I agree, unfortunately. I've seen some packages that just have no business being released - the original packager threw something together in 10 minutes then went AWOL and ignored all the bug reports.
eg. Bug 280859. The packager forgot to package the runtime library FFS. 138 days old. Unfixed.
There's just no quality control on the packages, and that brings the whole distro down.
I had one instance where some clown had packaged a dev package so that it pulled in most of gnome 2. The library wasn't GUI related, the include files definately weren't GUI related... it was down to one optional binary that few people used anyway... I suggested weakening this to 'suggests' as it made the library essentially unusable to me (since to compile my app people would have had to install 50MB of junk) and just a got torrent of abuse back from the maintainer telling me I was 'stupid' for not having gnome (on my headless fileserver with no X).
Couple that with the X debacle (where debian is usually 6-12 months behind in releases, even in unstable) and I'm really looking for something better... unfortunately there are few other server distros out there (especially not using apt, which I wouldn't do without having tried others).
Have you even looked at both links? They're entirely different.
I even believed you for a minute... it's not like dups are uncommon on slashdot, but WTF has storing data on your own hard disk got to do with a dodgy research paper?
Macs are the wieredest ones for crashing... Mine sits on the desk doing nothing most of the time (it's mostly a build server for the osx versions of my software.. it's got nothing installed but the basic osx + xcode) - I'll come to it after maybe a fortnight and the finder has gone into 100% CPU mode and taken most of the rest of the system with it... luckily the power plug is only a couple of inches away:)
The Win boxes stay up when not in use. Of course when you start to use them the story changes... it's *so* easy to bring down a windows box (my favourite is crashing the LSA.. it gives a 30 second countdown before falling over). [btw. before anyone asks that's not a virus I write software that integrates at quite a low level and the Win32 API isn't error checked at all at that level - the slightest error brings the whole thing crashing down].
Linux is *hard* to crash. Not impossible (fork bomb, even with ulimit, can sometimes cause the autokill routines to kill system processes like inetd). A runaway app at 100% CPU though doesn't do it, unlike Win and Mac.
The prize goes to AS400 which I haven't even been able to make break its stride even after trying hard. I can lock an individual login, but never had any effect on the stability of the system. Pity AS400 is such a damn awful piece of crap to work with most of the time...
You can't - only the original authors can sign over their copyright.
If you fork then assign to fork to the FSF, only the code that you own would then belong to them (and since that's likely to be lots of individual lines of code within the codebase then good luck trying to get the FSF to defend the entire app...)
I've never seen a monitor with dead pixels.. are they really that common?
I just tested this laptop and the main desktop with the dead pixel test site and it proves there are definately none on either. I've also setup an office with 35 dell monitors without a single dead pixel on any.
OTOH it's not unusual to find zero dead pixel policies in this country.. I certainly wouldn't buy from anyone that didn't have one.
I prefer the laws we have here. It doesn't matter what you do with your money.. the state can take all of it and sell your assets to the highest bidder to recover it (the principle being money gained illegaly does not belong to you). They also have the right to empty bank accounts, retirement plans, etc. and for the big cases can sqeeze the offshore accounts too (easy against some countries, less so against others).
They use it a lot against drug dealers etc. using that against spammers (who are also gaining money by in an illegal manner) would be really nice.
Unfortunately Richter is in the US where all he has to do is claim he's compliant with the (I) CAN-SPAM (AS-MUCH-AS-I-LIKE) act and he's home free.
Corporate america is who buys Cisco kit which is expensive with a capital E, X and P!
(Their hardware is reasonable... want the firmware for your hardware? That's extra - the cost of the item again or (for their routers) many times that. Want *support*? Well, how big is your bank account? Oh dear. Can you get a mortgage?)
Yes they are.. it's quite common for packages to rely on other packages to be useful.
Any system that does not work with the existing dependency system is just not useful to me. Nice try, but no cigar.
At no point has anyone explained what the *point* of this is? Whoopee it's a frontend. Never use them myself, but what makes your system better than red carpet or aptitude?
And *don't* say read the frikkin FAQ. Your server is dead, I can't, and I'm damned if I'm coming back here every 20 minutes for the next two days until it comes back again..
If your bank is doing this, move you account. Seriously. I've *never* been charged for an ATM transaction, and I've used nearly every kind of ATM.
(There are some privately run ATMs around.. you can always tell them as they're freestanding and often have a modem wire sticking out of the back. Avoid these like the plague, because they do charge you).
You mean one that plays the songs in exactly the same order every day with no variation?
Smart playlists aren't so hot when you take this into account - the only way to re-randomize them is to plug them into itunes, and if I wanted to do that every day I'd have bought a flash based player rather than a 60GB Ipod.
I love it when americans say Minidisc didn't catch on. They mean they didn't catch on in America.
In the rest of the world they're really popular.
I liked my HiMD player... might go back to it as the audio quality is streets ahead of the ipod (due to the uneven bass response on the ipod there are some tunes I refuse to listen to on it). The thing that stops me is HiMD disks are damn near impossible to get, and the old 80minute ones aren't really long enough for an extended journey (you can maybe get 2 hours at decent quality).
If I'm listening on random (the shuffle doesn't support any other kind of course) and I like a tune I've not noticed before.. how do I find out who it is/what album/etc.? - by looking at the screen.
On the flipside Ipods aren't particularly good at using in your pocket - that damned wheel is a nightmare - no tactile feedback and it sometimes does odd things because it interprets your fumbling as an attempt to go into some random menu item.. they're designed to be pulled out of your pocket and used in daylight.
Damn I criticised apple on Slashdot - that's my karma blown.
Rechargables die. Recharging them from the USB port is a bad idea - they prefer a full recharge/discharge cycle (any rechargable can only be charged a limited number of times (around 2000, but it varies and if you get a bad one it could be around 1/10th of that).. if you're always charging it from 50% capacity you've effectively halved the useful life of the battery). If you plug into a dock 3 times a day thay's your Ipod bricked after 2 years.
I've gone the other way and only connect when it's absolutely necessary.. about once every two weeks when the battery dies. That should hopefully help it last a lot longer. I'd prefer an ipod that took normal batteries though.
I would not be prepared to allow my code to be relicensed with such a clause.
The BSD advertising clause was far less onerous - it only involved a line of copyright.
*forcing* source distribution even to those that have no access to binaries means someone could post a slashdot story like 'foosoft has a GPL light saber in development' and they would be forced to send their incomplete development code to 5 million slashdotters who asked for it. That could easily kill the project.
I'm actually really worried by the GPLv3 because it stands a good chance of defeating the whole purpose of copyleft.
eg. Someone takes my GPLv2 code with the 'or later' clause in it, and makes a GPLv3 derivative. GPLv2 is very likely to be incompatible with GPLv3, so I can't then take that code and put it back in my GPLv2 work... I've lost control of that particular code unless I move wholesale to GPLv3 (which I'm not inclined to do based on the information available so far.. as a protection I've started removing the 'or later' clause already).
This is in fact slightly worse than the BSD scenario because it allows someone to fork a new opensource project with your code and you can't do anything but stand and watch.
You can submit your code under GPLv2 or any later version... the project maintainer chooses to use GPLv2 and *only* GPLv2. That doesn't change the status of your code, only of the project that it is a part of.
If as seems likely GPLv3 is incompatible with GPLv2 (thereby forcing projects that use it to use it exclusively) the only issue would be submitting GPLv3 code to a GPLv2 only project, which would not be allowed.
Any substantial project is going to use a mixture of libraries from different sources. It probably has multiple contributors - many of whom you have no contact with any more.
Moving to GPLv3 is limited in that:
1. *All* the libraries you link with must be v3 compatible. 2. *All* the users that use your software must also be prepared to jump to v3 (otherwise you lose a large part of your userbase). 3. You must contact everyone who has ever contributed to your source code and ask their permission. If a single person objects you cannot change.
Dogs are similar. If a dog is about to run into you stand still... they instinctively run around static object (trees, etc.) but expect moving objects (rabbits, people, etc.) to get out of the way.
That saved me a few times when my dog had decided to dive into the river and then come back and say hi...
And those of us who have *paid* also have the right to remove the DRM once it gets to us. Sounds fair to me.
If you don't want to then fine... wait until you upgrade your computer and find that DRM has locked you out because you 'copied' the files to the new one.
*are* there any alternatives to/.? K5 went downhill and is now not worth looking it... anything else is just going to have all the slashdot groupthink and resident trolls all over again, so even if the articles were better the site may not improve.
Slashdot Mirror
I agree, unfortunately. I've seen some packages that just have no business being released - the original packager threw something together in 10 minutes then went AWOL and ignored all the bug reports.
eg. Bug 280859. The packager forgot to package the runtime library FFS. 138 days old. Unfixed.
There's just no quality control on the packages, and that brings the whole distro down.
I had one instance where some clown had packaged a dev package so that it pulled in most of gnome 2. The library wasn't GUI related, the include files definately weren't GUI related... it was down to one optional binary that few people used anyway... I suggested weakening this to 'suggests' as it made the library essentially unusable to me (since to compile my app people would have had to install 50MB of junk) and just a got torrent of abuse back from the maintainer telling me I was 'stupid' for not having gnome (on my headless fileserver with no X).
Couple that with the X debacle (where debian is usually 6-12 months behind in releases, even in unstable) and I'm really looking for something better... unfortunately there are few other server distros out there (especially not using apt, which I wouldn't do without having tried others).
No it isn't.
Have you even looked at both links? They're entirely different.
I even believed you for a minute... it's not like dups are uncommon on slashdot, but WTF has storing data on your own hard disk got to do with a dodgy research paper?
That was a dup too... they presumably deleted it and replaced it with another dup.
Only on slashdot....
Macs are the wieredest ones for crashing... Mine sits on the desk doing nothing most of the time (it's mostly a build server for the osx versions of my software.. it's got nothing installed but the basic osx + xcode) - I'll come to it after maybe a fortnight and the finder has gone into 100% CPU mode and taken most of the rest of the system with it... luckily the power plug is only a couple of inches away :)
The Win boxes stay up when not in use. Of course when you start to use them the story changes... it's *so* easy to bring down a windows box (my favourite is crashing the LSA.. it gives a 30 second countdown before falling over). [btw. before anyone asks that's not a virus I write software that integrates at quite a low level and the Win32 API isn't error checked at all at that level - the slightest error brings the whole thing crashing down].
Linux is *hard* to crash. Not impossible (fork bomb, even with ulimit, can sometimes cause the autokill routines to kill system processes like inetd). A runaway app at 100% CPU though doesn't do it, unlike Win and Mac.
The prize goes to AS400 which I haven't even been able to make break its stride even after trying hard. I can lock an individual login, but never had any effect on the stability of the system. Pity AS400 is such a damn awful piece of crap to work with most of the time...
You can't - only the original authors can sign over their copyright.
If you fork then assign to fork to the FSF, only the code that you own would then belong to them (and since that's likely to be lots of individual lines of code within the codebase then good luck trying to get the FSF to defend the entire app...)
I've never seen a monitor with dead pixels.. are they really that common?
I just tested this laptop and the main desktop with the dead pixel test site and it proves there are definately none on either. I've also setup an office with 35 dell monitors without a single dead pixel on any.
OTOH it's not unusual to find zero dead pixel policies in this country.. I certainly wouldn't buy from anyone that didn't have one.
I prefer the laws we have here. It doesn't matter what you do with your money.. the state can take all of it and sell your assets to the highest bidder to recover it (the principle being money gained illegaly does not belong to you). They also have the right to empty bank accounts, retirement plans, etc. and for the big cases can sqeeze the offshore accounts too (easy against some countries, less so against others).
They use it a lot against drug dealers etc. using that against spammers (who are also gaining money by in an illegal manner) would be really nice.
Unfortunately Richter is in the US where all he has to do is claim he's compliant with the (I) CAN-SPAM (AS-MUCH-AS-I-LIKE) act and he's home free.
Corporate america is who buys Cisco kit which is expensive with a capital E, X and P!
(Their hardware is reasonable... want the firmware for your hardware? That's extra - the cost of the item again or (for their routers) many times that. Want *support*? Well, how big is your bank account? Oh dear. Can you get a mortgage?)
Yes they are.. it's quite common for packages to rely on other packages to be useful.
Any system that does not work with the existing dependency system is just not useful to me. Nice try, but no cigar.
At no point has anyone explained what the *point* of this is? Whoopee it's a frontend. Never use them myself, but what makes your system better than red carpet or aptitude?
And *don't* say read the frikkin FAQ. Your server is dead, I can't, and I'm damned if I'm coming back here every 20 minutes for the next two days until it comes back again..
Huh?
If you ignore security, stability and some flexibility
The first two are *critical* to a webserver, and rule out IIS in the first sentence.
Where are you going that charges you?
If your bank is doing this, move you account. Seriously. I've *never* been charged for an ATM transaction, and I've used nearly every kind of ATM.
(There are some privately run ATMs around.. you can always tell them as they're freestanding and often have a modem wire sticking out of the back. Avoid these like the plague, because they do charge you).
playlists emulates a sorf of perfect radio DJ
You mean one that plays the songs in exactly the same order every day with no variation?
Smart playlists aren't so hot when you take this into account - the only way to re-randomize them is to plug them into itunes, and if I wanted to do that every day I'd have bought a flash based player rather than a 60GB Ipod.
I love it when americans say Minidisc didn't catch on. They mean they didn't catch on in America.
In the rest of the world they're really popular.
I liked my HiMD player... might go back to it as the audio quality is streets ahead of the ipod (due to the uneven bass response on the ipod there are some tunes I refuse to listen to on it). The thing that stops me is HiMD disks are damn near impossible to get, and the old 80minute ones aren't really long enough for an extended journey (you can maybe get 2 hours at decent quality).
A screen is definately a must.
If I'm listening on random (the shuffle doesn't support any other kind of course) and I like a tune I've not noticed before.. how do I find out who it is/what album/etc.? - by looking at the screen.
On the flipside Ipods aren't particularly good at using in your pocket - that damned wheel is a nightmare - no tactile feedback and it sometimes does odd things because it interprets your fumbling as an attempt to go into some random menu item.. they're designed to be pulled out of your pocket and used in daylight.
Damn I criticised apple on Slashdot - that's my karma blown.
Rechargables die. Recharging them from the USB port is a bad idea - they prefer a full recharge/discharge cycle (any rechargable can only be charged a limited number of times (around 2000, but it varies and if you get a bad one it could be around 1/10th of that).. if you're always charging it from 50% capacity you've effectively halved the useful life of the battery). If you plug into a dock 3 times a day thay's your Ipod bricked after 2 years.
I've gone the other way and only connect when it's absolutely necessary.. about once every two weeks when the battery dies. That should hopefully help it last a lot longer. I'd prefer an ipod that took normal batteries though.
So you're saying it reads ~50MB of data into memory in under 10 seconds? Allowing for 2-3 seconds spinup time that's a max. of 8 seconds.
That's pretty damned efficient - even my old Pentium doesn't get that data rate.
Odyssey5 awesome?? I'd like some of that crack you're smoking...
Its pretext was nice, but it should have been a mini series - after about the 3rd episode it had run out of ideas.
I would not be prepared to allow my code to be relicensed with such a clause.
The BSD advertising clause was far less onerous - it only involved a line of copyright.
*forcing* source distribution even to those that have no access to binaries means someone could post a slashdot story like 'foosoft has a GPL light saber in development' and they would be forced to send their incomplete development code to 5 million slashdotters who asked for it. That could easily kill the project.
I'm actually really worried by the GPLv3 because it stands a good chance of defeating the whole purpose of copyleft.
eg. Someone takes my GPLv2 code with the 'or later' clause in it, and makes a GPLv3 derivative. GPLv2 is very likely to be incompatible with GPLv3, so I can't then take that code and put it back in my GPLv2 work... I've lost control of that particular code unless I move wholesale to GPLv3 (which I'm not inclined to do based on the information available so far.. as a protection I've started removing the 'or later' clause already).
This is in fact slightly worse than the BSD scenario because it allows someone to fork a new opensource project with your code and you can't do anything but stand and watch.
There's a list of kernel developers but that isn't anything like the number of submitters.
Hell, there are a few lines of code in there that I wrote and I'm not listed.
To change the license you would have to get permission from *everyone* including people like me that once sent a patch via email 2 years ago.
You can submit your code under GPLv2 or any later version... the project maintainer chooses to use GPLv2 and *only* GPLv2. That doesn't change the status of your code, only of the project that it is a part of.
If as seems likely GPLv3 is incompatible with GPLv2 (thereby forcing projects that use it to use it exclusively) the only issue would be submitting GPLv3 code to a GPLv2 only project, which would not be allowed.
How?
Any substantial project is going to use a mixture of libraries from different sources. It probably has multiple contributors - many of whom you have no contact with any more.
Moving to GPLv3 is limited in that:
1. *All* the libraries you link with must be v3 compatible.
2. *All* the users that use your software must also be prepared to jump to v3 (otherwise you lose a large part of your userbase).
3. You must contact everyone who has ever contributed to your source code and ask their permission. If a single person objects you cannot change.
Dogs are similar. If a dog is about to run into you stand still... they instinctively run around static object (trees, etc.) but expect moving objects (rabbits, people, etc.) to get out of the way.
That saved me a few times when my dog had decided to dive into the river and then come back and say hi...
And those of us who have *paid* also have the right to remove the DRM once it gets to us. Sounds fair to me.
If you don't want to then fine... wait until you upgrade your computer and find that DRM has locked you out because you 'copied' the files to the new one.
*are* there any alternatives to /.? K5 went downhill and is now not worth looking it... anything else is just going to have all the slashdot groupthink and resident trolls all over again, so even if the articles were better the site may not improve.