Back in college (95, can it really be 10 years ago?) we ran a mac lab that had Word, Wordperfect, and an interesting WP called Nisus. Nisus was cool because it had regular expression searches. You could do normal search/replace, use a partially visual pseudo regexp search, or use full backslash \1 search replace. My boss used the viusal one to learn regular expressions, then used full to do some text file => HTML conversions. I used WordPerfect usually, even though it was a bit less stable on the mac than word was at the time (corrupted preferences meant crashing WordPerfect) because it had "Reveal Codes". There's nothing I've seen that can help you fix garbled layout quite like reveal codes can. The most innovative things microsoft has done? Squiggly lines and a talking, sulking paperclip. The spellchecker squiggles were at least partially useful, though sometimes got in the way and sometimes distracted you from your primary task. The grammar squiggles, never useful, always turned off. As for usefulness of Clippy, well....
The stupid thing about the guys rant against OpenDoc is that there's NOTHING preventing MS from using opendocument. Hell, they're on the OASIS board. Massacussetts is dictating file format, but not saying no MS. Since Office 12 isn't out yet, it shouldn't be too hard to write a inport/output filter for it, especially since they have a format that is similar in concept (XML based files). If a few OOo guys can do it, the mighty MS can do it as well. They probably have a library someplace internally that already does it, just in case, but they need to rail against it because they understand better than anyone else the power of their network effects. Word Doc requires Word which requires Windows.
And why would anybody with half a brain trust encryption software they can't audit if need be? C'mon: the cryptosystem has to mathematically so whoopass that wether or not an attacker has the source makes no difference. Proprietary is talking the talk. FOSS is walking the walk. How many IT departments have that level of skill to perform an audit for algorithm holes in FOSS? the fact that real cryptographers are still finding (occasional) bugs in openssl code that's been out for years doesn't give me a lot of confidence a server admin would be able to find any. Cryptography in particular has very subtle bugs. Where would you look to find subtle timing bugs that give off info that don't blast the algorithm wide open, but cut down keyspace? Auditing source is a cool thing to have, but I doubt it has that much use in the real world. There's always going to be the assumption that someone else has done it.
We're a large'ish organization, at least in our field. We don't have any SSH.com software around, we use openssh on the servers, and putty on the Windows clients, openssh on the servers. We integrated RSA SecureID into the system, then we scotched that initiative. We're in finance, and in general we shy away from OpenSource, preferring someone to partner with for support, but there's never been any plans to replace OpenSSH. It just works.
Though i agree with you in that a lot of people don't understand that getting software into an enterprise is a complex process, and simply saying "this roxx0rz!" won't cut it, OpenSSH does fit a lot of organizations. It just works.
Yeah but read the license agreement (available if you put in fake info for a trial download on http://www.ssh.com/support/downloads/tectia-client /evaluation.mpl). It clears them of pretty much everything. IANAL, but I don't think this is any more or less protection than what you get from most typical OSI licenses.
Though TFA mentions extra protection for rule sets like SOX and others, actually checking the license shows them pretty fairly lacking. Like most EULAs, you give up pretty much everything. This is what you get from: http://www.ssh.com/support/downloads/tectia-client /evaluation.mpl It looks like it is their normal license, plus an amendment for the temporary license period. I extracted some parts on liability, yadda yadda.
8. WARRANTY
LICENSOR EXPRESSLY DISCLAIMS, TO THE EXTENT PERMITTED BY APPLICABLE LAW, ALL WARRANTIES, WHETHER EXPRESS, IMPLIED OR STATUTORY, INCLUDING, WITHOUT LIMITATION, ANY IMPLIED WARRANTIES OF MERCHANTABILITY, OF FITNESS FOR A PARTICULAR PURPOSE, NONINFRINGEMENT OF THIRD PARTY INTELLECTUAL PROPERTY RIGHTS, AND ANY WARRANTY THAT MAY ARISE BY REASON OF TRADE USAGE, CUSTOM OR COURSE OF DEALING. LICENSOR DOES NOT WARRANT THAT THE SOFTWARE WILL BE FREE FROM BUGS OR THAT ITS USE WILL BE UNINTERRUPTED NOR THAT THE SOFTWARE WILL OPERATE WITH ANY HARDWARE AND/OR OTHER SOFTWARE OR REGARDING THE USE, OR THE RESULTS OF THE USE, OF THE SOFTWARE OR DOCUMENTATION IN TERMS OF CORRECTNESS, ACCURACY, RELIABILITY OR OTHERWISE. WITHOUT LIMITING THE FOREGOING, YOU ACKNOWLEDGE THAT THE SOFTWARE IS PROVIDED "AS IS," WITHOUT WARRANTY OF ANY KIND.
9. LIMITATION OF LIABILITY
THE ENTIRE RISK AS TO RESULTS AND PERFORMANCE OF THE SOFTWARE IS ASSUMED BY YOU. ANY LIABILITY OF LICENSOR WITH RESPECT TO THE SOFTWARE, THE PERFORMANCE THEREOF OR DEFECTS THEREIN, OR UNDER THIS AGREEMENT, UNDER ANY WARRANTY, NEGLIGENCE, STRICT LIABILITY OR OTHER LEGAL THEORY SHALL BE LIMITED EXCLUSIVELY TO PRODUCT REPLACEMENT OR, IF REPLACEMENT IS INADEQUATE AS A REMEDY, OR, IN LICENSOR'S SOLE OPINION, IMPRACTICAL, TO A REFUND OF THE ACTUAL AMOUNT PAID BY YOU TO LICENSOR, IF ANY, FOR THE SOFTWARE OR SERVICES GIVING RISE TO THE CLAIM.
10. DISCLAIMER OF DAMAGES
UNDER NO CIRCUMSTANCES WILL LICENSOR OR ITS LICENSORS BE LIABLE FOR ANY SPECIAL, INDIRECT, INCIDENTAL, EXEMPLARY OR CONSEQUENTIAL DAMAGES OF ANY KIND OR NATURE WHATSOEVER, WHETHER BASED ON CONTRACT, WARRANTY, TORT (INCLUDING NEGLIGENCE), STRICT LIABILITY OR OTHERWISE, ARISING OUT OF OR IN ANY WAY RELATED TO THE SOFTWARE, THIS AGREEMENT, WHETHER DUE TO A BREACH OF LICENSOR'S OBLIGATIONS HEREUNDER OR OTHERWISE, EVEN IF LICENSOR OR ITS LICENSORS HAVE BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGE OR IF SUCH DAMAGE COULD HAVE BEEN REASONABLY FORESEEN, AND NOTWITHSTANDING ANY FAILURE OF ESSENTIAL PURPOSE OF ANY EXCLUSIVE REMEDY PROVIDED IN THIS AGREEMENT. SUCH LIMITATION ON DAMAGES INCLUDES, BUT IS NOT LIMITED TO, DAMAGES FOR LOSS OF GOODWILL, LOST PROFITS, LOSS OF DATA OR SOFTWARE, WORK STOPPAGE, COMPUTER FAILURE OR MALFUNCTION OR IMPAIRMENT OF OTHER GOODS. IN NO EVENT WILL LICENSOR OR ITS LICENSORS BE LIABLE FOR THE COSTS OF PROCUREMENT OF SUBSTITUTE SOFTWARE OR SERVICES.
YOU ACKNOWLEDGE THAT THIS SOFTWARE IS NOT DESIGNED OR LICENSED FOR USE IN ON-LINE EQUIPMENT IN HAZARDOUS ENVIRONMENTS SUCH AS OPERATION OF NUCLEAR FACILITIES, AIRCRAFT NAVIGATION OR CONTROL, OR LIFE-CRITICAL APPLICATIONS. LICENSOR EXPRESSLY DISCLAIMS ANY LIABILITY RESULTING FROM USE OF THE SOFTWARE IN ANY SUCH ON-LINE EQUIPMENT IN HAZARDOUS ENVIRONMENTS AND ACCEPTS NO LIABILITY IN RESPECT OF ANY ACTIONS OR CLAIMS BASED ON THE USE OF THE SOFTWARE IN ANY SUCH ON-LINE EQUIPMENT IN HAZARDOUS ENVIRONMENTS BY YOU. FOR PURPOSES OF THIS PARAGRAPH, THE TERM "LIFE-CRITICAL APPLICATION" MEANS AN APPLICATION IN WHICH THE FUNCTIONING OR MALFUNCTIONING OF THE SOFTWARE MAY RESULT DIRECTLY OR INDIRECTLY IN PHYSICAL INJURY OR LOSS OF HUMAN LIFE.
Not sure what Online in Hazardous environments means. There's only a partial explanation; one additional interpretaion would have all of the Internet hazardous because of crackers. I like how some companies beat you over the head with "you can't sue anybody" then neglect to meantion you can't really sue them either. It's a true statement of most OSI licenses, but it's no worse than theirs in that regard.
Interesting, but doesn't solve the problem of port knocking. A rootkit would have kernel access, so would be able to set up a port knocking situation. You might be able to detect a certain subclass of exploits, but never be able to say you caught the full set.
First, GIF can only handle 8-bit (256 color) pallettes. Granted, each line can have it's own pallette, but it's still fairly restrictive, compared to PNG's 24-bit color. GIFs can handle up to 8 bit (any power of 2 palette), with a single palette per image, not per line.
Not as useless, since Opera and Google have a business relationship now. I think market share won't be the deciding factor whether or not to spend development effort on this. Besides, if it was all about market share, they wouldn't bother with spending money and time developing for Firefox, just spend money developing IE. There's obviously other reasons for development.
* Software that depends on a working internet connection Windows XP activation? Windows Media player that needs to phone home to check DRM on a song you already bought?
* Service outages completely out of your control What service outages can you control? Would BSODing your computer be in your control? How about all the times you have to uninstall/reinstall your app because of it crashing? Isn't opening a new window easier?
* Platform issues all over again (Mac vs Linux vs Windows 2k vs Windows XP, Firefox vs Explorer vs Opera, JVM issues, etc.) Google works fine on mac vs. Linux. XPCOM apps work just as well on all Mozilla platforms - can run code on Linux, Mac, IE. Part of the platform issues was MSs deliberate use of non-standards. If you have a platform monoculture, of course you won't have platform issues, but you do lose choice.
* No customer-controlled version control (want to stay on Powerpoint 2007 Service Pack 1 because SP2 breaks your slides? Too bad! Not upgrading your app because in the next 24 hours you have a million dollar client proposal? Sorry, your app service provider wants to silently roll out a "bugfix" that causes problems for you) I wish I could stay on Word 97, Word 95 even if it handled the scroll wheel natively. I can't, i can't buy a new license for it, I won't be compatible with all of the other people out there. I think somewhere, someplace, I have a disk with some Word 6.0 files on it. Will Word 12 be able to read it? * Having to license software yearly, or go through byzantine activation procedures (Quark XPress 6.0 activation, anyone?) Are you aware of MS License 6.0?
I agree with some of your points, but these are symptoms of complexity, not anything specific with the Web. Microsoft would have you do many of these things (they'd love to have Word as a service, make you pay per word?) just under their control. You're talking about some complexities as if they are specific to the web model, but they are not. Versions are a problem no matter how your app is delivered.
I don't want a web word processor. That would not work at all. But certain things do lend themselves to the web, and we should be open to that, when it makes sense.
BeOS also chopped-and-changed who it supported, which didn't help. It was originally for the PPC,... Actually, PPC was the second chip. It's a rare company that can bring a user base through not one but two migrations. Apple's 68K => PPC was one of the smoothest ever, they still lost huge marketshare (for this and other reasons).
TOE would help free up some cycles on an individual machine, but I'm guessing the question doesn't ask that. It's a little unclear, but I'm guessing it's "If already I can hit a Virtual machine, why can't I code to a virtual machine who's real machine happens to be the cluster?"
Generally it looks like he's asking for a VM to be an abstraction layer over a cluster. The problem is abstractions are simplifications, and you can't just simplify away the real problems of a cluster. There are some solutions that can always help simplify (OpenMPI sounds closest with what little research I've done) but clusters are hard, and simplification layers on top of it can't delete all complexity.
truss/strace is a syscall tracer. Anything in your app that makes a syscall gets it's arguments and return values logged. ltrace adds the ability to do the same with dynamic library calls.
dtrace is much different, you have areas of your kernel that have probes, places that accumulate data. dtrace is a language where you can read these probe areas (including the syscall interface) and print them out to user level and figure out whats going on (wrong) in your kernel.
For the people who say Sun isn't real about open source, they should realize this is a differentiating technology, years ahead of what anything in Linux/bsd or commercial linuxes have. If it's going into the BSD kernel, it's probably also BSD licensed, meaning all UNIXes can take this.
Ironically, WinNT was the first OS to have POSIX compliance. MS was the first company to bother with the cerification. The UNIX companies saw the fact that they were POSIX as blatantly obvious, and didn't both initially. They came around when they saw they were losing "POSIX" contracts to WinNT.
Originally, WinNT was a Microkernel, with OS2 and POSIX support. Both of the latter were bare minimums, to satisfy contractual obligations (IBM and OS/2) or checklists for new contracts (POSIX). Neither worked well. As tiem went on, more and more things ended up in the kernel (graphics, apps and servers) it would be hard to call it a microkernel anymore, more like some kind of hybrid.
That may be a semantic argument. The definition of "the Internet" (with capitalization) I used is the global TCP/IP network we're all familiar with. My parent post used Internet, and it colored my response. There have been many global internets (BITNET, uunet, even a global AppleTalk net tunneled through TCP/IP, etc.) but for most people, Internet is TCP/IP.
I don't think they've ever really made a finished product. DOS was pretty good for what it did, espectially since it wasn't natively internet friendly. Since DOS 1.0 came out in 1981, predating the Internet by a few years, it would be hard for it to be natively Internet friendly on release. Neither UNIX (being the platform the Internet was created on) nor MacOS had Internet connectivity natively at initial release either. Eventually DOS did offer Internet connectivity, my first browsing of ftpspace and gopherspace was on DOS, but the Internet is essentially synonymous with webspace now, and the WWW really cries out for a bitmapped not a character driven display.
On another note, people keep wondering why they should upgrade to another version of Word, since there are no new compelling features. Word seems to be, for a vast majority of users "done".
Marketshare != mindshare. Apple definitely punches above it's weight, more than any other tech company probably. I'm sure they did, just as Intel fingerprints are all over Steve Jobs' sniping at IBM over PowerPC performance (right before the dual core 970s were launched).
Not necessarily. x86 is just ABI compatibility. The original Pentiums were "from scratch" to some degree, they took ABi compatibility but impemented in a new way that had real performance benefits. The Pentium pro did as well, which doesn't run x86 code directly, just decodes them into RISC type micro-ops and scheduled those instead. The cores never really differed until NetBurst, which had some innovative ideas about caching micro-ops, but it turns out it's not as efficient as Pentium-Pro core (as seen in Pentium 3, pentium M) on some metrics. So it's obvious Pentium 4 core is bad, the question is can they hit one out of the park with a new core, or will the just extend the old pentium-pro core? we'll see.
The cross licensing deal was interesting. Led to a few thigns...
Originally Intel wasn't sure it could pump out enough 386's, so it licensed with AMD to pick up the slack. It saw how much money it was leaving on the table, so when the 486's came around, it tried freezing AMD out, including calling them i486, and tell AMD they can't use the i. Case went to court, you can't trademark a letter (Zilog tried, with Z). The next generation came out with a silly marketing name, Pentium.
Significantly now, it was a cross licensing deal, giving Intel hands on AMD products as well. That's how they can do EM64T, a copy of AMD64, and give AMD no royalties.
I'm not sure how much this "new chip" you want differs than Itanium, or PowerPC for that matter (though I believe PowerPC is much better engineered than Itanium).
Itanium, current marketing aside, was never intended to be a niche chip. It was supposed to be so radical that it replaced the x86 ecosystem with a new one (and happily for Intel, one much more under it's control than x86). Itanium execution mistakes aside, the success of AMD64/EM64T over ia64 shows that the market does want backwards compatibility.
IBM tried this with PowerPC. It was supposed to be a new RISC ecosystem to compete against and eventually replace x86. Some big names had a port - MacOS, WindowsNT, Solaris. Apple had too - the 68x00 line was running out of gas fast, and the transition was actually very smooth, though it still lost massive amounts of market share. Sun pulled back, probably caught with the same "why port to another chip if software is just a way to move our hardware" problem APple has faced. Windows NT never really sold, and running under emulation sucked. Standard chicken and egg problem there.
Even Linux, which because of being open source is relatively simpler too bootstrap a PowerPC system, is still predominantly x86. More people on x86 means more binaries targetting x86, more testing, meaning x86 is more polished, people faced with a choice pick x86.
Slashdot Mirror
I just miss OGG the caveman... THe good old days of slashdot trolls...
Back in college (95, can it really be 10 years ago?) we ran a mac lab that had Word, Wordperfect, and an interesting WP called Nisus. Nisus was cool because it had regular expression searches. You could do normal search/replace, use a partially visual pseudo regexp search, or use full backslash \1 search replace. My boss used the viusal one to learn regular expressions, then used full to do some text file => HTML conversions. I used WordPerfect usually, even though it was a bit less stable on the mac than word was at the time (corrupted preferences meant crashing WordPerfect) because it had "Reveal Codes". There's nothing I've seen that can help you fix garbled layout quite like reveal codes can. The most innovative things microsoft has done? Squiggly lines and a talking, sulking paperclip. The spellchecker squiggles were at least partially useful, though sometimes got in the way and sometimes distracted you from your primary task. The grammar squiggles, never useful, always turned off. As for usefulness of Clippy, well....
The stupid thing about the guys rant against OpenDoc is that there's NOTHING preventing MS from using opendocument. Hell, they're on the OASIS board. Massacussetts is dictating file format, but not saying no MS. Since Office 12 isn't out yet, it shouldn't be too hard to write a inport/output filter for it, especially since they have a format that is similar in concept (XML based files). If a few OOo guys can do it, the mighty MS can do it as well. They probably have a library someplace internally that already does it, just in case, but they need to rail against it because they understand better than anyone else the power of their network effects. Word Doc requires Word which requires Windows.
And why would anybody with half a brain trust encryption software they can't audit if need be? C'mon: the cryptosystem has to mathematically so whoopass that wether or not an attacker has the source makes no difference. Proprietary is talking the talk. FOSS is walking the walk.
How many IT departments have that level of skill to perform an audit for algorithm holes in FOSS? the fact that real cryptographers are still finding (occasional) bugs in openssl code that's been out for years doesn't give me a lot of confidence a server admin would be able to find any. Cryptography in particular has very subtle bugs. Where would you look to find subtle timing bugs that give off info that don't blast the algorithm wide open, but cut down keyspace? Auditing source is a cool thing to have, but I doubt it has that much use in the real world. There's always going to be the assumption that someone else has done it.
We're a large'ish organization, at least in our field. We don't have any SSH.com software around, we use openssh on the servers, and putty on the Windows clients, openssh on the servers. We integrated RSA SecureID into the system, then we scotched that initiative. We're in finance, and in general we shy away from OpenSource, preferring someone to partner with for support, but there's never been any plans to replace OpenSSH. It just works.
Though i agree with you in that a lot of people don't understand that getting software into an enterprise is a complex process, and simply saying "this roxx0rz!" won't cut it, OpenSSH does fit a lot of organizations. It just works.
Yeah but read the license agreement (available if you put in fake info for a trial download on http://www.ssh.com/support/downloads/tectia-client /evaluation.mpl). It clears them of pretty much everything. IANAL, but I don't think this is any more or less protection than what you get from most typical OSI licenses.
Not sure what Online in Hazardous environments means. There's only a partial explanation; one additional interpretaion would have all of the Internet hazardous because of crackers. I like how some companies beat you over the head with "you can't sue anybody" then neglect to meantion you can't really sue them either. It's a true statement of most OSI licenses, but it's no worse than theirs in that regard.
Interesting, but doesn't solve the problem of port knocking. A rootkit would have kernel access, so would be able to set up a port knocking situation. You might be able to detect a certain subclass of exploits, but never be able to say you caught the full set.
and me with no mod points.....
First, GIF can only handle 8-bit (256 color) pallettes. Granted, each line can have it's own pallette, but it's still fairly restrictive, compared to PNG's 24-bit color.
GIFs can handle up to 8 bit (any power of 2 palette), with a single palette per image, not per line.
I wish I would apply the concept of labels to files on my harddisk.
MacOS has had since at least since System 6 days, early 90's or so.
Not as useless, since Opera and Google have a business relationship now. I think market share won't be the deciding factor whether or not to spend development effort on this. Besides, if it was all about market share, they wouldn't bother with spending money and time developing for Firefox, just spend money developing IE. There's obviously other reasons for development.
Sounds pretty damn scary to me, too.
* Software that depends on a working internet connection
Windows XP activation? Windows Media player that needs to phone home to check DRM on a song you already bought?
* Service outages completely out of your control
What service outages can you control? Would BSODing your computer be in your control? How about all the times you have to uninstall/reinstall your app because of it crashing? Isn't opening a new window easier?
* Platform issues all over again (Mac vs Linux vs Windows 2k vs Windows XP, Firefox vs Explorer vs Opera, JVM issues, etc.)
Google works fine on mac vs. Linux. XPCOM apps work just as well on all Mozilla platforms - can run code on Linux, Mac, IE. Part of the platform issues was MSs deliberate use of non-standards. If you have a platform monoculture, of course you won't have platform issues, but you do lose choice.
* No customer-controlled version control (want to stay on Powerpoint 2007 Service Pack 1 because SP2 breaks your slides? Too bad! Not upgrading your app because in the next 24 hours you have a million dollar client proposal? Sorry, your app service provider wants to silently roll out a "bugfix" that causes problems for you)
I wish I could stay on Word 97, Word 95 even if it handled the scroll wheel natively. I can't, i can't buy a new license for it, I won't be compatible with all of the other people out there. I think somewhere, someplace, I have a disk with some Word 6.0 files on it. Will Word 12 be able to read it?
* Having to license software yearly, or go through byzantine activation procedures (Quark XPress 6.0 activation, anyone?)
Are you aware of MS License 6.0?
I agree with some of your points, but these are symptoms of complexity, not anything specific with the Web. Microsoft would have you do many of these things (they'd love to have Word as a service, make you pay per word?) just under their control. You're talking about some complexities as if they are specific to the web model, but they are not. Versions are a problem no matter how your app is delivered.
I don't want a web word processor. That would not work at all. But certain things do lend themselves to the web, and we should be open to that, when it makes sense.
BeOS also chopped-and-changed who it supported, which didn't help. It was originally for the PPC,...
Actually, PPC was the second chip. It's a rare company that can bring a user base through not one but two migrations. Apple's 68K => PPC was one of the smoothest ever, they still lost huge marketshare (for this and other reasons).
A lot of the command line net tools are still BSD though....
TOE would help free up some cycles on an individual machine, but I'm guessing the question doesn't ask that. It's a little unclear, but I'm guessing it's "If already I can hit a Virtual machine, why can't I code to a virtual machine who's real machine happens to be the cluster?"
Generally it looks like he's asking for a VM to be an abstraction layer over a cluster. The problem is abstractions are simplifications, and you can't just simplify away the real problems of a cluster. There are some solutions that can always help simplify (OpenMPI sounds closest with what little research I've done) but clusters are hard, and simplification layers on top of it can't delete all complexity.
truss/strace is a syscall tracer. Anything in your app that makes a syscall gets it's arguments and return values logged. ltrace adds the ability to do the same with dynamic library calls.
dtrace is much different, you have areas of your kernel that have probes, places that accumulate data. dtrace is a language where you can read these probe areas (including the syscall interface) and print them out to user level and figure out whats going on (wrong) in your kernel.
For the people who say Sun isn't real about open source, they should realize this is a differentiating technology, years ahead of what anything in Linux/bsd or commercial linuxes have. If it's going into the BSD kernel, it's probably also BSD licensed, meaning all UNIXes can take this.
Ironically, WinNT was the first OS to have POSIX compliance. MS was the first company to bother with the cerification. The UNIX companies saw the fact that they were POSIX as blatantly obvious, and didn't both initially. They came around when they saw they were losing "POSIX" contracts to WinNT.
Originally, WinNT was a Microkernel, with OS2 and POSIX support. Both of the latter were bare minimums, to satisfy contractual obligations (IBM and OS/2) or checklists for new contracts (POSIX). Neither worked well. As tiem went on, more and more things ended up in the kernel (graphics, apps and servers) it would be hard to call it a microkernel anymore, more like some kind of hybrid.
That may be a semantic argument. The definition of "the Internet" (with capitalization) I used is the global TCP/IP network we're all familiar with. My parent post used Internet, and it colored my response. There have been many global internets (BITNET, uunet, even a global AppleTalk net tunneled through TCP/IP, etc.) but for most people, Internet is TCP/IP.
I don't think they've ever really made a finished product. DOS was pretty good for what it did, espectially since it wasn't natively internet friendly.
Since DOS 1.0 came out in 1981, predating the Internet by a few years, it would be hard for it to be natively Internet friendly on release. Neither UNIX (being the platform the Internet was created on) nor MacOS had Internet connectivity natively at initial release either. Eventually DOS did offer Internet connectivity, my first browsing of ftpspace and gopherspace was on DOS, but the Internet is essentially synonymous with webspace now, and the WWW really cries out for a bitmapped not a character driven display.
On another note, people keep wondering why they should upgrade to another version of Word, since there are no new compelling features. Word seems to be, for a vast majority of users "done".
Marketshare != mindshare. Apple definitely punches above it's weight, more than any other tech company probably.
I'm sure they did, just as Intel fingerprints are all over Steve Jobs' sniping at IBM over PowerPC performance (right before the dual core 970s were launched).
Not necessarily. x86 is just ABI compatibility. The original Pentiums were "from scratch" to some degree, they took ABi compatibility but impemented in a new way that had real performance benefits. The Pentium pro did as well, which doesn't run x86 code directly, just decodes them into RISC type micro-ops and scheduled those instead. The cores never really differed until NetBurst, which had some innovative ideas about caching micro-ops, but it turns out it's not as efficient as Pentium-Pro core (as seen in Pentium 3, pentium M) on some metrics. So it's obvious Pentium 4 core is bad, the question is can they hit one out of the park with a new core, or will the just extend the old pentium-pro core? we'll see.
The cross licensing deal was interesting. Led to a few thigns...
Originally Intel wasn't sure it could pump out enough 386's, so it licensed with AMD to pick up the slack. It saw how much money it was leaving on the table, so when the 486's came around, it tried freezing AMD out, including calling them i486, and tell AMD they can't use the i. Case went to court, you can't trademark a letter (Zilog tried, with Z). The next generation came out with a silly marketing name, Pentium.
Significantly now, it was a cross licensing deal, giving Intel hands on AMD products as well. That's how they can do EM64T, a copy of AMD64, and give AMD no royalties.
I'm not sure how much this "new chip" you want differs than Itanium, or PowerPC for that matter (though I believe PowerPC is much better engineered than Itanium).
Itanium, current marketing aside, was never intended to be a niche chip. It was supposed to be so radical that it replaced the x86 ecosystem with a new one (and happily for Intel, one much more under it's control than x86). Itanium execution mistakes aside, the success of AMD64/EM64T over ia64 shows that the market does want backwards compatibility.
IBM tried this with PowerPC. It was supposed to be a new RISC ecosystem to compete against and eventually replace x86. Some big names had a port - MacOS, WindowsNT, Solaris. Apple had too - the 68x00 line was running out of gas fast, and the transition was actually very smooth, though it still lost massive amounts of market share. Sun pulled back, probably caught with the same "why port to another chip if software is just a way to move our hardware" problem APple has faced. Windows NT never really sold, and running under emulation sucked. Standard chicken and egg problem there.
Even Linux, which because of being open source is relatively simpler too bootstrap a PowerPC system, is still predominantly x86. More people on x86 means more binaries targetting x86, more testing, meaning x86 is more polished, people faced with a choice pick x86.
No massive shift to new ecosystem.
There's a "10 hot comments" box you can enable for your home page. This wasn't on it though.
the hidden dupe code of the Slashcode...