> BeOS would have made a great OS X. It was designed from the ground up to be incredibly quick and robust.
Mmm. NeXTstep was designed from the ground up to be (somewhat) quick and robust.
BeOS, is very inferior to NeXTstep, and pretty late in the game (I used both, and developed on both. Please, take time to comment if you tried both of them)
BeOS strong point was the multiprocessor design from the ground up (which gave him excellent multitasking possibilities). But on almost all aspects, it is techincally very poor compared to NeXTstep (C++ interfaces, Huge 'NIH' syndrom, lack of most basic utilities)
NeXT bought Apple in december 1996 for -400M$ (fell free to change the order of the words and the sign of the price for the official version), but it took them *4* years to get to OSX beta. And they probably took their fastest path.
> I'm sure if Apple picked up BeOS back in 1996, everything I described above would have been in place by now.
I don't see why you think it would have taken less time than doing it from NeXTstep. (Well, I suspect it is because you lack clues of the involved work and the respective level of NeXTstep and BeOS. But I would not say this:-) )
My bet, is that it would never have seen the day. Never. It would have been a Copland/Pink project again. (ie: Coolest thing going to get out of the door real soon now. Wait, it'll kick ass. Oh, you will not beleive it. It is going to be the best thing since sliced bread. Oops, cancelled.)
> if they need to address the Solaris and NT market, use the portability stuff inherent in Cocoa (OPENSTEP)
Unfortunately, you cannot do any YB stuff on windows since the end of last month. Fini. Termine. Nada. The license does not enalbe deplyment of AppKit applications on NT. Period.
> I sure wonder how things would have worked out if Apple had evangelized Cocoa a LOT more strongly pushed the cross-platform features, and ease of programming, and not done the work on Carbon, and not made Classic so easy to fall back on
Oh, yes. I beleive they tried to kill Cocoa, but the it wouldn't die. I hope that *now* they will push cocoa as much as they can. It is rather easy:
* Evangelize
* Do a lot of examples/documentation
* Get correct support for obj from metrowerks
* Do the new APIs on Cocoa first. (Ie: "if you want to get more out of this, you'll have to learn a bit of cocoa")
* Resurect the NT version
And the opensource community could do something nice too by contributing to GNUstep. Think about it:
* Darwin already boot on x86
* XFree86 already runs on darwin
* GNUstep already sort-of run on X
Mmmm. This looks like a open source MacOS X clone on the horizon... maybe in 3 or 4 years.
> This is why WebObjects is moving away from ObjectiveC.
In fact not exactly. But I'll only tell the true story on my death bed. Let's pretend this is right.
> To these people Apple has stepped back from supporting ObjC where they need it
The problem started before Apple bought NeXT. NeXTstep, then OPENSTEP was (and still is) used by very big non-disclosed corporation, that basically used NeXT as their outsourced R&D staff. Those guys have millions and millions of line of code, involving Distributed Objects and a lot of EOF.
Then, half accidentally, NeXT stumbled on WebObjects, and decided that it would be the next big thing. (This kind of shitf was classic in the NeXT community, as we already had the 'education market' as the orignal big thing, abandonned to 'the desktop publishing market' (or was it the opposite?), killed for 'interpersonal computing', itself replaced by 'mission-critical custom apps', while the hardware line got a chaotic course [m68k->i386->hppa->sparc], then the OS disapeared replaced by the yellow box, while the whole libraries were rewritten for OpenStep, fundamentally breaking *every* application and existing source code)
At this point of time (about 96), we had lost our Hardware platform, and our Operating System.
NeXT started to tell all developers that the future was java, but promised that Objective-C would stay. And the reason Objective-C would stay was:
"We cannot kill objective-c because WebObjects is based on it"
Then apple bought NeXT, and started reviving a defunct operating system.
Then the re-write of WebObjects on Java appeared, along the java version of EOF.
It looked like we were loosing our language too. Apple was not communicating on ObjC, most new cocoa documentation was about java, because MacOS developpers don't want to hear about objc. Internally, Apple decided to get rid of ObjC too.
So, it is rather refreshing to find that they changed their mind (probably because they had no choice, IMHO)
> in fact most of the apps that were ported to Java for early DP's were moved back to ObjC for DP4 and beyond...
Yes. I am French, a language were there is no distinction between 'he' and 'it' ('il' means both 'he' and 'it'). I can't help making this error again and again.
The Mac OS X equivalent of the classic Mac OS Control Panel, the System Preferences application, has been revamped for Beta. True to Apple's "we eat our own dog food" motto, System Preferences was a Java application (using the Cocoa API) in DP4. Unfortunately, it was also one of the slowest launching applications on the system due to the overhead of loading the VM. In Beta, it uses the Objective C interface to the Cocoa API instead, and performance is now acceptable.
This looks real good for ObjC. You now what ? NeXTstep really did have 10 years of advance. He's only getting mainstream now...:-)
> The Desktop uses the NeXT Workspace Manager code, mostlike
No. It shares zero code. The Desktop is written in C++/Carbon and have no Objc. Hey, I provided you the way to check (with nm) if you have a developer system, or if you get the tools from darwin.
The Desktop have nowhere the 'feel' of the original Workspace. And he is crippled with bugs (NFS mounts randomly appearing as links that points to nowhere, frequent temporary freezes, I could cite dozens like that). I hope it will not be too much tied into the OS, so we can write an ObjC replacement (from a developer point of view, the question is knowing if the NSWorkspace protocol will be implemented by the Dock [as in NeXTstep], or by the Desktop. I would prefer the former)
> Sadly, ObjC is being pretty aggressively phased out by Apple in favor of Java
This is what I thought, so I checked all the fundamental binaries of OSX. And it is quite the opposite.
IIRC, only the calculator and the Setup assistant are in java.
Only the Desktop and a handfull of (crappy) demos are in Carbon.
Even the Dock is in ObjC. And the MP3 player or the Clock (must check again, I don't have my MOSX box handy, but I'm pretty sure of that). And preferences. I think the loginPanel is objc too.
It is very refreshing to see that ObjC applications are of higher quality. When you get hooked to cocoa, it is hard to get back. I am incredibely happy that apple have now teams of engineers back to ObjC. This is, IMHO, the best new of this public beta. (In a side note, the new ProjectBuilder don't correctly support java yet. Funny as hell, if you want my opinion)
> The next rev of WebObjects, I believe, won't support ObjC, for example.
This is right. But this decision have been taken years ago, when they wanted to phase out all ObjC. But it looks like they are going back on that, which is a very very good thing.
Unfortunately, WebObject Objc is probably lost forever. And EOF (not present in the beta) will be missed too...
Applications (.app) are really folders. Drag and drop them in a shell to look inside them.
Cd to the Content/MacOS directory. There is the real application exectuable.
Then, if you have the developer tools (from darwin, for instance), you can look at executable content with 'nm'.
If you have things like:
0000f078 t -[MyClass _myMethod:]
then it is a Cocoa application.
You'll find that most of them are really cocoa applications.
The Desktop seems (to me) a poor rip of the NeXT Workspace Manager. It is real crappy, have many bugs, and is nowhere as usable as the program he tries to clone.
It is interesting to see that it is (almost) the only important Carbon app shipped with the system.
Yeah. I post this from a Mac OS X Server too. It is correct (I hated it at first, because of the frankenstein merge of Mac and NeXT, but well, I am used to it now. Not as clean as the NeXT, but nowehere as crappy as the mac)
OS X gives a different feeling, but seems very solid too.
Ooops, I re-read your post. Herm. You didn't forgot about multipass.
> What, the Xbox can only do four?
No. The XBox will do 4 at each pass. So with two passes, you've got 8.
Strangely, carmack says 8 passes for Doom (he said that 30 would give renderman-like quality, and that it'll be possible in a near future.), while Abrash says "4 textures" * "shadows done on a second pass".
Fun is that both make 8. Sounds like Carmack and Abrash may have worked together:-)
You totally forgot multipass. Carmack said once in his.plan that he envision doing 20 to 30 passes over all the geometry in a near future.
3D is not only pushing textured lighted triangle. It is often about pushing many of them.
Btw, the second option may or may not be nicer to programmer. It depend on the used algorithm. It will rock if developers uses all the power of the graphic pipeline, but it may suck badly if the configuration of it forces the developer to make an extra pass. The raw-power-make-many-pass option, would probably do better in that case.
The MS bet, is that, as the hardware will be fixed, programmers will develop algorithm very well adapted to it. But they'd better not miss anything important. Strangely, I can take a bet that there will be a flaw somewhere, and that every X-Box game will suffer from it (ie: just at looking at the screenshot, you'll be able to say: oh-oh, an Xbox. Look: the whatever mode is missing, so they are forced to do an extra pass and the pixels are washed out cause the precision of is too low. Or maybe you won't:-) )
> taking the generous work of a lot of other people and sticking their "Netscape" badge
Maybe you should read yours. You obviously *never* looked at mozilla. Go to bugzilla. Look at who does what. Go to bonzai. Look at who does what.
Netscape gives back mozilla to the community. Not the opposite. The community will get back marketshare, mindshare, and ubiquitous presence of mozilla-derived products.
> I wish that Linux extremists actually read their facts
I wish everyone read their facts. For instance I ever went to your home page before replying. You put Objective-C as one of your favorite language, so you are problably not that stupid, but only playing one on slashdot...
> Easter Eggs are a time hounoured tradition of putting your name on something which otherwise wouldn't have your name anywhere near it
The name of developpers used to appear in About boxes.
There is now a tendancy in software (as usual, apple have the lead here, check OSX) to give zero credit to the developers of the software. This is a business decision. A lot of companies hired the most talented people of their competitors by simply getting their names from the about boxes.
But as, both from an ego standpoint and because having its name in a recognised product is important for later jobs, programmer, all product have now easter eggs.
Sure MS easter eggs are bigger than copmpetitors...
> Amazon puts it around their core sales strategy. So, they try to patent it.
So only one buisiness WORLD-WIDE (because the WTO is here to ensure that us europeean will have to recognize american patents in a few years, and the rest of the world will have too) can do one-click shopping ?
In brick-and-mortar, this would means that someone could have patented the idea of going into a store, choosing what you want to buy on shelves, but them into a sort of thingy wiuth little wheels, and paying the goods before living. Would you have granted that patent ? Would you want that there will be only *ONE* store where you can buy stuff this way ? How can you be so clueless ?
> Who here hasn't pushed the rules to check how far they could go, with sports, work, overclocking, etc?
This have nothing in common with the patent case. If you want to push the rules by running faster, more power to you. If you overclock your 486 DX/33 to 1.3Ghz, more power to you.
Those cases have *no* influence on my life, or on anybody else life. Zero. Nada. You have the freedom to do whatever you want but, as we say in french 'La liberte individuelle s'arrete la ou commence celle des autres' (paraphrased: your freedom stops where the freedom of others starts)
The problem with 'amazon bending the rules' is that they have real impact on other people life, and this at two levels:
* First everybody that want to do something as trivial as one-click shopping is fucked
* Second, and much more importantly, trivial patents and IP opn software is moving into a acceptable idea (hey, you have +4 insighfull for such a lame protection of amazon patent)
> They're trying to discover their boundaries.
They are trying to discover *ours*. And they are trying to set those boundaries as far in our ass as we'll let them do.
I've been searching the bugzilla database, and there doesn't seem to be 2500 bugs, but 250. Of them, many are closed or duplicate. That make about 100 bugs.
Many install bugs, most of them due to *bad* CDs (and not yet closed)
Many minor bugs (kind of log of crond is redirected to/var/log/message AND/var/log/crond.log). Ie: harmless
Few missing drivers bugs (eata seems a major one, here)
Few non-distro related bugs (ie: gnome bugs, etc, etc)
Some related to the upgrade of gcc, which now refuses to compile some bad c++ code.
Sure, there are problem, there are people that cannot install, but nothing huge.
Btw, the guy that is linked from the slashdot story appear on several bug reports doing some finger-pointing. Strange, isnt'it ?
NeXT software Inc (okay, it started in the late 80's). The web have been developped on NeXT machines. NeXT have been the first Object Oriented system out there. NeXTstep have been considered "the most respected piece of software on the planet" (don't remember from who the quote was).
The influence of NeXTstep on today computing is much greater than you might suspect (and, it is not only because the close box of Windows 95 is a exact pixel copy of the NeXT one)
They had 10 years of advance. Modern operating systems (even Mac OS X) are not yet up to what NeXTstep was.
First, RedHat is the major linux distributor. For many people linux==RH.
Second, RedHat have both an i386 and a sparc version.
The net result is that, in people brain, the drop of sparc by redhat will be interpreted by linux==i386.
Look: all the major software vendors that go to linux closed-source build their product for the RH distribution. RH dropping sparc means much less chances to have support outside i386
I, for one, have given a lot of credit to redhat only because of their sparc distribution (btw, do you know that Bjarne Stroustrup runs sparc-linux ?).
I had the (stupid) feeling that linux would become somthing like NeXTstep, where there was 4 versions (m68k,i386,hppa,sparc) absolutely identical. You seamlessly cross compiled from one version to another by checking a box. Binaries were fats (ie: the same binary contained version for several archs). I thought that RH would do more and more versions of linux on more and more different hardware, to get the ultimate portable ubiquitous operating system Vendors of applications would just get an i386 version, build with "-arch sparc -arch alpha -arch -m68k -arch arm -arch ppc -arch mips" and get their app on every RH platform.
Looks like they missed the grand scheme. But the most irritating thing is reading those kind of comments on slashdot:
> is making the smart decision to drop it
> it's a business decision, and a good one
I am pretty sure they made a lot of money on sparc, but *indirectly*. With your kind of reasonning, they should fire Alan Cox and the others, as they obviously don't *directly* make money from their work. You know, sometimes a vendor have to keep something for the 'prestige', or for having an edge against the competition.
Someday I'll have to tell y'all the funny story of how my workstation accidentally started routing between our ethernet and token ring networks for our entire corporate WAN.
I've been told that an old version of netware did IPX/IP routing out of the box by default, and many sites had problem.
Slashdot Mirror
> BeOS would have made a great OS X. It was designed from the ground up to be incredibly quick and robust.
:-) )
Mmm. NeXTstep was designed from the ground up to be (somewhat) quick and robust.
BeOS, is very inferior to NeXTstep, and pretty late in the game (I used both, and developed on both. Please, take time to comment if you tried both of them)
BeOS strong point was the multiprocessor design from the ground up (which gave him excellent multitasking possibilities). But on almost all aspects, it is techincally very poor compared to NeXTstep (C++ interfaces, Huge 'NIH' syndrom, lack of most basic utilities)
NeXT bought Apple in december 1996 for -400M$ (fell free to change the order of the words and the sign of the price for the official version), but it took them *4* years to get to OSX beta. And they probably took their fastest path.
> I'm sure if Apple picked up BeOS back in 1996, everything I described above would have been in place by now.
I don't see why you think it would have taken less time than doing it from NeXTstep. (Well, I suspect it is because you lack clues of the involved work and the respective level of NeXTstep and BeOS. But I would not say this
My bet, is that it would never have seen the day. Never. It would have been a Copland/Pink project again. (ie: Coolest thing going to get out of the door real soon now. Wait, it'll kick ass. Oh, you will not beleive it. It is going to be the best thing since sliced bread. Oops, cancelled.)
Cheers,
--fred
> You could if you were less stupid.
Etre traite d'imbecile par un cretin est un plaisir de gourmet
--fred
> if they need to address the Solaris and NT market, use the portability stuff inherent in Cocoa (OPENSTEP)
Unfortunately, you cannot do any YB stuff on windows since the end of last month. Fini. Termine. Nada. The license does not enalbe deplyment of AppKit applications on NT. Period.
> I sure wonder how things would have worked out if Apple had evangelized Cocoa a LOT more strongly pushed the cross-platform features, and ease of programming, and not done the work on Carbon, and not made Classic so easy to fall back on
Oh, yes. I beleive they tried to kill Cocoa, but the it wouldn't die. I hope that *now* they will push cocoa as much as they can. It is rather easy:
* Evangelize
* Do a lot of examples/documentation
* Get correct support for obj from metrowerks
* Do the new APIs on Cocoa first. (Ie: "if you want to get more out of this, you'll have to learn a bit of cocoa")
* Resurect the NT version
And the opensource community could do something nice too by contributing to GNUstep. Think about it:
* Darwin already boot on x86
* XFree86 already runs on darwin
* GNUstep already sort-of run on X
Mmmm. This looks like a open source MacOS X clone on the horizon... maybe in 3 or 4 years.
Cheers,
--fred
Gosh. I hate google. Now you want me to back up my claims !
Seeking. Mmm. You are right. I screwed it badly. Don't know where this '30' figure came from.
Added a bit of extra misinformation into slashdot. Sorry for that.
Cheers,
--fed
> This is why WebObjects is moving away from ObjectiveC.
In fact not exactly. But I'll only tell the true story on my death bed. Let's pretend this is right.
> To these people Apple has stepped back from supporting ObjC where they need it
The problem started before Apple bought NeXT. NeXTstep, then OPENSTEP was (and still is) used by very big non-disclosed corporation, that basically used NeXT as their outsourced R&D staff. Those guys have millions and millions of line of code, involving Distributed Objects and a lot of EOF.
Then, half accidentally, NeXT stumbled on WebObjects, and decided that it would be the next big thing. (This kind of shitf was classic in the NeXT community, as we already had the 'education market' as the orignal big thing, abandonned to 'the desktop publishing market' (or was it the opposite?), killed for 'interpersonal computing', itself replaced by 'mission-critical custom apps', while the hardware line got a chaotic course [m68k->i386->hppa->sparc], then the OS disapeared replaced by the yellow box, while the whole libraries were rewritten for OpenStep, fundamentally breaking *every* application and existing source code)
At this point of time (about 96), we had lost our Hardware platform, and our Operating System.
NeXT started to tell all developers that the future was java, but promised that Objective-C would stay. And the reason Objective-C would stay was:
"We cannot kill objective-c because WebObjects is based on it"
Then apple bought NeXT, and started reviving a defunct operating system.
Then the re-write of WebObjects on Java appeared, along the java version of EOF.
It looked like we were loosing our language too. Apple was not communicating on ObjC, most new cocoa documentation was about java, because MacOS developpers don't want to hear about objc. Internally, Apple decided to get rid of ObjC too.
So, it is rather refreshing to find that they changed their mind (probably because they had no choice, IMHO)
> in fact most of the apps that were ported to Java for early DP's were moved back to ObjC for DP4 and beyond...
Yep. This worth every press release of the earth.
Cheers,
--fred
> when you say "he" do you mean "it"
Yes. I am French, a language were there is no distinction between 'he' and 'it' ('il' means both 'he' and 'it'). I can't help making this error again and again.
Cheers,
--fred
From the article:
:-)
The Mac OS X equivalent of the classic Mac OS Control Panel, the System Preferences application, has been revamped for Beta. True to Apple's "we eat our own dog food" motto, System Preferences was a Java application (using the Cocoa API) in DP4. Unfortunately, it was also one of the slowest launching applications on the system due to the overhead of loading the VM. In Beta, it uses the Objective C interface to the Cocoa API instead, and performance is now acceptable.
This looks real good for ObjC. You now what ? NeXTstep really did have 10 years of advance. He's only getting mainstream now...
--fred
> It doesn't. To run OS X Beta you have to disable the second processor through Open Firmware
Are you sure of that ? I know someone running the beta on a bi G4, and he would be mad if he had to turn one ofthem off.
Can someone shed the light on this issue ?
Cheers,
--fred
> The Desktop uses the NeXT Workspace Manager code, mostlike
No. It shares zero code. The Desktop is written in C++/Carbon and have no Objc. Hey, I provided you the way to check (with nm) if you have a developer system, or if you get the tools from darwin.
The Desktop have nowhere the 'feel' of the original Workspace. And he is crippled with bugs (NFS mounts randomly appearing as links that points to nowhere, frequent temporary freezes, I could cite dozens like that). I hope it will not be too much tied into the OS, so we can write an ObjC replacement (from a developer point of view, the question is knowing if the NSWorkspace protocol will be implemented by the Dock [as in NeXTstep], or by the Desktop. I would prefer the former)
> Sadly, ObjC is being pretty aggressively phased out by Apple in favor of Java
This is what I thought, so I checked all the fundamental binaries of OSX. And it is quite the opposite.
IIRC, only the calculator and the Setup assistant are in java.
Only the Desktop and a handfull of (crappy) demos are in Carbon.
Even the Dock is in ObjC. And the MP3 player or the Clock (must check again, I don't have my MOSX box handy, but I'm pretty sure of that). And preferences. I think the loginPanel is objc too.
It is very refreshing to see that ObjC applications are of higher quality. When you get hooked to cocoa, it is hard to get back. I am incredibely happy that apple have now teams of engineers back to ObjC. This is, IMHO, the best new of this public beta. (In a side note, the new ProjectBuilder don't correctly support java yet. Funny as hell, if you want my opinion)
> The next rev of WebObjects, I believe, won't support ObjC, for example.
This is right. But this decision have been taken years ago, when they wanted to phase out all ObjC. But it looks like they are going back on that, which is a very very good thing.
Unfortunately, WebObject Objc is probably lost forever. And EOF (not present in the beta) will be missed too...
Cheers,
--fred
Applications (.app) are really folders. Drag and drop them in a shell to look inside them.
Cd to the Content/MacOS directory. There is the real application exectuable.
Then, if you have the developer tools (from darwin, for instance), you can look at executable content with 'nm'.
If you have things like:
0000f078 t -[MyClass _myMethod:]
then it is a Cocoa application.
You'll find that most of them are really cocoa applications.
The Desktop seems (to me) a poor rip of the NeXT Workspace Manager. It is real crappy, have many bugs, and is nowhere as usable as the program he tries to clone.
It is interesting to see that it is (almost) the only important Carbon app shipped with the system.
Long life to ObjC.
Cheers,
--fred
Yeah. I post this from a Mac OS X Server too. It is correct (I hated it at first, because of the frankenstein merge of Mac and NeXT, but well, I am used to it now. Not as clean as the NeXT, but nowehere as crappy as the mac)
OS X gives a different feeling, but seems very solid too.
Cheers,
--fred
From Name.Space, Incorporated:
.sucks
From Rathbawn Computers Limited
.sansansan
Nice TLDs, indeed
Cheers,
--fred
Ooops, I re-read your post. Herm. You didn't forgot about multipass.
:-)
> What, the Xbox can only do four?
No. The XBox will do 4 at each pass. So with two passes, you've got 8.
Strangely, carmack says 8 passes for Doom (he said that 30 would give renderman-like quality, and that it'll be possible in a near future.), while Abrash says "4 textures" * "shadows done on a second pass".
Fun is that both make 8. Sounds like Carmack and Abrash may have worked together
Cheers,
--fred
You totally forgot multipass. Carmack said once in his .plan that he envision doing 20 to 30 passes over all the geometry in a near future.
:-) )
3D is not only pushing textured lighted triangle. It is often about pushing many of them.
Btw, the second option may or may not be nicer to programmer. It depend on the used algorithm. It will rock if developers uses all the power of the graphic pipeline, but it may suck badly if the configuration of it forces the developer to make an extra pass. The raw-power-make-many-pass option, would probably do better in that case.
The MS bet, is that, as the hardware will be fixed, programmers will develop algorithm very well adapted to it. But they'd better not miss anything important. Strangely, I can take a bet that there will be a flaw somewhere, and that every X-Box game will suffer from it (ie: just at looking at the screenshot, you'll be able to say: oh-oh, an Xbox. Look: the whatever mode is missing, so they are forced to do an extra pass and the pixels are washed out cause the precision of is too low. Or maybe you won't
Cheers,
--fred
> taking the generous work of a lot of other people and sticking their "Netscape" badge
Maybe you should read yours. You obviously *never* looked at mozilla. Go to bugzilla. Look at who does what. Go to bonzai. Look at who does what.
Netscape gives back mozilla to the community. Not the opposite. The community will get back marketshare, mindshare, and ubiquitous presence of mozilla-derived products.
> I wish that Linux extremists actually read their facts
I wish everyone read their facts. For instance I ever went to your home page before replying. You put Objective-C as one of your favorite language, so you are problably not that stupid, but only playing one on slashdot...
Cheers,
--fred
> Easter Eggs are a time hounoured tradition of putting your name on something which otherwise wouldn't have your name anywhere near it
The name of developpers used to appear in About boxes.
There is now a tendancy in software (as usual, apple have the lead here, check OSX) to give zero credit to the developers of the software. This is a business decision. A lot of companies hired the most talented people of their competitors by simply getting their names from the about boxes.
But as, both from an ego standpoint and because having its name in a recognised product is important for later jobs, programmer, all product have now easter eggs.
Sure MS easter eggs are bigger than copmpetitors...
Cheers,
--fred
> Amazon puts it around their core sales strategy. So, they try to patent it.
So only one buisiness WORLD-WIDE (because the WTO is here to ensure that us europeean will have to recognize american patents in a few years, and the rest of the world will have too) can do one-click shopping ?
In brick-and-mortar, this would means that someone could have patented the idea of going into a store, choosing what you want to buy on shelves, but them into a sort of thingy wiuth little wheels, and paying the goods before living. Would you have granted that patent ? Would you want that there will be only *ONE* store where you can buy stuff this way ? How can you be so clueless ?
> Who here hasn't pushed the rules to check how far they could go, with sports, work, overclocking, etc?
This have nothing in common with the patent case. If you want to push the rules by running faster, more power to you. If you overclock your 486 DX/33 to 1.3Ghz, more power to you.
Those cases have *no* influence on my life, or on anybody else life. Zero. Nada. You have the freedom to do whatever you want but, as we say in french 'La liberte individuelle s'arrete la ou commence celle des autres' (paraphrased: your freedom stops where the freedom of others starts)
The problem with 'amazon bending the rules' is that they have real impact on other people life, and this at two levels:
* First everybody that want to do something as trivial as one-click shopping is fucked
* Second, and much more importantly, trivial patents and IP opn software is moving into a acceptable idea (hey, you have +4 insighfull for such a lame protection of amazon patent)
> They're trying to discover their boundaries.
They are trying to discover *ours*. And they are trying to set those boundaries as far in our ass as we'll let them do.
Cheers,
--fred
I've been searching the bugzilla database, and there doesn't seem to be 2500 bugs, but 250. Of them, many are closed or duplicate. That make about 100 bugs.
/var/log/message AND /var/log/crond.log). Ie: harmless
Many install bugs, most of them due to *bad* CDs (and not yet closed)
Many minor bugs (kind of log of crond is redirected to
Few missing drivers bugs (eata seems a major one, here)
Few non-distro related bugs (ie: gnome bugs, etc, etc)
Some related to the upgrade of gcc, which now refuses to compile some bad c++ code.
Sure, there are problem, there are people that cannot install, but nothing huge.
Btw, the guy that is linked from the slashdot story appear on several bug reports doing some finger-pointing. Strange, isnt'it ?
Cheers,
--fred
Do you think that the guy Apple is suing for leaking pictures of the cube will be forced to buy apple shares ?
That would be painfull for him.
Cheers,
--fred
Yeah. It deserves the +5.
--fred
Bruce Perens never die. They are just casts to void
I love slashdot for this. News for the nerds, stuff that matters. How true...
> But in the 1990s?
NeXT software Inc (okay, it started in the late 80's). The web have been developped on NeXT machines. NeXT have been the first Object Oriented system out there. NeXTstep have been considered "the most respected piece of software on the planet" (don't remember from who the quote was).
The influence of NeXTstep on today computing is much greater than you might suspect (and, it is not only because the close box of Windows 95 is a exact pixel copy of the NeXT one)
They had 10 years of advance. Modern operating systems (even Mac OS X) are not yet up to what NeXTstep was.
I call that a technical achievment.
Cheers,
--fred
Well.
First, RedHat is the major linux distributor. For many people linux==RH.
Second, RedHat have both an i386 and a sparc version.
The net result is that, in people brain, the drop of sparc by redhat will be interpreted by linux==i386.
Look: all the major software vendors that go to linux closed-source build their product for the RH distribution. RH dropping sparc means much less chances to have support outside i386
I, for one, have given a lot of credit to redhat only because of their sparc distribution (btw, do you know that Bjarne Stroustrup runs sparc-linux ?).
I had the (stupid) feeling that linux would become somthing like NeXTstep, where there was 4 versions (m68k,i386,hppa,sparc) absolutely identical. You seamlessly cross compiled from one version to another by checking a box. Binaries were fats (ie: the same binary contained version for several archs). I thought that RH would do more and more versions of linux on more and more different hardware, to get the ultimate portable ubiquitous operating system Vendors of applications would just get an i386 version, build with "-arch sparc -arch alpha -arch -m68k -arch arm -arch ppc -arch mips" and get their app on every RH platform.
Looks like they missed the grand scheme. But the most irritating thing is reading those kind of comments on slashdot:
> is making the smart decision to drop it
> it's a business decision, and a good one
I am pretty sure they made a lot of money on sparc, but *indirectly*. With your kind of reasonning, they should fire Alan Cox and the others, as they obviously don't *directly* make money from their work. You know, sometimes a vendor have to keep something for the 'prestige', or for having an edge against the competition.
It is a sloppy road they took.
Cheers,
--fred
Yeah. The correct wording is:
There's no culpability, and above all, no one to *blame* when it all goes south
Cheers,
--fred
I've been told that an old version of netware did IPX/IP routing out of the box by default, and many sites had problem.
Cheers
--fred