But what we need is one package format that everyone can use. RPM's are a pain in the ass to put onto a distribution that wasn't built from the ground up from RPM's. They have almost no standard naming system for packages. They have to be linked against the exact same minor version of libs once they are installed. RPM's should be dumped. Deb's and Tgz's have the same problem. We need a fresh package format that, instead of keeping a database of installed packages, uses a./configure/autoconf like script to actually check for what's installed. Also, it needs NOT to use M4! It needs to be able to link against at least all minor versions it was linked against (maybe produce a warning to tell you).
These little things would take care of all the major issues with RPM/Deb/Tgz packages. A measure of backward compatability might be nice, but in all honesty, we are only inches further with binary package formats than we were 10 years ago!!!!
While we're on the subject about things we don't like with Gentoo;)
It is very annoying that Gentoo's rc program was written in a compiled language. I feel as much as possible relating to init should be written in bash. I also don't like rc-update. Gentoo's init scripts are the most annoying things in the world. They aren't even bash scripts iirc. If you look at the top of one it's shebang is to another program (not running gentoo right now so I couldn't tell you which).
Best init scripts I think are the ones from Linux from Scratch. It uses symlinking (/etc/rc.d/rc3.d etc. etc.). It is basically all written in bash. The init scripts are simple as hell. It comes with a templete init script so you can write your own in the same style. Has external functions for cleanly loading and killing the process (loadproc and killproc). Gentoo's overly cluttered init system just leads to very messy init scripts. K.I.S.S.
What the hell are you talking about? Please the article, ok. make xconfig is just a way to configure your kernel options at compile time. THAT'S IT. Some people prefer having a graphical interface to it. It's COMPLETELY seperate and has nothing to do with the kernel itself. Most people have Xfree86 running on their box with either gtk or QT. Thus, the kernel developers gave those people another option for configuring their kernel. Most people have Ncurses too, so they gave those people make menuconfig. Some people don't have any of those, so they gave us make config. None of those have ANYTHING to do with the "over GUIfication of linux". Frankly, I don't know how the hell your parent post was ever modded up because you obviously have not RTFA.
There's actually a new kernel option that you can put the.config right into your kernel. That way say you delete your kernel source tree and forgot to save your.config (or just plain forgot where you saved it), or working on another system, you can just configure it based on your current running kernel.
Before you go off ranting and raving maybe you should read the article. What he's talking about is the addition of gtk and qt interfaces to make {x,g}config. Instead of using tk for xconfig, it now uses QT. It also adds the use of make gconfig which uses Gtk. omfgz these new additions are going to ruin Linux and there will be rape and pillage and GUIfication and it's not going to be an OS anymore because it switched from tk to qt to configure it while leaving config and menuconfig untouched and instead of legos it's going to be linkin logs omgz!!!!! k thx la~
I think this is the bigger story of the day. Basically, SCO is now claiming only those 17 header files as the "stolen" SysV code. They are now claiming they need to see the AIX source to be able to site exact copyright infringements.
IBM pointed out that SCO has publicly claimed HP-UX and Tru64 do not have infringing code. SCO then countered with "HP/Compaq have not contributed code to linux" to which IBM produced documents showing they did contribute. IBM went on to say, how could they prove they were not contributing code without seeing HP-UX/Tru64? ( I think they were trying to make the argument that if they can say what isn't infringing without seeing the source then they should be able to say exactly what is infringing without needing the AIX source)
Also interesing pointed out by IBM is during Darl's Harvard speech, he claimed that
"...[T]here is roughly a million lines of code that tie into contributions that IBM has made and that's subject to litigation that's going on. We have basically supplied that. In fact, that is going to be the subject of a hearing that comes up Friday..." (emphasis added.)
Yet they have only produced 17 files and will not actually say which lines in those files are infringing, but does say only parts of those files are infringing.
I agree. Viruses are not hard to write. Espically the ones discussed in the article. It's really amauter hour. A virus (the type discussed in the article anyway) is just a program. The ones talking about in the article run just like any other legit program and require less skill to make one than a real program.
The REAL masters are those who find _new_ things. The MIT student who broke the xbox encryption with a belt sander. The ones whom run a program hundreds of times through gdb to find exploits. The ones who write a backdoor that can evade port scanners. Those are the real masters of the art. These are bored kids. They can come back to me when they've got the craftsmanship of a group like tcniso.
I can't access the trolltech right now (too slow) but I'm 99% sure that was true at one point, but NOT true anymore. The timeline iirc was it was free on windows, then not free, and sometime in the 3.x series it became free again. I can't confirm this though til their site lets up some.
So if they don't have any production ready server OS with 64-bit support, why even bother with desktops??? Desktop systems will see the least benefit by far. If I were MS, espically since they have no production ready win2k/2k3, I would be working damn hard to get a server version out.
I'm a little confused. Does MS have 64-bit versions of Win2k and Win2k3 available? I've heard plenty about this XP version, but nothing about their server versions (not to say it doesn't exist, I just haven't heard anything).
Instead of giving me cameras, games, hand jobs, with my phone; why can't you just give me a phone that works almost all the time and gets great reception?
I live in a moderatly sized city with plenty of cell phone vendors and (supposedly) a lot of towers. I can't even get reception with my phone in my own home, let alone when I'm at work inside a building built to withstand a hurricane (thick cement walls).
All this extra stuff is really an effort to divert the attention away from the base product, which, for most people, is still a sub standard product. Everyone I know (except those with Nextel's) has to jump through hoops to even write a grocery list down on their phone, let alone hold a real conversation. My phone was 215 dollars, has a big lcd screen, polyphonic ring tones, games, screen savers, internet access, but can not do what it was designed to do properly.
Last time I checked, I bought a phone, NOT A PDA! So why can't you give me a phone?
Should be noted Oregon State's involvement in this. Oregon State has provided A LOT of bandwidth to Linux projects including Linux from Scratch and Gentoo. They've got a very very big pipe there and have been more than generous with it there.
I can pretty much tell you who's going to be the fastest and slowest without a doubt. Gentoo fastest OpenBSD slowest (I don't even use Gentoo btw). All those are for the most part general purpose distros meant to run on a lot of hardware. They are not meant to run as fast as possible. They basically are compiled -O -march=i386. Gentoo being the exception since it's a source compiled distro and are going to tweak the hell out of the cc flags. I know this is going to make a lot of people with distro partialness mad, but general purpose distributions were not meant for speed. Gentoo wasn't designed with ease in mind, it was designed for speed.
As for OpenBSD, it's focus is on security. From personal experience, it's noticibly slower than any *nix I've ever used. Not to say it's bad, it's just not built for speed.
I think that list needs to be reworked considerbly. Maybe put them into categories such as source compiled distros, general purpose distros, security focused distros. Because comparing a source distro to a general purpose one is no contest. Not to say speed is everything, but it's really apples an oranges. Comparing one source distro to many general purpose ones is like racing a porshe and a honda. The porshe is going to win, but that's not what Hondas are built for.
I know a lot of people suggest switching to Mozilla, but it's not even about Mozilla. Almost every other browser is better than IE these days. Opera, Konqueror, Mozilla, (insert other browsers here). IE hasn't had anything significant in years. It lacks basic even basic things other browsers have. Pop up blocking, tabbed browsing, the ability to stop gif ads from looping, much better basic security policies, more w3c compiant, they don't make up their own html tags, plus many many more.
IE is so horrible. It's gotten to the point that just by using IE you are pretty much guaranteed to get spyware/adware/virus. Most of the people I know who use IE have their homepage changed daily, get a new toolbar every two days, so many pop ups they have to reboot weekly, their email stolen from cookies hourly, and a partridge in a pear tree.
If they really want to do something, fines are not the way to go. The only effective measure would be to make them modify Windows. In the US, because they weren't forced to modify windows' bundling, the fines levied really had no effect. In fact it was a positive experience for them because they paid their fines mostly in free software to schools, furthering their monopoly. If IE was seperated from the OS, then something might have actually change.
Reports were that the EU was considering seperating Windows Media Player from the OS. I think this would be a good start, but not enough. And a mere fine wouldn't be enough either.
What I think really needs to be done to break their monopoly.
a) Seperate Outlook Express, IE, Media Player, and Windows Messenger from the OS. Make them free downloads from Microsoft.com
b) OEM's shouldn't have such a hefty penalty for also selling competing OS's. All MS OEM contracts for desktop systems contain provisions that keep you from selling any other OS than windows. If you don't sign a contact, you pay retail. This needs to be done away with.
c) All file formats non-patentable and documents for the file format specs need to be made available publicly so competing software can be made interoperable.
I don't know German, but it sure looks like it was once there then redirects to a generic page. Maybe Bill Gates slammed his cock down on the table and demanded it be taken off. Said one man "Sure was tiny, and a hell of a laugh, but we all decided we should take it off anyway. You'd think with all that money he could afford to be circumsized".
I don't get this overwhelming support for gnome. On just technical merit, KDE is at least as good as Gnome, so all the fuss is basically over stupid QT license issues. KDE is a very well oiled project with regular releases and an awesome product. So why are only a select few distro's using it as the default WM?
At one point they did, but they put it back
on
SCO Offline
·
· Score: 1
Sometime last night they did change it actually. And then in the middle of the night (3AM EST or somewhere about) they added it again. So for a few hours it was gone. But for whatever reason brought it back.
You missed the point. Binary packages as a base install are fine, but NOT for upgrading software from 3rd parties. If you installed your kde from the CD then it's fine as a binary, but when you grab an rpm from freshmeat that was compilied against gcc 2.95 and you are running 3.4, there are going to be problems. Problems that do not occour when compiling from source. RPM's have not standard on naming packages either. Sometimes the same package is split up into 5 parts. How are you supposed to resolve dependencies then? With almost every binary package system, the second you install software not supplied by the vendor or from source, you are almost guaranteed to break dependencies. The gnu build system does not suffer from this. I think portage is an acceptable wrapper for it.
SCO website just a symbol...
on
SCO Offline
·
· Score: 5, Insightful
Realistically, who the hell even goes to the SCO website. They've got so few new potential customers anyway (I would put the number at zero). Current UnixWare users doubtfully visit their website very much anyway.
Their website being down is more of a symbol. A symbol to them of "Look at what they are doing to us". It's obviously not very important to them anyway seeing as how in the past they've taken it down for hours to days at a time for "server upgrades". If it were that critical to them, they wouldn't have had downtime. But it was cheaper to take it down and do what they needed to do to spend the money to keep it up during upgrades.
Anyway, SCO can eat apple sauce out of my ass with a spoon.
The problem is, many OEM's _can't_ sell Linux preinstalled on desktops without the strong hand of MS slamming down upon them. It's all or nothing. You either sell all Window's PC's or you're paying retail for for each copy. No OEM discount. and if you have a contract, we will sue you for breech of contract.
Not saying there's a huge demand, but if they offered it, people would buy it. Espically if it were the same PC but minus the cost of windows and office bundles they many times include. This would be very significant because for cheap PC's (some as low as a couple hundred bucks), shaving the price off of software could lower the price of a 300 dollar PC by as much as 50 dollars. So for many, the option of a PC that's around 20 percent less in price but is the same hardware, it would be a no brainer. Espically if OEM's started to give Linux a little nudge and put those PC's on the forefront.
These are all dreams though because OEM's have signed long term contracts with MS and when it's re-negotation time, it's either you run an all MS shop or you pay retail.
Slashdot Mirror
But what we need is one package format that everyone can use. RPM's are a pain in the ass to put onto a distribution that wasn't built from the ground up from RPM's. They have almost no standard naming system for packages. They have to be linked against the exact same minor version of libs once they are installed. RPM's should be dumped. Deb's and Tgz's have the same problem. We need a fresh package format that, instead of keeping a database of installed packages, uses a ./configure/autoconf like script to actually check for what's installed. Also, it needs NOT to use M4! It needs to be able to link against at least all minor versions it was linked against (maybe produce a warning to tell you).
These little things would take care of all the major issues with RPM/Deb/Tgz packages. A measure of backward compatability might be nice, but in all honesty, we are only inches further with binary package formats than we were 10 years ago!!!!
Am I going crazy or was this already posted like a week ago.
While we're on the subject about things we don't like with Gentoo ;)
It is very annoying that Gentoo's rc program was written in a compiled language. I feel as much as possible relating to init should be written in bash. I also don't like rc-update. Gentoo's init scripts are the most annoying things in the world. They aren't even bash scripts iirc. If you look at the top of one it's shebang is to another program (not running gentoo right now so I couldn't tell you which).
Best init scripts I think are the ones from Linux from Scratch. It uses symlinking (/etc/rc.d/rc3.d etc. etc.). It is basically all written in bash. The init scripts are simple as hell. It comes with a templete init script so you can write your own in the same style. Has external functions for cleanly loading and killing the process (loadproc and killproc). Gentoo's overly cluttered init system just leads to very messy init scripts. K.I.S.S.
What the hell are you talking about? Please the article, ok. make xconfig is just a way to configure your kernel options at compile time. THAT'S IT. Some people prefer having a graphical interface to it. It's COMPLETELY seperate and has nothing to do with the kernel itself. Most people have Xfree86 running on their box with either gtk or QT. Thus, the kernel developers gave those people another option for configuring their kernel. Most people have Ncurses too, so they gave those people make menuconfig. Some people don't have any of those, so they gave us make config. None of those have ANYTHING to do with the "over GUIfication of linux". Frankly, I don't know how the hell your parent post was ever modded up because you obviously have not RTFA.
There's actually a new kernel option that you can put the .config right into your kernel. That way say you delete your kernel source tree and forgot to save your .config (or just plain forgot where you saved it), or working on another system, you can just configure it based on your current running kernel.
Before you go off ranting and raving maybe you should read the article. What he's talking about is the addition of gtk and qt interfaces to make {x,g}config. Instead of using tk for xconfig, it now uses QT. It also adds the use of make gconfig which uses Gtk. omfgz these new additions are going to ruin Linux and there will be rape and pillage and GUIfication and it's not going to be an OS anymore because it switched from tk to qt to configure it while leaving config and menuconfig untouched and instead of legos it's going to be linkin logs omgz!!!!! k thx la~
I think this is the bigger story of the day. Basically, SCO is now claiming only those 17 header files as the "stolen" SysV code. They are now claiming they need to see the AIX source to be able to site exact copyright infringements.
IBM pointed out that SCO has publicly claimed HP-UX and Tru64 do not have infringing code. SCO then countered with "HP/Compaq have not contributed code to linux" to which IBM produced documents showing they did contribute. IBM went on to say, how could they prove they were not contributing code without seeing HP-UX/Tru64? ( I think they were trying to make the argument that if they can say what isn't infringing without seeing the source then they should be able to say exactly what is infringing without needing the AIX source)
Also interesing pointed out by IBM is during Darl's Harvard speech, he claimed that
"...[T]here is roughly a million lines of code that tie into contributions that IBM has made and that's subject to litigation that's going on. We have basically supplied that. In fact, that is going to be the subject of a hearing that comes up Friday..." (emphasis added.)
Yet they have only produced 17 files and will not actually say which lines in those files are infringing, but does say only parts of those files are infringing.
Even better quote...
''This guy,'' he proclaimed, ''is the best at Visual Basic.'
The funny thing is, in some circles this is considered a good thing.
I agree. Viruses are not hard to write. Espically the ones discussed in the article. It's really amauter hour. A virus (the type discussed in the article anyway) is just a program. The ones talking about in the article run just like any other legit program and require less skill to make one than a real program.
The REAL masters are those who find _new_ things. The MIT student who broke the xbox encryption with a belt sander. The ones whom run a program hundreds of times through gdb to find exploits. The ones who write a backdoor that can evade port scanners. Those are the real masters of the art. These are bored kids. They can come back to me when they've got the craftsmanship of a group like tcniso.
I can't access the trolltech right now (too slow) but I'm 99% sure that was true at one point, but NOT true anymore. The timeline iirc was it was free on windows, then not free, and sometime in the 3.x series it became free again. I can't confirm this though til their site lets up some.
So if they don't have any production ready server OS with 64-bit support, why even bother with desktops??? Desktop systems will see the least benefit by far. If I were MS, espically since they have no production ready win2k/2k3, I would be working damn hard to get a server version out.
I'm a little confused. Does MS have 64-bit versions of Win2k and Win2k3 available? I've heard plenty about this XP version, but nothing about their server versions (not to say it doesn't exist, I just haven't heard anything).
Instead of giving me cameras, games, hand jobs, with my phone; why can't you just give me a phone that works almost all the time and gets great reception?
I live in a moderatly sized city with plenty of cell phone vendors and (supposedly) a lot of towers. I can't even get reception with my phone in my own home, let alone when I'm at work inside a building built to withstand a hurricane (thick cement walls).
All this extra stuff is really an effort to divert the attention away from the base product, which, for most people, is still a sub standard product. Everyone I know (except those with Nextel's) has to jump through hoops to even write a grocery list down on their phone, let alone hold a real conversation. My phone was 215 dollars, has a big lcd screen, polyphonic ring tones, games, screen savers, internet access, but can not do what it was designed to do properly.
Last time I checked, I bought a phone, NOT A PDA! So why can't you give me a phone?
Should be noted Oregon State's involvement in this. Oregon State has provided A LOT of bandwidth to Linux projects including Linux from Scratch and Gentoo. They've got a very very big pipe there and have been more than generous with it there.
Yes, but he will still have to masterbate.
I can pretty much tell you who's going to be the fastest and slowest without a doubt. Gentoo fastest OpenBSD slowest (I don't even use Gentoo btw). All those are for the most part general purpose distros meant to run on a lot of hardware. They are not meant to run as fast as possible. They basically are compiled -O -march=i386. Gentoo being the exception since it's a source compiled distro and are going to tweak the hell out of the cc flags. I know this is going to make a lot of people with distro partialness mad, but general purpose distributions were not meant for speed. Gentoo wasn't designed with ease in mind, it was designed for speed.
As for OpenBSD, it's focus is on security. From personal experience, it's noticibly slower than any *nix I've ever used. Not to say it's bad, it's just not built for speed.
I think that list needs to be reworked considerbly. Maybe put them into categories such as source compiled distros, general purpose distros, security focused distros. Because comparing a source distro to a general purpose one is no contest. Not to say speed is everything, but it's really apples an oranges. Comparing one source distro to many general purpose ones is like racing a porshe and a honda. The porshe is going to win, but that's not what Hondas are built for.
I know a lot of people suggest switching to Mozilla, but it's not even about Mozilla. Almost every other browser is better than IE these days. Opera, Konqueror, Mozilla, (insert other browsers here). IE hasn't had anything significant in years. It lacks basic even basic things other browsers have. Pop up blocking, tabbed browsing, the ability to stop gif ads from looping, much better basic security policies, more w3c compiant, they don't make up their own html tags, plus many many more.
IE is so horrible. It's gotten to the point that just by using IE you are pretty much guaranteed to get spyware/adware/virus. Most of the people I know who use IE have their homepage changed daily, get a new toolbar every two days, so many pop ups they have to reboot weekly, their email stolen from cookies hourly, and a partridge in a pear tree.
If they really want to do something, fines are not the way to go. The only effective measure would be to make them modify Windows. In the US, because they weren't forced to modify windows' bundling, the fines levied really had no effect. In fact it was a positive experience for them because they paid their fines mostly in free software to schools, furthering their monopoly. If IE was seperated from the OS, then something might have actually change.
Reports were that the EU was considering seperating Windows Media Player from the OS. I think this would be a good start, but not enough. And a mere fine wouldn't be enough either.
What I think really needs to be done to break their monopoly.
a) Seperate Outlook Express, IE, Media Player, and Windows Messenger from the OS. Make them free downloads from Microsoft.com
b) OEM's shouldn't have such a hefty penalty for also selling competing OS's. All MS OEM contracts for desktop systems contain provisions that keep you from selling any other OS than windows. If you don't sign a contact, you pay retail. This needs to be done away with.
c) All file formats non-patentable and documents for the file format specs need to be made available publicly so competing software can be made interoperable.
I don't know German, but it sure looks like it was once there then redirects to a generic page. Maybe Bill Gates slammed his cock down on the table and demanded it be taken off. Said one man "Sure was tiny, and a hell of a laugh, but we all decided we should take it off anyway. You'd think with all that money he could afford to be circumsized".
The point I never get is... how does all this relate to KDE? KDE is OSS and falls under the GPL version of QT. It's not commercial in one bit.
I don't get this overwhelming support for gnome. On just technical merit, KDE is at least as good as Gnome, so all the fuss is basically over stupid QT license issues. KDE is a very well oiled project with regular releases and an awesome product. So why are only a select few distro's using it as the default WM?
Sometime last night they did change it actually. And then in the middle of the night (3AM EST or somewhere about) they added it again. So for a few hours it was gone. But for whatever reason brought it back.
You missed the point. Binary packages as a base install are fine, but NOT for upgrading software from 3rd parties. If you installed your kde from the CD then it's fine as a binary, but when you grab an rpm from freshmeat that was compilied against gcc 2.95 and you are running 3.4, there are going to be problems. Problems that do not occour when compiling from source. RPM's have not standard on naming packages either. Sometimes the same package is split up into 5 parts. How are you supposed to resolve dependencies then? With almost every binary package system, the second you install software not supplied by the vendor or from source, you are almost guaranteed to break dependencies. The gnu build system does not suffer from this. I think portage is an acceptable wrapper for it.
Realistically, who the hell even goes to the SCO website. They've got so few new potential customers anyway (I would put the number at zero). Current UnixWare users doubtfully visit their website very much anyway.
Their website being down is more of a symbol. A symbol to them of "Look at what they are doing to us". It's obviously not very important to them anyway seeing as how in the past they've taken it down for hours to days at a time for "server upgrades". If it were that critical to them, they wouldn't have had downtime. But it was cheaper to take it down and do what they needed to do to spend the money to keep it up during upgrades.
Anyway, SCO can eat apple sauce out of my ass with a spoon.
The problem is, many OEM's _can't_ sell Linux preinstalled on desktops without the strong hand of MS slamming down upon them. It's all or nothing. You either sell all Window's PC's or you're paying retail for for each copy. No OEM discount. and if you have a contract, we will sue you for breech of contract.
Not saying there's a huge demand, but if they offered it, people would buy it. Espically if it were the same PC but minus the cost of windows and office bundles they many times include. This would be very significant because for cheap PC's (some as low as a couple hundred bucks), shaving the price off of software could lower the price of a 300 dollar PC by as much as 50 dollars. So for many, the option of a PC that's around 20 percent less in price but is the same hardware, it would be a no brainer. Espically if OEM's started to give Linux a little nudge and put those PC's on the forefront.
These are all dreams though because OEM's have signed long term contracts with MS and when it's re-negotation time, it's either you run an all MS shop or you pay retail.