As usual, there's more than one candidate. Both d-i and PGI are in development, and sarge will release with whatever works at the time. Even if that is neither of them.
Unstable has never traditionally been "leading edge", although in more recent years some packages have become so; this is largely the fault of newer developers who forget the adage:
You upload stable software to unstable.
Unstable is where the next release is constructed. No more or less than that.
Minix may be simple and small... but that doesn't make up for the fragile code and unhelpful design of the system. To do anything practical you have to modify most of the large tasks; this is the antithesis of modern modular design.
But certainly "awesome". I'm filled with awe if it even compiles.
I'd say that microkernels have finally proven themselves to be ineffective at the job of being the basis of serious operating systems (e.g. BeOS, QNX, ChorusOS, OS/2 Merlin, Hurd, MacOSX)
Having spent a few weeks adding some code to minix as part of a course in operating systems design, I can safely say that it is a sterling example of how not to design an operating system.
The code ranges from messy to obscure; it is difficult to maintain, and rather fragile.
Where's the point of teaching students something that isn't relevant to the real world?
boot-floppies, not "boot floppies". It's the name of the debian installation system. CDs, network boot images, etc. are all variants on the boot-floppies system (at present).
Care to be a little more specific? OSS projects tend to be *more* responsive, rather than less, afaict, since you have pretty direct access to the developers.
Of course, it has to be a *good* idea. If it's not a dramatic one, then the project leader may not be willing to spend time working on it. This doesn't mean they won't include it if you send them a patch, just that they have other priorities.
Read the first article in the series, JonKatz decided to do this for a reason - he originally wanted (and still wants) to publish it, but people objected to the reuse of slashdot posts without permission, which would be hard to get.
It seems to me like Microsoft really pissed him off during the trial.
I'm not surprised - wasn't he really close to declaring them in contempt at one point? And do I recall correctly something about a video which was more or less falsified?
A testament to how fscking smart Don Knuth really, is I suppose.
More like a testament to how stubborn he is. If you batter away at that brick wall with your head for long enough, it'll give.;)
/dev/hda1 is a file. Think about it. Instant encrypted filesystem? Certainly less overheads than conventional loopback. Anyone know if this would work?
Actually, it does. Although I never used them until I had learned C, and then had to go back and do a little dabbling in BASIC. Remember that nasty , obscure ! operator? (There is at least one other, but this was about 6 or 7 years ago - I can't remember exactly) It takes an integer argument (can be a variable) and behaves as if it was the memory address given by that argument - that's a pointer. It's called "peek" and "poke", but that doesn't stop it being pointers by another name.
"Real programmers can write assembler/Fortran code in any language"
C++ is converted into assembler by gcc, then assembled
Compile your OO program with gcc, then work back from the assembly into the language of your choice. I suggest Logo as a possibly interesting choice. Or COBOL. Or Fortran itself. Or just decompile back to C.
In short, you aren't open sourcing the *GAME* per se, but the *ENGINE* that drives the game. Let's face it - you don't really want other people monkeying with your data files, maps, etc. Yes, they can copy them, but it's not legal, because those are components you pay for, and the license is written appropriately.
No, this is no barrier to the clients, but it does prevent another corporation from setting up a competing server. They can write another game based on the same engine. So can you - the real question then is, who can write the better game?
Actually being British, I can safely say that Americans are generally thought of as being loud, irritating, obsessed with themselves and their country, assume that America is the only place on the planet where anything useful happens, and none too bright. (It's an opinion - I make no suggestions as to it's accuracy)
Well, it looks like Blizzard changed their position on lost items - guess they lied when they said they couldn't.
Hey, I wonder if they'll give me back that +2 to all skills, mana leech, life leech, fast cast, +20 str, +20 dex amulet that I, er, "lost" when I got lagged out?
Try doing that for a few hundred people, with uncooperative ISPs - and besides, what they really want to do is ban the users permanently, and that's one thing dynamic IP defeats.
Slashdot Mirror
As usual, there's more than one candidate. Both d-i and PGI are in development, and sarge will release with whatever works at the time. Even if that is neither of them.
Nope. No, I don't have to agree to that.
Minix may be simple and small... but that doesn't make up for the fragile code and unhelpful design of the system. To do anything practical you have to modify most of the large tasks; this is the antithesis of modern modular design.
But certainly "awesome". I'm filled with awe if it even compiles.
I'd say that microkernels have finally proven themselves to be ineffective at the job of being the basis of serious operating systems (e.g. BeOS, QNX, ChorusOS, OS/2 Merlin, Hurd, MacOSX)
Having spent a few weeks adding some code to minix as part of a course in operating systems design, I can safely say that it is a sterling example of how not to design an operating system.
The code ranges from messy to obscure; it is difficult to maintain, and rather fragile.
Where's the point of teaching students something that isn't relevant to the real world?
Joey Hess implemented this for Debian in about 400 lines of perl, some time ago. Hardly a "major" feature.
It reminds me more of the little old lady who swallowed a fly...
Funny how Debian users attack RPM on technical grounds, and RPM users attack Debian _users_ on ad hominem grounds...
Maybe this says something.
boot-floppies, not "boot floppies". It's the name of the debian installation system. CDs, network boot images, etc. are all variants on the boot-floppies system (at present).
The mode changes are possibly redundant, but they allow users a bit more flexibility. I don't see that as a particular problem.
Chanserv only de-ops users when it is broken. This is hopefully fixed.
ident is not required, although the rules on what usernames are valid are a bit stricter.
Uhh, no it wasn't. The base system was frozen, not the entire release. Have you been reading slashdot again?
The Earth was commissioned and run by mice.
Mice have 3 toes on each foot, plus a tail, total 13.
Since we count in base 10 because that's how many fingers we have, it stands to reason that mice count in base 13.
In base 13, 6 * 9 == 42
Care to be a little more specific? OSS projects tend to be *more* responsive, rather than less, afaict, since you have pretty direct access to the developers. Of course, it has to be a *good* idea. If it's not a dramatic one, then the project leader may not be willing to spend time working on it. This doesn't mean they won't include it if you send them a patch, just that they have other priorities.
Read the first article in the series, JonKatz decided to do this for a reason - he originally wanted (and still wants) to publish it, but people objected to the reuse of slashdot posts without permission, which would be hard to get.
I'm not surprised - wasn't he really close to declaring them in contempt at one point? And do I recall correctly something about a video which was more or less falsified?
A testament to how fscking smart Don Knuth really, is I suppose. More like a testament to how stubborn he is. If you batter away at that brick wall with your head for long enough, it'll give. ;)
Actually, it does. Although I never used them until I had learned C, and then had to go back and do a little dabbling in BASIC. Remember that nasty , obscure ! operator? (There is at least one other, but this was about 6 or 7 years ago - I can't remember exactly) It takes an integer argument (can be a variable) and behaves as if it was the memory address given by that argument - that's a pointer. It's called "peek" and "poke", but that doesn't stop it being pointers by another name.
C++ is converted into assembler by gcc, then assembled
Compile your OO program with gcc, then work back from the assembly into the language of your choice. I suggest Logo as a possibly interesting choice. Or COBOL. Or Fortran itself. Or just decompile back to C.
In short, you aren't open sourcing the *GAME* per se, but the *ENGINE* that drives the game. Let's face it - you don't really want other people monkeying with your data files, maps, etc. Yes, they can copy them, but it's not legal, because those are components you pay for, and the license is written appropriately.
No, this is no barrier to the clients, but it does prevent another corporation from setting up a competing server. They can write another game based on the same engine. So can you - the real question then is, who can write the better game?
I believe, a little under 4 times longer back than America has existed for (don't quote me on the math).
Bah! We never wanted them anyway... ;)
Actually being British, I can safely say that Americans are generally thought of as being loud, irritating, obsessed with themselves and their country, assume that America is the only place on the planet where anything useful happens, and none too bright. (It's an opinion - I make no suggestions as to it's accuracy)
Hey, I wonder if they'll give me back that +2 to all skills, mana leech, life leech, fast cast, +20 str, +20 dex amulet that I, er, "lost" when I got lagged out?
Try doing that for a few hundred people, with uncooperative ISPs - and besides, what they really want to do is ban the users permanently, and that's one thing dynamic IP defeats.