Application Program Interface. This is how the application program interfaces to the operating system. AFAIK, *BSD can be compiled to accept system calls from native SystemV and Linux binaries.
The problem is they need to use a term for publically acessible software, without regard for the specifics of licensing terms, and public domain (no caps) is probably the only viable choice to minimize confusion with PHBs and the general public. Further 'public domain' is quoted, so the reader is alerted to the fact that the usage does not necessarily conform to the readers understanding of the definition of the term. From HP to its intended audience, the usage is correct. From Linux-oriented, *BSD-oriented, or GPL-oriented to any audience, the usage would not be correct, but an explanation of licensing terminology would be in order.
Right on target. Maybe FUD, but I like it. BTW, I haven't moved to Linux (yet). Golden rule. Who has the gold makes the rules.
Do your planning based on where Linux will be two years or so from now. With heavyweights like IBM behind it, it will grow in ways not yet imagined. Still a lot of stuff to sort out, in particular how to mix Customer-Proprietary, Vendor-Proprietary, BSD-licensed, LGPL-licensed, GPL-licensed, Public-Domain software into the systems infrastructure. (I think I missed some in that list, but you get the idea.)
Just think. Whenever Windows2000 does come out, Instant Legacy. HeHe;-)
There may be a tacit assumption that the web site will be released as Open Source. The requirements for submission are in line with that assumption. However, if that is the intention, it needs to be explicitly stated, otherwise it looks like trying to gather Open Source into Closed Source, which is an emphatic no-no.
OK, I'll bite. From Wichita Falls, Texas. Set to -10 threshold with highest posts first. Rereading, I will go to the bottom and work my way back up. At this point, your post was moderated to +2. As "news and commentary", the earlier posts will get the most attention, both readers and moderators. Anyone very interested will read and reread, so later posts are far from irrelevant. Probably the major value of later posts comes a hundred years or so in the future, when historians research the evolution of whatever they call the rennasaince (sp?) of whatever it is we are going through. This is more fun to watch than a soap opera.:-)
Errr.... no. I haven't seen *any*. I've seen several posts on the meaning of hacker. I do claim to know something of what a hacker is. This is not enough to make me one. Keypunching raw a EBCDIC IPL loader probably qualifies me as a "Real Programmer". It does not qualify me as a hacker. If I did it without access to a green card or the "Principals of Operation" manual, I would consider myself a hacker.
I think the media are clueless about both terms. I think we should regain control of them both.
>:cracker:/n./ One who breaks security on a system.
>:hacker:/n./ [originally, someone who makes furniture with an axe]
Think of cracker as in cracking a code, a cypher. This takes skill or patience or both.
Making furniture with an axe takes skill. Lots of skill. The key is that the results are out of proportion to the tools. A hacker is opposite to the poor workman who blames his tools.
A handmade chair, made with axe is a hack. Chippendale is not. C is a hack. Algol68 is not. Dunno about B. UNIX is a hack, at least the original. Multics is not. Mountain man is a hack. Flatlander is not. Climbing Everest without supplemental oxygen is a hack. Riding to the top in a helicopter is not.
Microsoft windows is not a hack. The tools were there. The results are poor considering the tools. Fails both counts.
Hackers build rather than tear down. Do not use the term for wanabes.
Hackers cannot be controlled. I think this is really why the suits, media, corporate etc. get scared.
We need to regain control of the terms, not for the benefit of hackers, but for the washed mass of J Random (L)Users out there.
A new name for us - Real Programmers
on
CNN on "hackers"
·
· Score: 1
Does keypunching (multi-punch) raw EBCDIC IPL loader count? After that, hex is easy.
"rich latencies"?!?!, "Communitarian Software"
on
The Power of Openness
·
· Score: 1
>rich latencies You ain't seen nuttin yet. And if it gets screwed up, you won't even suspect what you're missing.
>An advocacy group which sincerely tries to represent the community-of-many-names. They are more advocating and representing the lUsers (ie public) than the hackers.
>"Communitarian Software"... puts the emphasis where it belongs. Right on. BTW, this article is aimed at Harvard, the next generation of CEOs, Congressmen, Senators, etc.
>What if the primary beneficiary of oss software is the society as a whole? That is the intention. How? Who pays? I do not know. It is not simple, that I do know. Keeping track of who benefits how much from what seems counter-productive. Too complicated. Too invasive. Imagine toll streets and sidewalks. One toll per block. Pay with credit card. Worse.
Harvard's H2O http://www.opencode.org/h2o/ (/. today: The Power of Openness) seems to be taking several steps in the right direction. It is well worth reading. I don't think anyone has the answers yet. At least they seems to understand the questions. For society at large the solution has to be sociological, political.
http://cyber.law.harvard.edu/eon/center_fr.html "The Berkman Center for Internet & Society at Harvard Law School, for the whole University. The Center's mission is to explore and understand Cyberspace, its development, dynamics, norms, standards, and need or lack of need for laws and sanctions. We are a research Center, an action Center. We study Law and its complement in norms and values. What we seek to learn is not already recorded. Our method is to build out into Cyberspace, record data as we go, self-study and publish. Our mode is entrepreneurial non-profit. Our goal is to bring knowledge to the world."
Yes they do have an ulterior motive. Something like *not* killing the goose that lays the golden eggs. Do you want to be surrounded by Windows. If not, something better (ie Linux, Open Source) needs to be accessible, usable, and trusted by the masses. H2O seems more a consumer advocacy thing, with knowledge that Open Source programmers need to eat too.
> Brand names are a form of cultural credibility. If so, tell me what is the brand of the scientific method? The credentials, reputation, etc. of who proclaims the "science".
The article seems very well-researched and perceptive. It is important to consider where they are coming from -- they are on the side of the public, the users. In the area of public policy, Harvard (or Yale) should have significant expertise. They do seem to have a good grasp of the issues, much more than this poor head of mine. It *is* complex, like macroeconomics compared to microeconomics. Good programs are beautiful, to programmers. Good public policy is beautiful, to politicians. I, for one, wish them luck.
Point made. Loud and clear. I am sure you can build a transmitter/receiver from a 5-tube superhetrodyne radio and communicate around the world. (No, I'm not a ham, but I've known some.) CW is doable. If you can do voice with that you are *really* good.;-) There is a camaraderie among people who have survived, be it Slackware or Canadian winters. This tends to exclude people who haven't shared the experience. The trap is to assume that everybody should share that experience. Few men climb Everest. Extremely few do it without supplemental oxygen. Real men climb Everest without supplemental oxygen? Too high a barrier. Most people don't climb mountains, but most people do walk and even climb small hills. The point I am trying to make is that as the use of Linux spreads, the barrier to entry must fall, and the clueless and the newbies are not to be despised. The clueless will be able to do a few things, sometimes very useful things, but will be helpless if anything goes wrong or is out of place. Applies to us all one way or another, doesn't it. BTW, if you can build a PC from bare board and loose IC's, you are *not* a newbie. A Linus newbie, possibly. A newbie, no.
It shouldn't, but for historical reasons it is (difficult). The standard assumptions allow someone to just plug in a toaster and have it work. The assumptions required to just plug in a modem are not in place. Plug and play (plug and pray?) is an attempt to solve the problem.
BTW, installing American toaster (120v AC) in Europe (220v DC?) doesn't work.
Installing a toaster. 1) Find flat ledge space. 2) Put toaster on ledge. 3) Plug power cord into convenient outlet. Assumes ledges are flat and not too slippery Assumes toaster made for domestic power.
Installing a modem. 1) Find a spare slot. 2) Find a spare IO address (so cpu can talk to modem) 3) Find a spare IRQ (so modem can wake up cpu) 4) Convince OS that modem exists. 5) Convince Application that modem exists.
This is an example of where Standards *should* be applied. Imagine if each Power&Light company distributed its own favorite voltage. Then modems would be easy and toasters hard.
This is in response to the whole thread, primarily in disagreement with and rebuttal to AC.
>metaphor: Debian, Redhat, Slackware Entertaining and useful. Each distribution tries to do what is best according to the way it sees things. What works and what doesn't is a combination of a lot of little things and a few big things. Default assumptions about hardware also matter. It will be interesting how you will characterize Corel/Linux when it comes out.
>Conclusion: Clueless Newbies. Average Joe User... Let him suffer Windows... I understand your point, but should he have to suffer so much?
Right tool for the right job. Plan A. A different screwdriver for each screw. Plan B. One screwdriver for all screws. Neither extreme works. You need something in the middle. There are always tradeoffs. There are *lots* of ways to mess it up.
>In order to make a program easier to use (more user-friendly), you must make it less powerful for the advanced user. In order to make a program more powerful, it must become more complex, and therefore harder to use. Ultimately true, but it is better if simple things can be done simply, even with the power tools.
>...configurable complexity adds another layer of complexity. In spades. Ultimately, configurability is indistinguishable from chaos. AC apparently has configured something and thinks he understands configuration. Wrong. The *first* point of user configuration is the choice of which program to use. The simplest text editor is (DOS) COPY CON filename. Of course correcting mistakes is a pain.
>Ada... turned out to be a hideous language... If I remember correctly, the design of Ada makes what would be semantic (ie undetectable) errors in say PL/I or Fortran syntactic (ie detectable) errors in Ada. In some respects, Ada is more rigorous than Algol68, with reason. Depending on consequences, not killing troops, this can be a good tradeoff. But if you don't need it, you don't want to pay the penalty. Imagine the horrors if this is a configuration item.
>Putting in dynamically alterable toolbars, menus, popup boxes, etc. puts a huge strain on the computer system (hard disk, ram, processor) >>Bull. -- Wrong again. Compare hardware for DOS vs Windows.
>... modularity... opinion is that Linux is the worst-implemented of the bunch. AC is confusing modularity with a count of "modules". The proper metric counts interfaces and the amount of state information that must be known on both sides. If the Linux kernel is small, monolithic, and with well defined interfaces, it counts as *highly* modular. "NT may be layed out beautifully from the inside". Video drivers in kernel mode and crashes under heavy load say otherwise. Its all the state information that exists on both sides of the interfaces. If efficiency is ever important, it is far easier and cheaper to do the required analysis and design up front, and do it at the lowest level. Adding an efficiency layer on top just does not work.
>...the flashy configurable interface gets in my way a whole lot more than... Exactly. The cure. Configure the configuration,... , ad nauseum. Welcome to exponential complexity. If it gets everything you need or care about right, it is a good thing. If not, then it is a problem of where did they hide the.... An important point is that the configurable interface adds complexity and, since it is incomplete, reduces competence.
>Three great virues of a programmer (as defined by Larry Wall) Laziness, Impatience, and Hubris. Aren't these supposed to be vices, not virtues? Yeah, I know. Add this for Laziness: There *has* to be a easier (better, faster) way to do this. And expend more effort to indulge the laziness.
>half of all people in the world were of average or below average intelligence. To be nit-picky about it, I think you are right. The measures of IQ are at best an ordering of intelligence. If there is an equivalence between two people of 110 IQ each and two people, one with 100 IQ and one with 120 IQ, then it makes sense to use the mean. If it does not make sense, in my opinion it does not, then the mean is not useable, and the generic term average has to mean (bad pun) median. Your statement to your mother-in-law is a tautology (kinda like x=x). Your mother-in-law wants to believe that most people are smarter than most people, which is logically impossible.
Slashdot Mirror
Try:
C:\> DIR Program Files
--doesn't work.
C:\> DIR "Program Files"
--works.
C:\> COPY "Program Files\stuff" stuff
--quotes are required.
Thanks for the correction.
Application Program Interface.
This is how the application program interfaces to the operating system.
AFAIK, *BSD can be compiled to accept system calls from native SystemV and Linux binaries.
Agreed.
*BSD is an extremely valuable resource.
One size does NOT fit all.
Expect to see a transition: NT to Linux to *BSD.
The problem is they need to use a term for publically acessible software, without regard for the specifics of licensing terms, and public domain (no caps) is probably the only viable choice to minimize confusion with PHBs and the general public. Further 'public domain' is quoted, so the reader is alerted to the fact that the usage does not necessarily conform to the readers understanding of the definition of the term. From HP to its intended audience, the usage is correct. From Linux-oriented, *BSD-oriented, or GPL-oriented to any audience, the usage would not be correct, but an explanation of licensing terminology would be in order.
LINUX == shouting Linux, pay attention.
Right on target. Maybe FUD, but I like it. BTW, I haven't moved to Linux (yet). Golden rule. Who has the gold makes the rules.
;-)
Do your planning based on where Linux will be two years or so from now. With heavyweights like IBM behind it, it will grow in ways not yet imagined. Still a lot of stuff to sort out, in particular how to mix Customer-Proprietary, Vendor-Proprietary, BSD-licensed, LGPL-licensed, GPL-licensed, Public-Domain software into the systems infrastructure. (I think I missed some in that list, but you get the idea.)
Just think. Whenever Windows2000 does come out, Instant Legacy. HeHe
So you fix it so it doesn't break so easily.
Heh, heh. Good! (I need them myself.)
;-(
Consistency of interface. This is extremely important. (exit emacs via reset button
Now you know why you want to install NT twice into different directories. Boot the spare one to recover the primary one.
Lets see.... with 600+ DLLs, most of which have (had) different versions, Windows is several orders of magnitude more fragmented than *nix.
Open Source. All rights reserved. Oxymoron.
Good idea.
Good site.
Well said. One point I would like to add.
There may be a tacit assumption that the web site will be released as Open Source. The requirements for submission are in line with that assumption. However, if that is the intention, it needs to be explicitly stated, otherwise it looks like trying to gather Open Source into Closed Source, which is an emphatic no-no.
OK, I'll bite. :-)
From Wichita Falls, Texas.
Set to -10 threshold with highest posts first.
Rereading, I will go to the bottom and work my way back up.
At this point, your post was moderated to +2.
As "news and commentary", the earlier posts will get the most attention, both readers and moderators. Anyone very interested will read and reread, so later posts are far from irrelevant. Probably the major value of later posts comes a hundred years or so in the future, when historians research the evolution of whatever they call the rennasaince (sp?) of whatever it is we are going through. This is more fun to watch than a soap opera.
Errr.... no. I haven't seen *any*.
I've seen several posts on the meaning of hacker. I do claim to know something of what a hacker is. This is not enough to make me one. Keypunching raw a EBCDIC IPL loader probably qualifies me as a "Real Programmer". It does not qualify me as a hacker. If I did it without access to a green card or the "Principals of Operation" manual, I would consider myself a hacker.
I think the media are clueless about both terms.
/n./ One who breaks security on a system.
/n./ [originally, someone who makes furniture with an axe]
I think we should regain control of them both.
>:cracker:
>:hacker:
Think of cracker as in cracking a code, a cypher. This takes skill or patience or both.
Making furniture with an axe takes skill. Lots of skill. The key is that the results are out of proportion to the tools. A hacker is opposite to the poor workman who blames his tools.
A handmade chair, made with axe is a hack. Chippendale is not.
C is a hack. Algol68 is not. Dunno about B.
UNIX is a hack, at least the original. Multics is not.
Mountain man is a hack. Flatlander is not.
Climbing Everest without supplemental oxygen is a hack. Riding to the top in a helicopter is not.
Microsoft windows is not a hack. The tools were there. The results are poor considering the tools. Fails both counts.
Hackers build rather than tear down. Do not use the term for wanabes.
Hackers cannot be controlled. I think this is really why the suits, media, corporate etc. get scared.
We need to regain control of the terms, not for the benefit of hackers, but for the washed mass of J Random (L)Users out there.
Does keypunching (multi-punch) raw EBCDIC IPL loader count? After that, hex is easy.
>rich latencies
... puts the emphasis where it belongs.
You ain't seen nuttin yet. And if it gets screwed up, you won't even suspect what you're missing.
>An advocacy group which sincerely tries to represent the community-of-many-names.
They are more advocating and representing the lUsers (ie public) than the hackers.
>"Communitarian Software"
Right on. BTW, this article is aimed at Harvard, the next generation of CEOs, Congressmen, Senators, etc.
>What if the primary beneficiary of oss software is the society as a whole?
That is the intention. How? Who pays? I do not know. It is not simple, that I do know.
Keeping track of who benefits how much from what seems counter-productive. Too complicated. Too invasive. Imagine toll streets and sidewalks. One toll per block. Pay with credit card. Worse.
Harvard's H2O http://www.opencode.org/h2o/ (/. today: The Power of Openness) seems to be taking several steps in the right direction. It is well worth reading. I don't think anyone has the answers yet. At least they seems to understand the questions. For society at large the solution has to be sociological, political.
http://cyber.law.harvard.edu/eon/center_fr.html "The Berkman Center for Internet & Society at Harvard Law School, for the whole University. The Center's mission is to explore and understand Cyberspace, its development, dynamics, norms, standards, and need or lack of need for laws and sanctions. We are a research Center, an action Center. We study Law and its complement in norms and values. What we seek to learn is not already recorded. Our method is to build out into Cyberspace, record data as we go, self-study and publish. Our mode is entrepreneurial non-profit. Our goal is to bring knowledge to the world."
Yes they do have an ulterior motive. Something like *not* killing the goose that lays the golden eggs. Do you want to be surrounded by Windows. If not, something better (ie Linux, Open Source) needs to be accessible, usable, and trusted by the masses. H2O seems more a consumer advocacy thing, with knowledge that Open Source programmers need to eat too.
> Brand names are a form of cultural credibility. If so, tell me what is the brand of the scientific method?
The credentials, reputation, etc. of who proclaims the "science".
The article seems very well-researched and perceptive.
It is important to consider where they are coming from -- they are on the side of the public, the users. In the area of public policy, Harvard (or Yale) should have significant expertise. They do seem to have a good grasp of the issues, much more than this poor head of mine. It *is* complex, like macroeconomics compared to microeconomics.
Good programs are beautiful, to programmers.
Good public policy is beautiful, to politicians.
I, for one, wish them luck.
Point made. Loud and clear. ;-)
I am sure you can build a transmitter/receiver from a 5-tube superhetrodyne radio and communicate around the world. (No, I'm not a ham, but I've known some.) CW is doable. If you can do voice with that you are *really* good.
There is a camaraderie among people who have survived, be it Slackware or Canadian winters. This tends to exclude people who haven't shared the experience. The trap is to assume that everybody should share that experience.
Few men climb Everest. Extremely few do it without supplemental oxygen. Real men climb Everest without supplemental oxygen? Too high a barrier. Most people don't climb mountains, but most people do walk and even climb small hills. The point I am trying to make is that as the use of Linux spreads, the barrier to entry must fall, and the clueless and the newbies are not to be despised. The clueless will be able to do a few things, sometimes very useful things, but will be helpless if anything goes wrong or is out of place. Applies to us all one way or another, doesn't it.
BTW, if you can build a PC from bare board and loose IC's, you are *not* a newbie. A Linus newbie, possibly. A newbie, no.
It shouldn't, but for historical reasons it is (difficult). The standard assumptions allow someone to just plug in a toaster and have it work. The assumptions required to just plug in a modem are not in place. Plug and play (plug and pray?) is an attempt to solve the problem.
BTW, installing American toaster (120v AC) in Europe (220v DC?) doesn't work.
Installing a toaster.
1) Find flat ledge space.
2) Put toaster on ledge.
3) Plug power cord into convenient outlet.
Assumes ledges are flat and not too slippery
Assumes toaster made for domestic power.
Installing a modem.
1) Find a spare slot.
2) Find a spare IO address (so cpu can talk to modem)
3) Find a spare IRQ (so modem can wake up cpu)
4) Convince OS that modem exists.
5) Convince Application that modem exists.
This is an example of where Standards *should* be applied. Imagine if each Power&Light company distributed its own favorite voltage. Then modems would be easy and toasters hard.
This is in response to the whole thread, primarily in disagreement with and rebuttal to AC.
... Let him suffer Windows ...
... turned out to be a hideous language ...
... opinion is that Linux is the worst-implemented of the bunch.
... ... , ad nauseum. Welcome to exponential complexity. .... An important point is that the configurable interface adds complexity and, since it is incomplete, reduces competence.
>metaphor: Debian, Redhat, Slackware
Entertaining and useful. Each distribution tries to do what is best according to the way it sees things. What works and what doesn't is a combination of a lot of little things and a few big things. Default assumptions about hardware also matter. It will be interesting how you will characterize Corel/Linux when it comes out.
>Conclusion: Clueless Newbies. Average Joe User
I understand your point, but should he have to suffer so much?
Right tool for the right job.
Plan A. A different screwdriver for each screw.
Plan B. One screwdriver for all screws.
Neither extreme works. You need something in the middle. There are always tradeoffs. There are *lots* of ways to mess it up.
>In order to make a program easier to use (more user-friendly), you must make it less powerful for the advanced user. In order to make a program more powerful, it must become more complex, and therefore harder to use.
Ultimately true, but it is better if simple things can be done simply, even with the power tools.
>...configurable complexity adds another layer of complexity.
In spades. Ultimately, configurability is indistinguishable from chaos. AC apparently has configured something and thinks he understands configuration. Wrong.
The *first* point of user configuration is the choice of which program to use. The simplest text editor is (DOS) COPY CON filename. Of course correcting mistakes is a pain.
>Ada
If I remember correctly, the design of Ada makes what would be semantic (ie undetectable) errors in say PL/I or Fortran syntactic (ie detectable) errors in Ada. In some respects, Ada is more rigorous than Algol68, with reason. Depending on consequences, not killing troops, this can be a good tradeoff. But if you don't need it, you don't want to pay the penalty. Imagine the horrors if this is a configuration item.
>Putting in dynamically alterable toolbars, menus, popup boxes, etc. puts a huge strain on the computer system (hard disk, ram, processor)
>>Bull. -- Wrong again. Compare hardware for DOS vs Windows.
>... modularity
AC is confusing modularity with a count of "modules". The proper metric counts interfaces and the amount of state information that must be known on both sides. If the Linux kernel is small, monolithic, and with well defined interfaces, it counts as *highly* modular. "NT may be layed out beautifully from the inside". Video drivers in kernel mode and crashes under heavy load say otherwise. Its all the state information that exists on both sides of the interfaces. If efficiency is ever important, it is far easier and cheaper to do the required analysis and design up front, and do it at the lowest level. Adding an efficiency layer on top just does not work.
>...the flashy configurable interface gets in my way a whole lot more than
Exactly. The cure. Configure the configuration,
If it gets everything you need or care about right, it is a good thing. If not, then it is a problem of where did they hide the
>Three great virues of a programmer (as defined by Larry Wall) Laziness, Impatience, and Hubris.
Aren't these supposed to be vices, not virtues? Yeah, I know. Add this for Laziness: There *has* to be a easier (better, faster) way to do this. And expend more effort to indulge the laziness.
>half of all people in the world were of average or below average intelligence.
To be nit-picky about it, I think you are right.
The measures of IQ are at best an ordering of intelligence. If there is an equivalence between two people of 110 IQ each and two people, one with 100 IQ and one with 120 IQ, then it makes sense to use the mean. If it does not make sense, in my opinion it does not, then the mean is not useable, and the generic term average has to mean (bad pun) median. Your statement to your mother-in-law is a tautology (kinda like x=x). Your mother-in-law wants to believe that most people are smarter than most people, which is logically impossible.