[sarcasm] Great solution. Now that I think of it, I guess you solved most problems on earth. Have a way of autenticating everyone, so they will be responsible for their actions.[/sarcasm].
Really, if you could in theory develop such scheme, why would you need to be charged on an email basis? Any offending sender would be identified and fined to death.
While I agree email is useless until we get rid of spam (and I can be done) I don't think money solves anything. It will only make the problem worst by switching the incentive to send free mails into promoting email fraud (that meaning you get a bill of $1.000 because someone, somehow found a way arround the system to impersonate you).
Gentoo is not easy to a newby, and dependecies while most of the time work right, some times they do not. For instance, when I tried my first (and only) gentoo install, emerge gnome2 would require certain library that was masqued. As a result, nothing worked (yes, I needed a GUI, because I wanted a desktop machine not a server), a specific media library dependency could not be satisfied and that was it.
Skimming though documentation did not help, I just forced the system to install the latest version of the library and edited all the ebuilds that would require such dependency (that was already satistied, but not exactly the ultra-minor revision number that was deprecated).
So no, the techinique you are refering is not the only source of programs when you are using gentoo. I am not blaming gentoo, I really like gentoo, but if you are building from a ports based system, even them most minute error could halt your progress and make you say "what the hell, Red Hat just works...gentoo is too difficult"...
95% of the people have their machines preinstalled and NEVER EVER try to upgrade it. If they have to, they'd call you or me...or if it is an office, whoever is in charge.
I agree with you 100%, but then again, a semi newby friend of mine stumbled upon MANY problem trying to install gentoo.
1) CD was not working, so he had to rely on a disquette. Gentoo does not have their own intall from net disquettes. So you have to rely on what? A Red Hat or Slack boot disk? tombstr does not work (chroot problem), and is the suggested method (if you managed to reach the part of the forum that explins that it, YES, can be installed from the net).
2) GRUB, he never used grub, he didn't know anyone that could help him with that. I don't use GRUB myself (and it's NOT the standard in any way) but I quickly responded that YES, he can install LILO. But Gentoo does not help you with lilo like every other distro (including slackware, that's supposed to be what...difficult?). It doesn't help you with fstab either. Great to learn, but if you are NEW, I would probably guess that forcing people to write their own fstab is kind of unhelpfull.
3) Kernel. Ok, this was a net install, so your next step is to compile a custom kernel. Well, the gentoo sources failed to produce a working kernel, and it was not his faul. I know because I have the same machine as his does (identical notebook), and only a vanilla kernel did the job.
So he is still trying to get gentoo to work in his spare time, but he certainly isn't thinking gentoo is for the faint of heart.
On the other hand, I totaly enjoy gentoo, though sometimes I relly feel the urge to just drop something in/usr/local/src and install it the unforgivable way. I mean, sometimes I really HAVE to make a lot of changes to the way certain things are compiled, and at those times I feel I am not in love anymore with my gentoo (but the feeling doesn't last much...i totally like gentoo).
What if you application may be used by some other application? You need a registry or something? You are switching hells...how are programs supposed to find the binaries? Put everything into the path?
I think a nice solution would be the already invented filesystem with arbitrary attributes. Like if it was a database. So you have the filesystem view, but you also have "packages view". You could for instance (example only) mount/packages, and everything would be arranged in categories by default, and as you entered dirs, you'd be inside specific packages (collections of sofware, be it docs, applications, etc.) and inside there you would find specific revisions. Or you could just make a GUI version to browse that intuitively. I mean, everything should be labeled and possibly versioned. Even text files. It should be easy to use...
I don't know, it should be too clutered or strict that it gets in your way, an easy interface to manage the extra info should not be a problem. And, for instance, every distro should have information and the power to know what's in the system without having to issue -v, --version -V and the endless list of varing names just to know the version of a library/binary.
Hans has tris problem where he sees open source needs some retribution to some authors. Specifically, a filesystem or a hidden library that is crucial for the wellbehavedness of any OS doesn't get much credit while other end users programs like Mplayer do. So there is much less incentive to develop the boring part where you also don't get any visible rewards.
And it turns out there is no real way to reward OSS authors, no credit system, nobody (except some few developers) really knows what's the real value added by OSS coders. No way to reward with money nor even credits.
That's the problem hans sees, he dropped the towel to find a working way to fund OSS developements in a way that makes sense (ie: money goes where is needed, not where it is most visible) so he is now targeting a previous step. Trying to figure out who must be credited and making sure people know about it.
Of course, that's nearly imposible and intrusive. I want people rewarded, but I don't want a mess of a thouthand names in front of my face.
There sure is a solution, this ones does not look like one.
I have galeon, in the menu it read "Galeon Web Browser"...I also think this is a good idea, and that programs should be categorized AND described or rated by the distribution. Maybe a user survey could be the rating system. Gentoo pretty much helps in that direction, because things that people use regularly have ebuilds, things that do not have outdated ebuilds or no ebuilds at all.
But what are you trying to solve? As an admin, have you ever tried a distro with package management? In that case, you only care about what you installed, not individual files. Last time I made ls/usr/bin was in 1997...
2: You cannot have a Gentoo style community unless you are a distro that caters to people who are willing to go to great lengths to learn more about their OS and computer hardware.
Actually, you cannot learn a bit about Linux if you are a gentoo user. You might learn a bit at install time, but after that point, you learning skill will be a non-requirement. You only need to know the name of app you need (think of it: to compile something it's on the list, you have to do nothing but emerge, you don't even learn what the dependencies are because the system does not force you nor need you to do so. To compile whatever is not on the list the good way, you need to learn how to make ebuilds (bash+the gentoo way). And if you compile the "cowboy" way, well, that's NOT gentoo, and you are messing up the system.
On the other hand, it's a great way for distributor to learn to keep it simple, elegant, well though out, transparent, and user friendly, without having to try to restart everything from scratch.
Gentoo is a great source of user feedback for developers, an a solution to end users that do not want to babysit their OS. It's also a great source for a Metadistribution...
hard links are not used very much. Granted, the poster puts a stupid example (if he made the link he shoould know, and if he didn't, he shouldn't be deleting it, nor should it be special irreproduceable content).
Now, you could have been polite as the user is stating an uninformed opinion, but a honest one. Respect for each other is one of the things what makes us different from machines and animals.
Having a single place, shared library allows you to know exactly what library you have, and how vulnerable you are, in a single location. And if a vulnerability is found, you fix the problem in one shot.
That's the desktop enviroment job. It associates ONE web browser as the web browser. The distro job is to give you option at install time, explaning what the options are. Naming it "web browser" is going to complicate things further...
It obviously needs clients to setup the tunneling client. Under windows, it think it can be any ssh client that supports tunnel (all of them should).
What you could do is point them to any free ssh client that supports forwarding along with a ready-made configuration file (so that they don't even need to think or learn anything).
The trick for your users will be connecting to localhost on port whatever knowing that that traffic will be tunneled though ssh to your FTP server of choice. For a more in depth reading try this whitepaper (or visit ssh.com for other sources of information).
https://secure.vandyke.com/vandyke-php-bin/white pa pers.php
Paranoid in true fashion, they will exclude MAC even when they know MAC addresses can be spoofed really easy. If I can learn MAC addresses in your network then there is nothing that will prevent me from trying to be one of them.
It's Humanlike Inteligence now. I don't know why don't they just ditch the name and replace it with HI. Anyway, I seem to be compleining, but I am not really, I do no think chess program have an AI inside, not anything that could be labeled AI today This happens because most people, including me, think only human way of learning and acting can be labeled as "inteligent".
Certainly, we don't label certain insects which behave as machines as inteligent. In fact, everything that resembles machine like behaveour can't be AI by definition. In the essense of our definition is that we think of ourselves as non-machines, when in reality we are some kind of elegant and powerfull machine, with some special characteristics (we are not hardcoded that much)...
Re:AAC is pretty weak, no marketing can change tha
on
AAC vs. OGG vs. MP3
·
· Score: 1
But it does prove it didn't sound any better than anything. It "could" sound is not a great technical achievement, is it?
It's possible that the number of bugs per line of code has actually been going down.
You've found the solution! Let's just add ^M or spaces whenever we need to get the bugs per line lower. Great for stability.
Really, I don't think anybody cares about bugs per line, they care about crashes per day or maybe lost revenue per year (goverment should see it as lost taxes per budget year also).
It is not even funny. IE can crash, but the OS should not crash. Stop optimizing and comparing based on marginal speed gains. That beign said, xmame is kinda slow for my celeron 433...so i tend to optimize it as much as I can, so the thing is having optimization where you don't care about it being stable.
Your drugs / spam analogy may seem to hold, but does not really. The economics of drugs is that if you ban it, but a lot of CONSUMER desperately want the product, then you'll STILL have drugs beign sold. Note I am not saying that if the drug dealer wants to sell it, they will sell it, the one that's putting the money, the consumer, will be the one rasing the offer so that he can have it.
Spam, on the other hand, only pleases the distributor (the spam services provider) and the producer (seller of the penis enlargement program). It does NOT pleases the consumer, they are not going to pay more for the penis enlargement program just because they seem to start receiving the email less often.
So, in the one hand, you have unstopable offending marketing, on the other hand you have a product that people will literaly kill so the can get doped. Stoping spam is much easier economically. Just make the spammers revenues harder. If you want to stop drug dealers, then burn colombia, bolivia and your corrupted agents (among other places / guys).
Airplanes can kill people, rm cannot. And if there ANY situation where it could kill, I'm pretty sure they don't just login as root and start rming things.
In any case, you could write a simple (or complex if you like) wrapper script for rm...and then check if/,/boot or whatever is being dangerously addressed. It should be trivial to implement.
Now, your only problem would beconvincing people to use your command. The problem rm in it's currect fashion is like it is, is because people like it that way. Nobody is babysitting people, as they are the ones making the choise.
So your argument is not against rm as it is, but against the people using unix command. And you have a point, but don't blame it on the tools.
Can't afford? The largest part of the microsoft market share is pirated copies of MS Whatever. They don't have to afford it.
I'd say mostly anybody using Windows ever though of moving to Linux because of the cheap factor. ALL of the moves i've seen fall in unrelated categories: - Geek / coolness factor - Institutional, tired of MS - Wanted to tried / liked it - Needed stability
Nobody learns Linux because of the chepo factor, I can understand some companies liking the cheap factor, but those companies hire guys that do like unixes, the boss in never ever going to learn unix, they will just use gnome and openoffice (or crossover).
I think the reason so many people profess their love for Unix now is that the remaining alternative is pretty godawful
I thought of that why too, and having decided against studing anything computer related formally because i found it to "volatile", I have to confess I like Unix because facts warranty me that grep, find and mostly everything will be arround in the years to come. Also, because unix was unknown to me at the time, I never though of learning Visual Basic, as it seemd to change every year. I have limited time and especially, limited memory. I cannot erase stuff, so when I learn something I want it to be as generic as it gets. If it's not generic, then it's a very short termed knoledge and after two years, I will be knowing the same as a kid that's just starting.
That's what I like most about unix, they seem to evolve where it is really or cool, but they don't just change every interface to the OS tools because someone just felt like it their to decide. Look at DOS, some of us learned a great deal of stuff that got "deprecated" for a GUI, because MS GUI decided to kill the command line. HELLO?
I like unix because it is there and the basics evolve in a fashion compatible with my style. I want to learn stuff and get better, not ditch everything every two years.
C# being a common interface to everything in windows is a great step in that direction for MS. But maybe too late.
Slashdot Mirror
[sarcasm] Great solution. Now that I think of it, I guess you solved most problems on earth. Have a way of autenticating everyone, so they will be responsible for their actions.[/sarcasm].
Really, if you could in theory develop such scheme, why would you need to be charged on an email basis? Any offending sender would be identified and fined to death.
While I agree email is useless until we get rid of spam (and I can be done) I don't think money solves anything. It will only make the problem worst by switching the incentive to send free mails into promoting email fraud (that meaning you get a bill of $1.000 because someone, somehow found a way arround the system to impersonate you).
So no thanks.
1) Partition != Filesystem
2) Free Software cannot use the documentation that Microsoft only shares on a "want to tell" basic.
Gentoo is not easy to a newby, and dependecies while most of the time work right, some times they do not. For instance, when I tried my first (and only) gentoo install, emerge gnome2 would require certain library that was masqued. As a result, nothing worked (yes, I needed a GUI, because I wanted a desktop machine not a server), a specific media library dependency could not be satisfied and that was it.
Skimming though documentation did not help, I just forced the system to install the latest version of the library and edited all the ebuilds that would require such dependency (that was already satistied, but not exactly the ultra-minor revision number that was deprecated).
So no, the techinique you are refering is not the only source of programs when you are using gentoo. I am not blaming gentoo, I really like gentoo, but if you are building from a ports based system, even them most minute error could halt your progress and make you say "what the hell, Red Hat just works...gentoo is too difficult"...
95% of the people have their machines preinstalled and NEVER EVER try to upgrade it. If they have to, they'd call you or me...or if it is an office, whoever is in charge.
I agree with you 100%, but then again, a semi newby friend of mine stumbled upon MANY problem trying to install gentoo.
/usr/local/src and install it the unforgivable way. I mean, sometimes I really HAVE to make a lot of changes to the way certain things are compiled, and at those times I feel I am not in love anymore with my gentoo (but the feeling doesn't last much...i totally like gentoo).
1) CD was not working, so he had to rely on a disquette. Gentoo does not have their own intall from net disquettes. So you have to rely on what? A Red Hat or Slack boot disk? tombstr does not work (chroot problem), and is the suggested method (if you managed to reach the part of the forum that explins that it, YES, can be installed from the net).
2) GRUB, he never used grub, he didn't know anyone that could help him with that. I don't use GRUB myself (and it's NOT the standard in any way) but I quickly responded that YES, he can install LILO. But Gentoo does not help you with lilo like every other distro (including slackware, that's supposed to be what...difficult?). It doesn't help you with fstab either. Great to learn, but if you are NEW, I would probably guess that forcing people to write their own fstab is kind of unhelpfull.
3) Kernel. Ok, this was a net install, so your next step is to compile a custom kernel. Well, the gentoo sources failed to produce a working kernel, and it was not his faul. I know because I have the same machine as his does (identical notebook), and only a vanilla kernel did the job.
So he is still trying to get gentoo to work in his spare time, but he certainly isn't thinking gentoo is for the faint of heart.
On the other hand, I totaly enjoy gentoo, though sometimes I relly feel the urge to just drop something in
What if you application may be used by some other application? You need a registry or something? You are switching hells...how are programs supposed to find the binaries? Put everything into the path?
/packages, and everything would be arranged in categories by default, and as you entered dirs, you'd be inside specific packages (collections of sofware, be it docs, applications, etc.) and inside there you would find specific revisions. Or you could just make a GUI version to browse that intuitively. I mean, everything should be labeled and possibly versioned. Even text files. It should be easy to use...
I think a nice solution would be the already invented filesystem with arbitrary attributes. Like if it was a database. So you have the filesystem view, but you also have "packages view". You could for instance (example only) mount
I don't know, it should be too clutered or strict that it gets in your way, an easy interface to manage the extra info should not be a problem. And, for instance, every distro should have information and the power to know what's in the system without having to issue -v, --version -V and the endless list of varing names just to know the version of a library/binary.
Hans has tris problem where he sees open source needs some retribution to some authors. Specifically, a filesystem or a hidden library that is crucial for the wellbehavedness of any OS doesn't get much credit while other end users programs like Mplayer do. So there is much less incentive to develop the boring part where you also don't get any visible rewards.
And it turns out there is no real way to reward OSS authors, no credit system, nobody (except some few developers) really knows what's the real value added by OSS coders. No way to reward with money nor even credits.
That's the problem hans sees, he dropped the towel to find a working way to fund OSS developements in a way that makes sense (ie: money goes where is needed, not where it is most visible) so he is now targeting a previous step. Trying to figure out who must be credited and making sure people know about it.
Of course, that's nearly imposible and intrusive. I want people rewarded, but I don't want a mess of a thouthand names in front of my face.
There sure is a solution, this ones does not look like one.
A IBM microdrive (compactflash) should be quiet enough. Granted, it has moving parts, but who cares? With 2 GB you can do a lot of things.
I have galeon, in the menu it read "Galeon Web Browser"...I also think this is a good idea, and that programs should be categorized AND described or rated by the distribution. Maybe a user survey could be the rating system. Gentoo pretty much helps in that direction, because things that people use regularly have ebuilds, things that do not have outdated ebuilds or no ebuilds at all.
But what are you trying to solve? As an admin, have you ever tried a distro with package management? In that case, you only care about what you installed, not individual files. Last time I made ls /usr/bin was in 1997...
2: You cannot have a Gentoo style community unless you are a distro that caters to people who are willing to go to great lengths to learn more about their OS and computer hardware.
Actually, you cannot learn a bit about Linux if you are a gentoo user. You might learn a bit at install time, but after that point, you learning skill will be a non-requirement. You only need to know the name of app you need (think of it: to compile something it's on the list, you have to do nothing but emerge, you don't even learn what the dependencies are because the system does not force you nor need you to do so. To compile whatever is not on the list the good way, you need to learn how to make ebuilds (bash+the gentoo way). And if you compile the "cowboy" way, well, that's NOT gentoo, and you are messing up the system.
On the other hand, it's a great way for distributor to learn to keep it simple, elegant, well though out, transparent, and user friendly, without having to try to restart everything from scratch.
Gentoo is a great source of user feedback for developers, an a solution to end users that do not want to babysit their OS. It's also a great source for a Metadistribution...
hard links are not used very much. Granted, the poster puts a stupid example (if he made the link he shoould know, and if he didn't, he shouldn't be deleting it, nor should it be special irreproduceable content).
Now, you could have been polite as the user is stating an uninformed opinion, but a honest one. Respect for each other is one of the things what makes us different from machines and animals.
Having a single place, shared library allows you to know exactly what library you have, and how vulnerable you are, in a single location. And if a vulnerability is found, you fix the problem in one shot.
That's the desktop enviroment job. It associates ONE web browser as the web browser. The distro job is to give you option at install time, explaning what the options are. Naming it "web browser" is going to complicate things further...
It obviously needs clients to setup the tunneling client. Under windows, it think it can be any ssh client that supports tunnel (all of them should).
e pa pers.php
What you could do is point them to any free ssh client that supports forwarding along with a ready-made configuration file (so that they don't even need to think or learn anything).
The trick for your users will be connecting to localhost on port whatever knowing that that traffic will be tunneled though ssh to your FTP server of choice. For a more in depth reading try this whitepaper (or visit ssh.com for other sources of information).
https://secure.vandyke.com/vandyke-php-bin/whit
WEP and MAC exclusions are for the paranoid.
Paranoid in true fashion, they will exclude MAC even when they know MAC addresses can be spoofed really easy. If I can learn MAC addresses in your network then there is nothing that will prevent me from trying to be one of them.
Tunnel it though ssh. The FTP client does not need to know it is going tunneled, so it will work with your legacy apps.
It's Humanlike Inteligence now. I don't know why don't they just ditch the name and replace it with HI. Anyway, I seem to be compleining, but I am not really, I do no think chess program have an AI inside, not anything that could be labeled AI today This happens because most people, including me, think only human way of learning and acting can be labeled as "inteligent".
Certainly, we don't label certain insects which behave as machines as inteligent. In fact, everything that resembles machine like behaveour can't be AI by definition. In the essense of our definition is that we think of ourselves as non-machines, when in reality we are some kind of elegant and powerfull machine, with some special characteristics (we are not hardcoded that much)...
But it does prove it didn't sound any better than anything. It "could" sound is not a great technical achievement, is it?
It's possible that the number of bugs per line of code has actually been going down.
You've found the solution! Let's just add ^M or spaces whenever we need to get the bugs per line lower. Great for stability.
Really, I don't think anybody cares about bugs per line, they care about crashes per day or maybe lost revenue per year (goverment should see it as lost taxes per budget year also).
It is not even funny. IE can crash, but the OS should not crash. Stop optimizing and comparing based on marginal speed gains. That beign said, xmame is kinda slow for my celeron 433...so i tend to optimize it as much as I can, so the thing is having optimization where you don't care about it being stable.
Your drugs / spam analogy may seem to hold, but does not really. The economics of drugs is that if you ban it, but a lot of CONSUMER desperately want the product, then you'll STILL have drugs beign sold. Note I am not saying that if the drug dealer wants to sell it, they will sell it, the one that's putting the money, the consumer, will be the one rasing the offer so that he can have it.
Spam, on the other hand, only pleases the distributor (the spam services provider) and the producer (seller of the penis enlargement program). It does NOT pleases the consumer, they are not going to pay more for the penis enlargement program just because they seem to start receiving the email less often.
So, in the one hand, you have unstopable offending marketing, on the other hand you have a product that people will literaly kill so the can get doped. Stoping spam is much easier economically. Just make the spammers revenues harder. If you want to stop drug dealers, then burn colombia, bolivia and your corrupted agents (among other places / guys).
Very true, though some scripts/programs expect the files to be somewhere in particular (yeah, they shouldn't!).
Airplanes can kill people, rm cannot. And if there ANY situation where it could kill, I'm pretty sure they don't just login as root and start rming things.
/, /boot or whatever is being dangerously addressed. It should be trivial to implement.
In any case, you could write a simple (or complex if you like) wrapper script for rm...and then check if
Now, your only problem would beconvincing people to use your command. The problem rm in it's currect fashion is like it is, is because people like it that way. Nobody is babysitting people, as they are the ones making the choise.
So your argument is not against rm as it is, but against the people using unix command. And you have a point, but don't blame it on the tools.
Can't afford? The largest part of the microsoft market share is pirated copies of MS Whatever. They don't have to afford it.
I'd say mostly anybody using Windows ever though of moving to Linux because of the cheap factor. ALL of the moves i've seen fall in unrelated categories:
- Geek / coolness factor
- Institutional, tired of MS
- Wanted to tried / liked it
- Needed stability
Nobody learns Linux because of the chepo factor, I can understand some companies liking the cheap factor, but those companies hire guys that do like unixes, the boss in never ever going to learn unix, they will just use gnome and openoffice (or crossover).
I think the reason so many people profess their love for Unix now is that the remaining alternative is pretty godawful
I thought of that why too, and having decided against studing anything computer related formally because i found it to "volatile", I have to confess I like Unix because facts warranty me that grep, find and mostly everything will be arround in the years to come. Also, because unix was unknown to me at the time, I never though of learning Visual Basic, as it seemd to change every year. I have limited time and especially, limited memory. I cannot erase stuff, so when I learn something I want it to be as generic as it gets. If it's not generic, then it's a very short termed knoledge and after two years, I will be knowing the same as a kid that's just starting.
That's what I like most about unix, they seem to evolve where it is really or cool, but they don't just change every interface to the OS tools because someone just felt like it their to decide. Look at DOS, some of us learned a great deal of stuff that got "deprecated" for a GUI, because MS GUI decided to kill the command line. HELLO?
I like unix because it is there and the basics evolve in a fashion compatible with my style. I want to learn stuff and get better, not ditch everything every two years.
C# being a common interface to everything in windows is a great step in that direction for MS. But maybe too late.