1. RH, in its current state of transition and confusion, is aiming Fedora at the non-existent "bleeding-edge corporate desktop" market and thus continuing the RH8 and 9 trend of dumbing it down. Returning to techie roots should involve returning to the more sensible and *more usable* kind of interface we saw in earlier series before people got into the "usability" kicks.
2. Package management. Good grief. It starts with the install: there is no way to select individual packages for installation. After installation, the graphical "install packages" program, which has all sorts of problems up to and including frequent segfaults for many users, still doesn't allow individual package selection either (and hasn't since RH8). Its segfaults and the increasingly common RPM hangs result in locking problems and rpm db corruptions which require, at best, a rpm db rebuild, and at worst enough repair that a total reinstall is recommended instead. Ah, for the days of rpm 3.x and gnorpm! (Man, I miss that program- I've felt for some time that RH failing to put resources into it to keep gnorpm up with the move to gtk2 and later rpm versions was their worst move at least since GCC "2.96".)
I'm pretty sure there isn't yet any software for Windows which understands the ext3 journal, so you'll only find read only drivers for it. By far the best driver I've seen is the one I link to in my sig- ext2fsd. Takes a bit of tweaking to install, but works quickly, seamlessly, and, in my (readonly) experience, flawlessly. Other drivers I tried would crash all the time. Remember- before you use any windows ext2 driver in rw mode, be sure to back up your data.
Why? It's RO by default. If you are going to set it RW then you will definitely want a backup. BTW it doesn't understand journaling, so it won't let you try to write to ext3.
He linked to the 49G+'s page when he was saying that dedicated calculators offered only slim pickings; that's why I referenced it. Of course, the 49G does just about everything the 49G+ does, it just takes 3x the time to do it.
If the HP 49G+ is insufficient for your handheld computation needs, you're in a really unusual position. What exactly are you wanting your calculator to do?
The Win32 GTK 1.x port had lots of serious issues. However, I haven't noticed any serious issues with the GTK 2 port (which is used by just about all the win32 gtk apps except the stable version of the gimp). Care to elaborate?
I've defended Hatch on/. before, saying that the recent mistakes and outright idiotic actions of his which get Slashdot coverage are not indicative of his overall record; he represented Utah very well, overall, for many years. However, this is the last straw. He's gone senile. Justifying putting special interests above the law by citing "market realities"? What on earth does he think the law was passed for in the first place? Perhaps it hasn't occured to him that it was to correct such lopsided and unhealthy market conditions? Geez.
The GPL basically says "if you distribute software with this license and you have patents governing methods used in the software, you grant a perpetual patent license to anybody who obtains, in accordance with this license, the software or any legitimate derivative works, which patent license extends at least to the use and/or distribution of such."
In other words, if you use the gpl'd delta debugging software or its derivatives in a way which complies with the GPL, Passau University can't sue you for patent infringement because they distributed it under the GPL. However, if you violate the GPL they can sue you for patent infringement as well as copyright violation.
As to the benefits to Passau: if you want to use the patented process in proprietary or other non-gpl software you have to work out a patent license with Passau U (for which they may want you to pay them). In addition, the patent ensures the safety from third-party patent claims of the gpl'd software and any other software Passau makes or decides to license the patent for.
You also get to watch Yasser Seriawan sit and shake his head as his yappy co-commentator talks continually about his opinions that Kasparov must be scared out of his wits because of the computer's latest move and that computers are destined to beat human players just about 100% of the time within 2 years.
You seem to have confused ownership with 0wnz0r5h1p. The US may have 0wn3d Saddam's government, but the only part of Iraq the US owns is the.3 acres or so of the US embassy. Force does play a part in property ownership, but justice does as well.
1. The new SCO is not Caldera. Ransom Love and company, though they made plenty of bad business decisions (incl. the per-seat licensing fiasco), were actually a pretty good group.
2. Using Novell's agreement with SCO to order SCO to put up and/or shut up or even revoking the agreement would not be "descending to their level", it would just be moving this forward- SCO, who promised to show the evidence in June, realized long ago that they have no case, and isn't going to let evidence come out to that effect in court until they've finished pumping the stock for all it's worth and shipped the execs to places the US has no extradition treaties with.
This is why I said you can run a text environment using redhat packages on less, not that you can install with less. However, you can use the installers or anaconda patches provided by RULE if you want to try RedHat on such a system.
You can still run a text installation using RH packages with quite a bit less than 64 MB. Graphical environments for Linux have become much more resource-intensive than they used to be, but an X 4.3 install with a less-resource hungry DE than Gnome or KDE, such as XFCE, can still do fine with 64 MB. RedHat is here giving numbers which they will support and which ought to provide plenty of performance, not the "actually tolerable with" or much less the "bare minimum" numbers.
then you probably ought to become familiar with compiling a kernel from the same kernel series as the one you're running before doing a 2.6 kernel. Many drivers in 2.6 haven't been changed to the new driver specs yet, and 2.6 needs new userspace utils (module tools are the most crucial) to work. When you decide to compile a 2.6 kernel, read Dave Jones' What to expect from 2.6 which hasn't been updated in a couple months but is still very much a must-read.
Every joe off the street coming in and announcing to others (in this case to thousands of readers! how's that for publicity?) that they know how Mozilla should look, what it should be, and how the project should be run better than the Mozilla team does, they just can't code/do whatever other work themselves. The noise from these "armchair developers" as well as clueless users in this and other major projects has increased to the point that the usefulness of some of the primary communication channels for larger projects (bugzilla.mozilla.org, the main cygwin mailing list, several of the openoffice mailing lists, etc) has been decreasing rapidly. The signal to noise ratio just gets worse and worse.
In general, I think one of the things which would be a large factor in the future success of many large open-source projects would be strongly discouraging most input in crucial communication channels while finding better ways for non-developers to channel their input. In the case of a community as large and often discordant as that of Mozilla, it may even be a good idea to hold elections for community/end-user representatives who present concisely the concerns of the community without taking an "armchair developer" approach to things.
BTW, I was once the "armchair developer" type myself before I saw the light- I think I was part of the reason for Gerv's Unofficial Mozilla Spell-checker FAQ (bugzilla doesn't take/. hits, copy link location to instead of using the link).
A replacement for the current init system, while necessary, should have fewer, if possible, dependencies than the current init, not more. Unices are being deployed across more and more diverse kinds of systems, and dependencies on python and d-bus, both of which projects I support in themselves, are not going to be welcome in the init of the majority of unix systems today, especially in servers or embedded systems.
Of course there is a difference between ethics and law. However, one does have a duty to uphold the law unless there are reasons to disobey the law which are more ethically compelling than this duty. US copyright law in its current form isn't unethical per se (though, for instance, the actions of interests like Disney in influencing it certainly are), it's just stupid, so there's no compelling general ethical reason to break it. So while I'm all for working to get the law changed, I won't disobey it.
BTW, I would definitely say that it is ethical for the law to place restrictions on what you can legally do even when the prohibited actions don't harm anyone else in particular. But that's a conversation for some other day.
Some people actually try to be ethical, and there are even those who do so when they can see how they would be benefitted by an unethical action and can't see who would be harmed by it. I know it boggles the average slashdotter's mind, but try to imagine it.
Would a redesigned Debian logo get coverage on slashdot?
You mean like this, this, and this? A lot of/. users' feelings about particular projects or companies gets invested in the logo as well, so while the logo changes of Apple and Debian may not be of particular interest for some of us, I see no reason why such news shouldn't be on slashdot.
He says "might be" because he's not a lawyer, not because he thinks the GPL actually is vulnerable in court; otherwise he probably wouldn't have complied with the license when he was in the business. His only beef with the GPL appears to be that he thinks the "simple copyleft" licenses like the LGPL and the "variations on a theme of public domain" licenses like BSD are more condusive to different business models, and thus also more condusive to non-volunteer development of projects under those licenses, than the GPL.
Contrary to what others here (your usual/. cranks with rose-colored nostalgic memories of ancient hardware) may tell you, the 49G is a great calculator despite the flaws in the design of the keys and display as well as the "new-wave" case. However, don't run out and buy a 49G now; HP is finally gearing up to release new calcs: the 49G+ and 48gII replace the aging Saturn processor (2-4 mhz, I forget what the process was but it probably was 3 microns or so) of older HP's like the 48S/SX/G/G+/GX and 49G with an ARM9 processor (48-75 mhz, probably a.18 micron process). (I think the ARM9 doesn't do quite as many IPC as the Saturn, and low-level saturn emulation will give some performance penalty, but still, this is going to be a lot faster- hpcalc.org just states it will be 3-7 times faster than the 49G, that's a pretty wide range). In addition, they appear to have fixed many of the case/key/display problems which plagued the 49G, and battery life should be about the same as it was for the earlier 48/49 series. I have absolutely no doubt that the new calcs will be fantastic deals.
I'll ignore the fact that you've confused Gamespot and Gamespy. However, Gamespot is here praising just one boss- they in fact say many of the other enemies were pretty poor quality and easily defeatible by cheap stupid tactics- and one boss does not a game make. Furthermore, even if it were a good game, plenty of Gamespy's "most overrated" games were actually pretty good- it's just that the hype, the reviews, or the anticipation that they'd be even better than the great games they were successors to pointed towards a game light-years beyond the "OK, not great" games they turned out to be.
If you take that with a smile, you're both better-natured and thicker-skinned than a lot of people guessed from the hiring policy incident. Good luck with your vaporware endeavors;)
Slashdot Mirror
My two biggest complaints with it are:
1. RH, in its current state of transition and confusion, is aiming Fedora at the non-existent "bleeding-edge corporate desktop" market and thus continuing the RH8 and 9 trend of dumbing it down. Returning to techie roots should involve returning to the more sensible and *more usable* kind of interface we saw in earlier series before people got into the "usability" kicks.
2. Package management. Good grief. It starts with the install: there is no way to select individual packages for installation. After installation, the graphical "install packages" program, which has all sorts of problems up to and including frequent segfaults for many users, still doesn't allow individual package selection either (and hasn't since RH8). Its segfaults and the increasingly common RPM hangs result in locking problems and rpm db corruptions which require, at best, a rpm db rebuild, and at worst enough repair that a total reinstall is recommended instead. Ah, for the days of rpm 3.x and gnorpm! (Man, I miss that program- I've felt for some time that RH failing to put resources into it to keep gnorpm up with the move to gtk2 and later rpm versions was their worst move at least since GCC "2.96".)
You replied to him by linking to the same driver he linked to (also in my sig). Heh.
I'm pretty sure there isn't yet any software for Windows which understands the ext3 journal, so you'll only find read only drivers for it. By far the best driver I've seen is the one I link to in my sig- ext2fsd. Takes a bit of tweaking to install, but works quickly, seamlessly, and, in my (readonly) experience, flawlessly. Other drivers I tried would crash all the time. Remember- before you use any windows ext2 driver in rw mode, be sure to back up your data.
Why? It's RO by default. If you are going to set it RW then you will definitely want a backup. BTW it doesn't understand journaling, so it won't let you try to write to ext3.
He linked to the 49G+'s page when he was saying that dedicated calculators offered only slim pickings; that's why I referenced it. Of course, the 49G does just about everything the 49G+ does, it just takes 3x the time to do it.
If the HP 49G+ is insufficient for your handheld computation needs, you're in a really unusual position. What exactly are you wanting your calculator to do?
The Win32 GTK 1.x port had lots of serious issues. However, I haven't noticed any serious issues with the GTK 2 port (which is used by just about all the win32 gtk apps except the stable version of the gimp). Care to elaborate?
I've defended Hatch on /. before, saying that the recent mistakes and outright idiotic actions of his which get Slashdot coverage are not indicative of his overall record; he represented Utah very well, overall, for many years. However, this is the last straw. He's gone senile. Justifying putting special interests above the law by citing "market realities"? What on earth does he think the law was passed for in the first place? Perhaps it hasn't occured to him that it was to correct such lopsided and unhealthy market conditions? Geez.
The GPL basically says "if you distribute software with this license and you have patents governing methods used in the software, you grant a perpetual patent license to anybody who obtains, in accordance with this license, the software or any legitimate derivative works, which patent license extends at least to the use and/or distribution of such."
In other words, if you use the gpl'd delta debugging software or its derivatives in a way which complies with the GPL, Passau University can't sue you for patent infringement because they distributed it under the GPL. However, if you violate the GPL they can sue you for patent infringement as well as copyright violation.
As to the benefits to Passau: if you want to use the patented process in proprietary or other non-gpl software you have to work out a patent license with Passau U (for which they may want you to pay them). In addition, the patent ensures the safety from third-party patent claims of the gpl'd software and any other software Passau makes or decides to license the patent for.
You also get to watch Yasser Seriawan sit and shake his head as his yappy co-commentator talks continually about his opinions that Kasparov must be scared out of his wits because of the computer's latest move and that computers are destined to beat human players just about 100% of the time within 2 years.
You seem to have confused ownership with 0wnz0r5h1p. The US may have 0wn3d Saddam's government, but the only part of Iraq the US owns is the .3 acres or so of the US embassy. Force does play a part in property ownership, but justice does as well.
1. The new SCO is not Caldera. Ransom Love and company, though they made plenty of bad business decisions (incl. the per-seat licensing fiasco), were actually a pretty good group.
2. Using Novell's agreement with SCO to order SCO to put up and/or shut up or even revoking the agreement would not be "descending to their level", it would just be moving this forward- SCO, who promised to show the evidence in June, realized long ago that they have no case, and isn't going to let evidence come out to that effect in court until they've finished pumping the stock for all it's worth and shipped the execs to places the US has no extradition treaties with.
This is why I said you can run a text environment using redhat packages on less, not that you can install with less. However, you can use the installers or anaconda patches provided by RULE if you want to try RedHat on such a system.
You can still run a text installation using RH packages with quite a bit less than 64 MB. Graphical environments for Linux have become much more resource-intensive than they used to be, but an X 4.3 install with a less-resource hungry DE than Gnome or KDE, such as XFCE, can still do fine with 64 MB. RedHat is here giving numbers which they will support and which ought to provide plenty of performance, not the "actually tolerable with" or much less the "bare minimum" numbers.
Go to your /. Preferences and check the box which says "Collapse Sections". Then all the stories will make it on to your front page.
then you probably ought to become familiar with compiling a kernel from the same kernel series as the one you're running before doing a 2.6 kernel. Many drivers in 2.6 haven't been changed to the new driver specs yet, and 2.6 needs new userspace utils (module tools are the most crucial) to work. When you decide to compile a 2.6 kernel, read Dave Jones' What to expect from 2.6 which hasn't been updated in a couple months but is still very much a must-read.
Every joe off the street coming in and announcing to others (in this case to thousands of readers! how's that for publicity?) that they know how Mozilla should look, what it should be, and how the project should be run better than the Mozilla team does, they just can't code/do whatever other work themselves. The noise from these "armchair developers" as well as clueless users in this and other major projects has increased to the point that the usefulness of some of the primary communication channels for larger projects (bugzilla.mozilla.org, the main cygwin mailing list, several of the openoffice mailing lists, etc) has been decreasing rapidly. The signal to noise ratio just gets worse and worse.
/. hits, copy link location to instead of using the link).
In general, I think one of the things which would be a large factor in the future success of many large open-source projects would be strongly discouraging most input in crucial communication channels while finding better ways for non-developers to channel their input. In the case of a community as large and often discordant as that of Mozilla, it may even be a good idea to hold elections for community/end-user representatives who present concisely the concerns of the community without taking an "armchair developer" approach to things.
BTW, I was once the "armchair developer" type myself before I saw the light- I think I was part of the reason for Gerv's Unofficial Mozilla Spell-checker FAQ (bugzilla doesn't take
A replacement for the current init system, while necessary, should have fewer, if possible, dependencies than the current init, not more. Unices are being deployed across more and more diverse kinds of systems, and dependencies on python and d-bus, both of which projects I support in themselves, are not going to be welcome in the init of the majority of unix systems today, especially in servers or embedded systems.
Of course there is a difference between ethics and law. However, one does have a duty to uphold the law unless there are reasons to disobey the law which are more ethically compelling than this duty. US copyright law in its current form isn't unethical per se (though, for instance, the actions of interests like Disney in influencing it certainly are), it's just stupid, so there's no compelling general ethical reason to break it. So while I'm all for working to get the law changed, I won't disobey it.
BTW, I would definitely say that it is ethical for the law to place restrictions on what you can legally do even when the prohibited actions don't harm anyone else in particular. But that's a conversation for some other day.
Some people actually try to be ethical, and there are even those who do so when they can see how they would be benefitted by an unethical action and can't see who would be harmed by it. I know it boggles the average slashdotter's mind, but try to imagine it.
He says "might be" because he's not a lawyer, not because he thinks the GPL actually is vulnerable in court; otherwise he probably wouldn't have complied with the license when he was in the business. His only beef with the GPL appears to be that he thinks the "simple copyleft" licenses like the LGPL and the "variations on a theme of public domain" licenses like BSD are more condusive to different business models, and thus also more condusive to non-volunteer development of projects under those licenses, than the GPL.
Contrary to what others here (your usual /. cranks with rose-colored nostalgic memories of ancient hardware) may tell you, the 49G is a great calculator despite the flaws in the design of the keys and display as well as the "new-wave" case. However, don't run out and buy a 49G now; HP is finally gearing up to release new calcs: the 49G+ and 48gII replace the aging Saturn processor (2-4 mhz, I forget what the process was but it probably was 3 microns or so) of older HP's like the 48S/SX/G/G+/GX and 49G with an ARM9 processor (48-75 mhz, probably a .18 micron process). (I think the ARM9 doesn't do quite as many IPC as the Saturn, and low-level saturn emulation will give some performance penalty, but still, this is going to be a lot faster- hpcalc.org just states it will be 3-7 times faster than the 49G, that's a pretty wide range). In addition, they appear to have fixed many of the case/key/display problems which plagued the 49G, and battery life should be about the same as it was for the earlier 48/49 series. I have absolutely no doubt that the new calcs will be fantastic deals.
I'll ignore the fact that you've confused Gamespot and Gamespy. However, Gamespot is here praising just one boss- they in fact say many of the other enemies were pretty poor quality and easily defeatible by cheap stupid tactics- and one boss does not a game make. Furthermore, even if it were a good game, plenty of Gamespy's "most overrated" games were actually pretty good- it's just that the hype, the reviews, or the anticipation that they'd be even better than the great games they were successors to pointed towards a game light-years beyond the "OK, not great" games they turned out to be.
If you take that with a smile, you're both better-natured and thicker-skinned than a lot of people guessed from the hiring policy incident. Good luck with your vaporware endeavors ;)