Utter bullshit. Whether or not they *maintain* their dominant position through "monopolistic" practices, they got on top through purely legal means.
Doubtful, but we'll probably never know the truth. All I can say is that if the "Dos isn't done until Lotus won't run" stories are true, then that monopoly is totally ill-gotten. And the way they boxed out DR Dos by tying Windows to it, well they got caught, sued, and settled, so how again was that totally legally gained? And what about the *massive* astro-turffing campaign by MS against OS/2 4.0? That's anti-competitive, surely. And what they did to Grid. And Apple. And Stac. Central Point. The stories just go on and on. Most of those instances cost MS some money and some bad press, but guess what? They kept the monopoly, so to them, it was worth it.
I guess, though, that you've forgotten that the last anti-trust suit against Microsoft was brought in part due to MS' violation of two, count them two, previous Consent Decrees, each the result of Justice Department investigation. Originating back in the elder Bush adminisrtation.
And you're forgeting the other salient point: you could get the prevalent operating system (DOS) from someone other than IBM. Since Microsoft's deal with Big Blue was non-exclusive, they could, would, and did sell DOS to anyone and their brother. Thus Joe Blow in his dorm room could buy processors from Intel, motherboards with Phoenix Bios' to fire the thing up, running DOS (and thus Lotus 1-2-3, which was what really mattered), and assemble computers able to use the literally thousands of ISA cards that had been developed for use on IBM PC's. Wow, ain't non-exclusive deals and competition grand!
Never forget that MS was there and is partly responsible. And Joe Blow in his dorm room was really named Michael Dell.
Re:Postfix shortcomings
on
Postfix
·
· Score: 4, Informative
Postfix is both well documented, and well supported. From the well commented main.cf:
# The mailbox_command parameter specifies the optional external
# command to use instead of mailbox delivery. [some snipping] #mailbox_command =/some/where/procmail #mailbox_command =/some/where/procmail -a "$EXTENSION"
So not enabled by default, but easily remedied if you absolutely MUST have procmail. You can also enable it on a per-user basis by leaving those lines commented, and then using a.forward file in your home directory that calls procmail.
As for playing with spamassassin or other 3rd party programs, no problem. A quick check of the Documentation page at www.postfix.org reveals all kinds of good info. The consensus on postfix-users is to use amavisd-new, and then call antivirus and/or spam filters from there.
Now, answer this: the Navy has been able to knock down incoming anti-ship missiles for years now. The technology has gotten to the point where the chance of a missile impacting one of our ships is miniscule. How is that fundamentally different from shooting down an ICBM? Answer: it's not, it's only a question of scale.
Since/. loves analogies, here's one for you:
We have fabulous technology that allows us to keep people out of a given building, right? The vaults at Fort Knox, CIA Headquarters, the Whitehouse, whatever. But for the life of us we can't keep people out of the country. The borders, both land and sea, are porous. This is the difference between protecting a single ship, and protecting the continental United States. Yes it's just a matter of scale, but the orders of magnitude may take generations to overcome, and, personally, I don't think that it's a forgone conclusion that it will happen.
Incremental improvements are nothing. That program needs substantial, dramatic, improvements just to prove it's feasability, much less that it's achievable within out lifetimes.
Maybe he just got done with some bagels and loxen, sitting across from some old-school girls wearing bobby-soxen, chatting about all those poor people who died from the small poxen.
Then he asked, "if you're last name is Cox, do you refer to your family as 'Coxen'?"
On the other hand, anthropomorphizing computer boxes into the one 'o' 'x' word that ends with 'en' may mean he harbors a secret wish regarding oxen.
You did great! More time than I might have spent on it actually, but good info. It's been a few years! My only gripe has to with the 32 bit vs 16 bit thunking merely because something was running over DOS. Many, many applications of the time sidestepped DOS in order to get better performance, and thus didn't necessarily thunk to 16 bits for all operations. The win32s API is one example, and another is the Western Digital 32bit hard drive driver released at a similar time.
And yes, the 286 and 386 may have been close in performance, but my point remains: that the P4 was the first Intel processor to actually run slower, clock-for-clock.
As for who took the 286 to 25 Mhz, I bow to your resesarch; one of the links I had read made the claim that Intel did so, but they could have been total chuckleheads:). Maybe that makes me one too, for taking their word for it.
Maybe my mempry is fuzzy. OTOH, neither your link nor a quick Google search turn up anything substantive.
However, you're argument re. 386's running mostly DOS is questionable...Windows 3.0 could take partial advantage of the 386, and Win3.1 could, after a year, utilize Microsoft's brand spanking new win32s addition. And DOS 5.0 could access the greater memory allowed by the 386, so that's in the timeframe. Games sure as shit used all they could get, often making you use boot disks that provided an environment optimized for them (or at least not running anything else you might care to run).
Intel made 286's up to 25 Mhz, and 386's only up to 33 Mhz. And I'll stick by my recollection that the 386 was substantially faster out of the gate.
This may not be true in specific applications that support both SSE-2 and hyperthreading. Though I don't have a spcific head-to-head matchup, one particular program I might expect to see higher clock-for-clock performance on the P4 than the P3 is Cinebench 2003 [aceshardware.com].
Which is exactly what I said re. very specific, vertical market apps. I wasn't trying to diss the P4, so much as point out that it's a big break with how Intel went about processor design previously. It's not so much what's better, but that they're different.
You also forgot to mention that there was no "086".:)
When the 386 was introduced it was, clock-for-clock, twice as fast as the 286, no recompiling necessary. But it helped...the move to 32 bit code had begun. When the 486 was introduced, it smoked the 386, once again by being just about twice as fast as it's predecessor, again clock-cycle for clock-cycle. Without recompiling. But it helped, now that an fpu was standard (ignoring the crippled 486sx).
The Pentium was the first to really require recompiling to fully take advantage of the new architecture, but even without that it was not quite twice as fast as the 486. Intel claimed that an additional 30% or so gain (IIRC) could be had by recompiling, but no one really bothered. Why lose the legacy market? And remember that being optimized for the Pentium was originally one of Mandrake's reasons for being.
Well, then the Pentium Pro (later devolved to the Pentium 2) was introduced, and the increase broke stride a bit. Faster but maybe by only 1.5x or so. The P3 wasn't a new architecture so much as a new marketing exercise and a chance to unveil SSE, which provided some flashy performance increases in Photoshop, but that's about it. But that did require a recompile.
Keeping score? The Pentium 4 was the first new Intel architecture that didn't increase performance accross the board, clock-for-clock. Of course it's clock cycles faster, so it is faster, but no real drama in it. And you're claim that, "Then compilers and software in general adapted and it became faster" is baloney. So little software is specifically optimized for the P4 as to be negligible outside of specific vertical markets. And, even with optimized code, clock-for-clock, the P3 is still faster.
Whether revisionism or ignorance, you're wrong.
The parent post's comment about history repeated itself was regarding the intro of the P4, not any generation previous. Hence the criticism then and now of Intel for going for the marketing "Gee Whiz!" effect of maximum gigahertz, rather than actual superlative performance.
When it comes to interoperability between Windows and *nix, the answer is usually Samba. For you, you need Winbind, which will authenticate against a Windows Domain's PDC, and can be hooked into PAM.
One of these days, I'll rewire my house, but in the meantime, as I ponder what's involved, I've been wondering what you run all the coax to? Is there something equivalent to a patch panel, or even a switch, to collect coax?
r84x: So, I guess what he is saying is that radio is dying? People have listened to radio broadcasts of music for decades, and continue to, without being able to keep the music. What is different now?
What's stopping me from keeping the music? (Besides the lame-excuse-for-music that is broadcast these days.) Reel-to-reel tape, then cassettes, have allowed that possibility for years. That was the original reason for voice over introductions, and fade-out overlaps, etc., wasn't it? To taint the recordings that the masses were making. Once upon a time a few rogue stations broadcast entire albums, only to have the RIAA clench their iron fist.
My response to this is always, "Then why would I buy your widget instead of anyone else's?" I've never gotten a good answer; usually it's an embarrassed silence, as the other person realizes that they've talked themselves into a corner. About the best that they can do at that point is jump back into marketing mode and regurgitate their spiel.
But I totally applaud taking the time to tell the company that they've (already!) lost a sale, and why. Even though my (small) employer is an insignificant gnat in the face of the likes of Dell, I still took the time to tell them why we stopped purchasing from them, really hitting that some special one-time deal wouldn't make me stay a customer, as the problems would still remain.
Got any source re. the origin of chopsticks? I think you're full of it. Chopsticks are known to have been in use in China since at least 1000 BC. Hardly a gimmick seen 3000 years later.
For example, see this Japanese page which claims that chopsticks came to Japan from China in the 6th or 7th century, and that bronze chopsticks have been found in archealogical digs dating from just before Christ (the chopsticks, not the dig!).
This site goes back further, and includes a couple pictures:
I agree with your main idea, though I think that the last sentence is a big reach.
While it's true that linux dominated the CG render farms, don't write off it's presence on the content creation side of things too soon. Lucasarts announced a year ago that they were making a complete switch to linux on the desktop for content creation, and I think are about a third of the way into that. Disney is also moving that way. From what (little) I understand, access to code, and strong customisablility is REALLY improtatn to these folks, and as linux matures, the fit works well. Maya for linux was produced at the request of the cd industry, not because Alias thought it was "neat".
Raster files by definition are jsut a collection of dots, and so have no inherent meaning.
Vector files are collections of vectors, objects with properties. Those objects can have meaning to other programs, and it's a GOOD idea to use XML as a way to communicate what object types and properties are present. Ascii data, accessible by any number of already existent parsers. And the fact that XML is verbose, well who gives a shit. Ascii can be compressed, either as it's stored ala OpenOffice's file format, or transparently when sent across the wire.
What I did was to use Dawn to produce an initial.ldif file. Then I wrote a little script to massage entries based on the dn's and ou's that I wanted to use. Create said dn's and ou's via a seperate, hand crafted.ldif, a quick little exercise, bring in your massaged.ldif, and you're set.
To manage, phpLDAPadmin is the best tool I've found so far.
Outlook, Mozilla, etc. can all access as clients.
I also recommend LDAP System Administration by Gerald Carter, though with some reservations. It provides a decent grounding in LDAP, but won't be an end-all-be-all definitive resource.
Sorry. The tin-foil hat theory sounds good against The Man and especially against The Corporation, but in this case its looks like yet another urban legend. Check Snopes.
Pertinent quote from aforementioned Snopes article:
The change in sweetener wasn't anything that diabolical. Corn syrup was cheaper than cane sugar; that's what it came down to. In 1980 -- five years before the introduction of New Coke -- half the cane sugar in Coca-Cola had been replaced with high fructose corn syrup. By six months prior to New Coke's knocking the original Coca-Cola off the shelves, there was no cane sugar in American Coca-Cola. Whether they knew it or not, what consumers were drinking then was 100% sweetened by high fructose corn syrup.
Slashdot Mirror
Utter bullshit. Whether or not they *maintain* their dominant position through "monopolistic" practices, they got on top through purely legal means.
Doubtful, but we'll probably never know the truth. All I can say is that if the "Dos isn't done until Lotus won't run" stories are true, then that monopoly is totally ill-gotten. And the way they boxed out DR Dos by tying Windows to it, well they got caught, sued, and settled, so how again was that totally legally gained? And what about the *massive* astro-turffing campaign by MS against OS/2 4.0? That's anti-competitive, surely. And what they did to Grid. And Apple. And Stac. Central Point. The stories just go on and on. Most of those instances cost MS some money and some bad press, but guess what? They kept the monopoly, so to them, it was worth it.
I guess, though, that you've forgotten that the last anti-trust suit against Microsoft was brought in part due to MS' violation of two, count them two, previous Consent Decrees, each the result of Justice Department investigation. Originating back in the elder Bush adminisrtation.
And you're forgeting the other salient point: you could get the prevalent operating system (DOS) from someone other than IBM. Since Microsoft's deal with Big Blue was non-exclusive, they could, would, and did sell DOS to anyone and their brother. Thus Joe Blow in his dorm room could buy processors from Intel, motherboards with Phoenix Bios' to fire the thing up, running DOS (and thus Lotus 1-2-3, which was what really mattered), and assemble computers able to use the literally thousands of ISA cards that had been developed for use on IBM PC's. Wow, ain't non-exclusive deals and competition grand!
Never forget that MS was there and is partly responsible. And Joe Blow in his dorm room was really named Michael Dell.
4239 [enter] 12382 [enter] 147324 [+] [*] 2342 [+]
Anyone else would do:
12382 [+] 147324 [=] [*] 4239 [+] 2342 [=]
Since Pintos started torching their occupants.
So not enabled by default, but easily remedied if you absolutely MUST have procmail. You can also enable it on a per-user basis by leaving those lines commented, and then using a
As for playing with spamassassin or other 3rd party programs, no problem. A quick check of the Documentation page at www.postfix.org reveals all kinds of good info. The consensus on postfix-users is to use amavisd-new, and then call antivirus and/or spam filters from there.
Good luck!
Wow. So instead of Jail, you have MS Bob?
This is a great theme, actually! Free Parking would then become.....you guessed it! LINUX!!!!
Since
We have fabulous technology that allows us to keep people out of a given building, right? The vaults at Fort Knox, CIA Headquarters, the Whitehouse, whatever. But for the life of us we can't keep people out of the country. The borders, both land and sea, are porous. This is the difference between protecting a single ship, and protecting the continental United States. Yes it's just a matter of scale, but the orders of magnitude may take generations to overcome, and, personally, I don't think that it's a forgone conclusion that it will happen.
Incremental improvements are nothing. That program needs substantial, dramatic, improvements just to prove it's feasability, much less that it's achievable within out lifetimes.
showed the number of MegaBytes transferred as 1,000,400,143
That's a lot of copies of MyDoom!!!!!!
With apologies....I'll STFU now.
Maybe he just got done with some bagels and loxen, sitting across from some old-school girls wearing bobby-soxen, chatting about all those poor people who died from the small poxen.
Then he asked, "if you're last name is Cox, do you refer to your family as 'Coxen'?"
On the other hand, anthropomorphizing computer boxes into the one 'o' 'x' word that ends with 'en' may mean he harbors a secret wish regarding oxen.
"Jon, we will compensate you in remaindered copies of 'Little Nicky', 'Crossroads', and 'Ishtar' in DVD."
...sorry they're all Region 1, but we figured you'd be ok with that...
You did great! More time than I might have spent on it actually, but good info. It's been a few years! My only gripe has to with the 32 bit vs 16 bit thunking merely because something was running over DOS. Many, many applications of the time sidestepped DOS in order to get better performance, and thus didn't necessarily thunk to 16 bits for all operations. The win32s API is one example, and another is the Western Digital 32bit hard drive driver released at a similar time.
And yes, the 286 and 386 may have been close in performance, but my point remains: that the P4 was the first Intel processor to actually run slower, clock-for-clock.
As for who took the 286 to 25 Mhz, I bow to your resesarch; one of the links I had read made the claim that Intel did so, but they could have been total chuckleheads:). Maybe that makes me one too, for taking their word for it.
Maybe my mempry is fuzzy. OTOH, neither your link nor a quick Google search turn up anything substantive.
However, you're argument re. 386's running mostly DOS is questionable...Windows 3.0 could take partial advantage of the 386, and Win3.1 could, after a year, utilize Microsoft's brand spanking new win32s addition. And DOS 5.0 could access the greater memory allowed by the 386, so that's in the timeframe. Games sure as shit used all they could get, often making you use boot disks that provided an environment optimized for them (or at least not running anything else you might care to run).
Intel made 286's up to 25 Mhz, and 386's only up to 33 Mhz. And I'll stick by my recollection that the 386 was substantially faster out of the gate.
This may not be true in specific applications that support both SSE-2 and hyperthreading. Though I don't have a spcific head-to-head matchup, one particular program I might expect to see higher clock-for-clock performance on the P4 than the P3 is Cinebench 2003 [aceshardware.com].
:)
;)
Which is exactly what I said re. very specific, vertical market apps. I wasn't trying to diss the P4, so much as point out that it's a big break with how Intel went about processor design previously. It's not so much what's better, but that they're different.
You also forgot to mention that there was no "086".
And the 80186
You don't know you're history so well.
When the 386 was introduced it was, clock-for-clock, twice as fast as the 286, no recompiling necessary. But it helped...the move to 32 bit code had begun. When the 486 was introduced, it smoked the 386, once again by being just about twice as fast as it's predecessor, again clock-cycle for clock-cycle. Without recompiling. But it helped, now that an fpu was standard (ignoring the crippled 486sx).
The Pentium was the first to really require recompiling to fully take advantage of the new architecture, but even without that it was not quite twice as fast as the 486. Intel claimed that an additional 30% or so gain (IIRC) could be had by recompiling, but no one really bothered. Why lose the legacy market? And remember that being optimized for the Pentium was originally one of Mandrake's reasons for being.
Well, then the Pentium Pro (later devolved to the Pentium 2) was introduced, and the increase broke stride a bit. Faster but maybe by only 1.5x or so. The P3 wasn't a new architecture so much as a new marketing exercise and a chance to unveil SSE, which provided some flashy performance increases in Photoshop, but that's about it. But that did require a recompile.
Keeping score? The Pentium 4 was the first new Intel architecture that didn't increase performance accross the board, clock-for-clock. Of course it's clock cycles faster, so it is faster, but no real drama in it. And you're claim that, "Then compilers and software in general adapted and it became faster" is baloney. So little software is specifically optimized for the P4 as to be negligible outside of specific vertical markets. And, even with optimized code, clock-for-clock, the P3 is still faster.
Whether revisionism or ignorance, you're wrong.
The parent post's comment about history repeated itself was regarding the intro of the P4, not any generation previous. Hence the criticism then and now of Intel for going for the marketing "Gee Whiz!" effect of maximum gigahertz, rather than actual superlative performance.
Dot matrix!?!!?!!
Too much of a pussy to read punch cards???? Kids today! Hell in a hand basket, I tell ya!
When it comes to interoperability between Windows and *nix, the answer is usually Samba. For you, you need Winbind, which will authenticate against a Windows Domain's PDC, and can be hooked into PAM.
Browsing the docs is a very good idea. And, you can read The Official Samba-3 HOWTO and Reference Guide online. In particular, see Chapter 21. Winbind: Use of Domain Accounts.
Good luck.
One of these days, I'll rewire my house, but in the meantime, as I ponder what's involved, I've been wondering what you run all the coax to? Is there something equivalent to a patch panel, or even a switch, to collect coax?
r84x: So, I guess what he is saying is that radio is dying? People have listened to radio broadcasts of music for decades, and continue to, without being able to keep the music. What is different now?
What's stopping me from keeping the music? (Besides the lame-excuse-for-music that is broadcast these days.) Reel-to-reel tape, then cassettes, have allowed that possibility for years. That was the original reason for voice over introductions, and fade-out overlaps, etc., wasn't it? To taint the recordings that the masses were making. Once upon a time a few rogue stations broadcast entire albums, only to have the RIAA clench their iron fist.
So there goes your point...
"The CEO's excuse: everyones doing it."
My response to this is always, "Then why would I buy your widget instead of anyone else's?" I've never gotten a good answer; usually it's an embarrassed silence, as the other person realizes that they've talked themselves into a corner. About the best that they can do at that point is jump back into marketing mode and regurgitate their spiel.
But I totally applaud taking the time to tell the company that they've (already!) lost a sale, and why. Even though my (small) employer is an insignificant gnat in the face of the likes of Dell, I still took the time to tell them why we stopped purchasing from them, really hitting that some special one-time deal wouldn't make me stay a customer, as the problems would still remain.
Got any source re. the origin of chopsticks? I think you're full of it. Chopsticks are known to have been in use in China since at least 1000 BC. Hardly a gimmick seen 3000 years later.
For example, see this Japanese page which claims that chopsticks came to Japan from China in the 6th or 7th century, and that bronze chopsticks have been found in archealogical digs dating from just before Christ (the chopsticks, not the dig!).
This site goes back further, and includes a couple pictures:
I agree with your main idea, though I think that the last sentence is a big reach.
While it's true that linux dominated the CG render farms, don't write off it's presence on the content creation side of things too soon. Lucasarts announced a year ago that they were making a complete switch to linux on the desktop for content creation, and I think are about a third of the way into that. Disney is also moving that way. From what (little) I understand, access to code, and strong customisablility is REALLY improtatn to these folks, and as linux matures, the fit works well. Maya for linux was produced at the request of the cd industry, not because Alias thought it was "neat".
Apples:Oranges
Raster files by definition are jsut a collection of dots, and so have no inherent meaning.
Vector files are collections of vectors, objects with properties. Those objects can have meaning to other programs, and it's a GOOD idea to use XML as a way to communicate what object types and properties are present. Ascii data, accessible by any number of already existent parsers. And the fact that XML is verbose, well who gives a shit. Ascii can be compressed, either as it's stored ala OpenOffice's file format, or transparently when sent across the wire.
You know, there's this idea that geeks *never* use their testicles. I say baloney! They use them all the time!
Of course, no one ELSE does...
What I did was to use Dawn to produce an initial .ldif file. Then I wrote a little script to massage entries based on the dn's and ou's that I wanted to use. Create said dn's and ou's via a seperate, hand crafted .ldif, a quick little exercise, bring in your massaged .ldif, and you're set.
To manage, phpLDAPadmin is the best tool I've found so far.
Outlook, Mozilla, etc. can all access as clients.
I also recommend LDAP System Administration by Gerald Carter, though with some reservations. It provides a decent grounding in LDAP, but won't be an end-all-be-all definitive resource.
Pertinent quote from aforementioned Snopes article: