With Spam Kings, Mr. McWilliams has put together a book suitable for shelving next to The DaVinci Code and the Bat Book (Sendmail 2nd edition
The context implies that this is a Good Place For a Book To Be.
That strikes me as odd though - I recycled both: "DaVinci Code" for being a bunch of unfounded hokum, and Sendmail (the software, and therefore the book) for being too obfuscated for our simple few-dozen-domains setup (switched to Exim a few years ago, haven't looked back)
Maybe the implication is that I should do with Spam Kings what I do with spam... trash it (er, I mean, read it thoroughly and believe every word???)
Even "pressing the Save Button often" isn't a solution. Try opening a word document from a webpage in firefox, editing it for several hours, then close firefox. There goes all of your hard work that was actually a temporary file!
Not a Windows expert, but surely it's saved to %TEMP% (or whatever the variable is called), or maybe "Temporary Internet Files" (or whatever it's called)
GNOME's VFS tends to save stuff to/tmp, which is a PITA; surely Windows can't have contrived a more awkward solution?
Or - in summary - installing software is not trivial.
Why is this not f*cking obvious?
What do you expect to get paid for? The occasional "apt-get" doesn't exactly demand a high salary; installing the right software, correctly, is a valuable contribution.
Ahh, if only the distro's would honour it... my speedtouchconf.sf.net project would be so much simpler.
Alas, money is more important to them.
Strange that they can't see that my invested time keeps them their customers, and that without my time, they would have no money.
It's only a matter of time, of course....
Subsume my project, and support it yourselves, or don't CLAIM to support it when you don't. It only causes confusion.
And now, how do I uninstall it,since the Makefiles don't seem to have make remove/uninstall/etc. that would delete the installed binaries.
That is what really needs to be addressed: While RPM and other formats allow for the removal of packages (and, often, quite neatly - cleaning up changes to config files).
As for non-root installation of software, it can work in many cases, but if you want to install (say) an FTP server, it requires (and should require) root access. If you want to install an Office suite, it shouldn't *require* root acceess (depending on your local policy), and AFAIK, OOo 2 has changed the installation process appropriately, for example.
Some stuff does need root access, some stuff doesn't. In a home PC / university / commercial setting, the answers may be different according to the environment, which the software itself cannot contain: Should Foo.pkg be installed by users, or should it need to be installed by "root"? Home/Uni users might be happy with non-root priv's to install it, but a company might not want users to install it. The software can't determine local policy....
"Dallas courts have no presidence in this case because Lokitorrent.com was never hosted on Texas ISP servers. The owner of Lokitorrent is not a resident of Texas, and the MPAA's main offices are not in Texas. The MPAA has no reason in the world to sue in Dallas, but even if they did have presidence and sued in Dallas, you can be sure that the due dilligence on a case like this would take years. No way in the world could the MPAA have won a court case of this magnitude in just one month"
What kind of lawyer uses spelling like "presidence" and "dilligence"?!
If I buy a car, it comes with software in it, but they don't expect me to sign an EULA.
But what rights do you get over that software? You've already agreed that the car - as a whole - may kill you, and may contain flaws. Isn't that a EULA?
Do you get rights to the source code? For my previous car (Vectra, by GM/Opel/Vauxhall) the system had an interface under the handbrake... getting a device to interact with that cost a few hundred, with no rights to know what the code actually did.
I took it to a dealership when the engine warning light was on (which puts it into low-effiency, get-you-home-mode) and they were baffled, because it was reporting a fault in the aircon unit, but the car didn't have aircon! Without access to the source, they were no wiser than me as to the actual problem. A (binary) firmware update to the car fixed that. Still no wiser as to how the problem occured. Just meant that I couldn't go over 4500 RPM until it was "fixed".
Note that this wouldn't affect the GPL. The GPL is not an EULA, but a licence that provides additional rights above and beyond fair use.
Not strictly true. The GPL gives rights, and if you don't accept the terms, you don't get the rights.
So, by your reasoning, the GPL should be published on such boxes (which would be No Bad Thing). a big "Pure GPL (any version)" / "Pure GPL (v2)" / "OSI Certified Open Source Software" label on the box would preclude you from having to read the small-print.
I can't see many F/OSS vendors going for this, though - the few who could make this claim use downloads. RedHat can't claim Pure anything, surely, since the Entire Contents of the CD/DVD can't be distributed according to the GPL.
Debian could make the claim, but they don't tend to put big boxes in large PC retail outlets...
I put FC3 on my desktop and laptop. It seems to work pretty well on the laptop (Dell Latitude C640) but on the desktop (which dual-boots with Windows 2000), it'll often stall on the boot - dropping from the pretty graphical bootup process to tty1, saying "Enabling swap space", and just hang there. It's not crashed, as such - CTRL-ALT-DEL reboots it happily enough, and no keyboard LEDs blinking - but just doesn't get any further.
Also (and this may be a Firefox bug), after connecting to a customer's network (with DHCP for everything, and a.PAC file for Proxy config) it now takes a few minutes for a new Firefox window to cope with its life, and give refresh control back to the screen space it occupies, let alone accept input.
The laptop is mostly Intel chips; the Desktop which has problems, is as pure-Intel as possible - Intel Mobo with Intel chipset, Intel P4 CPU, using the onboard graphics, etc etc.... Rebooting normally seems to fix it.
Very odd, but not something I'm bothered about fixing. I'm downloading SuSE at the moment.
One thing I have to pay homage to Solaris - I really like providing NFS with Solaris. I always set and forget Solaris, its a pain in the arcane butt with a fairly austere userland, but once its configured it runs like a champion.
"vi/etc/dfstab" doesn't seem too arcane to me... but I do agree with you - NFS on Solaris "Just Works (TM)". NFS on Linux still seems strangely flakey, even though it seems to be a fully-documented protocol. I've never really understood this (and haven't ever tried to fix it, but since I have access to Solaris source code, and aren't really a low-level systems programmer, I've avoided trying to fix the Linux stuff).
I'd love to know why a Linux NFS server doesn't work as well as a Solaris one, though... I can nearly get a Linux box to work as a JumpStart server, but the NFS stuff just collapses really early on in the process...
Well that their tough 5h1t, isn't it? They want to make profit from free software, there's nothing written in stone that they have a right to succeed.
I don't care what RedHat want, I don't even care too much how much Linux gets used. I care about how secure the majority of internet-connected machines are (because that affects how much crap I get on my PC - spam from trojans, etc) - and the best way for that, is for more people to use Linux (or similar) and RedHat have the best bet of pushing that future, but to support RedHat's ego trip in the naive hope that they'll supplant Microsoft any time soon is, quite honestly, pathetic.
Accepting shitty attitudes from RedHat because "they're on our side" (hah!) will - in the best possible scenario - result in RedHat having the power that Microsoft currently have. It's an unlikely outcome, at best, so why even bother to support them?
If they're that threatened by the GPL, which they're making their money from, maybe they should rethink their business plan. Not just demand that everyone back down and leave the territory open for them! That ain't how the GPL works, my friend. Enforcing ridiculous restrictions which result in WhiteBoxLinux and CentOS is a mixed advertisement for F/OSS software; RedHat attacking the "clones" is a terrible advert, and actively restrictive of the continued migration towards F/OSS.
You're thinking of the suggested boilerplate text. My GPL software is relased under V2, no other license. The suggested boilerplate offers the software under the current GPL "or any subsequent version, at your discretion" (I paraphrase, since I've got a baby in my arms at the moment, it's hard to check the exact wording!)
Many GPL projects could be interpreted by GPLv3 when it is published; others (incl. Linux) explicitly choose to restrict themselves to licenses the authors have already read.
This is/. where it's convenient to ignore words when it feels appropriate.
If it were called FreeSolaris, then whining about it not being GPL'd might be valid - it's called OpenSolaris, it's OpenSource, distributed under an OSI-Approved license.
Exactly - I love RMS normally, but his comments about the CDDL ("nobody else uses it", or words to that effect) - of course nobody elses uses it, Sun have only just got it approved by the OSI, and released OpenSolaris using it!
Let's see the code opened up, and see what happens.
Where's the source to AIX? (as if anybody would want it?) Where's the source to SCO? (oh, wait, that's at kernel.org;-)
Hang on - are you really arguing for Free Software, or a Free Ride?
RedHat are complying, to that extent.
The problem with the SRPM's is not that they are source-code only, not binaries. The problem is that the binaries include "trademarked" (not copyrighted, therefore not covered by the GPL) images, so you have to rebuild the entire lot from SRPM's (v. easy) after removing their images.
That's a PITA; www.whiteboxlinux.org, and cents (or something similarly named; I forget off the top of my head) are projects which do this.
But then, what is RedHat support worth?! Shitty attitudes, bad manners, "works for me" attitude - I can get that from the F/OSS community, thanks:-)
Even on/., I'm amazed that this got +4'd
What, exactly, is the claim here?
"the scientific method and evolution" - I don't think that even the most eager evolutionist would put these two together as one-and-the same.
I don't think that most F/OSS developer would thank you for putting their hard work down as "at random", either.
After that (and I hope I've followed your positon correctly so far, though I'm not confident), it all falls apart terribly.
Not every package, I think you'll find, if you care to read the documentation.
On the one time I tried to get RedHat UK support to deal with a RedHat-written piece of software (RedHat Update Network, since you ask, not working through a MS Proxy Server) for a fully paid-up RedHat customer, they blamed it on MS Proxy Server - the GUI claims to deal with MS Proxy Server requirements, but doesn't work. Adding a tool from sf.net (I forget what; I'm sure I've mentioned this on/. before, so search for it) worked fine.
RedHat took no responsibility at all for this. I was VERY disappointed; I'd used their SW for free for years (I did buy RH5.1), but assumed that in 2004 (as it was) that their fully-paid-up commercial users would get better support than this. They'd gone from Debian to RedHat for the support, after all!
The software works (with additional software from sf.net), but that isn't the point - either RedHat fix the problem, or tell customers that the fix is to use the sf.net software, and that they're still offically supported afterr installing it.
Slashdot Mirror
The context implies that this is a Good Place For a Book To Be.
That strikes me as odd though - I recycled both: "DaVinci Code" for being a bunch of unfounded hokum, and Sendmail (the software, and therefore the book) for being too obfuscated for our simple few-dozen-domains setup (switched to Exim a few years ago, haven't looked back)
Maybe the implication is that I should do with Spam Kings what I do with spam... trash it (er, I mean, read it thoroughly and believe every word???)
Even "pressing the Save Button often" isn't a solution. Try opening a word document from a webpage in firefox, editing it for several hours, then close firefox. There goes all of your hard work that was actually a temporary file! /tmp, which is a PITA; surely Windows can't have contrived a more awkward solution?
Not a Windows expert, but surely it's saved to %TEMP% (or whatever the variable is called), or maybe "Temporary Internet Files" (or whatever it's called) GNOME's VFS tends to save stuff to
Are you sure you belong here? :-)
Why is this not f*cking obvious?
What do you expect to get paid for? The occasional "apt-get" doesn't exactly demand a high salary; installing the right software, correctly, is a valuable contribution.
Ho Hum.
Ahh, if only the distro's would honour it... my speedtouchconf.sf.net project would be so much simpler. Alas, money is more important to them. Strange that they can't see that my invested time keeps them their customers, and that without my time, they would have no money. It's only a matter of time, of course.... Subsume my project, and support it yourselves, or don't CLAIM to support it when you don't. It only causes confusion.
That is what really needs to be addressed: While RPM and other formats allow for the removal of packages (and, often, quite neatly - cleaning up changes to config files). As for non-root installation of software, it can work in many cases, but if you want to install (say) an FTP server, it requires (and should require) root access. If you want to install an Office suite, it shouldn't *require* root acceess (depending on your local policy), and AFAIK, OOo 2 has changed the installation process appropriately, for example. Some stuff does need root access, some stuff doesn't. In a home PC / university / commercial setting, the answers may be different according to the environment, which the software itself cannot contain: Should Foo.pkg be installed by users, or should it need to be installed by "root"? Home/Uni users might be happy with non-root priv's to install it, but a company might not want users to install it. The software can't determine local policy....
The bitten apple is a tribute to the late, great, Alan Turing (google: "apple logo turing" for info)
RTFM
Do they include spelling lessons?
What kind of lawyer uses spelling like "presidence" and "dilligence"?!
But what rights do you get over that software? You've already agreed that the car - as a whole - may kill you, and may contain flaws. Isn't that a EULA?
Do you get rights to the source code? For my previous car (Vectra, by GM/Opel/Vauxhall) the system had an interface under the handbrake... getting a device to interact with that cost a few hundred, with no rights to know what the code actually did.
I took it to a dealership when the engine warning light was on (which puts it into low-effiency, get-you-home-mode) and they were baffled, because it was reporting a fault in the aircon unit, but the car didn't have aircon! Without access to the source, they were no wiser than me as to the actual problem. A (binary) firmware update to the car fixed that. Still no wiser as to how the problem occured. Just meant that I couldn't go over 4500 RPM until it was "fixed".
That's as closed as Windows, IMNSVHO.
Not strictly true. The GPL gives rights, and if you don't accept the terms, you don't get the rights.
So, by your reasoning, the GPL should be published on such boxes (which would be No Bad Thing). a big "Pure GPL (any version)" / "Pure GPL (v2)" / "OSI Certified Open Source Software" label on the box would preclude you from having to read the small-print.
I can't see many F/OSS vendors going for this, though - the few who could make this claim use downloads. RedHat can't claim Pure anything, surely, since the Entire Contents of the CD/DVD can't be distributed according to the GPL.
Debian could make the claim, but they don't tend to put big boxes in large PC retail outlets...
It's worth staying away from people who put asterisks by their own spelling mistakes!
Also (and this may be a Firefox bug), after connecting to a customer's network (with DHCP for everything, and a .PAC file for Proxy config) it now takes a few minutes for a new Firefox window to cope with its life, and give refresh control back to the screen space it occupies, let alone accept input.
The laptop is mostly Intel chips; the Desktop which has problems, is as pure-Intel as possible - Intel Mobo with Intel chipset, Intel P4 CPU, using the onboard graphics, etc etc .... Rebooting normally seems to fix it.
Very odd, but not something I'm bothered about fixing. I'm downloading SuSE at the moment.
"vi /etc/dfstab" doesn't seem too arcane to me... but I do agree with you - NFS on Solaris "Just Works (TM)". NFS on Linux still seems strangely flakey, even though it seems to be a fully-documented protocol. I've never really understood this (and haven't ever tried to fix it, but since I have access to Solaris source code, and aren't really a low-level systems programmer, I've avoided trying to fix the Linux stuff).
I'd love to know why a Linux NFS server doesn't work as well as a Solaris one, though... I can nearly get a Linux box to work as a JumpStart server, but the NFS stuff just collapses really early on in the process...
The market didn't solve the problem - the community did.
There is a very significant difference here.
I don't care what RedHat want, I don't even care too much how much Linux gets used. I care about how secure the majority of internet-connected machines are (because that affects how much crap I get on my PC - spam from trojans, etc) - and the best way for that, is for more people to use Linux (or similar) and RedHat have the best bet of pushing that future, but to support RedHat's ego trip in the naive hope that they'll supplant Microsoft any time soon is, quite honestly, pathetic.
Accepting shitty attitudes from RedHat because "they're on our side" (hah!) will - in the best possible scenario - result in RedHat having the power that Microsoft currently have. It's an unlikely outcome, at best, so why even bother to support them?
If they're that threatened by the GPL, which they're making their money from, maybe they should rethink their business plan. Not just demand that everyone back down and leave the territory open for them! That ain't how the GPL works, my friend. Enforcing ridiculous restrictions which result in WhiteBoxLinux and CentOS is a mixed advertisement for F/OSS software; RedHat attacking the "clones" is a terrible advert, and actively restrictive of the continued migration towards F/OSS.
You're thinking of the suggested boilerplate text. My GPL software is relased under V2, no other license. The suggested boilerplate offers the software under the current GPL "or any subsequent version, at your discretion" (I paraphrase, since I've got a baby in my arms at the moment, it's hard to check the exact wording!)
Many GPL projects could be interpreted by GPLv3 when it is published; others (incl. Linux) explicitly choose to restrict themselves to licenses the authors have already read.
The DJ?
http://www.bbc.co.uk/radio2/shows/radcliffe/
He's a fantastic DJ, but I wouldn't trust him with legal matters!
If it were called FreeSolaris, then whining about it not being GPL'd might be valid - it's called OpenSolaris, it's OpenSource, distributed under an OSI-Approved license.
But hey, this is slashdot. Nuff said.
Let's see the code opened up, and see what happens.
Where's the source to AIX? (as if anybody would want it?) Where's the source to SCO? (oh, wait, that's at kernel.org ;-)
RedHat are complying, to that extent.
The problem with the SRPM's is not that they are source-code only, not binaries. The problem is that the binaries include "trademarked" (not copyrighted, therefore not covered by the GPL) images, so you have to rebuild the entire lot from SRPM's (v. easy) after removing their images.
That's a PITA; www.whiteboxlinux.org, and cents (or something similarly named; I forget off the top of my head) are projects which do this.
But then, what is RedHat support worth?! Shitty attitudes, bad manners, "works for me" attitude - I can get that from the F/OSS community, thanks :-)
Vendor quotes EULA; Insurance comapmny loses out, increases prices; company increases prices; client pays more for the same thing.
Client charges their customers more, making them less competetive.
What, exactly, is the claim here?
"the scientific method and evolution" - I don't think that even the most eager evolutionist would put these two together as one-and-the same.
I don't think that most F/OSS developer would thank you for putting their hard work down as "at random", either.
After that (and I hope I've followed your positon correctly so far, though I'm not confident), it all falls apart terribly.
Anyone else got an irrelevant axe to grind?
On the one time I tried to get RedHat UK support to deal with a RedHat-written piece of software (RedHat Update Network, since you ask, not working through a MS Proxy Server) for a fully paid-up RedHat customer, they blamed it on MS Proxy Server - the GUI claims to deal with MS Proxy Server requirements, but doesn't work. Adding a tool from sf.net (I forget what; I'm sure I've mentioned this on
RedHat took no responsibility at all for this. I was VERY disappointed; I'd used their SW for free for years (I did buy RH5.1), but assumed that in 2004 (as it was) that their fully-paid-up commercial users would get better support than this. They'd gone from Debian to RedHat for the support, after all!
The software works (with additional software from sf.net), but that isn't the point - either RedHat fix the problem, or tell customers that the fix is to use the sf.net software, and that they're still offically supported afterr installing it.