for developers to locate errors and warnings in code, as well as specify user-defined tasks. The task list is updated in "real time" as the developer completes tasks and generates new tasks.
while [ TRUE ] ; do sleep 1; grep..blah blah.. ; done
Take what is probably the best distro out there and preconfigure it and add some administration utilities and you get the best of both worlds.
I was thinking of something along these lines a while back. Something like a sane defaults Debian install where you pop in a CD, hit return once and an image installs to the first hard drive wiping it clean. The discover package would do all the heavy lifting of hardware configuration, use DHCP by defualt, X would be set to use VESA by default and GDM/KDM would be set to autologin (with a script to prompt for new user and root passwords on the first boot.)
Knoppix is great, but it's a PITA to pull all of the Knoppix specific packages out of an installed system so that you have a pure Debian box.
We sure won't get a great patent system out of these guys, but we may get a better one. After all, most companies don't bother enforcing patents violated by individuals - they enforce those violated by other companies.
Big companies with large patent portfolios have no business rewriting the patent laws since they have a vested interest in keeping some of the current flaws. In particular, they benefit from using cross licensing to form oligopolies and keep out new competition. The only flaw they'll be interested in fixing is the one that allows small independent developers to cash in on goofy patents at their expense.
This should be a lesson to VC funds and financing operations that finance companies whose business models are built upon legal action and sucking off the hard work and sweat of people who make ideas work through the creation of products that improve our lives.
Or it could be a lesson from Microsoft on how to pump money into a company then remove the resulting debt from the recipient's balance sheet.
I wouldn't be surprised. Crime is always considered high by the populace, and the most obvious solution is always to increase the penalty. Not that it always works.
I wouldn't say that increasing the penalties is the most obvious way of addressing crime (more money for envorcement would be the most obvious way), increasing penalties is simply the CHEAPEST way of addressing crime in the short run. If a politician decides to make not wearing your seat belt punishable by life in prison it will take years for the true cost of that change to start accumulating. The long run cost will be either more prison space or more paroles, neither of which the public will like, but the people changing the sentencing guidelines are not the same ones that will be around when those costs have to be addressed.
As soon as Monsantos seeds contaminate the world's food supply, they'll own it all.
That's the very thing I was thinking. If this ruling is valid then wouldn't it be in the best interest of patent holder to get the gene out in the wild?
Hmm. I don't have an IGP, but my understanding is that it should work properly with the dri-trunk-sid packages even though there have been some recent fixes. You did build the drm-trunk-module package and install it right?
OTOH, wait, maybe I don't want the 3D... when I used those drivers for ATI cards, I'd randomly wake up in the morning to find my or my wife's computer crashed hard in the midst of running one of the spiffy acceleratophilic screensavers.
Can anyone out there recommend a graphics card with decent performance and Open Source drivers?
To be honest, I had tried DRI many times in the past with similar problems. They usually came down to a rogue libGL.so getting into the mix somewhere, and ultimately the solution was to make absoloutely sure that all of the relevant components (drm, dri, xlibmesa) came from the exact same source. I would suggest installing from scratch and getting the minimum components to set up DRI properly. Some sort of sanity checker for DRI (like Nvidia has) would also be nice.
NVidia seems like the only choice for me, since only ATI's most recent line of products (ie: expensive products) are supported.
DRI covers virtually all of the ATI chips up to the 9200. FGLRX covers everything from the Radeon 8500 to 9800.
The only remaining problem is that some commercial developers just assume Nvidia is the standard. DRI in particular has come a long way. I've been using a Radeon 8500 with Michael Daenzer's DRI packages for Debian with few problems. Out of 30 commercial Linux games I have, 3 don't work properly: Descent 3, Savage and Heavy Gear 2. Savage's problem is trivial to fix, the others I'm not sure about.
I'd rather see ATI release the 3D specs on the R300 chips than see further improvements in the FGLRX driver.
First response, scrolled down a few pages till I saw 'back care' highlighted.. found the name, Janet Dean. Less than a minute! These people are not very good at their Google;-)
You should haved clicked on the "view as html" link then hit Ctrl-F then typed "back care".
Few seem to realize that the questions that start popping up at the end of installation are not coming from the installer. In Debian each package can ask questions to configuring the software. When someone goes hog-wild on TaskSel and marks 1000+ packages for installation, there are a lot of packages wanting to ask questions. The secret is to skip tasksel and use apt-get to pull in only the packages that are needed.
Needless to say, packages that abuse the Debconf system by popping up stupid or misleading questions are broken. There are often flame-wars on Debian-Devel about what types of questions are worth asking. In some cases (sendmail, xfree86, etc) the ability of packages to ask questions while configuring is a godsend. In other cases (lwp, etc) the questions are totally pointless and anyone who needs the functionality will know how to set it up.
What happens if you install a text-only RedHat or Mandrake system, then install XFree86 after the fact? Does X get configured properly? If you look at it from this perspective, allowing packages to ask questions does make a certain amount of sense.
Personally I don't really care about glitches, crashes and other problems with the machines. What I do care about is the use of uncertified software and the fact that these companies are more or less getting away with it. It sets a bad precedent for the future. Who cares if a few voting machines get decertified if you get to rig an election as a result? Any use of uncertified software should bar that company from ever producing voting machines in the US again. Do we really have to wait until someone is caught rigging a major election before real efforts are undertaken to stop it?
Next soundcard i'm going to buy will be by a company that actively supports linux or opens up the specifications: a product that can't work with both the OSs i use is a crippled one.
The only soundcard I ever use with Linux is the SoundBlaster Live. The first time you see goofy app #1 playing sound through artsd, goofy app #2 playing sound throguh esd, and goofy app #3 playing sound directly through/dev/dsp, all at the same time, you'll realize that the SB Live is the standard by which everything else should be measured. Are there any cards other than the SB Live and SB Audigy that can leave/dev/dsp unlocked no matter how many sounds are playing at the same time? Software mixing is terrible in comparison.
I believe that most of the old BSD Unix has been publicly released already, so any code from such a source is free and clear.
No. Anything before 4.4BSD-Lite is questionable. You might be risk using the older stuff on the basis that Caldera gave away the ancient UNIX code it was based on, but considering what bastards they are now it's probably not a good idea. If you're taking anything from BSD, make sure it's from a version released after the BSD settlement. Everything prior to that required a UNIX license.
I would very much like to know what methods they used to review the code. Interpreting the results of any comparison to remove false positives is a very subjective process. Do they have a method that doesn't generate false positives? If so, did their method find the code that SGI admitted to contributing by mistake?
To be honest, anyone claiming there is absoloutely no questionable code sounds as suspect as SCO claiming it's all questionable code. Reality is never black and white like that.
Of course, at this point it's obvious that any SCO code in Linux is trivial and unintentional. If there were millions of lines of SCO code in Linux it would be impossible to overlook. It remains to be seen how much GPL code has made its way into SCO's UNIX offerings.
This "amateur robotics" competition is really more about manufacturing skill and hand-eye coordination, not advancing technology or even implementing it in new or creatively useful ways. I see little practical difference between this and the Boy Scouts' Pine Derby.
I do wish they would stop calling anything remotely controlled a robot. By that definition "robots" have been common since the 1930s when the first RC model airplanes were created. Considering the price tags that are being mentioned, balsa wood planes actually sound like a better route since each student could draft, build, and pilot his or her own unique design for a small fraction of the cost.
Truth be told though, I was in Odyssey of the Mind, math relay and other goofy school competitions years back and I did enjoy it even if nothing substantial was learned in the process.
I design software for embedded devices and let me tell you, as soon as you add a password mechanism, then someone will lose the password within days. It's happened to me, and I finally had to put a global password in every machine.
I didn't intend for that comment to be a troll. I looked at the code because I was hoping it would be written in C++/MFC and thus be interesting to rebuild with WxWidgets, even if the end result is still only useful on Windows. Even though it's not useful to me, the Mono team might still be interested in using all or part of it as a test case for their C# implimentation.
I would be shocked if there were many large commercial projects that don't have stolen OSS code in them...
Only if it's free software that you wrote and you didn't assign the copyright to anyone else, reusing it in a commercial product is not stealing...the code still belongs to you and you can relicense it to your new employers under whatever terms you'd like.
Re:What, no editorial?
on
Red Hat Recap
·
· Score: 1
What that paragraph states is that you CAN put the software on as many computers as you want, but if you do you're not getting any tech support from Redhat. This is the support contract, not the software contract.
Right, but as I already pointed out, buying the software means purchasing a support contract. Imagine if RedHat didn't make the SRPMS available to anyone who hasn't paid for RHAS. If the support contract that each of those customers has agreed to prevents them from distributing the software, then an additional restriction has been placed on the software regardless of how far down in the chain of licenses it takes to reach that point.
Re:What, no editorial?
on
Red Hat Recap
·
· Score: 4, Informative
Read the EULA. It's all there.
You link to the wrong license agreement. RedHat uses the nice license for the software, and the nasty license for the support contract. Of course, RedHat will only sell you the software with a support contract, so the support contract terms apply to anyone wanting to purchase RHEL.
The support contract plainly states: "If Customer wishes to increase the number of Installed System, then Customer will purchase from Red Hat additional Services for each additional Installed System. During the term of this Agreement and for one (1) year thereafter, Customer expressly grants to Red Hat the right to audit Customer's facilities and records from time to time in order to verify Customer's compliance with the terms and conditions of this Agreement."
The end result, of course, is that you can't buy RHAS without giving up rights the GPL explicitly gives you.
Slashdot Mirror
for developers to locate errors and warnings in code, as well as specify user-defined tasks. The task list is updated in "real time" as the developer completes tasks and generates new tasks.
..blah blah.. ; done
while [ TRUE ] ; do sleep 1; grep
Take what is probably the best distro out there and preconfigure it and add some administration utilities and you get the best of both worlds.
I was thinking of something along these lines a while back. Something like a sane defaults Debian install where you pop in a CD, hit return once and an image installs to the first hard drive wiping it clean. The discover package would do all the heavy lifting of hardware configuration, use DHCP by defualt, X would be set to use VESA by default and GDM/KDM would be set to autologin (with a script to prompt for new user and root passwords on the first boot.)
Knoppix is great, but it's a PITA to pull all of the Knoppix specific packages out of an installed system so that you have a pure Debian box.
We sure won't get a great patent system out of these guys, but we may get a better one. After all, most companies don't bother enforcing patents violated by individuals - they enforce those violated by other companies.
Big companies with large patent portfolios have no business rewriting the patent laws since they have a vested interest in keeping some of the current flaws. In particular, they benefit from using cross licensing to form oligopolies and keep out new competition. The only flaw they'll be interested in fixing is the one that allows small independent developers to cash in on goofy patents at their expense.
This should be a lesson to VC funds and financing operations that finance companies whose business models are built upon legal action and sucking off the hard work and sweat of people who make ideas work through the creation of products that improve our lives.
Or it could be a lesson from Microsoft on how to pump money into a company then remove the resulting debt from the recipient's balance sheet.
I wouldn't be surprised. Crime is always considered high by the populace, and the most obvious solution is always to increase the penalty. Not that it always works.
I wouldn't say that increasing the penalties is the most obvious way of addressing crime (more money for envorcement would be the most obvious way), increasing penalties is simply the CHEAPEST way of addressing crime in the short run. If a politician decides to make not wearing your seat belt punishable by life in prison it will take years for the true cost of that change to start accumulating. The long run cost will be either more prison space or more paroles, neither of which the public will like, but the people changing the sentencing guidelines are not the same ones that will be around when those costs have to be addressed.
As soon as Monsantos seeds contaminate the world's food supply, they'll own it all.
That's the very thing I was thinking. If this ruling is valid then wouldn't it be in the best interest of patent holder to get the gene out in the wild?
Well, call me crazy, but how can surfing on lava not save a movie? Is there a better way short of two car chases in a single movie?
A surfing car chase scene like in Escape from LA?
Hmm. I don't have an IGP, but my understanding is that it should work properly with the dri-trunk-sid packages even though there have been some recent fixes. You did build the drm-trunk-module package and install it right?
You might also try here or here.
OTOH, wait, maybe I don't want the 3D... when I used those drivers for ATI cards, I'd randomly wake up in the morning to find my or my wife's computer crashed hard in the midst of running one of the spiffy acceleratophilic screensavers.
Can anyone out there recommend a graphics card with decent performance and Open Source drivers?
To be honest, I had tried DRI many times in the past with similar problems. They usually came down to a rogue libGL.so getting into the mix somewhere, and ultimately the solution was to make absoloutely sure that all of the relevant components (drm, dri, xlibmesa) came from the exact same source. I would suggest installing from scratch and getting the minimum components to set up DRI properly. Some sort of sanity checker for DRI (like Nvidia has) would also be nice.
NVidia seems like the only choice for me, since only ATI's most recent line of products (ie: expensive products) are supported.
DRI covers virtually all of the ATI chips up to the 9200.
FGLRX covers everything from the Radeon 8500 to 9800.
The only remaining problem is that some commercial developers just assume Nvidia is the standard. DRI in particular has come a long way. I've been using a Radeon 8500 with Michael Daenzer's DRI packages for Debian with few problems. Out of 30 commercial Linux games I have, 3 don't work properly: Descent 3, Savage and Heavy Gear 2. Savage's problem is trivial to fix, the others I'm not sure about.
I'd rather see ATI release the 3D specs on the R300 chips than see further improvements in the FGLRX driver.
First response, scrolled down a few pages till I saw 'back care' highlighted.. found the name, Janet Dean. Less than a minute! These people are not very good at their Google ;-)
You should haved clicked on the "view as html" link then hit Ctrl-F then typed "back care".
It's just the installer that gets me.
Few seem to realize that the questions that start popping up at the end of installation are not coming from the installer. In Debian each package can ask questions to configuring the software. When someone goes hog-wild on TaskSel and marks 1000+ packages for installation, there are a lot of packages wanting to ask questions. The secret is to skip tasksel and use apt-get to pull in only the packages that are needed.
Needless to say, packages that abuse the Debconf system by popping up stupid or misleading questions are broken. There are often flame-wars on Debian-Devel about what types of questions are worth asking. In some cases (sendmail, xfree86, etc) the ability of packages to ask questions while configuring is a godsend. In other cases (lwp, etc) the questions are totally pointless and anyone who needs the functionality will know how to set it up.
What happens if you install a text-only RedHat or Mandrake system, then install XFree86 after the fact? Does X get configured properly? If you look at it from this perspective, allowing packages to ask questions does make a certain amount of sense.
>> So? What's wrong with not using Linux?
> Nothing, as long as he's now using the HURD.
Or GNU/kFreeBSD.
Personally I don't really care about glitches, crashes and other problems with the machines. What I do care about is the use of uncertified software and the fact that these companies are more or less getting away with it. It sets a bad precedent for the future. Who cares if a few voting machines get decertified if you get to rig an election as a result? Any use of uncertified software should bar that company from ever producing voting machines in the US again. Do we really have to wait until someone is caught rigging a major election before real efforts are undertaken to stop it?
Next soundcard i'm going to buy will be by a company that actively supports linux or opens up the specifications: a product that can't work with both the OSs i use is a crippled one.
/dev/dsp, all at the same time, you'll realize that the SB Live is the standard by which everything else should be measured. Are there any cards other than the SB Live and SB Audigy that can leave /dev/dsp unlocked no matter how many sounds are playing at the same time? Software mixing is terrible in comparison.
The only soundcard I ever use with Linux is the SoundBlaster Live. The first time you see goofy app #1 playing sound through artsd, goofy app #2 playing sound throguh esd, and goofy app #3 playing sound directly through
I believe that most of the old BSD Unix has been publicly released already, so any code from such a source is free and clear.
No. Anything before 4.4BSD-Lite is questionable. You might be risk using the older stuff on the basis that Caldera gave away the ancient UNIX code it was based on, but considering what bastards they are now it's probably not a good idea. If you're taking anything from BSD, make sure it's from a version released after the BSD settlement. Everything prior to that required a UNIX license.
I would very much like to know what methods they used to review the code. Interpreting the results of any comparison to remove false positives is a very subjective process. Do they have a method that doesn't generate false positives? If so, did their method find the code that SGI admitted to contributing by mistake?
To be honest, anyone claiming there is absoloutely no questionable code sounds as suspect as SCO claiming it's all questionable code. Reality is never black and white like that.
Of course, at this point it's obvious that any SCO code in Linux is trivial and unintentional. If there were millions of lines of SCO code in Linux it would be impossible to overlook. It remains to be seen how much GPL code has made its way into SCO's UNIX offerings.
This "amateur robotics" competition is really more about manufacturing skill and hand-eye coordination, not advancing technology or even implementing it in new or creatively useful ways. I see little practical difference between this and the Boy Scouts' Pine Derby.
I do wish they would stop calling anything remotely controlled a robot. By that definition "robots" have been common since the 1930s when the first RC model airplanes were created. Considering the price tags that are being mentioned, balsa wood planes actually sound like a better route since each student could draft, build, and pilot his or her own unique design for a small fraction of the cost.
Truth be told though, I was in Odyssey of the Mind, math relay and other goofy school competitions years back and I did enjoy it even if nothing substantial was learned in the process.
(I will bow to any proof that MS gives away licenses to American Non-Profit Orgs.)
They sell to nonprofits through TechSoup.
I design software for embedded devices and let me tell you, as soon as you add a password mechanism, then someone will lose the password within days. It's happened to me, and I finally had to put a global password in every machine.
What company do you work for exactly?
I didn't intend for that comment to be a troll. I looked at the code because I was hoping it would be written in C++/MFC and thus be interesting to rebuild with WxWidgets, even if the end result is still only useful on Windows. Even though it's not useful to me, the Mono team might still be interested in using all or part of it as a test case for their C# implimentation.
I would be shocked if there were many large commercial projects that don't have stolen OSS code in them...
Only if it's free software that you wrote and you didn't assign the copyright to anyone else, reusing it in a commercial product is not stealing...the code still belongs to you and you can relicense it to your new employers under whatever terms you'd like.
The license looks good, but the code is C#.
Maybe the Mono folks will be interested.
What that paragraph states is that you CAN put the software on as many computers as you want, but if you do you're not getting any tech support from Redhat. This is the support contract, not the software contract.
Right, but as I already pointed out, buying the software means purchasing a support contract. Imagine if RedHat didn't make the SRPMS available to anyone who hasn't paid for RHAS. If the support contract that each of those customers has agreed to prevents them from distributing the software, then an additional restriction has been placed on the software regardless of how far down in the chain of licenses it takes to reach that point.
Read the EULA. It's all there.
You link to the wrong license agreement. RedHat uses the nice license for the software, and the nasty license for the support contract. Of course, RedHat will only sell you the software with a support contract, so the support contract terms apply to anyone wanting to purchase RHEL.
The support contract plainly states: "If Customer wishes to increase the number of Installed System, then Customer will purchase from Red Hat additional Services for each additional Installed System. During the term of this Agreement and for one (1) year thereafter, Customer expressly grants to Red Hat the right to audit Customer's facilities and records from time to time in order to verify Customer's compliance with the terms and conditions of this Agreement."
The end result, of course, is that you can't buy RHAS without giving up rights the GPL explicitly gives you.