Actually, the employees typically DON'T use the devices.
I have engineered features for a set-top & tv box -- and I don't have (probably never will have) that tv.
I have worked for computer companies whilst never owning ANY of their product.
I have just done some engineering work for a printer company, and while I *have* in the past owned the vendor product, I will never own this particular product (and, indeed have never SEEN the product).
I have worked with a major graphics board company, and, though I do own several of their products, I was never given one to "home test".
In other words, the engineers put in the features, but we DON'T actually "eat the dog food". That job is left to Product Managers who probably don't care, and Marketing who probably doesn't either (make sure it meets the requirements).
So, if a "one-hour battery life" was in the requirements (or worse, no mention of battery life at all), that's what gets delivered.
And the justification? The employees/contractors won't BUY the stuff (why would we?); the company feels it is too expensive to build extra prototypes -- and besides, what does the employee know anyway? Stick to engineering; that's what we pay you for.
First, multiple video streams can be interleaved into a single mpeg transport stream. Each of the streams has its own 33 bit decode/present timestamps, and you can (for this application), ensure that they are synchronized.
Next, the video output can be gen-locked. One free-running video, and feedback from the DAC to keep the other two video outputs synchronized. There are some chip solutions that can do this, and several board solutions as well. To make this viable, you probably want a chip-level solution.
Now, the client currently uses 3 dvds, synchronized with a remote? That's very, um, coarse, so gen-lock is probably not even needed -- just the DTC/PTC should do very nicely (and MUCH more accurately). Of course, we assume a "single" (stereo/5.1/whatever) sound stream in the mpeg.
For a board level solution, look at 3xXilleon, or something like that.
I would like to pursue this issue. Your claim is that OpenOffice.org doesn't reliably convert Microsoft Word documents.
I do contract work. So far, OO.o has opened every doc file I have come across, with one exception.
It was the technical programming (register) guide for the ATI Xilleon.
Over 1,000 pages, and formatted in tables. Then again, all tools in Redhat 9 refused to have anything sensible to do with the document.
But, so far, that has been the only problem case. And I have put a LOT of stuff into OO.o. For example -- my current gig involves a 10MB Excel spreadsheet. And yes, OO.o handles it, and the attendant doc files associated with the project.
Of course, most of your documents may be these very long , convoluted things -- in which case you have my sympathies. But, that kind of data shouldn't be locked away like that. At least the ATI chip reference was available in HTML as well (thank $DEITY.
If these are short documents that don't work, email me one of them. (you'll have to "unravel" my email address from the header).
As to "complaining oabout saving files in doc format"... I agree with you on that one. It is actually complaining about NOT saving in native format, and I find that bizarre.
GUIs are playing catch-up again. "locate" has been around on Unix platforms for, what, 20 or 30 years? And, it does wildcard searches for filenames, ready to jam into grep for content search.
Content could be indexed, but its a bit project specific (so us Unix heads only do it on specific projects, right?).
For the un-initiated, a process runs (typically once a day), and indexes all filenames on your system. You can then get instant answers to "Show me all Microsoft Word documents on my system"
file `locate *.doc` | grep Microsoft
and many other queries. This stuff is PLAIN ORDINARY UNIX/LINUX. Ah well, doesn't help the completely casual user. You know, "If the option isn't clearly presented, it can't possibly be done -- or I just don't want to bother".
More power to 'em, them -- but people PLEASE don't ask when this will be ported to Linux/Unix!
NT was not ported to big-endian machines. The machines it ran on were either little endian or bi-endian.
To make a reasonable transition, you would want big-endian support for legacy 68k, and legacy mac-os applications. But if the kernel runs little-endian? Its a problem.
Anyway, I believe that was the reason that there was never an NT port to Sparc.
Around a year ago, my wife and I were shopping for an HDTV ready wide screen TV. We had narrowed the choice to several models, including the 42" Hitatchi. The TV was advertised in several venues in our city including a shop named "TV City" (or something like that).
We did NOT want to purchase the unit over the Internet -- prefering to deal with a local vendor (just in case it wasn't suitable, or it had problems, whatever).
We went to see the retailer, and were told "Yes, we confirm the price". Also, we were NOT allowed to "test-drive" the TV (the manager refused to make the remote available, or feed any signal to the TV). We informed the manager that we would walk... still no satisfaction, and out the door we went.
To Future Shop (owned by Best Buy). They had the same TV, and honored the price. We used part of the difference to purchase the "Extended Warantee", including in-home alignment (coming up soon). One of the few times we have purchased such a programme, because it is generally not worth it, but we have small kids -- damage could easily happen.
The sales 'droid left us alone, making the shopping experience reasonably pleasurable.
So, good price, good service.
And, we are happy with the TV (although my wife now thinks we should have gone for a larger one).
The answer is that microsoft components are inappropriately shared. For example, the html render component used in userland is used to render help pages for system administration. Yes, the SAME binary.
And... there isn't any way to decouple these things. At least, not without redesign, and "opening" up the Windows platform to competition.
Now, it IS happening. but it must be coupled with an overhaul of the user experience (GUI). The user must perceive that certain things take more time and effort because certain ring barriers cannot be crossed (eg. more trusted code can NEVER call on less trusted code).
This stuff isn't new, and these "mis-features" are hardly limited to Windows. For example, I have seen a component in the Linux world that allows the VFS layer in the kernel to actually invoke user level programs to implement "convenient" file systems. Lots of fun for the experimenter, but a PRIME attack point. I mean, think, a crack in a user level component can now give greater than root (ei. kernel) access. Who's your daddy?
Since I (and I know I am a dinosaur) do my papers in troff (groff), I use refer.
I really don't know TeX, so can't comment on bibtex, but it seems to do the same job.
refer allows me to collect my references, and put them away into flat-file databases. I refer to them later by author name and year, and bingo! it does the rest.
How easy is that?
And, really, I find that vi *is* the fastest way to revise documents. pic handles most of my drawing, eqn for my equations, and standard macros (styles) for lay-out.
Now for an example of a drawing in "pic", for the un-initiated:
That's it - no "finding the ellipse tool", figuring out how to get text inside, how to draw arrows (with lines? how to do the ends?) and how to connect objects. It's just done.
Now, to get back to the topic. A linux laptop may be nice, but I find that my palm 505 with a keyboard gives me all I need, with NO boot-time. I would prefer a F/OSS platform, but am totally unwilling to sacrifice for it. I find ALL laptops to be a serious sacrifice. They don't fit in my pocket, and can't be hauled out easily to check someones address or my schedule. Battery life in insufficient for real work.
In other words, the only use a laptop has for me is as an "all-in-one" transportable desktop. Given I can find "internet cafes" in most places, a laptop isn't really all that compelling. Nice to see linux on one -- but I would rather see the a renewal of the Zaurus platform. "linuxcertified" (whoever they are) would better serve me by supplying and supporting Zaurus (or something like it) to the North American market. (PS. my price point is $300US).
So, to re-iterate: Mobile device must have "instant-on" (come on, even the Radio Shack M100 had THAT!). Solid state storage (no hard drive -- and, again, the M100 had it in 1982, my palm has it now). No larger than a binder, preferable to be able to put it in my pocket. Run F/OSS software. Easy to add serial, ethernet, and wireless (palm only has serial, USB and 802.11b, I think wired ethernet available, but I don't have it). The palm doesn't run "vi" (unfortunate -- that is what I would like to use for document editing with the keyboard).
Oh well -- I'll just have to wait and see...
I'm curious as to what you find difficult with TeX and bibs?
You don't need to block the client. Simply block 6881..6889 (incoming). Then, all downloads will happen, but slowly. BT won't be able to engage with other clients.
But, the service will still work. And, if someone does have a legitimate use, unblock specifically for that user.
"Other systems are available that recalculate a graphics image for each viewing frame, allowing considerable flexibility in movement of the image, but these systems are complex and expensive. "
Now, these systems are not so expensive! But... this statement implies that there is prior art.
"*NEVER* be made clear to the outside"? What *is* the "outside" when providing toolkit software?
"Why are you passing text directly to the database". Um... because the API to a SQL database is SQL which is... text?
You have eliminated the use of PHP (since if the interpreter fails, where is the error reported?) *and* SQL (because the API is text based) in your web page developement. Yes, I agree that would tend to increase security, but at what cost? Are you trying to eliminate your own job?
I am wondering about the obession that Microsoft has with Linux. I mean; COME ON; a free OS developed by mostly volunteers! How can that possibly be competition for the largest software company on the planet!
As one of the FOSS "minions", I am certainly interested in Microsoft - simply to ensure future interoperability. I want to ensure that future Windows will talk to my Apache Web Server, Email services, &etc.
I want to ensure that when someone brings a wireless laptop to my house, she can avail herself of the Internet, and local services.
But WTF does Microsoft care? Is the MS product so weak that FOSS actually hurts them? Then get 5 or 10 thousand developers working on the problem! And give us something better.
...What are you talking about? There is no problem getting a stent, or a bypass in Canada.
Hospitals don't routinely run out of money. There is nothing special about (say) November.
The problem is: most Canadians do live next to the Canada-USA border. Disallowing private clinics in Canada simply drives money across the border. I would like to see that money stay "at home".
And, I know many people who abuse our system. OHIP cards are in GREAT supply. Like, free healthcare for US border residents. Shame on the USA for that, but I (as a Canadian) am willing to tolerate leeching off of our system to provide poor people in the US with some health care.
Windows is monolithic. Microsoft said so. (Microsoft claimed that IE CANNOT be removed).
In linux, X *can* be replaced, and it isn't even needed (did you read the fine article?). Indeed, the *kernel* can be replaced (although calling the result "linux" would be... misleading).
Linux -> replace most EVERYTHING, (including GUI and kernel)
Ok, its a troll... but I'll bite. First, run libsafe on linux. That will offer buffer checking for the "common" cases -- at very little cost. No "recompile" needed.
I purchased a Xerox 820 with 10MB hard drive for $5000. Didn't blink an eye kicking out $400 for WordStar!
I purchased an IBM AT with 2MB memory, 20MB hard drive, moth co-pro and graphics for $12,000. Didn't blink an eye paying $1,000 in software.
If I purchase a computer for $100, I wouldn't blink an eye paying $10 for software...
It's interesting that the ratio that makes me comfortable is 90% hardware, 10% software, and that ratio has remained constant for over 20 years. I wonder if anyone has collected stats on that?
"...emulation, which is all about the MHz and basically non-parallelizable."
First, emulation IS "parallelizable". There is usually a decision: emulate, or translate, and if translating, how much optimization to apply. On a single processor machine, this is critical. It may take a great deal of time to translate; less time to emulate. If something is run once (or rarely), it doesn't make sense to translate. We can't afford the overhead.
On an MP (multi processor, or multi-core), we can emulate, and schedule translations. The translations don't have an immediate impact on run-time, but allow a future speed-up (assuming enough memory).
Secondly, it is very difficult (typically), to model things like exceptions. The choice is to (1) be accurate, but slow, or (2) to be sloppy, potentially breaking some code. On an MP platform, multiple methods can be executed. If an exception doesn't happen, the results from the slower methods can be simply discarded.
MP can also be exploited to allow ILP increases by speculative execution. Assuming fast inter-processor communication.
I find that a dual-CPU machine is a "sweet spot" for most of my needs. The GUI, etc. typically exeuctes on one CPU, and my actual application on the other. The system is then MUCH more responsive under "load". I would imagine that "MAME" would allow X to draw on one processor, while it utilized other processsors for the emulation. [or maybe not, I don't MAME as I have no interest in arade games].
Back in the mid-80's, I wrote a program in PROLOG. Rule sets for a character based "GUI" entry screen components were sent, based on the current task to be performed. Data validation, control layout and control enabling was done by executing the PROLOG predicates.
Simple idea... worked well (we only had 2400 baud modems, 9600 baud was the upper end; sending only entry rules and the PROLOG was a reasonable choice).
Took a 512K machine (at the time, a very big micro).
This was *never* used for "pricing" -- it was used to specify typographic instruction (a slightly more advanced task, IMNSHO).
Obviously, can be used for "pricing", "estimating", &etc. (estimating would have been the next logical use).
Still stands as my only commercial PROLOG program.
And the wheel goes around...
I would think that there are other examples -- the IBM 3270 field control protocol is almost there (and I bet that it has been extended to cover this use as well). Other interesting conflicts are with Smalltalk/Squeak, and even the TCL/TK toolkit.
So I don't think that Microsoft will dare enforce this one.
The "repost" was a (very gentle) reminder that the information was ALREADY public. Of course the blogger can remove the reminder... but the cat WAS out of the bag. There was no need to edit the past.
The fact is, a gentle reminder is a lot nicer than a good flaming. And funnier, to boot.
Slashdot Mirror
Actually, the employees typically DON'T use the devices.
I have engineered features for a set-top & tv box -- and I don't have (probably never will have) that tv.
I have worked for computer companies whilst never owning ANY of their product.
I have just done some engineering work for a printer company, and while I *have* in the past owned the vendor product, I will never own this particular product (and, indeed have never SEEN the product).
I have worked with a major graphics board company, and, though I do own several of their products, I was never given one to "home test".
In other words, the engineers put in the features, but we DON'T actually "eat the dog food". That job is left to Product Managers who probably don't care, and Marketing who probably doesn't either (make sure it meets the requirements).
So, if a "one-hour battery life" was in the requirements (or worse, no mention of battery life at all), that's what gets delivered.
And the justification? The employees/contractors won't BUY the stuff (why would we?); the company feels it is too expensive to build extra prototypes -- and besides, what does the employee know anyway? Stick to engineering; that's what we pay you for.
Does lead to Dilbert moments, though.
Ratboy.
Ok... here are some ideas.
First, multiple video streams can be interleaved into a single mpeg transport stream. Each of the streams has its own 33 bit decode/present timestamps, and you can (for this application), ensure that they are synchronized.
Next, the video output can be gen-locked. One free-running video, and feedback from the DAC to keep the other two video outputs synchronized. There are some chip solutions that can do this, and several board solutions as well. To make this viable, you probably want a chip-level solution.
Now, the client currently uses 3 dvds, synchronized with a remote? That's very, um, coarse, so gen-lock is probably not even needed -- just the DTC/PTC should do very nicely (and MUCH more accurately). Of course, we assume a "single" (stereo/5.1/whatever) sound stream in the mpeg.
For a board level solution, look at 3xXilleon, or something like that.
Ratboy
I would like to pursue this issue. Your claim is that OpenOffice.org doesn't reliably convert Microsoft Word documents.
I do contract work. So far, OO.o has opened every doc file I have come across, with one exception.
It was the technical programming (register) guide for the ATI Xilleon.
Over 1,000 pages, and formatted in tables. Then again, all tools in Redhat 9 refused to have anything sensible to do with the document.
But, so far, that has been the only problem case. And I have put a LOT of stuff into OO.o. For example -- my current gig involves a 10MB Excel spreadsheet. And yes, OO.o handles it, and the attendant doc files associated with the project.
Of course, most of your documents may be these very long , convoluted things -- in which case you have my sympathies. But, that kind of data shouldn't be locked away like that. At least the ATI chip reference was available in HTML as well (thank $DEITY.
If these are short documents that don't work, email me one of them. (you'll have to "unravel" my email address from the header).
As to "complaining oabout saving files in doc format"... I agree with you on that one. It is actually complaining about NOT saving in native format, and I find that bizarre.
ratboy
GUIs are playing catch-up again. "locate" has been around on Unix platforms for, what, 20 or 30 years? And, it does wildcard searches for filenames, ready to jam into grep for content search.
Content could be indexed, but its a bit project specific (so us Unix heads only do it on specific projects, right?).
For the un-initiated, a process runs (typically once a day), and indexes all filenames on your system. You can then get instant answers to "Show me all Microsoft Word documents on my system"
file `locate *.doc` | grep Microsoft
and many other queries. This stuff is PLAIN ORDINARY UNIX/LINUX. Ah well, doesn't help the completely casual user. You know, "If the option isn't clearly presented, it can't possibly be done -- or I just don't want to bother".
More power to 'em, them -- but people PLEASE don't ask when this will be ported to Linux/Unix!
Ratboy.
NT was not ported to big-endian machines. The machines it ran on were either little endian or bi-endian.
To make a reasonable transition, you would want big-endian support for legacy 68k, and legacy mac-os applications. But if the kernel runs little-endian? Its a problem.
Anyway, I believe that was the reason that there was never an NT port to Sparc.
ratboy.
Around a year ago, my wife and I were shopping for an HDTV ready wide screen TV. We had narrowed the choice to several models, including the 42" Hitatchi. The TV was advertised in several venues in our city including a shop named "TV City" (or something like that).
We did NOT want to purchase the unit over the Internet -- prefering to deal with a local vendor (just in case it wasn't suitable, or it had problems, whatever).
We went to see the retailer, and were told "Yes, we confirm the price". Also, we were NOT allowed to "test-drive" the TV (the manager refused to make the remote available, or feed any signal to the TV). We informed the manager that we would walk... still no satisfaction, and out the door we went.
To Future Shop (owned by Best Buy). They had the same TV, and honored the price. We used part of the difference to purchase the "Extended Warantee", including in-home alignment (coming up soon). One of the few times we have purchased such a programme, because it is generally not worth it, but we have small kids -- damage could easily happen.
The sales 'droid left us alone, making the shopping experience reasonably pleasurable.
So, good price, good service.
And, we are happy with the TV (although my wife now thinks we should have gone for a larger one).
Ratboy.
nate nice:
It's very clear why.
The answer is that microsoft components are inappropriately shared. For example, the html render component used in userland is used to render help pages for system administration. Yes, the SAME binary.
And... there isn't any way to decouple these things. At least, not without redesign, and "opening" up the Windows platform to competition.
Now, it IS happening. but it must be coupled with an overhaul of the user experience (GUI). The user must perceive that certain things take more time and effort because certain ring barriers cannot be crossed (eg. more trusted code can NEVER call on less trusted code).
This stuff isn't new, and these "mis-features" are hardly limited to Windows. For example, I have seen a component in the Linux world that allows the VFS layer in the kernel to actually invoke user level programs to implement "convenient" file systems. Lots of fun for the experimenter, but a PRIME attack point. I mean, think, a crack in a user level component can now give greater than root (ei. kernel) access. Who's your daddy?
Ratboy
bibtex?
Since I (and I know I am a dinosaur) do my papers in troff (groff), I use refer.
I really don't know TeX, so can't comment on bibtex, but it seems to do the same job.
refer allows me to collect my references, and put them away into flat-file databases. I refer to them later by author name and year, and bingo! it does the rest.
How easy is that?
And, really, I find that vi *is* the fastest way to revise documents. pic handles most of my drawing, eqn for my equations, and standard macros (styles) for lay-out.
Now for an example of a drawing in "pic", for the un-initiated:
ellipse "Need simple drawing?";
arrow;
ellipse "Use PIC!";
That's it - no "finding the ellipse tool", figuring out how to get text inside, how to draw arrows (with lines? how to do the ends?) and how to connect objects. It's just done.
Now, to get back to the topic. A linux laptop may be nice, but I find that my palm 505 with a keyboard gives me all I need, with NO boot-time. I would prefer a F/OSS platform, but am totally unwilling to sacrifice for it. I find ALL laptops to be a serious sacrifice. They don't fit in my pocket, and can't be hauled out easily to check someones address or my schedule. Battery life in insufficient for real work.
In other words, the only use a laptop has for me is as an "all-in-one" transportable desktop. Given I can find "internet cafes" in most places, a laptop isn't really all that compelling. Nice to see linux on one -- but I would rather see the a renewal of the Zaurus platform. "linuxcertified" (whoever they are) would better serve me by supplying and supporting Zaurus (or something like it) to the North American market. (PS. my price point is $300US).
So, to re-iterate: Mobile device must have "instant-on" (come on, even the Radio Shack M100 had THAT!). Solid state storage (no hard drive -- and, again, the M100 had it in 1982, my palm has it now). No larger than a binder, preferable to be able to put it in my pocket. Run F/OSS software. Easy to add serial, ethernet, and wireless (palm only has serial, USB and 802.11b, I think wired ethernet available, but I don't have it). The palm doesn't run "vi" (unfortunate -- that is what I would like to use for document editing with the keyboard).
Oh well -- I'll just have to wait and see...
I'm curious as to what you find difficult with TeX and bibs?
Ratboy.
3nuff:
I started my career programming on one of the SOLs. I *loved* that machine. Yes, I could/would give it a good home...
fred_weigel@hotmail.com
ratboy.
grolshie:
You don't need to block the client. Simply block 6881..6889 (incoming). Then, all downloads will happen, but slowly. BT won't be able to engage with other clients.
But, the service will still work. And, if someone does have a legitimate use, unblock specifically for that user.
Ratboy.
Yes.
0 mark
1 mark (0)
2 mark (0,1)
3 mark (0,1,3)
4 mark (0,1,4,6)
These are all OGRs, and complete. 5 is the first OGR where it is not complete. I don't think they stop being possible (but I could well be wrong).
Ratboy
"Other systems are available that recalculate a graphics image for each viewing frame, allowing considerable flexibility in movement of the image, but these systems are complex and expensive. "
Now, these systems are not so expensive! But... this statement implies that there is prior art.
There are companies that can help you...
Try nytric.com (just an example)
They will do hardware design, FPGA, programming, plastics, housing, &etc (marketing, offshoring). For a percentage, of course.
Make sure you get NDA & legal protections with them.
Ratboy.
"*NEVER* be made clear to the outside"? What *is* the "outside" when providing toolkit software?
"Why are you passing text directly to the database". Um... because the API to a SQL database is SQL which is... text?
You have eliminated the use of PHP (since if the interpreter fails, where is the error reported?) *and* SQL (because the API is text based) in your web page developement. Yes, I agree that would tend to increase security, but at what cost? Are you trying to eliminate your own job?
Ratboy.
I need some PORN... ooooh, donut!
I am wondering about the obession that Microsoft has with Linux. I mean; COME ON; a free OS developed by mostly volunteers! How can that possibly be competition for the largest software company on the planet!
As one of the FOSS "minions", I am certainly interested in Microsoft - simply to ensure future interoperability. I want to ensure that future Windows will talk to my Apache Web Server, Email services, &etc.
I want to ensure that when someone brings a wireless laptop to my house, she can avail herself of the Internet, and local services.
But WTF does Microsoft care? Is the MS product so weak that FOSS actually hurts them? Then get 5 or 10 thousand developers working on the problem! And give us something better.
Ratboy.
...What are you talking about? There is no problem getting a stent, or a bypass in Canada.
Hospitals don't routinely run out of money. There is nothing special about (say) November.
The problem is: most Canadians do live next to the Canada-USA border. Disallowing private clinics in Canada simply drives money across the border. I would like to see that money stay "at home".
And, I know many people who abuse our system. OHIP cards are in GREAT supply. Like, free healthcare for US border residents. Shame on the USA for that, but I (as a Canadian) am willing to tolerate leeching off of our system to provide poor people in the US with some health care.
Windows is monolithic. Microsoft said so. (Microsoft claimed that IE CANNOT be removed).
In linux, X *can* be replaced, and it isn't even needed (did you read the fine article?). Indeed, the *kernel* can be replaced (although calling the result "linux" would be... misleading).
Linux -> replace most EVERYTHING, (including GUI and kernel)
Windows -> cannot replace web browser.
So, which is monolithic and which is modular?
Ratboy.
Ok, its a troll... but I'll bite. First, run libsafe on linux. That will offer buffer checking for the "common" cases -- at very little cost. No "recompile" needed.
And, you can go more paranoid from there...
Ratboy.
Hat's off!
I purchased a Xerox 820 with 10MB hard drive for $5000. Didn't blink an eye kicking out $400 for WordStar!
I purchased an IBM AT with 2MB memory, 20MB hard drive, moth co-pro and graphics for $12,000. Didn't blink an eye paying $1,000 in software.
If I purchase a computer for $100, I wouldn't blink an eye paying $10 for software...
It's interesting that the ratio that makes me comfortable is 90% hardware, 10% software, and that ratio has remained constant for over 20 years. I wonder if anyone has collected stats on that?
Ratboy.
Kildall wrote PL/1 (PL/M) for the 8080 /first/. CP/M was written in PL/M (as were most of the tools).
Ratboy
"...emulation, which is all about the MHz and basically non-parallelizable."
First, emulation IS "parallelizable". There is usually a decision: emulate, or translate, and if translating, how much optimization to apply. On a single processor machine, this is critical. It may take a great deal of time to translate; less time to emulate. If something is run once (or rarely), it doesn't make sense to translate. We can't afford the overhead.
On an MP (multi processor, or multi-core), we can emulate, and schedule translations. The translations don't have an immediate impact on run-time, but allow a future speed-up (assuming enough memory).
Secondly, it is very difficult (typically), to model things like exceptions. The choice is to (1) be accurate, but slow, or (2) to be sloppy, potentially breaking some code. On an MP platform, multiple methods can be executed. If an exception doesn't happen, the results from the slower methods can be simply discarded.
MP can also be exploited to allow ILP increases by speculative execution. Assuming fast inter-processor communication.
I find that a dual-CPU machine is a "sweet spot" for most of my needs. The GUI, etc. typically exeuctes on one CPU, and my actual application on the other. The system is then MUCH more responsive under "load". I would imagine that "MAME" would allow X to draw on one processor, while it utilized other processsors for the emulation. [or maybe not, I don't MAME as I have no interest in arade games].
Ratboy.
Back in the mid-80's, I wrote a program in PROLOG. Rule sets for a character based "GUI" entry screen components were sent, based on the current task to be performed. Data validation, control layout and control enabling was done by executing the PROLOG predicates.
Simple idea... worked well (we only had 2400 baud modems, 9600 baud was the upper end; sending only entry rules and the PROLOG was a reasonable choice).
Took a 512K machine (at the time, a very big micro).
This was *never* used for "pricing" -- it was used to specify typographic instruction (a slightly more advanced task, IMNSHO).
Obviously, can be used for "pricing", "estimating", &etc. (estimating would have been the next logical use).
Still stands as my only commercial PROLOG program.
And the wheel goes around...
I would think that there are other examples -- the IBM 3270 field control protocol is almost there (and I bet that it has been extended to cover this use as well). Other interesting conflicts are with Smalltalk/Squeak, and even the TCL/TK toolkit.
So I don't think that Microsoft will dare enforce this one.
Ratboy.
Um... this is Smalltalk. The binary CAN be readily merged and modified.
It is rather difficult to distribute a Squeak application that DOESN'T allow modification and merging.
The only issue is the "distribution" part.
Ratboy.
The "repost" was a (very gentle) reminder that the information was ALREADY public. Of course the blogger can remove the reminder... but the cat WAS out of the bag. There was no need to edit the past.
The fact is, a gentle reminder is a lot nicer than a good flaming. And funnier, to boot.
Ratboy