I was under the impression that X.2 mostly did stuff to the UI,...
Mostly, but a fair amount of stuff is also happening at the Darwin/BSD level. One thing I know about in particular is that BSD-style signal handling is supposed to be in 10.2, which might allow porting of CMUCL (a kick-ass free-as-in-beer Common Lisp) with minimal pain and suffering.
"I tried to do an su. Wrong password. (My account has admin privs anyway; I shouldn't need to do an su at all.)"
OS X has three privilege levels, not two: superuser (root), admin, and regular user.
Close, but not quite. The admin stuff in OS X is one of the places where the NeXT people mapped Un*x to Mac conventions just right. Having admin privilege is exactly equivalent to being on the sudo list. This means you run as a normal user until (and only until you need special powers, you take those powers when you need them, and you give them up as soon as you can.
In short, on OS X you don't need a root user. Use sudo, not su.
* the iBook is tough, and I beat the crap out of portables. I've dropped my iBook off a table onto an industrial-grade carpet over concrete. The CD-ROM drive popped open, but otherwise no damage.
I envy the TiBook's larger screen, but not much else about it. However, I don't do video or graphic-intensive games, where the G4 would make a difference.
I run XFree86 and OrborOSX when I need to run X-based stuff. It has the usual propeller-key-click hack for the second mouse button, which I rarely need since I mostly use X to read my mail (Netscape running on a remote machine). Random multi-button USB mice work with OS X - this is more of an option for desktop machines, of course. I run OS X on a duel 450 MHz G4 tower, with a two-button Kensington mouse, and the second button works just fine under OrborOSX.
I do most of my hacking using emacs-on-aqua, an emacs port based on an old NeXT port. It runs all sorts of hairy elisp packages without a burp, and lets me do development without running X.
If so, you'll like the networking, power and sleep management in OS X. Location management is excellent. You can have arbitrarily many sets of networking environments, and switching between them takes two mouse clicks, no reboot required.
My 500 MHz iBook's wake from sleep time is two seconds, counting from the time I unlatch the lid. I reboot it only after major OS updates - the last one was July 8th. I have never lost work due to a faliure to wake from sleep.
My battery life in the field is about three hours, mostly running emacs and developing code. I can stop in the middle of anything, close the lid, and walk away confident that I won't lose work. The machine will sleep for about two weeks on a full charge (I lose ~7% battery power per 24 hours).
You can even safely run the battery completely dead - OS X's last gasp is to write the complete state of the machine to the hard drive, and when you find an outlet, plug in and reboot, you come back to exactly where you left off. My kids do this with full-screen games.
Of course, by "open" I mean non-DRM. I mean, come on, no sane military would trust its computing needs to a system with secret source code, proprietary undocumented file formats, and remote-control licensing, any one of which could stop things in mid-flight.
I think you're probably talking about NoteCards, an early hypertext system implemented in Interlisp-D. NoteCards was out in the research community at about the same time the Mac was first released (1984), and was a commercial product from Xerox in 1987.
Why doesn't Hollywood just buy up all the companies that make things it fears?
Because it can't afford them. The computer industry's revenues are bigger than the entertainment industry's revenues by about a factor of ten. {dammit, why can't I find a citation for this when I really need one? I've seen it any number of places on the web...} Suffice it to say crippling the US computer industry to protect the US media industry will be a net loss for the US.
In theory, you can give a Unix file a name of any arbitrary string of charcters. In reality, there are several characters that won't work, or will work at the raw file system level but be unreliable or plain broken above that. In rough order of obviousness, they are:
\000 - Putting a zero byte in a filename will break any program written in C.
'/' - A filename with an embedded slash will be unusable except by programs that walk directory contents very carefully.
white space - Filenames with embedded white space work with most basic system commands, but will break shell scripts that aren't prepared for them (which means most shell scripts).
I apologize for my tone, but this kind of misinformation is understandably a sore point with Lisp people. Paul Graham's On Lisp is a great book for explaining modern Lisp style, as well as dispelling Lisp myths. The dead trees version is out-of-print, but the ps/pdf versions are available for download. Also check out his Articles page.
The problem with using CL for heavy computing tasks is that Lisp is a linklist-based language.
Just curious, does all your knowledge about programming languages date from 1975, or just your prejudices about Lisp? Lisp has had arrays since about then, those arrays have the same O(1) access time as anyone else's arrays, and the performance of code using them is tuneable to FORTRAN speed or better. Whoever taught your "survey of programming languages" course did you a real disservice - maybe you should get them brought up on educational malpractice charges.
That said, FORTRAN can probabaly outrun Lisp on supercomputers, because of the effort put into parallel and vector optimizations on those platforms. I love Lisp, it's my preferred hacking environment, but I wouldn't propose it as the language of choice for big numerical applications unless there was a chance that hairy data structures might improve performance.
We both know the real problem is not little expressions like this, but multi-page nightmares that you would want editor support to check anyway. Hell, even vi has a lisp mode. Let me know when it sprouts an XML mode.
Yes, there are syntax hilighting editors... it seems odd to rely on one in a language that claims to have no syntax...
S-expressions don't have "no syntax" - they have a small, simple, regular syntax. Writing an S-expression reader/printer is an undergrad homework assignment. Writing an XML parser, on the other hand, is at least a semester's project, if not more. Here are some ot the the explicitly-stated design goals of XML:
4. It shall be easy to write programs which process XML documents.
6. XML documents should be human-legible and reasonably clear. 7. The XML design should be prepared quickly. 8. The design of XML shall be formal and concise.
All of these are affected by XML's baroque syntax, and I'd say it missed all of them, mostly because of this design goal:
...I usually kill X before I put it to sleep, just to be sure.
I DON'T have the machines boot into the GUI; it's trivial to type startx after logging in,...
That's why I love OS X. I can bring the machine with me when taking my kids to karate (say), sit down, open it, and be hacking in 5 seconds or so. When they come out, I can just close the machine, drop it in the case, walk away, and not worry about losing work. I typically live in emacs for days at a time this way.
Same for running low on battery power. You can safely run the battery to exhaustion, then go find an outlet, plug in and hit the boot button. The machine will come back up exactly where it went down, all applications still running, without missing a keystroke. No lie.
And, the white iBook is *tough*. I've dropped it off a table onto an indoor/outdoor carpeted floor. The CDROM door popped open, but no crash or other permanent damage.
These features are what Apple should be hyping, not the "megahertz myth".
... but I can't resist it. This laundry list is an instance of "all programming languages evolve toward Lisp, specifically Common Lisp. Of the ten points mentioned in the article, Common Lisp has had answers to almost all of them, ever since about 1985 or so.
10. Delete all deprecated methods, fields, classes, and interfaces.
When CL deprecated things, it did so by creating a new default top-level namespace (a "package" in CL terms) and leaving the deprecated stuff behind. The old namespace and the new one shared most symbols, so old and new stuff could coexist, but in general deprecated stuff would be flagged and had to be specifically requested for it to work.
9. Fix incorrect naming conventions.
CL is in general not case-sensitive, so it ducks the properStudlyCaps problem. CL names have always been unabbreviated except for a few historical symbols.
8. Eliminate primitive data types.
Never had 'em.
7. Extend chars to four bytes.
The exact size of the char datatype is not part of the CL specification. Implementations can and have made it bigger over the years, and programs that care can check at runtime if the implementation has big enough chars for their purposes by making a specified environmental inquiry.
6. Fix threads.
Threads aren't part of the CL language. Every major implementation has them, though, and all of them are easier on programmers than Java's threads (see here - scroll down about halfway to "You can't close over anything but final variables in an inner class!").
5. Convert file formats to XML.
Always had S-expressions as a default universal file format. See my.sig below.
4. Ditch the AWT.
CL doesn't have a specified GUI. The closest thing it has is CLIM, which had a lot of the speed and resource problems Java GUIs have. We had them ten years ago, though. The commercial CLIM implementations are quite useable these days.
3. Rationalize the collections.
CL's collections have always been rational. It mostly comes from #8 above - when objects are typed and variables are not required to be, most of your problems go away.
2. Redesign I/O.
CL's abstract interface to filesystems and pathnames has always been less Unix-centric.
1. Redesign class loading from scratch, this time with human interface factors in mind.
I gather a lot of his complaint here stems from the distinction in Java between an interface and a class. This distinction is there because of the Java designer's allergy to multiple inheritance, which CLOS has had since day one.
... which means you can set its energy-saving settings aggressively without worring about losing work or having to wait a long time for it to wake up.
I have a white iBook running OS X, which I only reboot when I install updates that require it. I have *never* had a crash due to failure to wake from sleep (which I've heard is still a problem with Linux and Windows laptops - please correct me if I'm wrong). Its wake from sleep time is typically 2 seconds, and it consumes ~7% of its battery power per day asleep.
At home I have a 466 G4 tower which is our house network gateway/router. Its hard drive and monitor are set to sleep early and often, though not the CPU. I've had trouble with it losing the cable modem connection when it sleeps completely, and dearly hope that 10.2 improves this situation.
Dean Ing wrote a short story called "Malf" in which he describes a legged vehicle designed for logging. One of the walkers gets stolen by a Mafia-linked guy who uses it to rob a bank (!) and the good guys have to chase it down with another walker.
The story isn't online anywhere, but it can be found here.
I wonder if that's possible here. Instead of fighting these ridiculous laws, is it possible to enforce them to such an extreme that the sponsors will be hoisted by their own petards?
Maybe. A big part of the problem is that lawyers think that we have the system we have because creators want it that way - that the only reason people create anything (art, patents, whatever) is so they can lock it up and make money by restricting access to it. See the Microsoft Lawyer post in this thread, and realize that it's not just MS lawyers, but basically all IP lawyers who think this way.
The only glimmer of hope I see is the possibility that legal maneuvers could be patented as business processes. If the lawyers suddenly have to worry about checking their briefs for patent infringement the same way we have to worry about checking our code, they might decide all this ownership stuff is not a good idea after all.
... is the backing of a company big enough to handle the potential legal expense, and not afraid of (or already aligned with) the media giants. They should be on at least speaking terms with the open source community, too.
Apple springs to mind. Anybody else?
www.xmlsucks.org
on
Perl & XML
·
· Score: 4, Insightful
It has been obvious ever since the first consent decree that Microsoft cannot be trusted to follow any agreement that isn't self-enforcing. The only remedy that has a chance of working is a structural one.
Our only hope is that the judge has the courage to propose a breakup, and the legal skills to write an opinion that convinces the Supreme Court (because that's the next stop from here if a breakup is decreed).
This book has several possible paths through it for different audiences. I would suggest the "Professional Programmer" path for this audience, with special attention to the "How to Find Your Way Around" chapters.
Also check out the "New to Lisp" section on the Franz Inc home page:
... and post it on apple.com. That way, if the data is any good, advocates and opponents could collectively argue about it, rather than the current "my anecdote is better than yours" stuff.
At a talk he gave at Worldcon in 1998. I can't find an online citation, so I hope he'll forgive me if I mangle his words:
Intenet censorship for China is like the old "marching Chinese" idea (if everyone in China marched 10 abreast past a given point, the line would never end because it would take more than one generation for the population to pass that point). Only now the question is, can the Chinese government shoot people faster than they can get on the Internet?
... and it's wonderful (for Microsoft, that is). Microsoft's line has always been "GPL bad and communistic, open-source (that we can steal from) OK." The article has Microsoft saying "open-source bad", period:
Microsoft's push is a new front in a long-running company assault on the open-source movement, which company officials have called "a cancer" and un-American.
Never underestimate the power of an out-of-context quotation.
Re:Design patterns and Lisp
on
Bitter Java
·
· Score: 2
A proper design pattern applies to all languages equally - because it's about design not about coding.
I believe we are actually in agreement. What I thought I was saying was that most of the stuff promoted as "patterns" is actually canonical ways of getting around limitations in Java/C++.
For example, the Visitor pattern is the canonical way of dealing with the belief/requirement in most OO langauges that writing a method that can dispatch on a published type somehow violates that type's encapsulation. In CLOS, you just write another generic function, which ends up being more efficient than the Visitor pattern (one dispatch versus two per visitation).
The claim in the presentation I referenced is that two-thirds of the patterns presented in the original Gang of Four book have this nature. They don't apply to all languages equally, because they're unnecessary in CLOS. Patterns would be more interesting to Lisp people (and more intellectually interesting in general) if they weren't primarily used to help people get things done despite using Java/C++.
Slashdot Mirror
Mostly, but a fair amount of stuff is also happening at the Darwin/BSD level. One thing I know about in particular is that BSD-style signal handling is supposed to be in 10.2, which might allow porting of CMUCL (a kick-ass free-as-in-beer Common Lisp) with minimal pain and suffering.
Close, but not quite. The admin stuff in OS X is one of the places where the NeXT people mapped Un*x to Mac conventions just right. Having admin privilege is exactly equivalent to being on the sudo list. This means you run as a normal user until (and only until you need special powers, you take those powers when you need them, and you give them up as soon as you can.
In short, on OS X you don't need a root user. Use sudo, not su.
I went with the iBook over a TiBook because:
* I'm cheap (my company is, anyway)
* the iBook is tough, and I beat the crap out of portables. I've dropped my iBook off a table onto an industrial-grade carpet over concrete. The CD-ROM drive popped open, but otherwise no damage.
I envy the TiBook's larger screen, but not much else about it. However, I don't do video or graphic-intensive games, where the G4 would make a difference.
I run XFree86 and OrborOSX when I need to run X-based stuff. It has the usual propeller-key-click hack for the second mouse button, which I rarely need since I mostly use X to read my mail (Netscape running on a remote machine). Random multi-button USB mice work with OS X - this is more of an option for desktop machines, of course. I run OS X on a duel 450 MHz G4 tower, with a two-button Kensington mouse, and the second button works just fine under OrborOSX.
I do most of my hacking using emacs-on-aqua, an emacs port based on an old NeXT port. It runs all sorts of hairy elisp packages without a burp, and lets me do development without running X.
If so, you'll like the networking, power and sleep management in OS X. Location management is excellent. You can have arbitrarily many sets of networking environments, and switching between them takes two mouse clicks, no reboot required.
My 500 MHz iBook's wake from sleep time is two seconds, counting from the time I unlatch the lid. I reboot it only after major OS updates - the last one was July 8th. I have never lost work due to a faliure to wake from sleep.
My battery life in the field is about three hours, mostly running emacs and developing code. I can stop in the middle of anything, close the lid, and walk away confident that I won't lose work. The machine will sleep for about two weeks on a full charge (I lose ~7% battery power per 24 hours).
You can even safely run the battery completely dead - OS X's last gasp is to write the complete state of the machine to the hard drive, and when you find an outlet, plug in and reboot, you come back to exactly where you left off. My kids do this with full-screen games.
Of course, by "open" I mean non-DRM. I mean, come on, no sane military would trust its computing needs to a system with secret source code, proprietary undocumented file formats, and remote-control licensing, any one of which could stop things in mid-flight.
Oh, wait...
I think you're probably talking about NoteCards,
an early hypertext system implemented in Interlisp-D. NoteCards was out in the research community at about the same time the Mac was first released (1984), and was a commercial product from Xerox in 1987.
Because it can't afford them. The computer industry's revenues are bigger than the entertainment industry's revenues by about a factor of ten. {dammit, why can't I find a citation for this when I really need one? I've seen it any number of places on the web...} Suffice it to say crippling the US computer industry to protect the US media industry will be a net loss for the US.
\000 - Putting a zero byte in a filename will break any program written in C.
'/' - A filename with an embedded slash will be unusable except by programs that walk directory contents very carefully.
white space - Filenames with embedded white space work with most basic system commands, but will break shell scripts that aren't prepared for them (which means most shell scripts).
I apologize for my tone, but this kind of misinformation is understandably a sore point with Lisp people. Paul Graham's On Lisp is a great book for explaining modern Lisp style, as well as dispelling Lisp myths. The dead trees version is out-of-print, but the ps/pdf versions are available for download. Also check out his Articles page.
Just curious, does all your knowledge about programming languages date from 1975, or just your prejudices about Lisp? Lisp has had arrays since about then, those arrays have the same O(1) access time as anyone else's arrays, and the performance of code using them is tuneable to FORTRAN speed or better. Whoever taught your "survey of programming languages" course did you a real disservice - maybe you should get them brought up on educational malpractice charges.
That said, FORTRAN can probabaly outrun Lisp on supercomputers, because of the effort put into parallel and vector optimizations on those platforms. I love Lisp, it's my preferred hacking environment, but I wouldn't propose it as the language of choice for big numerical applications unless there was a chance that hairy data structures might improve performance.
We both know the real problem is not little expressions like this, but multi-page nightmares that you would want editor support to check anyway. Hell, even vi has a lisp mode. Let me know when it sprouts an XML mode.
S-expressions don't have "no syntax" - they have a small, simple, regular syntax. Writing an S-expression reader/printer is an undergrad homework assignment. Writing an XML parser, on the other hand, is at least a semester's project, if not more. Here are some ot the the explicitly-stated design goals of XML:
All of these are affected by XML's baroque syntax, and I'd say it missed all of them, mostly because of this design goal:
Same for running low on battery power. You can safely run the battery to exhaustion, then go find an outlet, plug in and hit the boot button. The machine will come back up exactly where it went down, all applications still running, without missing a keystroke. No lie.
And, the white iBook is *tough*. I've dropped it off a table onto an indoor/outdoor carpeted floor. The CDROM door popped open, but no crash or other permanent damage.
These features are what Apple should be hyping, not the "megahertz myth".
When CL deprecated things, it did so by creating a new default top-level namespace (a "package" in CL terms) and leaving the deprecated stuff behind. The old namespace and the new one shared most symbols, so old and new stuff could coexist, but in general deprecated stuff would be flagged and had to be specifically requested for it to work.
CL is in general not case-sensitive, so it ducks the properStudlyCaps problem. CL names have always been unabbreviated except for a few historical symbols.
Never had 'em.
The exact size of the char datatype is not part of the CL specification. Implementations can and have made it bigger over the years, and programs that care can check at runtime if the implementation has big enough chars for their purposes by making a specified environmental inquiry.
Threads aren't part of the CL language. Every major implementation has them, though, and all of them are easier on programmers than Java's threads (see here - scroll down about halfway to "You can't close over anything but final variables in an inner class!").
Always had S-expressions as a default universal file format. See my
CL doesn't have a specified GUI. The closest thing it has is CLIM, which had a lot of the speed and resource problems Java GUIs have. We had them ten years ago, though. The commercial CLIM implementations are quite useable these days.
CL's collections have always been rational. It mostly comes from #8 above - when objects are typed and variables are not required to be, most of your problems go away.
CL's abstract interface to filesystems and pathnames has always been less Unix-centric.
I gather a lot of his complaint here stems from the distinction in Java between an interface and a class. This distinction is there because of the Java designer's allergy to multiple inheritance, which CLOS has had since day one.
There, have I pissed enough people off yet?
... which means you can set its energy-saving settings aggressively without worring about losing work or having to wait a long time for it to wake up.
I have a white iBook running OS X, which I only reboot when I install updates that require it. I have *never* had a crash due to failure to wake from sleep (which I've heard is still a problem with Linux and Windows laptops - please correct me if I'm wrong). Its wake from sleep time is typically 2 seconds, and it consumes ~7% of its battery power per day asleep.
At home I have a 466 G4 tower which is our house network gateway/router. Its hard drive and monitor are set to sleep early and often, though not the CPU. I've had trouble with it losing the cable modem connection when it sleeps completely, and dearly hope that 10.2 improves this situation.
Dean Ing wrote a short story called "Malf" in which he describes a legged vehicle designed for logging. One of the walkers gets stolen by a Mafia-linked guy who uses it to rob a bank (!) and the good guys have to chase it down with another walker.
The story isn't online anywhere, but it can be found here.
The only glimmer of hope I see is the possibility that legal maneuvers could be patented as business processes. If the lawyers suddenly have to worry about checking their briefs for patent infringement the same way we have to worry about checking our code, they might decide all this ownership stuff is not a good idea after all.
... is the backing of a company big enough to handle the potential legal expense, and not afraid of (or already aligned with) the media giants. They should be on at least speaking terms with the open source community, too.
Apple springs to mind. Anybody else?
For a more detailed, and more depressing, take on the above, see http://www.xmlsucks.org/but_you_have_to_use_it_any way/.
Yes, it's a PDF. Unroll it - it's worth the effort.
... because nothing short of a disaster with major associated loss of life will convince people that:
* It's not safe to use Microsoft "solutions" for anything remotely mission-critical
* The problems are caused by Microsoft's lack of attention to security
* The problems are made worse by their marketing that convinces people that Microsoft software is an acceptable substitute for thought about security
* All of this is enabled by their monopoly
* The only way to deal with the monopoly is to break it up
Note that I said "almost". If I must choose between an intact Microsoft and another WTC disaster, Microsoft can live on. But I would prefer neither.
It has been obvious ever since the first consent decree that Microsoft cannot be trusted to follow any agreement that isn't self-enforcing. The only remedy that has a chance of working is a structural one.
Our only hope is that the judge has the courage to propose a breakup, and the legal skills to write an opinion that convinces the Supreme Court (because that's the next stop from here if a breakup is decreed).
Available at:
http://psg.com/~dlamkins/sl/contents.html
This book has several possible paths through it for different audiences. I would suggest the "Professional Programmer" path for this audience, with special attention to the "How to Find Your Way Around" chapters.
Also check out the "New to Lisp" section on the Franz Inc home page:
http://www.franz.com
... and post it on apple.com. That way, if the data is any good, advocates and opponents could collectively argue about it, rather than the current "my anecdote is better than yours" stuff.
At a talk he gave at Worldcon in 1998. I can't find an online citation, so I hope he'll forgive me if I mangle his words:
Intenet censorship for China is like the old "marching Chinese" idea (if everyone in China marched 10 abreast past a given point, the line would never end because it would take more than one generation for the population to pass that point). Only now the question is, can the Chinese government shoot people faster than they can get on the Internet?
Never underestimate the power of an out-of-context quotation.
I believe we are actually in agreement. What I thought I was saying was that most of the stuff promoted as "patterns" is actually canonical ways of getting around limitations in Java/C++.
For example, the Visitor pattern is the canonical way of dealing with the belief/requirement in most OO langauges that writing a method that can dispatch on a published type somehow violates that type's encapsulation. In CLOS, you just write another generic function, which ends up being more efficient than the Visitor pattern (one dispatch versus two per visitation).
The claim in the presentation I referenced is that two-thirds of the patterns presented in the original Gang of Four book have this nature. They don't apply to all languages equally, because they're unnecessary in CLOS. Patterns would be more interesting to Lisp people (and more intellectually interesting in general) if they weren't primarily used to help people get things done despite using Java/C++.