If the x86 instruction set has as much cruft in it as I've heard some people say, due to keeping backwards compatibility from the 286 while adding new stuff on for each generation since, then writing a nice, new, shiny, cruft-free instruction set for the Crusoe (having learned what's good and what's not from the last x generations of processors) could be A Good Thing (tm).
It might be faster than x86 emulation (with fewer instructions to translate, the instructions in the translation cache might stay there longer with fewer misses or something) and compilers for it could probably be ported in a relatively short space of time if it was clean enough.
Hmmmm....a nice new 64-bit OS running on cheap hardware? Sounds good to me:)
distributed.net is a known and proven source for ranking the status of encryption. They are not a malicous orgainization, but rather a testbed, placing an encryption technology under intense scrutiny and forcing it to prove it's worth. Where CSS (or its relacement) placed under such scrutiny, the results may be enough to have future lawsuits thrown out of court. (and therefore lessening the slashdot posts/comments on this subject... no offense, but it is getting kind of old. There is very little new information on the subject left)
Huh? Places encryption technologies under intense scrutinies?
Surely all distributed.net does is brute force attack whatever is thrown at it by virtue of having an immense amount of computing power available to it worldwide. AFAIK it doesn't do cryptanalysis of the encryption algorithm used or place the algorithm under intense mathmatical scrutiny of any kind.
Even if we can't reverse engineer it, there's another alternative: A VQF plugin was recently released which simply used Yamaha's Windows.dll via wine, under XMMS. Why not do the same with MSs encoder?
Because that'll only work for GNU/Linux (or maybe *BSD) running on x86 boxen?
What about people running Linux on PPCs or Macs? Or Sun workstations? Or BeOS on some other hardware config.?
Just plugging in to some proprietary system ain't good enough if the proprietary system is only available to plug into on some target platforms. And if you're going to do x86 emulation to run something like Wine on (can Wine run on non-x86 hardware via some emulation layer?) I bet it ain't gonna be fast enough to decode & display streaming video in real time.
I think that having a non-object 'int' primitive is A Good Thing. Especially as Java doesn't implement operator overloading.
Imagine ints with no '+' operator. You'd have to call an 'add' member function (or similar) to do stuff with them.
Also, you want to add all the 'Object' base class stuff, and reference counting/garbage collection overhead to a temp var that's only going to be used as a loop counter?
These products *do not* violate their own license; this is legally impossible. Instead, the actions taken by the authors amend to purported license. With LyX, we made the qualifications explicit a couple of releases ago; KDE may be too far into flamewar for this to be possible.
No, the point I was trying to make is that you're violating the GPL copyright by modifying the GPL.
GNU GENERAL PUBLIC LICENSE
Version 2, June 1991
Copyright (C) 1989, 1991 Free Software Foundation, Inc. 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA
Everyone is permitted to copy and distribute verbatim copies of this license document, but changing it is not allowed.
I offer an alternative: Give us differential licenses! People like to refer to licenses using phrases like "GPL but you can't fork" or "BSD but you need to advertise in Playboy" or whatnot. If there's a pressing freedom or restriction that needs to be embedded in the license, ship the GPL with a patch. There will be some noise made, and your license should not include GPL in the name, but at leaast it would be simple to derive what specific changes you desired.
There is a slight problem with that. The GPL is copyrighted. The copyright does not say that the document itself falls under it's own copyright, so the GPL has not been GPLd. Hence you are not allowed to change or make modifications to the GPL (and hence not allowed to redistribute such changes either) for your own use.
Which is probably a good idea.
Imagine if someone forked a version of the GPL in such a way as it no longer forced the software it applied to to be free. Although the original GPL would have to still apply to the license itself, the license could be marketed as 'a GPL derivative' or 'an extension of the GPL'. If this were the case, the percieved 'power' of the GPL would be significantly weakened IMHO.
Or, as a friend of mine once said 'you can disguide a geek'.
Meaning geeks and nerds can have equivalent technical skills/inclinations, but that geeks also have a sociable side to them _that allows them to interact with non-geeks/nerds_. This also means that (at least in the circles I travel in) the geeks are slightly more vain (not vain in any absolute sense, naturally, just more vain than their nerdy counterparts:-) and have a wider circle of friends.
I agreed that three years at university definitely increases the likelyhood of a nerd transforming into a geek.
(Note : this is just the perceived meaning of those words in the circles I frequent. YMMV)
Can't speak for everyone else, but I mentioned my salary in another thread up there 'cos I feel safe in my relative anonymity. It's not like I'm leaving my 'real' name all over the place.
For the on-topic bit: I was suckered out to Virginia for $50k/year fresh out of college, but I've got lots of skill in realtime 3D which this company wanted, but they didn't know how to treat their employees and so I left after a few months.
Which do you prefer, money or happiness?
I prefer happiness. And I couldn't work anywhere I wasn't. (And for those that think that I can remain happy just thinking about money when all around me is crap, think again)
Fortunately, I managed to land a job doing what I want (C++ programming) in a really great company. It's small, and very non-corporate. By that I mean that the programmers don't have to wear suits. I often work in jeans, t-shirt & para boots, because that's what I'm comfortable in. Today I'm wearing 'proper' trousers with my t-shirt & para boots, but that's just because my jeans are in the wash:)
But that's not the best bit. The best bit is that the people I work with & around are very cool people that are fun to be around & who I enjoy working with & learning from. The atmosphere's great and very supportive and I make a difference.
And I know that not everywhere's like that. But the cool jobs _are_ out there 'in industry'. It might be tough finding them, but they're there.
One thing I have learned which surprised me. You don't have to be writing a cool bit of software to enjoy writing it. I'm writing a piece of software in possibly one of the dullest fields imaginable (apart from maybe payrolls:), but becuase of how complex & advanced this particular piece of software is, it's a cool thing to work on.
Oh, I live in the UK and I've been out of Uni for 2 years. I spent the first year in a start-up with friends (bad move, went nowhere), six months looking for work and landed this job as a Junior Programmer on £18K 7 months ago.
Apply the/. effect to the most offending (non-standard) web site of the week, followed by a barrage of (very polite, but somewhat disgruntled) emails mentioning 'how unfortunate it is that I can't access the site properly'.
Check back a month later to see if things have improved.
I know all too many bright and intelligent people who use Windows for the sole reason that all of their neighbors do.
That's a bad thing? Surely that's one of the best reasons to use Windows, and I personally can't really find any fault with that line of reasoning. If everyone else is using Windows and is passing around all their info in propietary MS Office formats, and they're partularly resistant to change, and there's a lot of them, then why not use Windows if (in that respect) it'll make your life easier?
Being bright enough to be truly individualistic (IMO) means not being non-conformist for the sake of being non-conformist, but being non-conformist when it suits you because the particular non-conformist attitude you've taken is a more logical than the course of action taken by sheep.
However, if the sheep happen to be doing the right thing, there's not point making life more difficult for yourself by doing something different for the sake of it. Similarly, if they're not doing the right thing, but swimming against the tide is just too much work for the benefit you expect to get out of it, go with the flow, man.:)
If it makes my life easier to use Windows for general desktop tomfoolery becuase 'everyone else is doing it', and what I do happens to not produce hourly BSODs, then I'll choose Windows for that very reason.
OTOH, having someone piss you off does not discharge *you* from taking responsibility for your actions.
If you're mature, have half a brain and keep a calm head, you should be able to refrain from 'getting pissed off' most of the time anyway.
(Note : being in fear of your life is somewhat different from being pissed off by something. Not that _this_ discharges you from taking responsibility for your actions, it just widens the range of what constitues a 'responsible' action under the circumstances)
Best idea I've read recently (can't remember where; probably on one of the boards here) is for HUD contacts.
Full '3D' (i.e. perspective altered for each eye) coverage of your eyes' viewable areas, with controls to make the 'background' variably translucent/transparent.
That would be cool. (If not without its problems, e.g. power/RGB input)
I didn't realise (having not looked incredibly closely at the Java threading spec) that it didn't even specify whether or not java threads were pre-emptive, co-op or a combination of the two.
In some cases, yes, implementation-defined behaviour is 'not too bad', but for something of that magnitude to be left out, especially in Java (where you're supposed to be able to 'write once, etc...), seems like such a cop-out.
(Almost as bad as the fact that operations on long (and double?) are not guaranteed to be atomic. What an ugly, unsymmetric piece of cruft that is.)
Thread scheduling, for some reason, is in the Java language specification specified to be left unspecified. This leaves the implementers of the virtual machines free to select whatever model they desire (and that is usually the simplest: non-preemptive fifo-scheduling), complicating evelopment of platform-independant multithreaded applications (where one often expects time-sliced, "fair" scheduling of some sort).
Surely that's just a J(V)M QOI issue. If the implementation you (or your customers) are using is crap, bitch to you J(V)M vendor about it (or tell your customers to) until it gets fucking fixed.
And if it doesn't, change vendors.
If your bosses (or your customers) ask you to write stuff that's going to run on a broken platform, what else can they reasonably expect but a correct, well written implementation that happens to perform badly on said pile-of-shite platform.
I also think that the language still has a way to go before being "mature" enough - part of that way would be completing the specification such that the virtual machines behave alike (and such that the "write once, run anywhere" can be delivered as promised by Sun...)
I thought that a version of the language was complete? (completely specified that is). Do you mean you think that current implementations of the J(V)M have a long way to go before being able to be classed as 'mature'?
Oh come on! I'm sure by 'Quake' she _meant_ 'Q3Atest2', it's just that 'Quake' is easier to type (no digits, and there's the deeply-entrenched 'muscular memory' that allows typing common things like 'Quake', 'finger johnc@idsoftware.com | less' and your login+pwd v. v. quick & easy:-P
K.
Destruction through stupidity rather than malice..
on
Rise of the Nanobots
·
· Score: 2
The one scenario that keeps coming to my mind is that early on in nanobot development some nanobot AI coder is going to be hacking some code late one night and introduce a small bug that just causes the bot to keep reproducing itself out of whatever's around. She'll tell her nano-bot-building-machine to build the first one, and becuase it all compiles cleanly, she'll go home for the night.
She'll wake up in the afternoon to find a huge sea-like nano-mass where her lab used to be. Which will be a huge swarm of nanobots deconstructing each other to build new copies of themselves except for the outer layer, which will be busy devouring the earth's crust in an attempt to do same.
And because the technology's new, we won't be able to stop them with other more advanced bots, 'cos they haven't been invented yet.
Slashdot Mirror
Someone moderate that up some more.
:)
If the x86 instruction set has as much cruft in it as I've heard some people say, due to keeping backwards compatibility from the 286 while adding new stuff on for each generation since, then writing a nice, new, shiny, cruft-free instruction set for the Crusoe (having learned what's good and what's not from the last x generations of processors) could be A Good Thing (tm).
It might be faster than x86 emulation (with fewer instructions to translate, the instructions in the translation cache might stay there longer with fewer misses or something) and compilers for it could probably be ported in a relatively short space of time if it was clean enough.
Hmmmm....a nice new 64-bit OS running on cheap hardware? Sounds good to me
And in the states, the LCD is a PFB segment of the population.
:-)
Uhhhh....last time I checked the LCD tends to encompass approximately 100% of any given population
distributed.net is a known and proven source for ranking the status of encryption. They are not a malicous orgainization, but rather a testbed, placing an encryption technology under intense scrutiny and forcing it to prove it's worth. Where CSS (or its relacement) placed under such scrutiny, the results may be enough to have future lawsuits thrown out of court. (and therefore lessening the slashdot posts/comments on this subject... no offense, but it is getting kind of old. There is very little new information on the subject left)
Huh? Places encryption technologies under intense scrutinies?
Surely all distributed.net does is brute force attack whatever is thrown at it by virtue of having an immense amount of computing power available to it worldwide. AFAIK it doesn't do cryptanalysis of the encryption algorithm used or place the algorithm under intense mathmatical scrutiny of any kind.
...but doesn't the article then go on to say that it *will* be the brightest full moon in 69 years (since 1930)?
So, it's not a *total* non-event.
Even if we can't reverse engineer it, there's another alternative: A VQF plugin was recently released which simply used Yamaha's Windows .dll via wine, under XMMS. Why not do the same with MSs encoder?
Because that'll only work for GNU/Linux (or maybe *BSD) running on x86 boxen?
What about people running Linux on PPCs or Macs? Or Sun workstations? Or BeOS on some other hardware config.?
Just plugging in to some proprietary system ain't good enough if the proprietary system is only available to plug into on some target platforms. And if you're going to do x86 emulation to run something like Wine on (can Wine run on non-x86 hardware via some emulation layer?) I bet it ain't gonna be fast enough to decode & display streaming video in real time.
K.
I think that having a non-object 'int' primitive is A Good Thing. Especially as Java doesn't implement operator overloading.
Imagine ints with no '+' operator. You'd have to call an 'add' member function (or similar) to do stuff with them.
Also, you want to add all the 'Object' base class stuff, and reference counting/garbage collection overhead to a temp var that's only going to be used as a loop counter?
No, IMO int primitives are definitely a plus.
K.
These products *do not* violate their own license; this is legally impossible. Instead, the actions taken by the authors amend to purported license. With LyX, we made the qualifications explicit a couple of releases ago; KDE may be too far into flamewar for this to be possible.
No, the point I was trying to make is that you're violating the GPL copyright by modifying the GPL.
GNU GENERAL PUBLIC LICENSE
Version 2, June 1991
Copyright (C) 1989, 1991 Free Software Foundation, Inc.
59 Temple Place - Suite 330, Boston, MA 02111-1307, USA
Everyone is permitted to copy and distribute verbatim copies of this license document, but changing it is not allowed.
I offer an alternative: Give us differential licenses! People like to refer to licenses using phrases like "GPL but you can't fork" or "BSD but you need to advertise in Playboy" or whatnot. If there's a pressing freedom or restriction that needs to be embedded in the license, ship the GPL with a patch. There will be some noise made, and your license should not include GPL in the name, but at leaast it would be simple to derive what specific changes you desired.
There is a slight problem with that. The GPL is copyrighted. The copyright does not say that the document itself falls under it's own copyright, so the GPL has not been GPLd. Hence you are not allowed to change or make modifications to the GPL (and hence not allowed to redistribute such changes either) for your own use.
Which is probably a good idea.
Imagine if someone forked a version of the GPL in such a way as it no longer forced the software it applied to to be free. Although the original GPL would have to still apply to the license itself, the license could be marketed as 'a GPL derivative' or 'an extension of the GPL'. If this were the case, the percieved 'power' of the GPL would be significantly weakened IMHO.
(IANAL)
K.
Or, as a friend of mine once said 'you can disguide a geek'.
:-) and have a wider circle of friends.
Meaning geeks and nerds can have equivalent technical skills/inclinations, but that geeks also have a sociable side to them _that allows them to interact with non-geeks/nerds_. This also means that (at least in the circles I travel in) the geeks are slightly more vain (not vain in any absolute sense, naturally, just more vain than their nerdy counterparts
I agreed that three years at university definitely increases the likelyhood of a nerd transforming into a geek.
(Note : this is just the perceived meaning of those words in the circles I frequent. YMMV)
Can't speak for everyone else, but I mentioned my salary in another thread up there 'cos I feel safe in my relative anonymity. It's not like I'm leaving my 'real' name all over the place.
K.
For the on-topic bit: I was suckered out to Virginia for $50k/year fresh out of college, but I've got lots of skill in realtime 3D which this company wanted, but they didn't know how to treat their employees and so I left after a few months.
:)
:), but becuase of how complex & advanced this particular piece of software is, it's a cool thing to work on.
Which do you prefer, money or happiness?
I prefer happiness. And I couldn't work anywhere I wasn't. (And for those that think that I can remain happy just thinking about money when all around me is crap, think again)
Fortunately, I managed to land a job doing what I want (C++ programming) in a really great company. It's small, and very non-corporate. By that I mean that the programmers don't have to wear suits. I often work in jeans, t-shirt & para boots, because that's what I'm comfortable in. Today I'm wearing 'proper' trousers with my t-shirt & para boots, but that's just because my jeans are in the wash
But that's not the best bit. The best bit is that the people I work with & around are very cool people that are fun to be around & who I enjoy working with & learning from. The atmosphere's great and very supportive and I make a difference.
And I know that not everywhere's like that. But the cool jobs _are_ out there 'in industry'. It might be tough finding them, but they're there.
One thing I have learned which surprised me. You don't have to be writing a cool bit of software to enjoy writing it. I'm writing a piece of software in possibly one of the dullest fields imaginable (apart from maybe payrolls
Oh, I live in the UK and I've been out of Uni for 2 years. I spent the first year in a start-up with friends (bad move, went nowhere), six months looking for work and landed this job as a Junior Programmer on £18K 7 months ago.
K.
Now _this_ is a good idea.
/. effect to the most offending (non-standard) web site of the week, followed by a barrage of (very polite, but somewhat disgruntled) emails mentioning 'how unfortunate it is that I can't access the site properly'.
Apply the
Check back a month later to see if things have improved.
Rinse, repeat.
I know all too many bright and intelligent people who use Windows for the sole reason that all of their neighbors do.
:)
That's a bad thing? Surely that's one of the best reasons to use Windows, and I personally can't really find any fault with that line of reasoning. If everyone else is using Windows and is passing around all their info in propietary MS Office formats, and they're partularly resistant to change, and there's a lot of them, then why not use Windows if (in that respect) it'll make your life easier?
Being bright enough to be truly individualistic (IMO) means not being non-conformist for the sake of being non-conformist, but being non-conformist when it suits you because the particular non-conformist attitude you've taken is a more logical than the course of action taken by sheep.
However, if the sheep happen to be doing the right thing, there's not point making life more difficult for yourself by doing something different for the sake of it. Similarly, if they're not doing the right thing, but swimming against the tide is just too much work for the benefit you expect to get out of it, go with the flow, man.
If it makes my life easier to use Windows for general desktop tomfoolery becuase 'everyone else is doing it', and what I do happens to not produce hourly BSODs, then I'll choose Windows for that very reason.
K.
Surely IE should be under 'consumer apps' instead of 'OS products' :)
OTOH, having someone piss you off does not discharge *you* from taking responsibility for your actions.
If you're mature, have half a brain and keep a calm head, you should be able to refrain from 'getting pissed off' most of the time anyway.
(Note : being in fear of your life is somewhat different from being pissed off by something. Not that _this_ discharges you from taking responsibility for your actions, it just widens the range of what constitues a 'responsible' action under the circumstances)
K.
Would Christopher have been hauled off to jail he if was the star quarterback on the high school football team? Not likely.
Would Christopher have written an "outstanding", 100-scoring essay if he was the star quarterback on the high school football team?
Not likely.
Goggles?
Best idea I've read recently (can't remember where; probably on one of the boards here) is for HUD contacts.
Full '3D' (i.e. perspective altered for each eye) coverage of your eyes' viewable areas, with controls to make the 'background' variably translucent/transparent.
That would be cool. (If not without its problems, e.g. power/RGB input)
No current version of the language?
Surely the most recent specification (i.e. the "old" one + appendix) of the language _is_ the current version. Isn't it??
Anything newer than that is surely just a draft (on paper) or a non-standard extension (in a J(V)M/compiler)
I am getting a little confused here.
K.
Cool. Thanks for that link to Holub's page.
I didn't realise (having not looked incredibly closely at the Java threading spec) that it didn't even specify whether or not java threads were pre-emptive, co-op or a combination of the two.
In some cases, yes, implementation-defined behaviour is 'not too bad', but for something of that magnitude to be left out, especially in Java (where you're supposed to be able to 'write once, etc...), seems like such a cop-out.
(Almost as bad as the fact that operations on long (and double?) are not guaranteed to be atomic. What an ugly, unsymmetric piece of cruft that is.)
Things have gotten better--J++ and the MS VM have been instrumental in this regard...
This "Microsoft VM" I keep hearing about - I keep thinking "What sort of Virtual Machine is it? What does it emulate?"
Could be anything. Playstation. DOS. Java. Washing machine. Toaster. Cray.
Thread scheduling, for some reason, is in the Java language specification specified to be left unspecified. This leaves the implementers of the virtual machines free to select whatever model they desire (and that is usually the simplest: non-preemptive fifo-scheduling), complicating evelopment of platform-independant multithreaded applications (where one often expects time-sliced, "fair" scheduling of some sort).
Surely that's just a J(V)M QOI issue. If the implementation you (or your customers) are using is crap, bitch to you J(V)M vendor about it (or tell your customers to) until it gets fucking fixed.
And if it doesn't, change vendors.
If your bosses (or your customers) ask you to write stuff that's going to run on a broken platform, what else can they reasonably expect but a correct, well written implementation that happens to perform badly on said pile-of-shite platform.
I also think that the language still has a way to go before being "mature" enough - part of that way would be completing the specification such that the virtual machines behave alike (and such that the "write once, run anywhere" can be delivered as promised by Sun...)
I thought that a version of the language was complete? (completely specified that is). Do you mean you think that current implementations of the J(V)M have a long way to go before being able to be classed as 'mature'?
Oh come on! I'm sure by 'Quake' she _meant_ 'Q3Atest2', it's just that 'Quake' is easier to type (no digits, and there's the deeply-entrenched 'muscular memory' that allows typing common things like 'Quake', 'finger johnc@idsoftware.com | less' and your login+pwd v. v. quick & easy :-P
K.
The one scenario that keeps coming to my mind is that early on in nanobot development some nanobot AI coder is going to be hacking some code late one night and introduce a small bug that just causes the bot to keep reproducing itself out of whatever's around. She'll tell her nano-bot-building-machine to build the first one, and becuase it all compiles cleanly, she'll go home for the night.
She'll wake up in the afternoon to find a huge sea-like nano-mass where her lab used to be. Which will be a huge swarm of nanobots deconstructing each other to build new copies of themselves except for the outer layer, which will be busy devouring the earth's crust in an attempt to do same.
And because the technology's new, we won't be able to stop them with other more advanced bots, 'cos they haven't been invented yet.
We might be able to nuke them....
K.