Hmmmm. That URL gives a Forbidden error. In fact, all the stories on the Seattle Times site give forbidden errors. While I'm sure this is a temporary glitch, I was able to get at the story using their search engine. Hopefully this URL will work for other people.
The "ever paranoid" Debian folks, who have been rather paranoid about RPM
Is there anything you would like to say to further justify this comment? Both rpm and dpkg are, after all, GPL'd.
Check the changelogs - the earliest dpkg changelog entry puts it 3 months earlier than the earliest rpm entry (Which means that both projects started at about the same time, not because one didn't like what the other was doing).
Also remember that hamm had about 1500 packages, 2 architectures, and perpetually hovered around the 80 release critical bugs mark until the final push to get it out the door was made. It shipped 5 months after freeze.
The major holdup with hamm was libc6 (it caused a lot of breakage with some of the more seldom-used packages) and multiple architectures (debian had up until that point been x86 only).
The major holdup with slink is that it wouldn't fit on one binary CD anymore, and dpkg methods and CD creation scripts had to be written from scratch to accomodate this.
I'm very optimistic about potato. There are a lot of people focused on all aspects of release this time around. We had CD scripts months before freeze this time, boot floppies are in active development, the release critical buglist went down by 75 bugs last weekend (it's extremely active - about a dozen bugs being opened and closed daily - which for a pool that small is amazing), and in general more people are getting involved.
And yes, hindsight is 20/20:)
Re:Package managment is the best
on
Debian Freezing
·
· Score: 1
Speed - RPM is an order of magnitude faster at doing queries
True. Debian's state structures need to be overhauled (they are currently rfc822-style plaintext records as opposed to redhat's hashed db) but that must be done the Right Way(tm) - the state information is simply too important.
Why does dpkg not offer a way to view untruncated package names with -l ?
-l isn't meant to be used in scripts and the like. Try --get-selections.
Package names are goofy. In redhat it's easy to guess what a package name would be. apt-get kicks ass but would be even better if I could say "apt-get install svgalib" rather than having to know that the svgalib package name is really "svgalibg1"
Debian took the approach of having separate packages for incompatible library versions, that can both be installed simultaneously. Redhat took the approach of creating separate install and upgrade commands, so you can have multiple versions of the same package installed at the same time. Personally I like Debian's approach, and it makes tools like apt-get easier to implement. You can always go into dselect, type/svga, and find it thataway.
The fact that the install is based on apt-get makes it somewhat uncomfortable to install on a space-limitted machine, since the install insists on downloading all packages before it starts installing them. Use CD, right? Ergh, what if this is a laptop or something w/o a CD-Rom.
...but if you don't have all those packages downloaded beforehand, the chances of something going wrong are much greater - and it can be a lot tougher to fix things without those packages, especially if ppp or dhcpcd or ftp or whatever breaks and you can't get back on the net. Partial-upgrade functionality would be nice but it doesn't exist yet. (although you can do apt-get install package1 package2 package3; apt-get autoclean; lather rinse repeat until your fingers rot away)
Also, anything that uses a file:/ url in sources.list won't be cached - that includes remote filesystems such as nfs or smb. I dunno, I've found laptops without either some form of network connectivity or cdrom drives rather useless.
#1. Not yet. But the idea is to make debconf be able to answer questions from an SQL database, a file accessible via HTTP, whatever, to replace asking questions at the console.
#2. This is open to discussion, although most people with an opinion on the subject say apt-get/dpkg eats rpm for lunch. Both have severe deficiencies, but to me dpkg has fewer of them.
For the equivalent solution with apt-get, you could make a directory on a server somewhere, run dpkg-scanpackages when you update something on that server (this creates the Packages file), and then have apt-get update; apt-get install packagename run every night from cron. That would update that package only.
Or, if you don't have any complicated dependancies requiring automatic installation ordering (it doesn't sound like that) you could replace the 'rpm -U *.rpm' in your cron script with 'dpkg -i *.deb':)
For years, Debian has maintained that it would be better for a developer to fork the distribution if you don't agree with them in a key political area (for example, the DFSG). However, it seems Corel was the only one who forked for this reason (Debian wouldn't have accepted wordperfect.debs into non-free without written permission that anyone -- like, oh, our mirror administrators -- can redistribute them).
Perhaps:
these other developers would just like their own sandbox to hack around in? This seems to be the case with Stormix, the developers have been associated with Debian for quite some time and the relationship between Stormix and Debian is fairly amicable.
they want to make and sell their own branded commercial distribution? (this would include both Stormix and Corel)
they want the fame of having their very own distribution?
some part of debian has become technically or politically hostile to them and they wish to fork because of that?
If it's #1, the best case would be to have the changes from their modifications put back into Debian itself, then everyone profits. If it's #2, more power to ya. If it's #3, sorry, you're going to be another also-ran. If it's #4, then communication with an aspect of the Debian project has quite possibly failed. Typically these things can be worked out and a concensus reached, unless it's one of those things that just isn't open for discussion (like the DFSG). And there are a LOT of things you can do without being a developer (for example, take a look at http://master.debian.org/~wakkerma/bugs/, nail a few of them, and email patches to nnnnn@bugs.debian.org)
I'm sure there's lots of other reasons too. Whatever they are, more discussion about them couldn't hurt.
Two things are gonna make this a royal pain for people trying to make their own DVD's:
Filesize limit on 32-bit linux means it gets tough to create 5GB dvd images. You'll have to do imaging from an Alpha, Sparc64, MIPS64, or Itanium box.
Outrageous prices on DVD-burning/pressing equipment
Also, to answer another poster, the potato freeze is being postponed until at least Nov 7, basically the holdup is the boot floppies - it's a bad idea to go into a freeze without working boot floppies.
56K is the maximum that the FCC (and, AFAIK, the rules of physics) allows on a normal analog line.
Both false. The FCC only regulates voltage.
The rules of physics say you can get at least 7 mbits/sec through a phone line (contact your favorite telcos for info on xDSL), and much more for short distances (raise your hand if you've ever made a working 100mbit ethernet cable out of cat1 or cat3 cable).
However, our current understanding of the rules of physics limits our ability to transmit more than 53000 symbols (bits) per second down the average extremely-poorly-strung non-shielded 26awg cable by modulating it at a limited amplitude (I believe it's about 5-6 volts)
Also, to respond to another poster, compression typically increases latency.
What happens down the road when twenty other special exemptions have to be granted, do we see the license extended by another length of the GPL?
The copyright holder can license the code in any way he/she wishes. In this case it appears Jason Gunthorpe considers being able to link to Qt a feature instead of a bug.
And this exemption is being granted only to a commercial entity.
No, the exemption was made so that anyone could link libapt-2.5 against Qt. It was not Corel-specific, that would be silly, if that was the case Debian itself wouldn't be able to use Corel's installer. (this is why DFSG rules #5, #6, and #8 exist).
So a corporation like Corel can just call up Debian on the phone and instantly get an exemption while an unknown student would not be able to freely study the library code to re-use in his projects throughout his latter career?
I don't get it. The exemption applies to everyone, including J. Random Student. Plus, even before the exemption, the library was GPL'd, so the student would have ample provisions to study and re-use it in his projects. He would just have to GPL them. In fact, he would still have to do that.
BTW, you don't just "call up Debian on the phone". You call up (or email) a developer. We're volunteers, we don't have a receptionist, let alone an office. And if there's any license changes to be done, it'll typically be done in writing, clearly documented in the package (/usr/doc/packagename/copyright) in a fresh new release.
20000 geeks' livelihoods depend directly upon Microsoft.
Hehehehehe. I live in that area, and I can say without a doubt that if Microsoft dropped off the face of the earth tomorrow, there'd still be jobs for all of them, and then some. There would also be a lot of suddenly-independant millionaires with nothing better to do with their time than start up their own company.
Hello, about a year ago my wrists started aching (sharp pains made it very uncomfortable to even drive home, I had no idea RSI could become a crippling problem in just one afternoon) I've switched to dvorak on several machines. Here's what I've experienced:
Fewer typos
About the same speed (I'm only a 70wpm typist in the first place)
Wrists have stopped hurting
It takes about 5 seconds for my brain to mentally switch from dvorak to qwerty and back again, it's not a big deal
It's a great way to keep people from messing with your machine
For the first week or so, I could only type about 5wpm on dvorak (look-at-chart-then-type-letter mode) but the second week I was at 10 and the third week I was at 20. A month after I switched I was at 35 (fast enough to be useful) and a month after that I was at full speed again.
After you reach this point it really doesn't matter what the keys on the keyboard are labeled - in fact I'm typing this right now on a dvorak-mapped qwerty-labeled Microsoft Natural keyboard.
Also, you'll learn much faster if you can resist the urge to switch back to qwerty for every huge task that comes up (just hang out on IRC for a while in dvorak mode, a few afternoons will suffice)
I've also strapped a foam pad to the back of my mouse to support my wrist (keep it level and stationary instead of on the desk), at least on the windows machines I have to use. Since I'm a console junkie the mouse doesn't get used in linux much.
Your license is valid, however you should be aware that it fails the Debian Free Software Guidelines, specifically clause #5:
5. No Discrimination Against Persons or Groups
The license must not discriminate against any person or group of persons.
Thank goodness we have this requirement, lest the distribution become unusable by anyone (government, outlaws, non-technical users, non-hams, business of any sort, you name it, there's a license out there that excludes it)
am I obligated to return the "changed" source at the end of each day? Or can I decide (okay, I've made enough progress to release my code).
If you distribute daily binaries, you should distribute daily source as well, but you don't have to - you can simply make it available upon request instead. Anyway, if you never distribute your changes as binaries, you don't have to ever distribute the source either.
By the way, the GPL doesn't claim to give you the right to 'release' things, it gives you the right to 'distribute' them.
For instance: Rob doesn't post a new tarball everyday when he fixes a minor glitch here or there. He makes releases as he has time and when he thinks he's made enough progress to merit it.
Rob is the copyright holder - he can do whatever he wants with it.
Both radio and lasers are electromagnetic energy. Radio is just much lower frequency (visible light is about 10 terahertz IIRC)
So far us humans haven't been able to control the modulation at anything near 10thz so we just do CW on it, which still gets a reasonable symbol rate.
The thing about that end of the electromagnetic spectrum is that it is extremely line-of-sight. You have direct experience with this, if you've ever seen a shadow.
Also even at much lower frequencies the line-of-sightness of radio becomes a problem. You will get blackouts from eclipses. We still don't have a way to get radio signals to the dark side of the moon.
You also run into the problem of even a perfect reciever not being able to detect a signal amongst the randomness of space, even with extremely directional antennas (you think they put dishes on those things cuz they look cool?:) you're attempting to radiate a signal to an area several million km^2 across and ten square meters of antenna gain area simply will not cut it.
Apparently this won't be a huge problem, since we are still recieving Voyager transmissions, but it might be an argument for lasers assuming the tight directional tolerances can be accomodated.
You can believe in evolution (as in macroevolution or Speciation), or you can believe in creation (as in god created stuff) but until either one can be proven, I think both should be kept out of our schools. It's simply not science.
I'm just concerned what the board who decided this will do next. This is a move towards neutrality, I'd hate to see them carry it too far.
I wasn't aware the ISO reference code had a license on it. Can someone point me to something that says people can use and distribute it or something that says it's public domain?
No matter what algorithm you use to hash the IP, it can be brute-force cracked fairly quickly.
The reason for this is that there are only 2^32 IP addresses with IPv4, and you don't have to search a good percentage of those (0.0.0.0/8 isn't allocated, 127.0.0.0/8 is reserved, 224.0.0.0/3 is reserved for experimental stuff including multicast...)
If it is indeed the case that it's software-flashed imagery, I'm going to laugh for hours when the Wired/CNet/Salon article comes out reporting that the image server has been hacked and tens of thousands of computers are now displaying whatever pics Hacking For Girlies thought looked cool at the time.
Slashdot Mirror
Hmmmm. That URL gives a Forbidden error. In fact, all the stories on the Seattle Times site give forbidden errors. While I'm sure this is a temporary glitch, I was able to get at the story using their search engine. Hopefully this URL will work for other people.
The kernel-2.2.13 images in potato currently ship with every sound driver known to man modularized.
In fact, the kernel-2.2.13 images in potato don't depend on anything special, so you can drop them right into a slink install. Just read http://www.debian.org/relea ses/slink/running-kernel-2.2 first.
The bug tracking system, especially the release-critical bug list, and the debian-boot mailing list.
where do I send updates?
Email them to bugnumber@bugs.debian.org
Is there anything you would like to say to further justify this comment? Both rpm and dpkg are, after all, GPL'd.
Check the changelogs - the earliest dpkg changelog entry puts it 3 months earlier than the earliest rpm entry (Which means that both projects started at about the same time, not because one didn't like what the other was doing).
The major holdup with hamm was libc6 (it caused a lot of breakage with some of the more seldom-used packages) and multiple architectures (debian had up until that point been x86 only).
The major holdup with slink is that it wouldn't fit on one binary CD anymore, and dpkg methods and CD creation scripts had to be written from scratch to accomodate this.
I'm very optimistic about potato. There are a lot of people focused on all aspects of release this time around. We had CD scripts months before freeze this time, boot floppies are in active development, the release critical buglist went down by 75 bugs last weekend (it's extremely active - about a dozen bugs being opened and closed daily - which for a pool that small is amazing), and in general more people are getting involved.
And yes, hindsight is 20/20 :)
True. Debian's state structures need to be overhauled (they are currently rfc822-style plaintext records as opposed to redhat's hashed db) but that must be done the Right Way(tm) - the state information is simply too important.
Why does dpkg not offer a way to view untruncated package names with -l ?
-l isn't meant to be used in scripts and the like. Try --get-selections.
Package names are goofy. In redhat it's easy to guess what a package name would be. apt-get kicks ass but would be even better if I could say "apt-get install svgalib" rather than having to know that the svgalib package name is really "svgalibg1"
Debian took the approach of having separate packages for incompatible library versions, that can both be installed simultaneously. Redhat took the approach of creating separate install and upgrade commands, so you can have multiple versions of the same package installed at the same time. Personally I like Debian's approach, and it makes tools like apt-get easier to implement. You can always go into dselect, type /svga, and find it thataway.
The fact that the install is based on apt-get makes it somewhat uncomfortable to install on a space-limitted machine, since the install insists on downloading all packages before it starts installing them. Use CD, right? Ergh, what if this is a laptop or something w/o a CD-Rom.
Also, anything that uses a file:/ url in sources.list won't be cached - that includes remote filesystems such as nfs or smb.
I dunno, I've found laptops without either some form of network connectivity or cdrom drives rather useless.
#2. This is open to discussion, although most people with an opinion on the subject say apt-get/dpkg eats rpm for lunch. Both have severe deficiencies, but to me dpkg has fewer of them.
For the equivalent solution with apt-get, you could make a directory on a server somewhere, run dpkg-scanpackages when you update something on that server (this creates the Packages file), and then have apt-get update; apt-get install packagename run every night from cron. That would update that package only.
Or, if you don't have any complicated dependancies requiring automatic installation ordering (it doesn't sound like that) you could replace the 'rpm -U *.rpm' in your cron script with 'dpkg -i *.deb' :)
For years, Debian has maintained that it would be better for a developer to fork the distribution if you don't agree with them in a key political area (for example, the DFSG). However, it seems Corel was the only one who forked for this reason (Debian wouldn't have accepted wordperfect .debs into non-free without written permission that anyone -- like, oh, our mirror administrators -- can redistribute them).
Perhaps:
If it's #1, the best case would be to have the changes from their modifications put back into Debian itself, then everyone profits.
If it's #2, more power to ya.
If it's #3, sorry, you're going to be another also-ran.
If it's #4, then communication with an aspect of the Debian project has quite possibly failed. Typically these things can be worked out and a concensus reached, unless it's one of those things that just isn't open for discussion (like the DFSG). And there are a LOT of things you can do without being a developer (for example, take a look at http://master.debian.org/~wakkerma/bugs/, nail a few of them, and email patches to nnnnn@bugs.debian.org)
I'm sure there's lots of other reasons too. Whatever they are, more discussion about them couldn't hurt.
Also, to answer another poster, the potato freeze is being postponed until at least Nov 7, basically the holdup is the boot floppies - it's a bad idea to go into a freeze without working boot floppies.
Both false. The FCC only regulates voltage.
The rules of physics say you can get at least 7 mbits/sec through a phone line (contact your favorite telcos for info on xDSL), and much more for short distances (raise your hand if you've ever made a working 100mbit ethernet cable out of cat1 or cat3 cable).
However, our current understanding of the rules of physics limits our ability to transmit more than 53000 symbols (bits) per second down the average extremely-poorly-strung non-shielded 26awg cable by modulating it at a limited amplitude (I believe it's about 5-6 volts)
Also, to respond to another poster, compression typically increases latency.
So far out of that list I've used frantica, maxima, abnorma, awfu, musica, termina, fata, norma, individua, geographica, idea, and sexua.
The possibilities are endless.
The copyright holder can license the code in any way he/she wishes. In this case it appears Jason Gunthorpe considers being able to link to Qt a feature instead of a bug.
And this exemption is being granted only to a commercial entity.
No, the exemption was made so that anyone could link libapt-2.5 against Qt. It was not Corel-specific, that would be silly, if that was the case Debian itself wouldn't be able to use Corel's installer. (this is why DFSG rules #5, #6, and #8 exist).
So a corporation like Corel can just call up Debian on the phone and instantly get an exemption while an unknown student would not be able to freely study the library code to re-use in his projects throughout his latter career?
I don't get it. The exemption applies to everyone, including J. Random Student. Plus, even before the exemption, the library was GPL'd, so the student would have ample provisions to study and re-use it in his projects. He would just have to GPL them. In fact, he would still have to do that.
BTW, you don't just "call up Debian on the phone". You call up (or email) a developer. We're volunteers, we don't have a receptionist, let alone an office. And if there's any license changes to be done, it'll typically be done in writing, clearly documented in the package (/usr/doc/packagename/copyright) in a fresh new release.
Redhat has similar guidelines.
I'd imagine SuSE and Caldera might wish to bundle such a thing if it was made available to them. Corel wouldn't, it'd be competing with Wine.
But that's not enough to "embrace" the linux distribution market.
Hehehehehe. I live in that area, and I can say without a doubt that if Microsoft dropped off the face of the earth tomorrow, there'd still be jobs for all of them, and then some. There would also be a lot of suddenly-independant millionaires with nothing better to do with their time than start up their own company.
For the first week or so, I could only type about 5wpm on dvorak (look-at-chart-then-type-letter mode) but the second week I was at 10 and the third week I was at 20. A month after I switched I was at 35 (fast enough to be useful) and a month after that I was at full speed again.
After you reach this point it really doesn't matter what the keys on the keyboard are labeled - in fact I'm typing this right now on a dvorak-mapped qwerty-labeled Microsoft Natural keyboard.
Also, you'll learn much faster if you can resist the urge to switch back to qwerty for every huge task that comes up (just hang out on IRC for a while in dvorak mode, a few afternoons will suffice)
I've also strapped a foam pad to the back of my mouse to support my wrist (keep it level and stationary instead of on the desk), at least on the windows machines I have to use. Since I'm a console junkie the mouse doesn't get used in linux much.
5. No Discrimination Against Persons or Groups
The license must not discriminate against any person or group of persons.
Thank goodness we have this requirement, lest the distribution become unusable by anyone (government, outlaws, non-technical users, non-hams, business of any sort, you name it, there's a license out there that excludes it)
for x in `cat binarydata`; do printf \\x`echo -e "obase=16\nibase=2\n$x" | bc`; done
am I obligated to return the "changed" source at the end of each day? Or can I decide (okay, I've made enough progress to release my code).
If you distribute daily binaries, you should distribute daily source as well, but you don't have to - you can simply make it available upon request instead. Anyway, if you never distribute your changes as binaries, you don't have to ever distribute the source either.
By the way, the GPL doesn't claim to give you the right to 'release' things, it gives you the right to 'distribute' them.
For instance: Rob doesn't post a new tarball everyday when he fixes a minor glitch here or there. He makes releases as he has time and when he thinks he's made enough progress to merit it.
Rob is the copyright holder - he can do whatever he wants with it.
So far us humans haven't been able to control the modulation at anything near 10thz so we just do CW on it, which still gets a reasonable symbol rate.
The thing about that end of the electromagnetic spectrum is that it is extremely line-of-sight. You have direct experience with this, if you've ever seen a shadow.
Also even at much lower frequencies the line-of-sightness of radio becomes a problem. You will get blackouts from eclipses. We still don't have a way to get radio signals to the dark side of the moon.
You also run into the problem of even a perfect reciever not being able to detect a signal amongst the randomness of space, even with extremely directional antennas (you think they put dishes on those things cuz they look cool? :) you're attempting to radiate a signal to an area several million km^2 across and ten square meters of antenna gain area simply will not cut it.
Apparently this won't be a huge problem, since we are still recieving Voyager transmissions, but it might be an argument for lasers assuming the tight directional tolerances can be accomodated.
Grabbing an entire CD is as simple as 'cdgrab'.
Not only does it use CDDB, it gives you a chance to correct its output as well.
Encodes in the background while ripping.
Includes SMP support (runs two or more encoders at once) for those who like to push the envelope. Just do 'cdgrab -j 2'.
Lets you customize the output scheme (defaults to Artist_Name/Track_Name.mp3)
Optionally creates playlists in original track order.
In Debian Potato NOW! :)
Of course, being the author, I'm biased
Check it out at http://packages.debian.org/cdgrab
I'm just concerned what the board who decided this will do next. This is a move towards neutrality, I'd hate to see them carry it too far.
Indeed, LAME rocks. cdgrab, the world's greatest cd ripping program :>, defaults to it :)
Just curious :)
The reason for this is that there are only 2^32 IP addresses with IPv4, and you don't have to search a good percentage of those (0.0.0.0/8 isn't allocated, 127.0.0.0/8 is reserved, 224.0.0.0/3 is reserved for experimental stuff including multicast...)
This is hilarious.