparmor syntax is easier to pick up than selinux so fixing it should be straightforward. Put your rules in apparor.d/local.

As for your tmp directory, chmod 777 is dumb. Make the directory jenkins:mysql or whatever instead.

Did this, except made the migration dir inside the workspace.

It wasn't apparmor in the first place, the default profile allows mysql to write under /tmp/ it was the group permission.

# ? Jan 26, 2015 15:29

Death Vomit Wizard: May 8, 2006; Bottom Feeder

Death Vomit Wizard posted:

I have successfully set up a bridge for my VM using Network Manager in Fedora21 using this guide. My question is, what do I do for my next VM? Add something to my bridge0? Follow the guide again and make a bridge1?

evol262 posted:

Add it to br0 or whatever you named it. Linux bridges are simple ARP things, but they can have practically unlimited slaves (practically unlimited in that you'll never hit the limit in practice)

Thanks. The problem wasn't my VM/bridge setup after all. Turns out my minimal CentOS install just needed ~~a "ifconfig eth0 up" and "dhclient -v eth0"~~ some changes to /etc/sysconfig/network-scripts/ifcfg-eth0.

I haven't learned anything about rpm's yet, but in the meantime I'm getting exactly what I wanted from

code:

./configure
make
checkinstall -R --install=no
yum install foo.rpm

e:f,b

Death Vomit Wizard fucked around with this message at 17:04 on Jan 26, 2015

# ? Jan 26, 2015 16:38

evol262: Nov 30, 2010; #!/usr/bin/perl

KozmoNaut posted:

The new 3rd-generation Carbon X1 moves back to physical function keys and physical buttons for the trackpad. It's pretty rad now.

That's great news, actually. I wish they would have done that a year ago when I needed a new laptop :negative:

Death Vomit Wizard posted:

Thanks. The problem wasn't my VM/bridge setup after all. Turns out my minimal CentOS install just needed a "ifconfig eth0 up" and "dhclient -v eth0".

Add:

ONBOOT=yes
BOOTPROTO=dhcp

To /etc/sysconfig/network-scripts/ifcfg-eth0

# ? Jan 26, 2015 16:49

SurgicalOntologist: Jun 17, 2004

evol262 posted:

How long have you waited? This sounds like rdns. Disable lookups?

It's been like 6 hours trying to pull from bitbucket. Isn't rdns a server-side thing? This is definitely a client-side problem.

# ? Jan 26, 2015 16:49

KozmoNaut: Apr 23, 2008; Happiness is a warm
Turbo Plasma Rifle

evol262 posted:

That's great news, actually. I wish they would have done that a year ago when I needed a new laptop

Correction: I meant physical buttons for the trackpoint. The trackpad is still button-less.

# ? Jan 26, 2015 16:54

evol262: Nov 30, 2010; #!/usr/bin/perl

SurgicalOntologist posted:

It's been like 6 hours trying to pull from bitbucket. Isn't rdns a server-side thing? This is definitely a client-side problem.

RDNS is usually server-side, but sometimes you get terrible lookups.

Also, the DNS timeout is more like 60 seconds. It's been 6 hours? Let me do some digging while I sit through a meeting...

E: before I look...

Are you using IPv6? Or ssh-agent?

Try turning off UseDNS. And if you're using the agent, try from a different user account or unset SSH_AUTH_SOCK

evol262 fucked around with this message at 17:03 on Jan 26, 2015

# ? Jan 26, 2015 17:00

wooger: Apr 16, 2005; YOU RESENT?

KozmoNaut posted:

The new 3rd-generation Carbon X1 moves back to physical function keys and physical buttons for the trackpad. It's pretty rad now.

The eBay market for thinlpads is weird due to the hosed up x1, x240 & t440 keyboard/trackpoint.

They can be had used, or new old stock for nearly as cheap as the *30 series.

# ? Jan 26, 2015 18:15

Gucci Loafers: May 20, 2006; Ask yourself, do you really want to talk to pair of really nice gaudy shoes?

KozmoNaut posted:

The new 3rd-generation Carbon X1 moves back to physical function keys and physical buttons for the trackpad. It's pretty rad now.

Heh, I don't mind the Trackpoint but without the buttons it's pretty darned difficult to use.

That's kinda of surprising that Linux isn't support that well on end-user hardware even in 2105. Didn't a year or two ago Linus said that he was gunning for the desktop?

# ? Jan 26, 2015 18:26

evol262: Nov 30, 2010; #!/usr/bin/perl

Tab8715 posted:

Heh, I don't mind the Trackpoint but without the buttons it's pretty darned difficult to use.

That's kinda of surprising that Linux isn't support that well on end-user hardware even in 2105. Didn't a year or two ago Linus said that he was gunning for the desktop?

Linux is extraordinarly well-supported on end-user hardware. You can basically put a current Fedora, Ubuntu, Arch, or whatever LiveUSB into any hardware and get every basic piece of hardware (wifi, power management, function keys on laptops, GPU -- maybe even with acceleration) working.

I know that's a low bar to set, but it's incredible progress even from a couple of years ago.

Most of the wonky "not supported well" stuff ends up being "gaming keyboards" which barely function in Windows without stupid utilities/drivers, function keys on the X1 (same), and other stuff which requires 8 different utility packs from your vendor, and which will probably stop working the next time Windows makes a major driver architecture revision, because the vendors CBF.

Plus there's the whole "Linux doesn't need the desktop because Android/embedded...". But I haven't met anyone who seriously cares about Linux "winning" since I used /. heavily a decade ago. It's usable for 90% of people, and that's enough

# ? Jan 26, 2015 18:38

Gucci Loafers: May 20, 2006; Ask yourself, do you really want to talk to pair of really nice gaudy shoes?

That is true, I remember trying to install linux in 2006 and things were much more difficult but what part of the linux power management portion is missing?

evol262 posted:

Plus there's the whole "Linux doesn't need the desktop because Android/embedded...". But I haven't met anyone who seriously cares about Linux "winning" since I used /. heavily a decade ago. It's usable for 90% of people, and that's enough

What do you mean?

# ? Jan 26, 2015 19:57

evol262: Nov 30, 2010; #!/usr/bin/perl

Tab8715 posted:

What do you mean?

Mobile is the new desktop, embedded devices are actually used by consumers instead of running coffee pots and cable boxes, many of these embedded/mobile/car/etc applications are running Linux anyway, maybe including Steamboxes, depending on how you classify "embedded".

# ? Jan 26, 2015 19:59

hooah: Feb 6, 2006; WTF?

evol262 posted:

You can basically put a current Fedora, Ubuntu, Arch, or whatever LiveUSB into any hardware and get every basic piece of hardware (wifi, power management, function keys on laptops, GPU -- maybe even with acceleration) working.

As recently as last summer this wasn't true for my laptop, a Yoga 2 Pro. I had to do sudo rmmod ideapad_laptop to get it to detect the wifi card.

# ? Jan 26, 2015 20:38

evol262: Nov 30, 2010; #!/usr/bin/perl

hooah posted:

As recently as last summer this wasn't true for my laptop, a Yoga 2 Pro. I had to do sudo rmmod ideapad_laptop to get it to detect the wifi card.

It's not perfect, but the fact that it works in 99% of cases is kind of the point. But there are two things there:

Ubuntu is a distro with terrible QA. I also have a Yoga 2 pro. Fedora doesn't have the world's best QE, but they managed to fix that default behavior last February.

Second, there will always be outliers. I can point to hardware that doesn't work out of the box on Windows, but nobody would argue that it isn't ready for the desktop.

# ? Jan 26, 2015 20:58

Polygynous: Dec 13, 2006; welp

Windows is not ready for the desktop. :colbert:

# ? Jan 26, 2015 21:06

Adbot: ADBOT LOVES YOU

# ? May 10, 2024 07:29

FISHMANPET: Mar 3, 2007; Sweet 'N Sour
Can't
Melt
Steel Beams

Lot's of stuff in Windows doesn't work out of the box but the way Windows drivers are, it's easy to just go to the manufacturer and download the driver and there you go. Linux is different and for whatever reason it's not that easy to install a new driver, so it has to work out of the box. It's a different standard but that's because it's a different system.

# ? Jan 26, 2015 21:20

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > The Linux Questions Thread: a bunch of pitfalls, but technically it's possible

«‹›1028 »