This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.
Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 »
526
Proposals / Re: Tierable Systems
« on: March 29, 2010, 12:52:31 pm »
Virtual accessor functions?
527
Proposals / Re: Tierable Systems
« on: March 29, 2010, 09:07:45 am »
How about multiple inheritance or a component system so you could combine them in more ways? Collisions, graphics, both or neither?
528
Proposals / Re: execute_string via Google V8
« on: March 28, 2010, 08:50:16 pm »
@retep: so you just like being macho?
Type declarations in Haskell are optional. You can give them, but they're not required. It works the same either way. Letting the compiler do things doesn't have anything to do with performance. Like I said, dependent typing can actually make things faster. Haskell's types are just as predictable as C++'s. They're just more specific and powerful. In fact, dependent typing lets you, the programmer, specify more.
Although I do get why you say it's beautiful. I think that way about C. I just think C++ is kind of in between C-like beauty and Haskell-like beauty.
Type declarations in Haskell are optional. You can give them, but they're not required. It works the same either way. Letting the compiler do things doesn't have anything to do with performance. Like I said, dependent typing can actually make things faster. Haskell's types are just as predictable as C++'s. They're just more specific and powerful. In fact, dependent typing lets you, the programmer, specify more.
Although I do get why you say it's beautiful. I think that way about C. I just think C++ is kind of in between C-like beauty and Haskell-like beauty.
529
Proposals / Re: execute_string via Google V8
« on: March 28, 2010, 07:41:33 pm »
It's got nothing to do with functional languages or amazing implicit list types. It's got to do with seeing other type systems. You can't assume C++'s types are perfect when your only point of reference is GM's.
C++ has verbose typing. You have to specify the type of every single variable even when you're initializing it in the same place. The compiler is smart enough to figure out the type of an expression, so it makes no sense to require that.
Templates are a sad excuse for parametric polymorphism. Haskell's type system, with types like [a] (list of any type), Eq a (any type for which equality can be checked) and Show a (any type which can be converted to a string) are much more powerful. Even Java's generics are better than C++ templates- they at least can limit parameters to types with the operators and functions the template uses.
Dependent typing is extremely useful. It can find, at compile time, possible bad array indexes and null pointers, as well as other bug conditions that usually require hope or runtime checks. In many situations there is absolutely no reason not to use it other than its absence from a language.
C++ has verbose typing. You have to specify the type of every single variable even when you're initializing it in the same place. The compiler is smart enough to figure out the type of an expression, so it makes no sense to require that.
Templates are a sad excuse for parametric polymorphism. Haskell's type system, with types like [a] (list of any type), Eq a (any type for which equality can be checked) and Show a (any type which can be converted to a string) are much more powerful. Even Java's generics are better than C++ templates- they at least can limit parameters to types with the operators and functions the template uses.
Dependent typing is extremely useful. It can find, at compile time, possible bad array indexes and null pointers, as well as other bug conditions that usually require hope or runtime checks. In many situations there is absolutely no reason not to use it other than its absence from a language.
530
Proposals / Re: execute_string via Google V8
« on: March 28, 2010, 06:36:17 pm »
That's because you've only ever seen C++'s types. They may be beautiful in comparison to GM's, but...
533
Just because we are free relative to the peepz in europe,Woah, woah, woah. I thought:
does not mean we are free in absolute terms.
There is no such thing as partial freedom.
Either you are free,
or you are not free.
537
Announcements / Re: Collisions
« on: March 26, 2010, 06:34:28 pm »
How about collision_polygon? It'd be just as easy, and you could give sprites polygonal masks rather than just squares.
538
Just thought this was interesting:
Here is cat from first edition unix: http://code.google.com/p/unix-jun72/source/browse/trunk/src/cmd/cat.s. Very short bit of assembly that I don't fully understand. What I do looks fine, it's just cat.
Here is cat from Plan 9: http://plan9.bell-labs.com/sources/plan9/sys/src/cmd/cat.c. Extremely simple C program, copies either stdin or a list of files to standard output. Can't compare it to original cat, but if it does any more it's just the list of files.
Now, here's GNU cat: http://git.savannah.gnu.org/gitweb/?p=coreutils.git;a=blob;f=src/cat.c;hb=HEAD. HOLY LONG CODE. This is cat, for crying out loud!
Here is cat from first edition unix: http://code.google.com/p/unix-jun72/source/browse/trunk/src/cmd/cat.s. Very short bit of assembly that I don't fully understand. What I do looks fine, it's just cat.
Here is cat from Plan 9: http://plan9.bell-labs.com/sources/plan9/sys/src/cmd/cat.c. Extremely simple C program, copies either stdin or a list of files to standard output. Can't compare it to original cat, but if it does any more it's just the list of files.
Now, here's GNU cat: http://git.savannah.gnu.org/gitweb/?p=coreutils.git;a=blob;f=src/cat.c;hb=HEAD. HOLY LONG CODE. This is cat, for crying out loud!
539
Announcements / Re: Another quickie
« on: March 26, 2010, 01:56:21 pm »
I was used to the old layout, but it was harder to see the more advanced features. It took me forever to realize the difference between font and paragraph formatting, while in 07 they're separated for you. It would be nice if they wouldn't make contextual controls invisible, but otherwise it's very easy to find things.
540
Announcements / Re: Another quickie
« on: March 26, 2010, 09:17:32 am »
Ribbon in Word is actually much easier to find things with. An extremely common experience: "Oh, that's new! ... Wait, that was in 03?"
On h.264: There would be a proprietary plugin of sorts, but it would be much better than the old WM plugin. You can use an existing codec and keep the rest of the tag in the browser. Of course it would be a pain on Linux where you don't have the codec, but Linux is a very small percentage of end users and I think people that care enough to try Linux care enough to download a codec.
Theora may improve, but there's still the submarine patent problem, same as gif. It may even get worse as they improve it. With h.264, you know who owns the patents and they seem inclined to let people use the codec. Of course that's not an ideal situation, but it's better than Theora. The least Firefox can do is support h.264 externally for now so that <video> survives. As it is, they're completely splitting the market and basically killing the tag.
Hopefully if MPEG-LA starts charging distributors for h.264 there will be a better codec available. I believe that's the only way to get people to switch- have a better codec available and then make h.264 cost to distribute.
On h.264: There would be a proprietary plugin of sorts, but it would be much better than the old WM plugin. You can use an existing codec and keep the rest of the tag in the browser. Of course it would be a pain on Linux where you don't have the codec, but Linux is a very small percentage of end users and I think people that care enough to try Linux care enough to download a codec.
Theora may improve, but there's still the submarine patent problem, same as gif. It may even get worse as they improve it. With h.264, you know who owns the patents and they seem inclined to let people use the codec. Of course that's not an ideal situation, but it's better than Theora. The least Firefox can do is support h.264 externally for now so that <video> survives. As it is, they're completely splitting the market and basically killing the tag.
Hopefully if MPEG-LA starts charging distributors for h.264 there will be a better codec available. I believe that's the only way to get people to switch- have a better codec available and then make h.264 cost to distribute.