I've searched systematically for something better, and haven't found it.
For writing your own documents, learn to use LaTeX. It's a learning curve that take a while to climb---you probably never run out of new things to learn, but reach a level of self-satisfaction---but the result is great-looking documents.
Also, since you'll be entering text plus markup, it's easy to put in a versioning system. It's also easy to enter math (it's based on TeX, by Don Knuth, so go figure).
With a bit of make-fu, it's also easy to insert figures made by gnuplot, graphviz and dia.
For reading m-sword files, I can recommend catdoc;)
McCreesh said 90% of Linux users traditionally receive OpenOffice.org updates straight from their Linux distribution's vendor
but that would still give windows >66% (assuming os x makes up 0%, which is possible due to neo office
Let's do the math. The official site sees (scaled down) 2 linux downloads and 8 windows downloads. For every 1 of these linux downloads, there's 9 downloading from the distro archives instead of the official site.
That gives us 20 linux downloads, 8 windows downloads. Or just above 25%. How did you come up with 66%?
Even if it's just 25%, that's a fair slice; this means that the plan of moving people over to open-source apps first and moving the OS out under them later has not been shown to be infeasible: windows users are moving to the open-source apps.
Only 221,000 downloads by Linux users were recorded
So just shy of 900,000 windows downloads? That's quite good.
I won't say "we're winning!!one!11ty", but some cautious optimism is probably in order.
What they do is that they compare the new drug being developped to a fake drug that has no effects.
And, if cures or alters a disease that we already have drugs for, one would hope they compare it to the best known drug(s) for treating that disease to see if it's worth using.
Add to the argument that [...] there is tech available that allowed the transport company to identify commuters carrying the illegal goods.
There isn't. Suppose I wear a hood plus striped shirt and carry a bag of money that says "loot". When the police officer on board starts questioning me, I show him the contract that documents my employment as a courier for Sillybank.
In the case we're talking about: the problem isn't detecting peer-to-peer. The problem is detecting copyright infringement.
You can have two identical bit streams where one infringes copyright and the other doesn't. It all depends on the humans at the ends of the bit streams, and the senders' agreements (or lack thereof) with the copyright holder.
(case in points: WoW, ubuntu, Creative Commons movies).
my TV will start in a second or two. So did my old Commodore 64. How is it that the more power we get, the -longer- this takes?
It's because we make our systems do more stuff. How much did those system do when they were started up?
Did they mount a couple of file systems, start cron+http+aptcache+distcc+cpufreqd+ntp daemons and wait for DHCP_ACK, then mount some more file systems and load up a highly configurable login screen?
I think it'd be easy to boot into a one-button gui saying "bring the system into a usable state now, please". XP does something like this.
Especially the TV comparison is unfair; the TV is a one-purpose box with the functionality done mostly in hardware. I'm sure one could write a minimalistic kernel that supports exactly one TV tuner card, exactly one graphics card and exactly one sound card, and have it boot into watch-TV-mode quite fast. That'd be closer to an apples-to-apples comparison [but not enough: the TV doesn't have a BIOS that supports general-purpose computing and does a lot of checks].
For example, my Mac will go from startup to login in half the time of either Vista -or- Ubuntu (not counting what happens -after- login [...])
Why do you omit counting what happens after login? Isn't the useful measure of boot time how long it takes from pressing the power button to having a computer that's usable?
By only measuring up to an arbitrary point, one can inflate boot speeds by deferring everything until after that arbitrary point. That doesn't give you a usable computer any sooner, it just cooks your numbers. See my one-button OS.
From what I hear, what OS X does right is deferring everything until after the login screen, plus lazily starting up services once you're logged in, such that the desktop is usable while the system is "post-login booting", and prioritizing well: if you ask for a networked file system, it'll do networking before, say, the frequency scaling daemon.
It's not just giving you a login screen. It's giving you a usable desktop with as short a wait as possible.
X.Org 7.4 [...]. Hotplugging support for input devices actually works now, so you can plug in mice and tablets and use them without having to reboot.
Having to reboot? Wouldn't that be a kernel issue and not an X.org issue? I can imagine why you'd have to restart the X server, but the kernel? Haven't the kernel had hotplugging support with hotplug or udev for a few years now?
Improvements to X.Org also allow for the easier to manage display control panel, which allows users to adjust resolutions and screen placement for single and multiple monitor displays easily.
The new Network Manager is a great improvement over the previous release. It allows your Ubuntu machine to connect to the network before a user logs in.
Still no easy bonding? I submitted a request for that [/me feels indignant].
A bit seriously though: bonding rocks. Wanna pick up your laptop and not break the sshfs connection to your file server? Sure. Wanna have bandwidth that doesn't suck while you're tethered down by the ethernet cable? You can have that too.
But not with NetworkManager unless you hack some of its dispatcher scripts. Only for the techies.
Better Support for Web Video and Audio Ubuntu now supports the high-quality setting in YouTube! We shall celebrate by watching videos of other people's animals at a better quality level. Additionally, now Ubuntu users can view the programming the BBC puts online in Totem. That's right, you can enjoy fine shows like Scotland Outdoors and The Archers from your Linux PC.
Cool! Uhh... what was updated again? Firefox? Flash? GStreamer? Totem? firefox-gstreamer-totemish-flv-plugin?
Type ecryptfs-setup-private in the Terminal, and you can hide and encrypt a folder in your Home directory. [...] This folder gives a secure location that you can use to store sensitive files, without paying the performance penalty that full-disk encryption incurs.
I wouldn't trust that. Applications may not know to keep data secret beyond umask, and so will store stuff in/tmp. Or your secret data will be put on the non-encrypted swap partition. And in my experience, full-disk encryption works fine, very little is noticable; a few.5s-delays when saving in emacs.
Config-less X.Org
Awesome!!1!
No seriously, I really think it is. Not much use to me now, but it'll probably be in the future.
[I'm still going to have an xorg.conf because it's a great place to cast spells that makes my trackball kick ass. EmulateWheel springs to mind, which is really a must with a Logitech Marble Mouse that has scroll _buttons_ instead of a wheel; no repeated scrolling otherwise, but with EmulateWheel I have it, and I have horizontal scrolling. Check out Battle for Wesnoth with horizontal scrolling, I wrote that:)]
Not the greatest written article. But I look forward to upgrading. Last time I did that, though, something broke. My plan is to pick a new package each day [or maybe every eight hours or so] and upgrade just that one. Then, when something breaks, I can limit it to one package plus dependencies, instead of all $BIGNUM packages.
You never just "click install" on windows xp.. 1) Pop in disk 2) First you have to [...]
Almost correct.
1-5) as you said. 6) Look at the popup that says lsass.exe will shut down your box in 30 seconds 7) Pull the box off the network 8-12) Do step 1-5 again 13) Download antivirus without connection to the network. Pixies and leprechauns are helpful here. 14) Install the antivirus 15-16) step 6-7
Based on a true story. I can't tell you how much I hated windows when I saw the sasser popup.
Is it really too much to ask of a user that they understand that it [...] generally does only what they tell it to do [...]
Yes it is too much to ask, because it's not true. Computers as machines do what developers tell them to do. The amount of instructions given by the programmer vis-a-vis object code vastly outnumbers the number of instructions a typical user is going to issue to the machine. [...]
True but irrelevant. Users aren't asked to do what the assembly instructions do, they're expected to understand how the world that's presented to them works. It can be simple or complex to a large degree independently of how much code is spent to built it. By and large, developers tell computers to do what the users of the application tell it to do (except for DRM).
Just like development; I'm not expected to understand the implementation of creat, open, read, write, close and so forth. I'm expected to understand how they interact. *I* am telling the computer to open a file (&c), not the kernel.
That's great and all, but what is someone who's a graphics designer, for example, going to do with their mastery over a Unix system? What good does it do a visual artist to know how to build a makefile or compile a kernel?
You can master Unix without knowing anything about makefiles or kernel builds; nice straw man, though. What are the benefits? Automating a lot of image handling tasks, say [look at sng and pildriver]. Or building their own image tagging/searching system [http/mail-header based key-value pairs in.txt files, find+grep]. Just off the top of my head.
but ATEOTD if you've bought a disk with N capacity and you only use N/2, it follows that the rest of it is essentially wasted. Might as well do something useful with it.
I come dangerously close to 100% before buying new storage. I even delete videos I'm done watching just to reclaim space [I can always download them again if I need them]. You're correct, but the condition of your if-statement is only true for 5% of the disk lifespan.
I'm not trying to say "wah wah it uses more space". I'm trying find out whether it's worth complaining about:)
You had to do some maintenance to reclaim that
Here's an interesting idea I just had: when the disk is full, it automatically runs through the non-newest versions of all files from oldest to newest, and deletes the old ones. Or perhaps some smart consideration of both age and freed-up size. But if the numbers say it doesn't give back much space...
I heard that the complete kernel revision history, as packed by git, takes up half the size of a single checkout. That's a very particular usage pattern, so be wary of generalizing this data.
"duh just type ~rf - m" or something something, because yeah, a menu to do that would kill someone
For no one thing would a menu item to do that thing be particularly bad. But you can't put _every_ task in a menu, because there are infinitely many tasks.
If you find people often tell you to type in commands you don't understand, it's probably because it's the most efficient way to do something once you do master it. See for instance http://www.linuxquestions.org/questions/linux-newbie-8/modifying-functions-678643/. I've built a 1650-byte podcatcher in #!sh [and that's including proper error-checking and all].
Well, there are some people who can't find Pacific Ocean on the map. I dont see map makers running around in panic, thinking how to make their maps more accessible to the general population...
I think it's just a shortage:
Well, I personally believe that U.S. Americans are unable to do so because uh, some people out there in our nation don't have maps
And we should all help Mark Shuttleworth make more maps!
and I believe that our education like such as South Africa [...] I believe that they should our education over here! in the U.S. should the the U.S., should help South Africa
I mean, think of the children! You don't hate children, do you?
so we will be able to build up our future for our children
To do even this simple thing with Linux, all of our applications would have to be re-written to enable a new file specification syntax,
Why? Hans Reiser demonstrated that files can be directories too, without breaking the VFS layer.
Say we bake versioning into the file system. You want the old versions of/home/user/shopping-list.txt; you go look in/home/user/shopping-list.txt/old/1. If you want the one from yesterday you go for/home/user/shopping-list.txt/old/bytime/2008-10-24.00:00:00, and the file system figures out which of the old versions was present at that time.
Same old syntax. The name resolution is handled differently, but that's all in the file system. You could probably even write a fuse file system that adds a layer of versioning on top of another file system. No need to ever touch the apps.
If you want the duct tape solution: write a shell script that checks whether anything changed every n minutes, then commit your home directory to subversion/git/....
Do you have any numbers on how much space was used on extra versions for a "typical" distribution of files and usage patterns? TANSTAAFL and all that;)
You're planning on giving _all_ the developers a console _each_? Some way that makes my head assplode. Isn't one of the points of consoles that you gather a bunch of people around the same machine and trash-talk while playing (i.e. the person-to-person socialization)? Oh well, if you think it's a good idea, go ahead.
I can tell you what my former employee did: put a big TV and a wii in the conference room. According to local tradition, we play a game or two after lunch, and a few friday afternoon while having a beer. If we spend too much time on the wii, the boss-man can probably see it on our weekly productivity reports.
In general, we were trusted to not slack off, which seemed to work fine. I saw the occasional emails saying "I'm sick, so I'm gonna work from home to the best of my ability today." When I was out of tobacco, I went to the nearby supermarket and bought some; I felt no need to tell the boss I'd be out for a few minutes.
Treat people like responsible adults and they will act responsibly.
but I think it is mostly that they are sick and tired of people voting with no clue who they are voting for.
There's a simpler explanation: more revenue from ad views.
They're a public company. They are obligated to make money. Making people vote based on an informed thought process--how exactly does that make Google money? Not all are going to vote for the candidate that helps Google. Meanwhile, the site is going to take time (i.e. money) to build and operate.
You can't really chalk it up as a branding or awareness thing either: the Google brand is strong and everybody knows it. And it's definitely not "the cost of doing business".
I think it's the money. For Google, that means ad views.
Let's divide people into three groups: those who buy, those who make the pirated (DRM-free) version, and those who pirate.
Those who buy will now get a better product.
Those who pirate never see the DRM in the first place.
Those who make the pirated version will have an easier time; this benefits the pirates ever so slightly, but DRM is often defeated faster than you can say Yo-Ho, so the benefit is ever so slight.
The real winners, whenever DRM is removed, are the honest consumers.
Slashdot Mirror
I've searched systematically for something better, and haven't found it.
For writing your own documents, learn to use LaTeX. It's a learning curve that take a while to climb---you probably never run out of new things to learn, but reach a level of self-satisfaction---but the result is great-looking documents.
Also, since you'll be entering text plus markup, it's easy to put in a versioning system. It's also easy to enter math (it's based on TeX, by Don Knuth, so go figure).
With a bit of make-fu, it's also easy to insert figures made by gnuplot, graphviz and dia.
For reading m-sword files, I can recommend catdoc ;)
McCreesh said 90% of Linux users traditionally receive OpenOffice.org updates straight from their Linux distribution's vendor
but that would still give windows >66% (assuming os x makes up 0%, which is possible due to neo office
Let's do the math. The official site sees (scaled down) 2 linux downloads and 8 windows downloads. For every 1 of these linux downloads, there's 9 downloading from the distro archives instead of the official site.
That gives us 20 linux downloads, 8 windows downloads. Or just above 25%. How did you come up with 66%?
Even if it's just 25%, that's a fair slice; this means that the plan of moving people over to open-source apps first and moving the OS out under them later has not been shown to be infeasible: windows users are moving to the open-source apps.
Only 221,000 downloads by Linux users were recorded
So just shy of 900,000 windows downloads? That's quite good.
I won't say "we're winning!!one!11ty", but some cautious optimism is probably in order.
Matthew 26:74
Then began he to curse and to swear, saying, I know not the man. And immediately the PENIS crew.
I want to be on the penis crew...
And the bash.org favorite, Genesis 22:3
And Abraham rose up early in the morning, and saddled his BUTT
Guy puts a saddle on his behind? He shoulda put it on his back if he wanted his kids to ride him.
What a bunch of poppydong! I've heard wangroaches speak better dickney in a schlongpit.
But "glButt tty" makes you think about fantbutttic graphics capabilities ;)
(http://www.catb.org/jargon/html/G/glass-tty.html)
What they do is that they compare the new drug being developped to a fake drug that has no effects.
And, if cures or alters a disease that we already have drugs for, one would hope they compare it to the best known drug(s) for treating that disease to see if it's worth using.
#!/usr/bin/python
marks = [0, 12, 29, 39, 72, 91, 146, 157, 160, 161, 166, 191, 207,
214, 258, 290, 316, 354, 372, 394, 396, 431, 459, 467, 480]
unmeasurable = set(range(1, 481))
for i in range(1, len(marks)):
for j in range(i):
unmeasurable.discard(marks[i] - marks[j])
print sorted(unmeasurable)
Output:
[81, 90, 93, 103, 110, 111, 120, 139, 153, 171, 172, 174, 176, 183, 184, 192, 196, 198, 200, 204, 210, 213, 216, 220, 221, 223, 227, 231, 232, 238, 241, 242, 243, 247, 249, 254, 255, 256, 257, 259, 262, 264, 267, 269, 272, 275, 279, 280, 283, 284, 286, 288, 291, 292, 294, 295, 296, 297, 308, 309, 311, 312, 317, 318, 326, 327, 328, 329, 330, 331, 332, 335, 336, 337, 338, 339, 341, 344, 345, 346, 347, 348, 349, 350, 351, 352, 353, 356, 358, 361, 362, 363, 364, 366, 369, 370, 371, 373, 374, 375, 377, 378, 379, 380, 381, 383, 385, 386, 388, 390, 391, 393, 397, 398, 399, 400, 401, 403, 404, 405, 406, 407, 409, 410, 411, 412, 413, 414, 415, 416, 417, 418, 421, 422, 423, 424, 425, 426, 427, 429, 432, 433, 434, 435, 436, 437, 439, 440, 442, 443, 444, 445, 446, 448, 449, 450, 452, 453, 454, 456, 457, 458, 460, 461, 462, 463, 464, 465, 466, 469, 470, 471, 472, 473, 474, 475, 476, 477, 478, 479]
Add to the argument that [...] there is tech available that allowed the transport company to identify commuters carrying the illegal goods.
There isn't. Suppose I wear a hood plus striped shirt and carry a bag of money that says "loot". When the police officer on board starts questioning me, I show him the contract that documents my employment as a courier for Sillybank.
In the case we're talking about: the problem isn't detecting peer-to-peer. The problem is detecting copyright infringement.
You can have two identical bit streams where one infringes copyright and the other doesn't. It all depends on the humans at the ends of the bit streams, and the senders' agreements (or lack thereof) with the copyright holder.
(case in points: WoW, ubuntu, Creative Commons movies).
Even if it boots up and shows the desktop quickly, the user will have to wait until all the startup programs finish loading.
The OS/deskenv/whatever could be smart about it and load things on a by-need basis, or keeping a sliding window of CPU free for the user to spend.
I think OS X does something like this: defer as much as possible, then load it when the user is here, in a smart way.
It's psychological - the user wants to see progress.
True, but...
If they can double-click on $BROWSER sooner they will be happy, even if the system is only semi-responsive.
False in my experience: you get annoyed with the system being so slow.
Being able to detect progress is good. Having a false sense of being done is bad.
my TV will start in a second or two. So did my old Commodore 64. How is it that the more power we get, the -longer- this takes?
It's because we make our systems do more stuff. How much did those system do when they were started up?
Did they mount a couple of file systems, start cron+http+aptcache+distcc+cpufreqd+ntp daemons and wait for DHCP_ACK, then mount some more file systems and load up a highly configurable login screen?
I think it'd be easy to boot into a one-button gui saying "bring the system into a usable state now, please". XP does something like this.
Especially the TV comparison is unfair; the TV is a one-purpose box with the functionality done mostly in hardware. I'm sure one could write a minimalistic kernel that supports exactly one TV tuner card, exactly one graphics card and exactly one sound card, and have it boot into watch-TV-mode quite fast. That'd be closer to an apples-to-apples comparison [but not enough: the TV doesn't have a BIOS that supports general-purpose computing and does a lot of checks].
For example, my Mac will go from startup to login in half the time of either Vista -or- Ubuntu (not counting what happens -after- login [...])
Why do you omit counting what happens after login? Isn't the useful measure of boot time how long it takes from pressing the power button to having a computer that's usable?
By only measuring up to an arbitrary point, one can inflate boot speeds by deferring everything until after that arbitrary point. That doesn't give you a usable computer any sooner, it just cooks your numbers. See my one-button OS.
From what I hear, what OS X does right is deferring everything until after the login screen, plus lazily starting up services once you're logged in, such that the desktop is usable while the system is "post-login booting", and prioritizing well: if you ask for a networked file system, it'll do networking before, say, the frequency scaling daemon.
It's not just giving you a login screen. It's giving you a usable desktop with as short a wait as possible.
Yes because it would be irresponsible not to download updates for Windows. It's so important that your box can get owned in less then 4 minutes
No, no, no. You have to use anecdotal evidence! http://linux.slashdot.org/comments.pl?sid=1007971&cid=25512997.
I'm of course completely impartial here ;)
All directories are files: sure.
What Hans showed was the converse: all files are directories (or can be made to be without breaking all sorts of shit).
Being-happy-to-teach-you-logic'ly yours,
--Jonas K
I think it was flamebait because people disagree with you. Not that this is how it should work, but this is how it works.
I'm too tired to address the point I made, but others have essentially done it for me.
Happy /dotting :)
Could someone explain to me why they think I'm being modded flamebait? I didn't aim for that, that's for sure...
X.Org 7.4 [...]. Hotplugging support for input devices actually works now, so you can plug in mice and tablets and use them without having to reboot.
Having to reboot? Wouldn't that be a kernel issue and not an X.org issue? I can imagine why you'd have to restart the X server, but the kernel? Haven't the kernel had hotplugging support with hotplug or udev for a few years now?
Improvements to X.Org also allow for the easier to manage display control panel, which allows users to adjust resolutions and screen placement for single and multiple monitor displays easily.
This is next to this image: http://www.maximumpc.com/files/u7/resolution.jpg. Who wants to bet that the control panel is part of GNOME, not X.org?
The new Network Manager is a great improvement over the previous release. It allows your Ubuntu machine to connect to the network before a user logs in.
Still no easy bonding? I submitted a request for that [/me feels indignant].
A bit seriously though: bonding rocks. Wanna pick up your laptop and not break the sshfs connection to your file server? Sure. Wanna have bandwidth that doesn't suck while you're tethered down by the ethernet cable? You can have that too.
But not with NetworkManager unless you hack some of its dispatcher scripts. Only for the techies.
Better Support for Web Video and Audio
Ubuntu now supports the high-quality setting in YouTube! We shall celebrate by watching videos of other people's animals at a better quality level. Additionally, now Ubuntu users can view the programming the BBC puts online in Totem. That's right, you can enjoy fine shows like Scotland Outdoors and The Archers from your Linux PC.
Cool! Uhh... what was updated again? Firefox? Flash? GStreamer? Totem? firefox-gstreamer-totemish-flv-plugin?
Type ecryptfs-setup-private in the Terminal, and you can hide and encrypt a folder in your Home directory. [...] This folder gives a secure location that you can use to store sensitive files, without paying the performance penalty that full-disk encryption incurs.
I wouldn't trust that. Applications may not know to keep data secret beyond umask, and so will store stuff in /tmp. Or your secret data will be put on the non-encrypted swap partition. And in my experience, full-disk encryption works fine, very little is noticable; a few .5s-delays when saving in emacs.
Config-less X.Org
Awesome!!1!
No seriously, I really think it is. Not much use to me now, but it'll probably be in the future.
[I'm still going to have an xorg.conf because it's a great place to cast spells that makes my trackball kick ass. EmulateWheel springs to mind, which is really a must with a Logitech Marble Mouse that has scroll _buttons_ instead of a wheel; no repeated scrolling otherwise, but with EmulateWheel I have it, and I have horizontal scrolling. Check out Battle for Wesnoth with horizontal scrolling, I wrote that :)]
Not the greatest written article. But I look forward to upgrading. Last time I did that, though, something broke. My plan is to pick a new package each day [or maybe every eight hours or so] and upgrade just that one. Then, when something breaks, I can limit it to one package plus dependencies, instead of all $BIGNUM packages.
Has it been half a year already? :)
-- Jonas K
You never just "click install" on windows xp..
1) Pop in disk
2) First you have to [...]
Almost correct.
1-5) as you said.
6) Look at the popup that says lsass.exe will shut down your box in 30 seconds
7) Pull the box off the network
8-12) Do step 1-5 again
13) Download antivirus without connection to the network. Pixies and leprechauns are helpful here.
14) Install the antivirus
15-16) step 6-7
Based on a true story. I can't tell you how much I hated windows when I saw the sasser popup.
You make some good points. I disagree with a few:
Is it really too much to ask of a user that they understand that it [...] generally does only what they tell it to do [...]
Yes it is too much to ask, because it's not true. Computers as machines do what developers tell them to do. The amount of instructions given by the programmer vis-a-vis object code vastly outnumbers the number of instructions a typical user is going to issue to the machine. [...]
True but irrelevant. Users aren't asked to do what the assembly instructions do, they're expected to understand how the world that's presented to them works. It can be simple or complex to a large degree independently of how much code is spent to built it. By and large, developers tell computers to do what the users of the application tell it to do (except for DRM).
Just like development; I'm not expected to understand the implementation of creat, open, read, write, close and so forth. I'm expected to understand how they interact. *I* am telling the computer to open a file (&c), not the kernel.
That's great and all, but what is someone who's a graphics designer, for example, going to do with their mastery over a Unix system? What good does it do a visual artist to know how to build a makefile or compile a kernel?
You can master Unix without knowing anything about makefiles or kernel builds; nice straw man, though. What are the benefits? Automating a lot of image handling tasks, say [look at sng and pildriver]. Or building their own image tagging/searching system [http/mail-header based key-value pairs in .txt files, find+grep]. Just off the top of my head.
but ATEOTD if you've bought a disk with N capacity and you only use N/2, it follows that the rest of it is essentially wasted. Might as well do something useful with it.
I come dangerously close to 100% before buying new storage. I even delete videos I'm done watching just to reclaim space [I can always download them again if I need them]. You're correct, but the condition of your if-statement is only true for 5% of the disk lifespan.
I'm not trying to say "wah wah it uses more space". I'm trying find out whether it's worth complaining about :)
You had to do some maintenance to reclaim that
Here's an interesting idea I just had: when the disk is full, it automatically runs through the non-newest versions of all files from oldest to newest, and deletes the old ones. Or perhaps some smart consideration of both age and freed-up size. But if the numbers say it doesn't give back much space...
I heard that the complete kernel revision history, as packed by git, takes up half the size of a single checkout. That's a very particular usage pattern, so be wary of generalizing this data.
"duh just type ~rf - m" or something something, because yeah, a menu to do that would kill someone
For no one thing would a menu item to do that thing be particularly bad. But you can't put _every_ task in a menu, because there are infinitely many tasks.
If you find people often tell you to type in commands you don't understand, it's probably because it's the most efficient way to do something once you do master it. See for instance http://www.linuxquestions.org/questions/linux-newbie-8/modifying-functions-678643/. I've built a 1650-byte podcatcher in #!sh [and that's including proper error-checking and all].
It's also dense communication-wise; compare "sudo ifconfig eth0 192.168.0.1" with "System -> Administration -> Network; unlock, wired connection, properties, enable, static ip, 192.168.0.1".
That being said, though, deskenvs should support the most common and important tasks in an easy-to-use way.
Well, there are some people who can't find Pacific Ocean on the map. I dont see map makers running around in panic, thinking how to make their maps more accessible to the general population...
I think it's just a shortage:
Well, I personally believe
that
U.S. Americans are unable to do so
because uh, some
people out there in our nation
don't have maps
And we should all help Mark Shuttleworth make more maps!
and I believe that our education like
such as South Africa
[...]
I believe that they should
our education over here!
in the U.S.
should the the U.S.,
should help South Africa
I mean, think of the children! You don't hate children, do you?
so we will be able to build up
our future
for our children
(watch more intellectual roadkill at http://www.youtube.com/watch?v=WALIARHHLII)
To do even this simple thing with Linux, all of our applications would have to be re-written to enable a new file specification syntax,
Why? Hans Reiser demonstrated that files can be directories too, without breaking the VFS layer.
Say we bake versioning into the file system. You want the old versions of /home/user/shopping-list.txt; you go look in /home/user/shopping-list.txt/old/1. If you want the one from yesterday you go for /home/user/shopping-list.txt/old/bytime/2008-10-24.00:00:00, and the file system figures out which of the old versions was present at that time.
Same old syntax. The name resolution is handled differently, but that's all in the file system. You could probably even write a fuse file system that adds a layer of versioning on top of another file system. No need to ever touch the apps.
If you want the duct tape solution: write a shell script that checks whether anything changed every n minutes, then commit your home directory to subversion/git/....
Do you have any numbers on how much space was used on extra versions for a "typical" distribution of files and usage patterns? TANSTAAFL and all that ;)
Have a look at Tracker and Beagle. There's desktop search on Linux too :)
You're planning on giving _all_ the developers a console _each_? Some way that makes my head assplode. Isn't one of the points of consoles that you gather a bunch of people around the same machine and trash-talk while playing (i.e. the person-to-person socialization)? Oh well, if you think it's a good idea, go ahead.
I can tell you what my former employee did: put a big TV and a wii in the conference room. According to local tradition, we play a game or two after lunch, and a few friday afternoon while having a beer. If we spend too much time on the wii, the boss-man can probably see it on our weekly productivity reports.
In general, we were trusted to not slack off, which seemed to work fine. I saw the occasional emails saying "I'm sick, so I'm gonna work from home to the best of my ability today." When I was out of tobacco, I went to the nearby supermarket and bought some; I felt no need to tell the boss I'd be out for a few minutes.
Treat people like responsible adults and they will act responsibly.
but I think it is mostly that they are sick and tired of people voting with no clue who they are voting for.
There's a simpler explanation: more revenue from ad views.
They're a public company. They are obligated to make money. Making people vote based on an informed thought process--how exactly does that make Google money? Not all are going to vote for the candidate that helps Google. Meanwhile, the site is going to take time (i.e. money) to build and operate.
You can't really chalk it up as a branding or awareness thing either: the Google brand is strong and everybody knows it. And it's definitely not "the cost of doing business".
I think it's the money. For Google, that means ad views.
No DRM? Good news for pirates!
No, good news for honest buyers.
Let's divide people into three groups: those who buy, those who make the pirated (DRM-free) version, and those who pirate.
Those who buy will now get a better product.
Those who pirate never see the DRM in the first place.
Those who make the pirated version will have an easier time; this benefits the pirates ever so slightly, but DRM is often defeated faster than you can say Yo-Ho, so the benefit is ever so slight.
The real winners, whenever DRM is removed, are the honest consumers.
A defect in the crystal structure results in the failure of every component that the defect is present in.
Every component, or one out of one. That's something like 100.00001863%