Huh? You can't prove a theory. I could say with quite a large amount of certainty, based on repeated observations, that the sun will rise tomorrow. But I still cannot prove it until tomorrow comes.
How would I remove all the other variables to the point where I can say, with certainty, that there really is something strange going on? It's just not possible. There's a hundred or so components that need to be aligned exactly, be totally clean, work perfectly, etc.
I work with holography. I shine a laser at a piece of film, then develop the film. And presto, I get no image. Do I throw out the theory that exposing film to light should produce an image? No, I assume that I screwed up and go back and start again. It's not uncommon for me to spend 3 months of cleaning, aligning, measuring and so on until I produce a proper image. I then throw away all the "bad" data. Maybe, theoretically, that data could be useful, but there's too many parameters to account for.
I think you're missing how beautifully google did this.
Google started scanning books, somewhat illegally. A _class action_ suit was then filed against google, by _some_ book writers but on behalf of _all_ book writers. This is the key point here. The purpose of a class action lawsuit is that you don't tie up the courts for years while every single person individually sues some company for something that they did wrong. Instead you have one big lawsuit and, and here's the kicker, no more lawsuits are allowed after that.
So, google gets sued in this class action lawsuit, and comes to an agreement that they can scan the books. Again, the class action lawsuit is speaking on behalf of _every_ person that ever has or will write a book. Think about that for a moment.
Say that you now write a book, and Google and some other company scan that book. You wrote this independently and have nothing to do with any guild or anything. You _cannot_ sue Google, but you _can_ sue the other company. This is the wonderful situation that Google is now in.
This is the strangeness of the situation. For another company to do the same thing, they would have to do something established as illegal, then try to entice a class action lawsuit, then try to come to the same arrangement. It's pretty risky!
Right, so if Windows contained a small bit of code to check that it was an intel machine, then it would be illegal for any company to remove that check and then sell the machine?
Say Microsoft added a clause that Microsoft Window could _only_ be run on Intel machines. Would this ruling make it truly illegal to sell AMD machines with Windows on?
You can compile a kernel without modules, but these days even most embedded systems don't bother with that. initrd is a collection of scripts, nothing to do with the kernel.
So.. you have no idea how the linux kernel works (hint - pretty much everything is a module) yet you somehow feel that you should have an opinion anyway?
Oh noes! It takes like 10 minutes to compile the default kernel for all those users that compile their own kernel! Clearly linux is going down the tubes! What insanity!
It's not a layer, it's a module. Even if distros compile it, the result is just an extra driver file in the directory. It is only loaded if you need it. How does that add any bloat?
Even at a source code level, it is completely self contained in a directory, other than a couple of one-line changes to tell the build system to compile.
And you are hardly going to be post pictures of when you're unhappy, and people tend to smile just for photos, so you get a selective image of someone.
I've been working on adding scripting support, so that these sorts of things can be added easily for other people, and done without increasing the memory footprint for those who don't use it.
If you're up for doing writing a bit of javascript to get all this information and display it nicely, I'll be happy to include it. Most of the work would be writing a nice UI using HTML or Qt Designer. (The scripting support supports either).
Actually I agree with you - but changing global shortcuts is hard - too much momentum:-) If I changed them, then we'd next have posts about how KDE is so damn annoying because they keep changing the global shortcuts..
I didn't like the look of automatically changing between KB/MB/GB per line because then you can't instantly see how much memory it is using. This way, the larger the text is displaying the number, the more memory.
I like the shortcut ideas - especially "/" to jump to the searchbox. I don't think using "k" for kill etc works, because typing "k" already causes it to jump to the process starting with the letter "k".
> The system monitor is much more buggy (crashed while i was look at it here, crashes if i change the monitor background color for networking
Heh, I get reports about that crash _daily_, despite fixing it months ago.:) I never realised so many people changed the background color. Maybe I need to fix my defaults.
> 1) it seams to take a lot of cpu, is it drawing graphs even if that tab is in the background and can that be avoided?
It doesn't draw graphs not being shown. The processes page takes around 3% CPU on my laptop (dual core 1.8Ghz AMD). About half the time is fetching all the info, and half for drawing. I'm running Qt 4.6 though, which is a bit faster.
> 2) the graphs don't have enough settings
Hmm, you'll hate me then:-D I just removed the custom-set colors thing, and instead set the colors based on the current KDE color scheme. People were complaining that the colors weren't set by their KDE color scheme.
> 3) You can't remove the default 2 tabs
Yeah, that's intentional:-D People accidentally delete them otherwise and then end up with a non-functioning app. You can manually unlock them - in ~/kde/share/apps/ksysguard/*.ksgrd and set "locked" to be 0.
> 4) ksysguardd seams to run all the time and i can't find out where to disable it (might be a fedora packaging issue)
It's being run by plasma, to provide data to its engines. It's only running when actually used, so maybe you have a CPU monitoring plasmoid thing?
> 5) status bar info could contain number of running processes
Trouble is, the more I put in the status bar, the wider it makes the window:-D I'll look into it.
Well, the System Activity thing is always loaded. Also the UI should be quite clean and not look like a desktop app. I'm looking for suggestions on how they could be merged and still look pretty:-)
I've implemented/io already, so it's in the very newest versions of KDE. Adding/netstat support would be interesting indeed.
* How would showing recently-terminated processes work UI wise? Maybe a context menu option. One possible program would be when a large number of processes are created/destroyed - e.g. when compiling. It might kinda 'flood' the display.
* I've started on adding scripting support (using javascript). I can add it for almost free (in terms of memory) since it's already loaded by KDE. I just need to write (or get others to write!) some scripts for it:-) This functionality should be in KDE 4.4
* That would be good:) Maybe done through the scripting support.
Slashdot Mirror
Huh? You can't prove a theory. I could say with quite a large amount of certainty, based on repeated observations, that the sun will rise tomorrow. But I still cannot prove it until tomorrow comes.
What has that got to do with LOCAL admin rights?
How would I remove all the other variables to the point where I can say, with certainty, that there really is something strange going on? It's just not possible. There's a hundred or so components that need to be aligned exactly, be totally clean, work perfectly, etc.
It's pretty rare for everything to go right.
I work with holography. I shine a laser at a piece of film, then develop the film. And presto, I get no image. Do I throw out the theory that exposing film to light should produce an image? No, I assume that I screwed up and go back and start again. It's not uncommon for me to spend 3 months of cleaning, aligning, measuring and so on until I produce a proper image. I then throw away all the "bad" data. Maybe, theoretically, that data could be useful, but there's too many parameters to account for.
I think you're missing how beautifully google did this.
Google started scanning books, somewhat illegally. A _class action_ suit was then filed against google, by _some_ book writers but on behalf of _all_ book writers. This is the key point here. The purpose of a class action lawsuit is that you don't tie up the courts for years while every single person individually sues some company for something that they did wrong. Instead you have one big lawsuit and, and here's the kicker, no more lawsuits are allowed after that.
So, google gets sued in this class action lawsuit, and comes to an agreement that they can scan the books. Again, the class action lawsuit is speaking on behalf of _every_ person that ever has or will write a book. Think about that for a moment.
Say that you now write a book, and Google and some other company scan that book. You wrote this independently and have nothing to do with any guild or anything. You _cannot_ sue Google, but you _can_ sue the other company. This is the wonderful situation that Google is now in.
This is the strangeness of the situation. For another company to do the same thing, they would have to do something established as illegal, then try to entice a class action lawsuit, then try to come to the same arrangement. It's pretty risky!
Right, so if Windows contained a small bit of code to check that it was an intel machine, then it would be illegal for any company to remove that check and then sell the machine?
Just for fun..
Say Microsoft added a clause that Microsoft Window could _only_ be run on Intel machines. Would this ruling make it truly illegal to sell AMD machines with Windows on?
You can compile a kernel without modules, but these days even most embedded systems don't bother with that. initrd is a collection of scripts, nothing to do with the kernel.
So.. you have no idea how the linux kernel works (hint - pretty much everything is a module) yet you somehow feel that you should have an opinion anyway?
Oh noes! It takes like 10 minutes to compile the default kernel for all those users that compile their own kernel! Clearly linux is going down the tubes! What insanity!
What's with all the idiotic posts?
It's not a layer, it's a module. Even if distros compile it, the result is just an extra driver file in the directory. It is only loaded if you need it. How does that add any bloat?
Even at a source code level, it is completely self contained in a directory, other than a couple of one-line changes to tell the build system to compile.
> 1. Please add "What's this" help entries to your widgets.
This is now done, just in time for the string freeze (Tomorrow!).
When you finally use 4.4, please let me know if you have any improvements for the the WhatsThis entries. ( johnflux at gmail dotty com)
You left-click on the column heading to sort. Is it not working?
For now, I added a tooltip saying that you can right click for more options. I'll try to improve the dialog box at some point.
And you are hardly going to be post pictures of when you're unhappy, and people tend to smile just for photos, so you get a selective image of someone.
It is, of course, a bad idea to run an app as supervisor.
Instead, when you try to kill or renice a process that you don't own, it will pop up a dialog box asking for the root password.
In older versions this was done with kdesu (using sudo) and in newer versions this is done with policykit.
I suppose it's possible that you have a really old version that doesn't have either though.
You can't, because the start up speed of the full version isn't acceptable - especially when the system is under full load.
I'm looking at trying to merge the two, but still unsure how exactly.
I've been working on adding scripting support, so that these sorts of things can be added easily for other people, and done without increasing the memory footprint for those who don't use it.
If you're up for doing writing a bit of javascript to get all this information and display it nicely, I'll be happy to include it. Most of the work would be writing a nice UI using HTML or Qt Designer. (The scripting support supports either).
This support will be in KDE 4.4.
Actually I agree with you - but changing global shortcuts is hard - too much momentum :-)
If I changed them, then we'd next have posts about how KDE is so damn annoying because they keep changing the global shortcuts..
In KDE 3 I tried to maintain the applets, but I simply did not have the time to do a decent job.
In KDE 4, I don't maintain any plasmoids, but instead rely on others to do that, and hopefully the better ones will float to the top.
> A way to group processes by library loaded and (private) memory allocated for those libraries would also be great.
How would that look UI wise?
I didn't like the look of automatically changing between KB/MB/GB per line because then you can't instantly see how much memory it is using. This way, the larger the text is displaying the number, the more memory.
I like the shortcut ideas - especially "/" to jump to the searchbox. I don't think using "k" for kill etc works, because typing "k" already causes it to jump to the process starting with the letter "k".
> The system monitor is much more buggy (crashed while i was look at it here, crashes if i change the monitor background color for networking
Heh, I get reports about that crash _daily_, despite fixing it months ago. :) I never realised so many people changed the background color. Maybe I need to fix my defaults.
> 1) it seams to take a lot of cpu, is it drawing graphs even if that tab is in the background and can that be avoided?
It doesn't draw graphs not being shown. The processes page takes around 3% CPU on my laptop (dual core 1.8Ghz AMD). About half the time is fetching all the info, and half for drawing. I'm running Qt 4.6 though, which is a bit faster.
> 2) the graphs don't have enough settings
Hmm, you'll hate me then :-D I just removed the custom-set colors thing, and instead set the colors based on the current KDE color scheme. People were complaining that the colors weren't set by their KDE color scheme.
> 3) You can't remove the default 2 tabs
Yeah, that's intentional :-D People accidentally delete them otherwise and then end up with a non-functioning app. You can manually unlock them - in ~/kde/share/apps/ksysguard/*.ksgrd and set "locked" to be 0.
> 4) ksysguardd seams to run all the time and i can't find out where to disable it (might be a fedora packaging issue)
It's being run by plasma, to provide data to its engines. It's only running when actually used, so maybe you have a CPU monitoring plasmoid thing?
> 5) status bar info could contain number of running processes
Trouble is, the more I put in the status bar, the wider it makes the window :-D I'll look into it.
> 6) Hmm, I'll look into that too.
Well, the System Activity thing is always loaded. Also the UI should be quite clean and not look like a desktop app. I'm looking for suggestions on how they could be merged and still look pretty :-)
I've implemented /io already, so it's in the very newest versions of KDE. Adding /netstat support would be interesting indeed.
* How would showing recently-terminated processes work UI wise? Maybe a context menu option. One possible program would be when a large number of processes are created/destroyed - e.g. when compiling. It might kinda 'flood' the display.
* I've started on adding scripting support (using javascript). I can add it for almost free (in terms of memory) since it's already loaded by KDE. I just need to write (or get others to write!) some scripts for it :-) This functionality should be in KDE 4.4
* That would be good :) Maybe done through the scripting support.