If we need a mirroring protocol, let's write one and make it a usable part of ftp/http/whatever. Inserting TXT records would be cumbersome, error-prone, and ultimately a massive mess.
Drop a note to the IETF, get yourself a proposed RFC number, and go for it.
Recognizing that some of these folks may have been around for a long time before I got around to noticing them, here (in no particular order) are my favorite newer writers.
Jeff Noon: Fiendishly clever, hard to find in USA. Read The Automated Alice and Vurt to see his breadth.
Iain (M.) Banks: Love the Culture, love his non-SF too. Brilliant eye for the human condition. The Player of Games and Use of Weapons blew me away.
Neal Stephenson: He's barely within my definition of new, but you all know who he is. He needs more discipline and/or a better editor if he is to last.
Sarah Zettel: New, good, and getting better. Try her novel Fool's War.
I hate to say this, but at least half of the folks you mention (Brin, Bear, Benford and Vinge) have all been writing popular SF for 20+ years. Vinge has always had very low output, but writing wasn't his real job until recently; Benford was winning awards in the early '70s; Bear published his first novel in 1979; Brin won the Hugo for novel with his 1982 novel "Startide Rising".
They're all great authors and have prominent places on eschasi's shelf, but they're almost all older than eschasi -- and he's old.
Someone should mod this guy up. The review was pretty much worthless as a comparison on SATA to ATA. Different manufacturer, different drive characteristics, no worthwhile description of how the drives were prepared. And the author is somehow surprised that the results were not what he expected? Duh!
People DO record albums in 60 minutes.. they call them 'Live" recordings. But most likely there will be mistakes made. and sound bad.
Having been an invited audience member for an artist who was doing a live recording, I must partially disagree. The artist put a great deal of additional work into the live show so that the recording quality would sufficiently good for use in the final disk.
This included not just setups, but working with the audience so that we'd be less likely to hose up the result.
That's one of the reasons a friendly audience was recruited.
It appears most of the responders (and the guy who put the headline on this/. article) didn't actually read the article.
The article says there is
A call to sign off on explicit rejection of "licenses that would prevent or discourage commercial adoption of promising cyber security technologies developed through federal R & D." has been issued by Adam Smith, Congressman for the Ninth District in the State of Washington.
This is pretty different than "Congress Members Oppose GPL for Govt. Research.'
It's much narrower, and the total number of congress members involved is 2.
That's 2, as in 2 out of 635.
And it's to be applied to the security software only.
The headline is much too broad, and therefore misleading.
An aside: yes, I know good headlines are short and should get the viewers attention.
But "Congressman Opposes GPL for Govt Security Software" is the same length and considerably more accurate.
End aside.
And it's a suggestion that licences be banned only if they "prevent or discourage commercial adoption" of the technologies.
Given the way most corporations have shied away from GNU licences, I think you can easily make the case that in practice the GPL discourages commercial applications.
There is one primary exception to that - standalone programs or systems.
Note, for example, that Linux and GNU emacs are wildly popular, but the various FSF C function libraries were not. The GNU library
licence was written because people were shying away from developing with GCC because FSF libc.a was required for gcc usage (I don't think that's true any more). Libc.a
was under GPL and that meant applications that were developed with gcc would come under GPL.
FSF created the library licence in an attempt to address the issue, but lately they seem to think that it was a mistake.
IMHO, they're confusing cause with effect.
Those libraries came into wider usage because the GPL didn't apply to software developed that used them, not because they were good libraries (though they are good libraries).
But IMHO if they weren't under the library licence, they would not have come into as common a use as you now see.
Let us also note that releasing the code to the public domain does not prevent applying the GPL to it by others! You can grab a copy, hack it up to your hearts content, slap the GPL on it, and go.
If your mods make it superior to the unrestricted original and the
public thinks the GPL restrictions aren't a problem, cool.
If not, well, the market has spoken.
IMHO, this proposal will simply prevent the GPL from being applied before the market has spoken.
I've owned Beta-II, Beta-III, VHS and SVHS decks. SVHS and Beta-III were never used for retail pre-recorded tapes, so what you could buy to watch was VHS and Beta-II. They were comparable in visual quality, though I'd give a nod to the Beta.
But when it came to recording your own, our Beta-III deck produced tapes which were indistinguishable from orginal broadcast. We owned a Beta-II/III and a VHS/SVHS deck. When
our friends asked why, we'd do show them a head-to-head comparison, Beta-II/VHS or Beta-III/SVHS. The betas won every time.
Is vastly the right word? Yeah, I think so. Look at it this way: my seat-of-the-pants estimate was that home VHS recording was 90 to 95% of broadcast quality, home Beta was 98 to 99%. That's only 3% or so. But it's vastly
closer to 100% than VHS or SVHS ever got.
I'm sooo glad the Tivos/etc are finally bringing in a next gen technology.
Last night just before heading for bed, it occurred to me that I'd not checked to see the latest version of Mozilla in a while. When the website said '1.1', I just assumed I'd been out of touch. Instead, it seems I'd gotten there just before the/. hit. *chortle* Lucky moi.
Re:PS/2 ports... WHY??
on
USB KVMs Compared
·
· Score: 4, Informative
Simple answer:
It was cheap when cheap mattered, and it
replaced a connection without changing the way it worked. And for the record, the old connections were not RS-232 ports, so the question is a bit off the mark. Yes, they had some RS-232 components, but they weren't full-featured.
If memory serves, IBM invented this cable format for the PS/2.
(You do remember IBM had a computer called the PS/2, right?)
This particular connector might have been adapted from something else (strong resemblance to a Mac kbd/mouse connector, as I recall),
but the big public play was with the PS/2.
It was smaller, cheaper, etc, than the AT keyboard connector, and therefore a win all around. IBM then quickly adapted it for use with their `standard' PCs, and everybody followed.
(Again, this was back in the days when IBM did something and everybody followed.)
It was (and is) signal-compatible with the old keyboard, and I've still got a few of those ATPS2 keyboard adapters around somewhere
in the dungeon.
At the time the AT-style keyboard connector was
put on PCs (early 1980s), nobody ever dreamed we'd be having roomfuls of these things all cabled up to a single monitor/keyboard/mouse.
The computers were too expensive for a person to own more than one, and not powerful or reliable enough to put into racks and stacks to provide specialized services.
The kind of switching talked about here just wasn't envisioned, and in fact, would have been considered insane. Putting more expense into the component so it could be hot-swapped (and making the appropriate O/S changes) was simply not worth doing.
But eventually technology catches up, and the
marriage of USB with keyboard/mouse permits
the right thing to happen.
One could argue that you ought to simply be
able to plug/unplug the keyboard from USB host
to USB host, and in fact, you can do that with some of them.
But that still doesn't solve your monitor connection problems, and last time I checked,
SVGA didn't work over USB.
So a single switch that moves 3 components simultaneously is a big win.
I looked over the article posted at the top,
and am pretty impressed.
USB KVM switches are a great idea, and after reading over the above articles, I want one.
In fact, I want both of 'em, because each has
some very cool features.
Hmmm... so I guess I'll wait until the next generation comes out, when each company adds features from the other.:-)
What percentage of the general public buys music CDs?
I bet it's significantly less than 90%.
Combine that with 90% of the downloaders buying CDs, and you can make a case that downloaders are more likely to buy CDs than the general populace.
Now, admittedly that's a bogus arguement.
Almost anyone who is downloading MP3s is doing so because they're a music fan,
and therefore is not representative of the US as a whole.
But it sure sounded good for a second, didn't it?
"There are lies, damned lies, and statistics."
-- Benjamin Disraeli.
And for instructions on how to do it,
see this. -- "97.45% of all statistics are made up." - me
Judas priest, what a bunch of whining wankers. If they'd put both cuts and all the material in a single six-disk set, you'd whine because you had to buy it all at once. If they packaged it in six different boxes, you'd whine because they'd be more expensive than a set as a whole.
Right now, Amazon is offering the first one at $17.97, the second at $25.99. So you can buy either one at a very reasonable price, or all six disks for $44.00.
Let me repeat that for all of you whiners who didn't understand it:
You can buy all six disks, including two full cuts, for only $44.00.
Damn, do I feel ripped off. Especially when the six-disk set of 'The Godfather', "remastered" for the umpteenth time, is $75. Or the single, no-specials, no-restoration DVS of 'Harold and Maude' is $25.50.
Get a clue. Jackson et. al. had to make a decision as to how to package it. They made their decision. As far as I can tell, they decided that they would not force folks to buy duplicate material if they wanted everything, and kept it all *very very cheap*. Looks to me like a damned fine choice.
Using a classic bit of social engineering and a photograph donated by a mutual, er, friend, we modified a directors web page at UUNET. If you click on just the right letter, it takes you to a photograph other than the one you would expect. I checked a few minutes ago, and it's still there....
This has been Suns semi-official line for a long, long time. See this netnews posting from 1989 for an example.
And as for licences staying with (or not with) the hardware, well, you can't have it both ways.
When I buy a copy of W2K, I put it on whatever machine I want to -- provided I only have it installed on one machine at a time. When I buy a copy of Solaris, I'm currently stuck to a given piece of hardware since Sun won't sell me new hardware to go with my old licence.
I think the days for these kind of limited handhelds are numbered. . . I would rather carry around a mini-cell phone and a lite-laptop . ..
I submit that you're a minority.
Ultimately I think we'll have voice-recognition technologies in our handhelds and some sort of
on-demand backwards connection to our `laptops'.
Both laptop and PDA/phone/etc will be `mere' peripherals to whatever our central processing and storgage system is. Yes, laptops will get smaller and smaller, but they're forever going to be constrained by screen size and (to a lesser extent) power. I don't think we'll ever see a laptop with a 17" screen that fits in your pocket.
Unless you've got hellacious pockets, of course.
Peripherals, unlike central systems, map well to to specialized uses. My PP just died, and I'm seriously considering the Treo. It means that I'd carry one peripheral rather than three (pager, phone, PDA). For me (and many others) that's a big win.
With any luck, the Treo will last three years, which is all I expect from a computer anyway. Even if you happen to be right and fully integrated laptops take over, it isn't gonna happen in the next three years. Thus again, the Treo is the right answer even for the mid-term.
In the short term, it means better integration between phone lists, email, notes, etc. And to me, that's more than worth the three year cost per year. At the end of three years . . . well, we'll see what's out there.
If this were happening to me, I'd change my netnews.sigfile to read
--
If the Organization line on this post says 'RoadRunner', then the opinions expressed here are the official opinions of 'RoadRunner'. They put their name on them, they must approve.
Want to see a living example of the difference between North and South Korea? Find South Korea on the nighttime map from NASA. Notice how it's separated from China by a dark gap? That gap is North Korea.
Gads, how ironic. Consider this prior statement from Gates:
"There are no significant bugs in our released software that any
significant number of users want fixed."
-- Bill Gates, 1995 interview in Focus Magazine (Germany)
Like he did with the Internet, Gates just ignored bugs for as long as possible.
Once they finally became a threat, he suddenly cares. I fully expect him to act against bugs with the same concern for interoperation, user support, and openness that he has with all Microsoft products which use the internet.
Our high-tech societies (particulary the US and Europe) are based on
the assumption that people will not willfully disrupt them. Power lines are
in the open, police give traffic tickets assuming that the offender will
show up and pay them, and so forth.
All of this is based on the assumption that ordinary citizens, criminals, and even foreign
adversaries are
worthy of a great deal of trust.
Our technological systems are based on the assumption of trust.
When you can no longer assume it,
things which were formerly technically or financially feasible
become infeasible.
The ultimate breakdown of this trust gives us societies which
bomb themselves back to the stone age.
Various countries in a state of permanent civil war are good
examples.
Bridges are built (expensive!) on the assumption that neither
the good guys nor the bad guys will blow them up because
that damage hurts society as a whole more than the benefit to either side.
Once that implicit trust fails,
neither side bothers rebuilding bridges.
The end result is a death spiral into the stone age, or a total victory by one side or the other,
or a recognition that both sides must act with trust.
Now we face enemies which cannot be trusted.
Some of them feel that this is the right way to fight, others think a stone-age end result is what should happen.
As a result, the old trust-based systems will have to be changed.
This risk to systems is going to change both our
systems and our society.
What change will depend on the difficulty of protection, the value of the system, and the cost if it fails.
Some systems can simply be made more robust
(cf Israels policies on El Al flights).
Some will be made more resilient in the face of intermittent failure
(cf the way so many companies no longer trust that electric companies will provide uninterrupted power).
Some will be abandoned as simply infeasible.
Pulling and pushing those decisions will be the societal urge to enforce trust (which, ultimately, isn't trust at all) via strong identity and continual authorization and monitoring.
No matter where it goes, it's going to be a different world.
-- eschasi, who's been mulling this over for far too many years
I certianly stand corrected on Intel vs. i86 clones, but still believe (seat of the pants) that 95% of the hosts running Linux are i86 or x86-clone based. If you've got numbers to contradict my opinion, I'll happily change it based on facts. Thus far your facts don't stand up. You say intel and amd are about neck and neck in sales, but according to the press releases from last week AMD now hit 20% of market share (up from 16%). That sure isn't neck and neck.
My other point still stands -- show people who care about performance a 45% increase in thruput, and they'll respond. Yes, the compiler is not free in the sense often used here. When you've got servers driven to the wall, buying the next increment of performance costs big bucks. If you can get another 10% throughput with a $500 compiler rather than hardware upgrade, it's a win. Especially when you can dist those compiled objects to all your hosts.
As for me, Joe Home User, I'm pretty interested in what the free (as in beer) version will do for me.
...if that's your normal installation procedure, why do you care if you have source at all?
For him, yeah, that example shows nothing about the utility of having open source. But my normal installation procedure (and the one he probably actually uses) is a bit different:
%./configureswitches&& make
. . . test . . .
% su # make install
Should any of those step fail and should I determine that the problem lies in the package,
I can fix it.
Aside from that, yeah, most folks just grab the RPMs and don't actually give a shit so long as they don't have to shell out dollars.
I'm sorry, but that's just plain silly. 90-95% of the freeware systems (Linux, *BSD, BEOS) run on Intel-based systems. If the Intel compiler is as good as claimed and is reasonably priced, every single significant *IX-based web server is going to be cut over to it in nothing flat.
Sure, GCC won't die. But if these numbers hold up, it's going to be second banana on Intel-based systems for a long time to come.
I've got news for you -- rms is a control freak and a fanatic. And the GPL is one of the most controlling licences in the world. It's not a licence written by someone who wants freedom, it's a licence written by someone who wants to ram his definition of freedom down peoples throats.
But aside from that I have no strong opinions on the GPL.:-)
If folks here are correct in implying he's running for the board so he can shut off discussion of non-GPL software (IMHO non-GPL != non-free), well, that's the actions of both a fanatic and a control freak. If the GNOME developers want to mention non-free software, they've already shown by their actions that's what they want to do. If someone thinks that discussion is a bad thing, then the right thing to do is to get in there as an equal and try to convince them.
Running for the board so you can override others' will is, IMHO, a violation of freedom. But thus far all I've heard is rumor of his reasons. Sourceforge reports it as "speculation" (tho they offer some reasons as to why that speculation is being made). Until I hear a statement from Stallman as to his reasons, I'll withhold judgement on this particular issue.
I read a friends borrowed copy of this
when I was about 40, and loved it. My kids (9 and 11) were alternately mystified and amused. I was continually amused. Unfortunately the copy belonged to a friend who was just having kids of his own, and the cruel bastard wanted it back.
At last, a copy of my own - and the kids are sophisticated enough mathematically to get it all. Maybe I'll get three, one for me and one for each of them.
Slashdot Mirror
Drop a note to the IETF, get yourself a proposed RFC number, and go for it.
Jeff Noon: Fiendishly clever, hard to find in USA. Read The Automated Alice and Vurt to see his breadth.
Iain (M.) Banks: Love the Culture, love his non-SF too. Brilliant eye for the human condition. The Player of Games and Use of Weapons blew me away.
Neal Stephenson: He's barely within my definition of new, but you all know who he is. He needs more discipline and/or a better editor if he is to last.
Sarah Zettel: New, good, and getting better. Try her novel Fool's War.
And that's about it. Sorry....
They're all great authors and have prominent places on eschasi's shelf, but they're almost all older than eschasi -- and he's old.
Lies, damned lies, statistics, and benchmarks.
The article says there is
This is pretty different than "Congress Members Oppose GPL for Govt. Research.' It's much narrower, and the total number of congress members involved is 2. That's 2, as in 2 out of 635. And it's to be applied to the security software only. The headline is much too broad, and therefore misleading. And it's a suggestion that licences be banned only if they "prevent or discourage commercial adoption" of the technologies. Given the way most corporations have shied away from GNU licences, I think you can easily make the case that in practice the GPL discourages commercial applications.There is one primary exception to that - standalone programs or systems. Note, for example, that Linux and GNU emacs are wildly popular, but the various FSF C function libraries were not. The GNU library licence was written because people were shying away from developing with GCC because FSF libc.a was required for gcc usage (I don't think that's true any more). Libc.a was under GPL and that meant applications that were developed with gcc would come under GPL. FSF created the library licence in an attempt to address the issue, but lately they seem to think that it was a mistake. IMHO, they're confusing cause with effect. Those libraries came into wider usage because the GPL didn't apply to software developed that used them, not because they were good libraries (though they are good libraries). But IMHO if they weren't under the library licence, they would not have come into as common a use as you now see.
Let us also note that releasing the code to the public domain does not prevent applying the GPL to it by others! You can grab a copy, hack it up to your hearts content, slap the GPL on it, and go. If your mods make it superior to the unrestricted original and the public thinks the GPL restrictions aren't a problem, cool. If not, well, the market has spoken. IMHO, this proposal will simply prevent the GPL from being applied before the market has spoken.
Feh, enough of that, I'm ranting.
But when it came to recording your own, our Beta-III deck produced tapes which were indistinguishable from orginal broadcast. We owned a Beta-II/III and a VHS/SVHS deck. When our friends asked why, we'd do show them a head-to-head comparison, Beta-II/VHS or Beta-III/SVHS. The betas won every time.
Is vastly the right word? Yeah, I think so. Look at it this way: my seat-of-the-pants estimate was that home VHS recording was 90 to 95% of broadcast quality, home Beta was 98 to 99%. That's only 3% or so. But it's vastly closer to 100% than VHS or SVHS ever got.
I'm sooo glad the Tivos/etc are finally bringing in a next gen technology.
Last night just before heading for bed, it occurred to me that I'd not checked to see the latest version of Mozilla in a while. When the website said '1.1', I just assumed I'd been out of touch. Instead, it seems I'd gotten there just before the /. hit. *chortle* Lucky moi.
It was cheap when cheap mattered, and it replaced a connection without changing the way it worked. And for the record, the old connections were not RS-232 ports, so the question is a bit off the mark. Yes, they had some RS-232 components, but they weren't full-featured.
If memory serves, IBM invented this cable format for the PS/2. (You do remember IBM had a computer called the PS/2, right?) This particular connector might have been adapted from something else (strong resemblance to a Mac kbd/mouse connector, as I recall), but the big public play was with the PS/2. It was smaller, cheaper, etc, than the AT keyboard connector, and therefore a win all around. IBM then quickly adapted it for use with their `standard' PCs, and everybody followed. (Again, this was back in the days when IBM did something and everybody followed.) It was (and is) signal-compatible with the old keyboard, and I've still got a few of those ATPS2 keyboard adapters around somewhere in the dungeon.
At the time the AT-style keyboard connector was put on PCs (early 1980s), nobody ever dreamed we'd be having roomfuls of these things all cabled up to a single monitor/keyboard/mouse. The computers were too expensive for a person to own more than one, and not powerful or reliable enough to put into racks and stacks to provide specialized services. The kind of switching talked about here just wasn't envisioned, and in fact, would have been considered insane. Putting more expense into the component so it could be hot-swapped (and making the appropriate O/S changes) was simply not worth doing.
But eventually technology catches up, and the marriage of USB with keyboard/mouse permits the right thing to happen. One could argue that you ought to simply be able to plug/unplug the keyboard from USB host to USB host, and in fact, you can do that with some of them. But that still doesn't solve your monitor connection problems, and last time I checked, SVGA didn't work over USB. So a single switch that moves 3 components simultaneously is a big win.
I looked over the article posted at the top, and am pretty impressed. USB KVM switches are a great idea, and after reading over the above articles, I want one. In fact, I want both of 'em, because each has some very cool features.
Hmmm... so I guess I'll wait until the next generation comes out, when each company adds features from the other. :-)
What percentage of the general public buys music CDs? I bet it's significantly less than 90%. Combine that with 90% of the downloaders buying CDs, and you can make a case that downloaders are more likely to buy CDs than the general populace.
Now, admittedly that's a bogus arguement. Almost anyone who is downloading MP3s is doing so because they're a music fan, and therefore is not representative of the US as a whole. But it sure sounded good for a second, didn't it?
"There are lies, damned lies, and statistics." -- Benjamin Disraeli.
And for instructions on how to do it, see this.
--
"97.45% of all statistics are made up." - me
Right now, Amazon is offering the first one at $17.97, the second at $25.99. So you can buy either one at a very reasonable price, or all six disks for $44.00.
Let me repeat that for all of you whiners who didn't understand it:
You can buy all six disks, including two full cuts, for only $44.00.
Damn, do I feel ripped off. Especially when the six-disk set of 'The Godfather', "remastered" for the umpteenth time, is $75. Or the single, no-specials, no-restoration DVS of 'Harold and Maude' is $25.50.
Get a clue. Jackson et. al. had to make a decision as to how to package it. They made their decision. As far as I can tell, they decided that they would not force folks to buy duplicate material if they wanted everything, and kept it all *very very cheap*. Looks to me like a damned fine choice.
And if you don't like it, don't buy it.
Easy. Try
Using a classic bit of social engineering and a photograph donated by a mutual, er, friend, we modified a directors web page at UUNET. If you click on just the right letter, it takes you to a photograph other than the one you would expect. I checked a few minutes ago, and it's still there....
And as for licences staying with (or not with) the hardware, well, you can't have it both ways. When I buy a copy of W2K, I put it on whatever machine I want to -- provided I only have it installed on one machine at a time. When I buy a copy of Solaris, I'm currently stuck to a given piece of hardware since Sun won't sell me new hardware to go with my old licence.
I submit that you're a minority. Ultimately I think we'll have voice-recognition technologies in our handhelds and some sort of on-demand backwards connection to our `laptops'. Both laptop and PDA/phone/etc will be `mere' peripherals to whatever our central processing and storgage system is. Yes, laptops will get smaller and smaller, but they're forever going to be constrained by screen size and (to a lesser extent) power. I don't think we'll ever see a laptop with a 17" screen that fits in your pocket. Unless you've got hellacious pockets, of course.
Peripherals, unlike central systems, map well to to specialized uses. My PP just died, and I'm seriously considering the Treo. It means that I'd carry one peripheral rather than three (pager, phone, PDA). For me (and many others) that's a big win.
With any luck, the Treo will last three years, which is all I expect from a computer anyway. Even if you happen to be right and fully integrated laptops take over, it isn't gonna happen in the next three years. Thus again, the Treo is the right answer even for the mid-term.
In the short term, it means better integration between phone lists, email, notes, etc. And to me, that's more than worth the three year cost per year. At the end of three years . . . well, we'll see what's out there.
Step back, boys, eschasi's goin' shopping!
If this were happening to me, I'd change my netnews .sigfile to read
--
If the Organization line on this post says 'RoadRunner', then the opinions expressed here are the official opinions of 'RoadRunner'. They put their name on them, they must approve.
Want to see a living example of the difference between North and South Korea? Find South Korea on the nighttime map from NASA. Notice how it's separated from China by a dark gap? That gap is North Korea.
The rest of the article is just as damning.
Like he did with the Internet, Gates just ignored bugs for as long as possible. Once they finally became a threat, he suddenly cares. I fully expect him to act against bugs with the same concern for interoperation, user support, and openness that he has with all Microsoft products which use the internet.
Our technological systems are based on the assumption of trust. When you can no longer assume it, things which were formerly technically or financially feasible become infeasible. The ultimate breakdown of this trust gives us societies which bomb themselves back to the stone age. Various countries in a state of permanent civil war are good examples. Bridges are built (expensive!) on the assumption that neither the good guys nor the bad guys will blow them up because that damage hurts society as a whole more than the benefit to either side. Once that implicit trust fails, neither side bothers rebuilding bridges. The end result is a death spiral into the stone age, or a total victory by one side or the other, or a recognition that both sides must act with trust.
Now we face enemies which cannot be trusted. Some of them feel that this is the right way to fight, others think a stone-age end result is what should happen. As a result, the old trust-based systems will have to be changed.
This risk to systems is going to change both our systems and our society. What change will depend on the difficulty of protection, the value of the system, and the cost if it fails. Some systems can simply be made more robust (cf Israels policies on El Al flights). Some will be made more resilient in the face of intermittent failure (cf the way so many companies no longer trust that electric companies will provide uninterrupted power). Some will be abandoned as simply infeasible.
Pulling and pushing those decisions will be the societal urge to enforce trust (which, ultimately, isn't trust at all) via strong identity and continual authorization and monitoring.
No matter where it goes, it's going to be a different world.
-- eschasi, who's been mulling this over for far too many years
My other point still stands -- show people who care about performance a 45% increase in thruput, and they'll respond. Yes, the compiler is not free in the sense often used here. When you've got servers driven to the wall, buying the next increment of performance costs big bucks. If you can get another 10% throughput with a $500 compiler rather than hardware upgrade, it's a win. Especially when you can dist those compiled objects to all your hosts.
As for me, Joe Home User, I'm pretty interested in what the free (as in beer) version will do for me.
For him, yeah, that example shows nothing about the utility of having open source. But my normal installation procedure (and the one he probably actually uses) is a bit different:
% ./configure switches && make
. . . test . . .
% su
# make install
Should any of those step fail and should I determine that the problem lies in the package, I can fix it.
Aside from that, yeah, most folks just grab the RPMs and don't actually give a shit so long as they don't have to shell out dollars.
I'm sorry, but that's just plain silly. 90-95% of the freeware systems (Linux, *BSD, BEOS) run on Intel-based systems. If the Intel compiler is as good as claimed and is reasonably priced, every single significant *IX-based web server is going to be cut over to it in nothing flat. Sure, GCC won't die. But if these numbers hold up, it's going to be second banana on Intel-based systems for a long time to come.
RonB seems to be the first guy who understood the question correctly. Congrats. :-)
But aside from that I have no strong opinions on the GPL.:-)
If folks here are correct in implying he's running for the board so he can shut off discussion of non-GPL software (IMHO non-GPL != non-free), well, that's the actions of both a fanatic and a control freak. If the GNOME developers want to mention non-free software, they've already shown by their actions that's what they want to do. If someone thinks that discussion is a bad thing, then the right thing to do is to get in there as an equal and try to convince them.
Running for the board so you can override others' will is, IMHO, a violation of freedom. But thus far all I've heard is rumor of his reasons. Sourceforge reports it as "speculation" (tho they offer some reasons as to why that speculation is being made). Until I hear a statement from Stallman as to his reasons, I'll withhold judgement on this particular issue.
At last, a copy of my own - and the kids are sophisticated enough mathematically to get it all. Maybe I'll get three, one for me and one for each of them.