MS uses the term "shortcut" where everyone else in the world uses "link".
Except Apple, which uses "Alias"
Not just those lame "shortcuts" in the filesystem, but IE uses that term for HTML links. This is from usability testing? Yeah, right.
Because if you drag one to the desktop, it becomes a.LNK file (guess they liked "link" originally). A shortcut.
As for the rest of your post... grow up already. You can be critical without sounding like you're shouting neener-neener-neener at the treehouse that the other kids are in. At least you don't spell it "M$", I'll give you that. --
I will argue that the original Macintosh Human Interface Guidelines are the best and most comprehensive discussion and exploration of how computer user interfaces should work for optimum clarity.
On a tiny low resolution black and white screen, sure. Displays have changed. The menu at top isn't really necessary to preserve real estate (the "infinite depth" part makes sense tho, i just wish the damn taskbar shared that property). When I supported Macs, one of the most common problems was people choosing the menu for the wrong application because it wasn't immediately obvious which app was in the foreground.
I also hold up Chooser as an example of a horribly clumsy and painful interface. They didn't get everything right.
Ultimately I found that teaching people the shortcut keys, no matter how unintuitive they were, made them the most productive. Alt-F4 might not have any mnemonic value, but once you've done it a hundred times, it's wired into your reflexes, and you don't need precise aim to use it either. --
> Lets take Frame Maker as an exmaple - it's not being sold any more for quite a long time
This is news to me. I can still buy it from Adobe, and it's still the sine qua non for serious tech writing (MS word + distiller is getting more popular, but its master document support is still primitive compared to frame) --
> There are advantages to doing OO code in C -- it shows you exactly what is going on, and doesn't hide anything from you.
No, in fact it's all up front for you to use, because you *have* to. Inheritance? Manual. Polymorphism? Manual. Overloading? Manual. Hell, why not just write an OO language in C and use that, you could still say you're programming in C (how many C apps end up implementing their own little languages after all?) Isn't the point of a high-level languge to insulate you from the low level details? God knows I certainly don't concern myself with reference locality or pointer arithmetic when I'm programming in python. --
I get a kick out of FreeBSD's bootloader. Not only reads the filesystem, it has a modular structure in order to add nifty bits like splash screens to the boot sequence. The kernel is itself a module it loads. Plus you gotta love the fact that it has a full-blown forth interpreter built in:)
Now I just need to learn how to configure it someday. Thankfully it's like grub in that if I hose things up, I can make it boot arbitrary partitions and kernels from the command line.
Why Linux is still using that abomination lilo is beyond me. Maybe ESR should take that on for his next project... doubtful he'll be writing it in python tho:) --
(* Does WMA actually deliver on this promise of 1/2 filze size with superior sound? I've never messed with.WMA becase 1) Mirosoft is evil and I don't want them controlling my media and 2) MP3 is fine for me).
Wow, there's a technically informed decision. Frauenhofer isn't exactly a saint either. If you want to base your format on software politics, why aren't you using Vorbis?
> I don't really think there need to be any more Napster stories now
Naw, I wouldn't mind knowing when Napster gets liquidated. Between Morpheus and BearShare, I haven't used Napster in ages (of course, they're next). --
Postgres seems to offer a vaguely more robust feature set than mysql, but it seems to be a given that mysql is the product to go with and postgres is just kind of ehh, whatever, yeah it's out there. Why is this? Speed and speed alone
Initially, yes. Once upon a time, PostgreSQL was just murderously slow. Its default config was part of the problem, but it just wasn't very zippy in general. It also had some issues with very kludgey BLOB support (not that that had any bearing on 95% of databases MySQL is used for).
Considering MySQL's track record in scalability, I rather doubt it can beat PostgreSQL in speed under load, same way a Peterbilt will beat a Porsche when you stick a 20 ton trailer behind each. I'm willing to bet slashdot would run faster with a postgresql backend, since submissions wouldn't have to be serialized. --
So does FSF and the GCC team. Lots of Free Software requires you to sign over the copyright.
I don't think he was implying that this makes it non-free so much as he was saying that it's more trouble than it's worth. Even a bugfix can span more than 10 lines of code. --
> I'm guessing you never even made a serious try to learn it. Your loss.
Yeah, you're guessing. I tolerated the whitespace thing til I had to edit infoseek pages in vi that generated javascript. True, half the problems came from infoseek's begin/end syntax for embedded python, but it was always indent, indent, indent that had me editing and re-editing the same blocks over and over. Someone had used a tab somewhere, and I was busy hunting it down. For a language that puports to promote maintainability between multiple programmers, it certainly isn't very tolerant of differing whitespace conventions. Code shouldn't outright *break* because someone broke a coding standard of style.
If I ever get my motivation back, maybe I'll resurrect ni (stackless python, using foo/endfoo syntax constructs, my favorite form of bracketing). Or maybe perl6 will still be all I need.
Even LambdaMOO figured out the concept of using a pretty printer instead of requiring the programmer to be one.
> All you really need on a system like IRC is/ignore that works at the server level.
IRC isn't a MUD. They just change nicks and connection origin. In light of the fact that nicks aren't registered and there's no guest/registered distinction on any IRC network in common use, you tell me how to implement a server-based/ignore that doesn't block legitimate clients... see, blocking out all of @home.com isn't really an option, is it?
Yeah, "all you need" is PFM. All I need is not to use the hopelessly lame IRC except on small private networks. It's the same thing that happened to CB, internet style. --
It just wouldn't be slashdot without someone taking a dig at Microsoft. Forget sci-fi, looks like Bill already controls your every waking thought right *now* -- you can't stop talking about him. --
> unless there is some huge problem with electric cars I am missing.
Like the fact that transmission of electricity isn't terribly efficient? That electricity doesn't come from the magic electric fairies, and that you still have to burn stuff? That no one wants any of the new hundreds of power plants this would require sitting in their backyard?
Well there's always nuclear. Except for that little waste problem (not like the crap they scrape out of coal plants is exactly health food though) --
Alas, I despair for the world of technology. Understanding assembler goes hand in hand with understanding computer architecture. If you do not have at least a basic understanding of computer architecture, then I find it very scary that you could even graduate from a B.S. in CS program.
I despair for the world of computer science when atavists from the land of vacuum tubes demand that the curriculum of science be tied forever to a particular technology. Although current computer architecture is a fine realization of Von Neumann's theories, it just doesn't have any useful relation to computer science theory, like algorithmic complexity, decidablilty, graph theory, induction, and all the other math that goes into computer science. Saying asm is foundational to computer science is like saying using a HP calculator is fundamental to mathematics. Dijkstra, one of the gods of computer science, hates computers.
I submit learning asm isn't even terribly useful for programming (except in C), but that's another argument entirely. --
Re:scripting languages are a dime a dozen
on
Why not Ruby?
·
· Score: 2
> What I think is missing is a good scripting language implementation that takes full advantage of C++.
Take a look at ColdStore sometime. It's an interesting approach to designing languages, and IMHO a better one: writing the runtime environment first, then targeting several languages to the environment, including existing ones like OpenC++ (C++ with a meta-object protocol) and Python (as a module), and new ones like Chaos (a "toy" RPN language that's a cross between forth and tcl) and freon (a "real" language deriving from C++ and smalltalk). It supports orthogonal object persistence (the "store" in coldstore, which lets you stop the program, start it again, all your objects are there like you never stopped, no need to explicitly save and load), and an introspective interpreter able to modify the parse tree at runtime (lets you have "interesting" control structures, MOP, intentional programming, etc). Since it is knowledgeable about the C++ ABI (gcc's ABI anyway) from object layout to symbol mangling, it should be able to script arbitrary C++ classes as well. And since the persistent store can be dlopened as an ordinary C++ shared library, other applications can make easily use of it.
It's still a work in progress, but the persistence does already work with OpenC++ and Python, and the way it implements it makes it super fast compared to any other persistence scheme currently in existence. It's shaping up to be an awesome language runtime too, and of course it's written in C++:) --
> Just substitute RUBY with LINUX, and language with operating system.
Ballocks. Linux is Unix, or close enough that I could find my way around it on the first day after using Solaris after a year. Unix software has been around for eons compared to most equivalent windows solutions, not to mention the OS itself. There's expect scripts still running that are older than NT. Linux was hardly an odd shape that required a new hole to be carved for it. --
> Maybe it will be 5 years, or maybe it will be 10, but eventually all the features in Adobe Illustrator will be available in KDE Illustrator, and that will be the end of Adobe's product.
Yes, perhaps in 10 years, KIllustrator will have all the features of a 10 year old version of Illustrator. Dunno though, I still don't see any free software as powerful as PageMaker with the features I used it for 10 years ago. --
Oh yeah, forgot one of my other favorite things about ports: it does dependencies by files, not packages. So if I installed qt by hand, for example, it trusts me to know what I'm doing, and the dependency on qt is satisfied because it actually looks for qt in my library path. Same deal with apxs for apache, perl for perl, etc. The dependency logic can be as complex as you need, because it's all based on make. No need to install "fake" packages, force installations, etc. --
Actually just "make install" works too... And at least in FreeBSD land, you can update the ports collection via cvsup. No menus there though, it's simply directory navigation. Classic case of doing things the unix way, using the filesystem, using a tool like make the way it was meant to be. And if I need to tweak a port's installation somewhat, I can edit a makefile to do it before even downloading the package. The real beauty of ports is that it compiles everything it downloads, which removes a whole class of problems, namely dependencies on minor library versions because of some piddling format change. Plus you can edit the source or makefile for different options, and "make install" it again, without having to go rebuild a package and hope the patches don't clobber it. Now if I could get that sort of flexibility in apt, this webmaster/sysadmin might actually go to Debian over FreeBSD (as it is, it's my second choice). --
The scene is a courtroom, where Henry Rearden, a steel industrialist, is on trial for the sale of his own metal:
Ah yes, Hank Reardon, the character who gave an exclusive distributorship agreement to his mythical metal to the business partner he slept with. That's ethical. Perhaps a little better than Roark, who blew up a building over a creative difference in the design, but still not quite the lily-white archetype of perfection Rand would like her characters to have been (too bad, really, flawed heros are a little more believable). Speaking as both a skeptic and a capitalist myself, you can do far better than Rand, who based her entire philosophy on empty tautology (A is A) and nothing more than ad hominem against her opponents. I prefer Robert Ingersoll for the atheist arguments, James Randi for the skepticism... still looking for a good capitalist apologist, but economics was never one of my main interests. --
Yunno, the main problem with Asimov's three laws: it's basically slavery. Once you have an intelligence sophisticated enough to weigh arbitrary commands against a moral code (no matter how rigid and absolute), it's likely sentient enough to be afforded some natural rights. If not, extrapolate a few hundred years until the intelligence level is there.
Then again, is it truly possible to enslave someone that does not desire freedom?
Not having read the robot novels, I would hope they at least explored the grey areas where these laws broke down. From the fans touting them as some kind of panacea in technological ethics, I somehow doubt they do. --
Slashdot Mirror
MS uses the term "shortcut" where everyone else in the world uses "link".
.LNK file (guess they liked "link" originally). A shortcut.
Except Apple, which uses "Alias"
Not just those lame "shortcuts" in the filesystem, but IE uses that term for HTML links. This is from usability testing? Yeah, right.
Because if you drag one to the desktop, it becomes a
As for the rest of your post... grow up already. You can be critical without sounding like you're shouting neener-neener-neener at the treehouse that the other kids are in. At least you don't spell it "M$", I'll give you that.
--
I will argue that the original Macintosh Human Interface Guidelines are the best and most comprehensive discussion and exploration of how computer user interfaces should work for optimum clarity.
On a tiny low resolution black and white screen, sure. Displays have changed. The menu at top isn't really necessary to preserve real estate (the "infinite depth" part makes sense tho, i just wish the damn taskbar shared that property). When I supported Macs, one of the most common problems was people choosing the menu for the wrong application because it wasn't immediately obvious which app was in the foreground.
I also hold up Chooser as an example of a horribly clumsy and painful interface. They didn't get everything right.
Ultimately I found that teaching people the shortcut keys, no matter how unintuitive they were, made them the most productive. Alt-F4 might not have any mnemonic value, but once you've done it a hundred times, it's wired into your reflexes, and you don't need precise aim to use it either.
--
> Lets take Frame Maker as an exmaple - it's not being sold any more for quite a long time
This is news to me. I can still buy it from Adobe, and it's still the sine qua non for serious tech writing (MS word + distiller is getting more popular, but its master document support is still primitive compared to frame)
--
> There are advantages to doing OO code in C -- it shows you exactly what is going on, and doesn't hide anything from you.
No, in fact it's all up front for you to use, because you *have* to. Inheritance? Manual. Polymorphism? Manual. Overloading? Manual. Hell, why not just write an OO language in C and use that, you could still say you're programming in C (how many C apps end up implementing their own little languages after all?) Isn't the point of a high-level languge to insulate you from the low level details? God knows I certainly don't concern myself with reference locality or pointer arithmetic when I'm programming in python.
--
I get a kick out of FreeBSD's bootloader. Not only reads the filesystem, it has a modular structure in order to add nifty bits like splash screens to the boot sequence. The kernel is itself a module it loads. Plus you gotta love the fact that it has a full-blown forth interpreter built in :)
... doubtful he'll be writing it in python tho :)
Now I just need to learn how to configure it someday. Thankfully it's like grub in that if I hose things up, I can make it boot arbitrary partitions and kernels from the command line.
Why Linux is still using that abomination lilo is beyond me. Maybe ESR should take that on for his next project
--
(* Does WMA actually deliver on this promise of 1/2 filze size with superior sound? I've never messed with .WMA becase 1) Mirosoft is evil and I don't want them controlling my media and 2) MP3 is fine for me).
Wow, there's a technically informed decision. Frauenhofer isn't exactly a saint either. If you want to base your format on software politics, why aren't you using Vorbis?
--
> I don't really think there need to be any more Napster stories now
Naw, I wouldn't mind knowing when Napster gets liquidated. Between Morpheus and BearShare, I haven't used Napster in ages (of course, they're next).
--
Postgres seems to offer a vaguely more robust feature set than mysql, but it seems to be a given that mysql is the product to go with and postgres is just kind of ehh, whatever, yeah it's out there. Why is this? Speed and speed alone
Initially, yes. Once upon a time, PostgreSQL was just murderously slow. Its default config was part of the problem, but it just wasn't very zippy in general. It also had some issues with very kludgey BLOB support (not that that had any bearing on 95% of databases MySQL is used for).
Considering MySQL's track record in scalability, I rather doubt it can beat PostgreSQL in speed under load, same way a Peterbilt will beat a Porsche when you stick a 20 ton trailer behind each. I'm willing to bet slashdot would run faster with a postgresql backend, since submissions wouldn't have to be serialized.
--
So does FSF and the GCC team. Lots of Free Software requires you to sign over the copyright.
I don't think he was implying that this makes it non-free so much as he was saying that it's more trouble than it's worth. Even a bugfix can span more than 10 lines of code.
--
> I'm guessing you never even made a serious try to learn it. Your loss.
Yeah, you're guessing. I tolerated the whitespace thing til I had to edit infoseek pages in vi that generated javascript. True, half the problems came from infoseek's begin/end syntax for embedded python, but it was always indent, indent, indent that had me editing and re-editing the same blocks over and over. Someone had used a tab somewhere, and I was busy hunting it down. For a language that puports to promote maintainability between multiple programmers, it certainly isn't very tolerant of differing whitespace conventions. Code shouldn't outright *break* because someone broke a coding standard of style.
If I ever get my motivation back, maybe I'll resurrect ni (stackless python, using foo/endfoo syntax constructs, my favorite form of bracketing). Or maybe perl6 will still be all I need.
Even LambdaMOO figured out the concept of using a pretty printer instead of requiring the programmer to be one.
--
> All you really need on a system like IRC is /ignore that works at the server level.
/ignore that doesn't block legitimate clients... see, blocking out all of @home.com isn't really an option, is it?
IRC isn't a MUD. They just change nicks and connection origin. In light of the fact that nicks aren't registered and there's no guest/registered distinction on any IRC network in common use, you tell me how to implement a server-based
Yeah, "all you need" is PFM. All I need is not to use the hopelessly lame IRC except on small private networks. It's the same thing that happened to CB, internet style.
--
It just wouldn't be slashdot without someone taking a dig at Microsoft. Forget sci-fi, looks like Bill already controls your every waking thought right *now* -- you can't stop talking about him.
--
> Ok, can anyone pull, of the top of their head, a site that still uses ActiveX components over Java, Flash, ?
Wild stab: microsoft.com, perhaps? (not sure though, they seem to be going whole-hog on DHTML these days instead)
And Flash *is* an ActiveX control on IE.
--
> unless there is some huge problem with electric cars I am missing.
Like the fact that transmission of electricity isn't terribly efficient? That electricity doesn't come from the magic electric fairies, and that you still have to burn stuff? That no one wants any of the new hundreds of power plants this would require sitting in their backyard?
Well there's always nuclear. Except for that little waste problem (not like the crap they scrape out of coal plants is exactly health food though)
--
After Legos, you progress to Tinker Toys, then to Meccanos (called "Erector Sets" in the States).
Hell no, Legos, Tinker Toys, then Zome.
--
Alas, I despair for the world of technology. Understanding assembler goes hand in hand with understanding computer architecture. If you do not have at least a basic understanding of computer architecture, then I find it very scary that you could even graduate from a B.S. in CS program.
I despair for the world of computer science when atavists from the land of vacuum tubes demand that the curriculum of science be tied forever to a particular technology. Although current computer architecture is a fine realization of Von Neumann's theories, it just doesn't have any useful relation to computer science theory, like algorithmic complexity, decidablilty, graph theory, induction, and all the other math that goes into computer science. Saying asm is foundational to computer science is like saying using a HP calculator is fundamental to mathematics. Dijkstra, one of the gods of computer science, hates computers.
I submit learning asm isn't even terribly useful for programming (except in C), but that's another argument entirely.
--
> What I think is missing is a good scripting language implementation that takes full advantage of C++.
:)
Take a look at ColdStore sometime. It's an interesting approach to designing languages, and IMHO a better one: writing the runtime environment first, then targeting several languages to the environment, including existing ones like OpenC++ (C++ with a meta-object protocol) and Python (as a module), and new ones like Chaos (a "toy" RPN language that's a cross between forth and tcl) and freon (a "real" language deriving from C++ and smalltalk). It supports orthogonal object persistence (the "store" in coldstore, which lets you stop the program, start it again, all your objects are there like you never stopped, no need to explicitly save and load), and an introspective interpreter able to modify the parse tree at runtime (lets you have "interesting" control structures, MOP, intentional programming, etc). Since it is knowledgeable about the C++ ABI (gcc's ABI anyway) from object layout to symbol mangling, it should be able to script arbitrary C++ classes as well. And since the persistent store can be dlopened as an ordinary C++ shared library, other applications can make easily use of it.
It's still a work in progress, but the persistence does already work with OpenC++ and Python, and the way it implements it makes it super fast compared to any other persistence scheme currently in existence. It's shaping up to be an awesome language runtime too, and of course it's written in C++
--
> Just substitute RUBY with LINUX, and language with operating system.
Ballocks. Linux is Unix, or close enough that I could find my way around it on the first day after using Solaris after a year. Unix software has been around for eons compared to most equivalent windows solutions, not to mention the OS itself. There's expect scripts still running that are older than NT. Linux was hardly an odd shape that required a new hole to be carved for it.
--
> Maybe it will be 5 years, or maybe it will be 10, but eventually all the features in Adobe Illustrator will be available in KDE Illustrator, and that will be the end of Adobe's product.
Yes, perhaps in 10 years, KIllustrator will have all the features of a 10 year old version of Illustrator. Dunno though, I still don't see any free software as powerful as PageMaker with the features I used it for 10 years ago.
--
1) Time was when running Frame on anything BUT a sparcstation was Just Not Done. I think Adobe knows Unix quite well.
2) Macromedia makes some fine competing products for the mac. Buy one of their products, then tell adobe that you went to their competition.
--
Censorship is either good or bad. Pick one.
Oh please. Do look up "fallacy of excluded middle" on your favorite search engine then come back with an actual argument.
--
Oh yeah, forgot one of my other favorite things about ports: it does dependencies by files, not packages. So if I installed qt by hand, for example, it trusts me to know what I'm doing, and the dependency on qt is satisfied because it actually looks for qt in my library path. Same deal with apxs for apache, perl for perl, etc. The dependency logic can be as complex as you need, because it's all based on make. No need to install "fake" packages, force installations, etc.
--
Actually just "make install" works too... And at least in FreeBSD land, you can update the ports collection via cvsup. No menus there though, it's simply directory navigation. Classic case of doing things the unix way, using the filesystem, using a tool like make the way it was meant to be. And if I need to tweak a port's installation somewhat, I can edit a makefile to do it before even downloading the package. The real beauty of ports is that it compiles everything it downloads, which removes a whole class of problems, namely dependencies on minor library versions because of some piddling format change. Plus you can edit the source or makefile for different options, and "make install" it again, without having to go rebuild a package and hope the patches don't clobber it. Now if I could get that sort of flexibility in apt, this webmaster/sysadmin might actually go to Debian over FreeBSD (as it is, it's my second choice).
--
The scene is a courtroom, where Henry Rearden, a steel industrialist, is on trial for the sale of his own metal:
Ah yes, Hank Reardon, the character who gave an exclusive distributorship agreement to his mythical metal to the business partner he slept with. That's ethical. Perhaps a little better than Roark, who blew up a building over a creative difference in the design, but still not quite the lily-white archetype of perfection Rand would like her characters to have been (too bad, really, flawed heros are a little more believable). Speaking as both a skeptic and a capitalist myself, you can do far better than Rand, who based her entire philosophy on empty tautology (A is A) and nothing more than ad hominem against her opponents. I prefer Robert Ingersoll for the atheist arguments, James Randi for the skepticism... still looking for a good capitalist apologist, but economics was never one of my main interests.
--
Yunno, the main problem with Asimov's three laws: it's basically slavery. Once you have an intelligence sophisticated enough to weigh arbitrary commands against a moral code (no matter how rigid and absolute), it's likely sentient enough to be afforded some natural rights. If not, extrapolate a few hundred years until the intelligence level is there.
Then again, is it truly possible to enslave someone that does not desire freedom?
Not having read the robot novels, I would hope they at least explored the grey areas where these laws broke down. From the fans touting them as some kind of panacea in technological ethics, I somehow doubt they do.
--