...software and video and music piracy is spiralling to ridiculous levels...
Do you mean the pirates in Washington, or do you mean the pirates in Hollywood? Note that the Disney oreganization started its life by stealing from Buster Keaton...
Suppose you put *two* dozen of them on a chip, and suppose they are *four* times faster. You still have less than a quarter the performance of a Conroe or POWER5 (both of which are dual-core, with each core sustaining more than 200 instructions in flight at a time), and you still have to manage that parallelism "by hand". Actually, the "four times faster" won't work, either -- remember that memory is still 200 times slower than Conroe or POWER5; if memory were 800 times slower than your processor, you'd really lose your performance!
This has been discussed ad nauseam in the computer architecture community, and I repeat: it's not a good idea!
The reason modern systems are so fast is that they hide a lot of fine grained parallelism behind the scenes. It is very hard to express this kind of parallelism in a way that it can be executed on a stack machine.
How important is this parallism? Consider that modern processors have 10-30 pipeline stages, 3-6 execution units that can have an instruction executing at each stage; moreover, most of them have out-of-order execution units that handle instructions more in the order that data is available for them rather than the order they are listed in the object file (and main memory is hundreds of times slower than the processors themselves, so this is important!). Typically, such processors can have more than 100 instructions in some stage of execution (more than 250 for IBM POWER5:-)
Consider, also, that the only pieces of anything-like-current stack hardware are Intel x87-style floating point units, that Intel is throwing away -- for good reason! -- in favor of (SSE) vector style units. In the current Intel processors, the vector unit emulates
an x87 if it needs to -- but giving only a quarter of the performance.
Someone made remarks about Java and.Net interpreters: in both cases, the interpreter is simulating a purely scalar machine with no fine grained parallelism; no wonder an extensible software-stack implementation is one of the simplest to implement. Stacks are not the way that true Java compilers like gjc generate code, though!
No, stack-based hardware is not a good idea. And haven't been since some time in the eighties, when processors started to be pipelined, and processor speed started outstripping memory speed.
...a program that *wants* click-to-top could easily get it by calling raise() on all clicks...
I HATE such uppity software.
acroread is a good example of a program that is damned hard to work with because it insists
on raising itself, obscuring everything else, any time the mouse passes over any part of its window.
I want software that obeys the instructions I have given it: only raise on titlebar-click; strict focus-follows-mouse. That's one of the reasons I dislike older netscape/mozilla/firefox (they disobey me, and insist on click-to-focus; seamonkey gets it right), and hate Gnome and M$, which insist on doing their own thing.
Message boards - unsupported web client
This feature requires a more recent version of Microsoft Internet Explorer or Mozilla Firefox. To download the latest version of Microsoft Internet Explorer, visit the Internet Explorer Web site.
Yeah, imagine a language that doesn't need braces because tabs already contain the same semantic (sic) information.
And another: make
Whether a makefile is correct depends upon hidden state: whether the rule-indentation is a tab or a sequence of spaces. It took Stu Feldman (the inventor of make) about three months to realize that this was a bad mistake, but at that point he had an active community of eight users, and didn't want to break backwards compatibility.
As any book lover knows, books in the plural lose their solidity of substance and become a gas, filling all available space... Henry Petroski, in The Book on the Bookshelf, recommends a ruthlessly imperialist approach to the problem: "Kitchen and pantry cabinets can be commandeered in the fight to find bookshelf space, and a family's eating habits can be changed. When the china is displaced by paper plates, there is no longer any reason why books cannot be stored in the dishwasher too."
My wife is an attorney, and she has to deal with documents
that repeatedly go through different versions of Word: at her clients, and at the other
side, and at the other side's attorneys. All
these different versions of Word
frequently corrupt documents so badly that Word throws
up its hands and says, "I can't deal with this.".
(Back and forth between '97 and 2000 or XP is particularly troublesome...)
And the fix is to run them through abiword and
save as rtf!
Kaspersky notes significant penetration of US military
networks (among others) by Sony spyware that was certainly
installed without authorization there, that opens security holes
in those systems, and that regularly "phones home."
(Snooping on how that latter behavior affects DNS servers
is how they made this discovery.)
That is clearly a violation of the National Security
parts of the Federal Computer Fraud and Abuse Act, which
calls for ten-year jail terms for those offenses (instead
of "only" five years for mere "federal interest computers.")
A bunch of Sony executives ought to be doing some hard jail
time! ...and in my opinion, the sentences ought
to be serial, rather than concurrent.
I'll wager you a Coke against a Pepsi that Mark Russinovich's computer was password-protected.
Sony deliberately and surreptitiously evaded that password
protection to invade and change settings on Mark's
computer. Tell me why he should not sue SONY for DMCA violation!
Veteran security expert Bruce Schneier confirmed reports that an anti-cheating tool called "The Warden," used by players of the popular network game World of Warcraft (WoW), collects information about all running processes in Windows, and reports back about those processes to the server of the game's publisher, Blizzard:
Open Source comes from a uniquely American idea now,
thanks to the Internet, exportable to the world at large:
voluntary community associations.
Alexis de Tocqueville (one of history's sharpest
observers of the American psyche) commented at length
upon voluntary community associations: when Americans
encounter a need in their communities they form(ed)
voluntary associations to deal with it rather than
saying "Let the government do it" and waiting forever
for an inadequate response. In this they are unlike
any other people on Earth: not the British, not the
French, Germans, Italians, or other Europeans, not
even like the Canadians.
The most obvious examples are the neighborhood
hospitals, fire-fighting associations and the
community schools that existed before the States
took them over.
I noted a more recent example on a recent trip to
Hawaii: that state refuses to recognize the village
of Volcano on the Big Island. The village funds its
fire department, its parks, and its community center
by monthly community cookouts! (I was there
for a marvelous Mongolian barbecue:-)
For that matter, what is a corporation but a voluntary
association for the purpose of doing business (and
generally--but not necessarily! --there are
not-for-profit corporations, particularly hospitals
and universities) making a profit.
And now the Internet has made possible globe-spanning
virtual communities that build Linux and a host of
other Open Source programs.
The fact that the numeric data test can be
expressed as a regular expression implies
obviousness (and that expression having
been described by a slashdot
reader within the first fifteen minutes
of posting); and
The fact that run-time (re-)configurable
highlighting has a long history (I point
to syntax highlighting in your favorite
programming editor; I know that at least
for nedit it can be turned on/off
by a click)
implies to me that this is a combination of obviousness
and prior art, hence should not be patentable.
Within a month after the Opteron release, computer
scientists at the Fraunhoffer Institute in Germany
had a "perl" script that would modify Intel-compiler
generated executables to bypass the "Genuine Intel"
test embedded in the executables, and use the optimized
code path.
Or has everyone forgotten Alpha? A totally kick ass CPU design killed by lack of native apps?
...and incompetent marketing and terrible management.
Consider that Apple wanted to go with the Alpha for
their Macs, and Digital wouldn't return their phone
calls. The PowerPC was definitely a "second choice"
for them. If there had been a major market for Alpha,
and if DEC^H^H^HCompaq^H^H^H^H^H^HHP had had competent
management, the computing world would be a very different
place today.
Slashdot Mirror
This has been discussed ad nauseam in the computer architecture community, and I repeat: it's not a good idea!
How important is this parallism? Consider that modern processors have 10-30 pipeline stages, 3-6 execution units that can have an instruction executing at each stage; moreover, most of them have out-of-order execution units that handle instructions more in the order that data is available for them rather than the order they are listed in the object file (and main memory is hundreds of times slower than the processors themselves, so this is important!). Typically, such processors can have more than 100 instructions in some stage of execution (more than 250 for IBM POWER5 :-)
Consider, also, that the only pieces of anything-like-current stack hardware are Intel x87-style floating point units, that Intel is throwing away -- for good reason! -- in favor of (SSE) vector style units. In the current Intel processors, the vector unit emulates an x87 if it needs to -- but giving only a quarter of the performance.
Someone made remarks about Java and .Net interpreters: in both cases, the interpreter is simulating a purely scalar machine with no fine grained parallelism; no wonder an extensible software-stack implementation is one of the simplest to implement. Stacks are not the way that true Java compilers like gjc generate code, though!
No, stack-based hardware is not a good idea. And haven't been since some time in the eighties, when processors started to be pipelined, and processor speed started outstripping memory speed.
acroread is a good example of a program that is damned hard to work with because it insists on raising itself, obscuring everything else, any time the mouse passes over any part of its window.
I want software that obeys the instructions I have given it: only raise on titlebar-click; strict focus-follows-mouse. That's one of the reasons I dislike older netscape/mozilla/firefox (they disobey me, and insist on click-to-focus; seamonkey gets it right), and hate Gnome and M$, which insist on doing their own thing.
Whether a makefile is correct depends upon hidden state: whether the rule-indentation is a tab or a sequence of spaces. It took Stu Feldman (the inventor of make) about three months to realize that this was a bad mistake, but at that point he had an active community of eight users, and didn't want to break backwards compatibility.
My wife is an attorney, and she has to deal with documents that repeatedly go through different versions of Word: at her clients, and at the other side, and at the other side's attorneys. All these different versions of Word frequently corrupt documents so badly that Word throws up its hands and says, "I can't deal with this.". (Back and forth between '97 and 2000 or XP is particularly troublesome...)
And the fix is to run them through abiword and save as rtf!
The peak of Word development occurred with Word 5 for the Mac, and it has been going downhill ever since. And feeping creaturitis is exactly why.
After all, it supports neither PostScript nor PDF.
That's what it sounds like to me!
So why should they not be prosecuted under the Computer Fraud and Abuse Act (US CODE TITLE 18 > PART I > CHAPTER 47 > 1030)? And why shouldn't a of their executives be in jail -- with ten-year terms instead of five, for invading national-security systems?
That is clearly a violation of the National Security parts of the Federal Computer Fraud and Abuse Act, which calls for ten-year jail terms for those offenses (instead of "only" five years for mere "federal interest computers.")
A bunch of Sony executives ought to be doing some hard jail time! ...and in my opinion, the sentences ought
to be serial, rather than concurrent.
I'll wager you a Coke against a Pepsi that Mark Russinovich's computer was password-protected. Sony deliberately and surreptitiously evaded that password protection to invade and change settings on Mark's computer. Tell me why he should not sue SONY for DMCA violation!
http://www.schneier.com/blog/archives/2005/10/bliz zard_entert.html
More commentary on Tom'sHardware: http://www.tgdaily.com/2005/10/24/world_of_warcraf t_warden_is_it_spyware/index.html
Alexis de Tocqueville (one of history's sharpest observers of the American psyche) commented at length upon voluntary community associations: when Americans encounter a need in their communities they form(ed) voluntary associations to deal with it rather than saying "Let the government do it" and waiting forever for an inadequate response. In this they are unlike any other people on Earth: not the British, not the French, Germans, Italians, or other Europeans, not even like the Canadians.
The most obvious examples are the neighborhood hospitals, fire-fighting associations and the community schools that existed before the States took them over.
I noted a more recent example on a recent trip to Hawaii: that state refuses to recognize the village of Volcano on the Big Island. The village funds its fire department, its parks, and its community center by monthly community cookouts! (I was there for a marvelous Mongolian barbecue :-)
For that matter, what is a corporation but a voluntary association for the purpose of doing business (and generally--but not necessarily! --there are not-for-profit corporations, particularly hospitals and universities) making a profit.
And now the Internet has made possible globe-spanning virtual communities that build Linux and a host of other Open Source programs.
- The fact that the numeric data test can be
expressed as a regular expression implies
obviousness (and that expression having
been described by a slashdot
reader within the first fifteen minutes
of posting); and
- The fact that run-time (re-)configurable
highlighting has a long history (I point
to syntax highlighting in your favorite
programming editor; I know that at least
for nedit it can be turned on/off
by a click)
implies to me that this is a combination of obviousness and prior art, hence should not be patentable.And I've been very happy with it.
...and you can disable this run-time test with the right PERL script...
Consider that Apple wanted to go with the Alpha for their Macs, and Digital wouldn't return their phone calls. The PowerPC was definitely a "second choice" for them. If there had been a major market for Alpha, and if DEC^H^H^HCompaq^H^H^H^H^H^HHP had had competent management, the computing world would be a very different place today.