You're free to change your mind. But if you change your mind at the same time as everyone else, without any new information coming to light, you risk being labelled a sheep.
I'd say the main features a first language must have are:
It must be interpreted: mucking about with compilers takes confidence in your code;
It must have a straightforward syntax;
Input and output must be straightforward: getting the program to do things quickly is what keeps up interest
It must have newbie-friendly error reporting to avoid initial frustration
I think in all of these respects Python is as good as BASIC. Plus it has GUI IDE's which really help with polishing I/O, plus one is learning good programming habits from day one. It doesn't hurt to have the powerful features there: on the contrary it's nice when one wants them not to have to learn a new language.
Coding machine language is nice, but it's not so easy now as it was back then. One of the satifscations of learning Z80 and 6502 was that you could grasp the architecture of the processor without outside help and you knew you had mastered the whole instruction set. Not so easy with multi-level caching, instruction pipelines, stack frames coprocessor pipelines, memory management, SMP, privileged modes... have I forgotten anything? Well, the point is, really grasping a modern PC is a much bigger deal than it was 20 years ago.
It's funny how different the views expressed on this item have a rather more anti-Napster quality than those in the original interview. Is the congregation obediently abosorbing Linus' view on the matter?
The advantage of the verbatim transcript is that not many people are gping to think that this is the 16th version that finally made it through a panel of lawyers review. I'm guessing Lars wanted it this way. I don't think Slashdot are guilty of treating this interview casually.
Of course you could have said: no, no, you're mistakened! It doesn't `Smell like the Scent of Freshly Cut Grass'TM, it `Smells just like Grass that has been Freshly Cut'(TM not yet granted).
divec's original post is simply ignorant, as I pointed out in this post. Goldbach's conjecture is first-order, but it is a first order sentence of a theory which cannot be given a primitive recursive axiomatisation. A lot of people here are confused about language vs. theory and decidability vs. completeness: back to the textbooks, I think.
This is not true. Gödel's completeness theorem only applies to primitively recursive axiomatisations in first-order logic. Arithmetic, as Gödel's incompletensss theorem's show, cannot be captured by by such an axiomatisation. Goldbach's conjecture is a Pi^0_1 statement of arithmetic (a measure of the intrinsic logical complexity of a statement that includes the Gödel sentence), so as far as we know, we might not know of a formalisation capable of proving it to be true.
It's worth remembering that the point of the remedy is to stop MS from using its unique position to leverage its monopolies, not to sate our vindictiveness. Separating the OS from the applications accomplishes this goal.
Remember being a monopoly is not the harm that antitrust law sets out to correct. It is using that monopoly power in restraint of competition. I'm in favour of the original DOJ submission (split MS in two). I think the more drastic remedies would hurt current MS consumers.
Obviously this renders the crypto export restrictions in the US redundant: you can export anywhere from the US in two hops. I see three main options for US policy makers (from least likely to most likely): drop their own export restrictions, reimpose crypto restrictions or pretend it is not happening.
It's a story about fighting injustice, and the idea that the good guys might be the bandits in the forest, and that's better captured by the `rob from the rich to feed the poor' than it is by your nationalist take on it.
The suspicion is that Microsoft put the extensions in purely to break interoperability. Since interoperability is one of the key points of Kerberos, if that is so, then they are trying to derail an open standard.
If that wasn't their aim, then why haven't they tried to defend themselves on this point, instead of the lunatic `trade secret' route they chose?
Well, like the joke says, that there's a sheep in the field is your opinion. The only factual evidence you have is that there appears to be something in the field that seems from one side to be a sheep. In the light of Michael's post, I'm inclined he's in the right to say what he does about TrustE. Drop the patronising `Please learn the difference between a fact and an opinion'.
The game is $42, and that gives you both client and server software. Most of the development was done on the linux platform. There's a developer journal at Station Terminus that gives lots of this kind of information.
They've gone to a lot of effort on this game. The idea has been around for four years, it's been developed properly with a team of 12 for over two years, and the beta release was last Septemeber. It looks like they've taken their time to get things right.
This games generated quite a bit of excitement, and deservedly so. From the inteviews at the fan sites, this is an example of a game where the developers decided to go it alone after poor initial reception from publishers (eg. publishers wouldn't do a lionux port, but insisted that noone else would be allowed to do one as a precondition), and have ended up in the position of being able to dictate terms to the publishers now.
Oh, and it isn't massively multi-player, and that's a *good* thing: they've gone for allowing people to create their own servers, and for a gameplay model that seamlessly moves from single player to multiplayer, whilst retaining lots of depth. Very ambitious, and it looks really successful. I'm impressed.
There's an in-depth preview at the Adrenaline Vault, which was published last month. It gives a pretty good idea as to why folks are excited about this one.
One of the reasons cited in the article for Boo's collapse is `overly complex websites', that most customers could not read. Some meat to the `Viewable with any browser' campaign methinks...
The point is that MS *used* to be very vocal in supporting standards making (rightly) a lot of noise about Netscape's proprietary extensions to HTML. But now they're the biggest browser...
Slashdot Mirror
You're free to change your mind. But if you change your mind at the
same time as everyone else, without any new information coming to
light, you risk being labelled a sheep.
confidence in your code;
do things quickly is what keeps up interest
frustration
I think in all of these respects Python is as good as BASIC. Plus it
has GUI IDE's which really help with polishing I/O, plus one is
learning good programming habits from day one. It doesn't hurt to
have the powerful features there: on the contrary it's nice when one
wants them not to have to learn a new language.
Coding machine language is nice, but it's not so easy now as it was
back then. One of the satifscations of learning Z80 and 6502 was that
you could grasp the architecture of the processor without outside help
and you knew you had mastered the whole instruction set. Not so easy
with multi-level caching, instruction pipelines, stack frames
coprocessor pipelines, memory management, SMP, privileged
modes... have I forgotten anything? Well, the point is, really
grasping a modern PC is a much bigger deal than it was 20 years ago.
Charles
It's funny how different the views expressed on this item have a
rather more anti-Napster quality than those in the original
interview. Is the congregation obediently abosorbing Linus' view on
the matter?
The advantage of the verbatim transcript is that not many people are
gping to think that this is the 16th version that finally made it
through a panel of lawyers review. I'm guessing Lars wanted it this
way. I don't think Slashdot are guilty of treating this interview
casually.
Of course you could have said: no, no, you're mistakened! It doesn't `Smell
like the Scent of Freshly Cut Grass'TM, it `Smells just like Grass
that has been Freshly Cut'(TM not yet granted).
divec's original post is simply ignorant, as I pointed out in this
post. Goldbach's conjecture is first-order, but it is a first
order sentence of a theory which cannot be given a primitive recursive
axiomatisation. A lot of people here are confused about language
vs. theory and decidability vs. completeness: back to the textbooks, I
think.
Goldbach's conjecture is first-order. There is no quantification over proerties or sets of numbers in it.
Hilbert a gay British atheist? I think you have him confused with someone else...
This is not true. Gödel's completeness theorem only applies to
primitively recursive axiomatisations in first-order logic.
Arithmetic, as Gödel's incompletensss theorem's show, cannot be
captured by by such an axiomatisation. Goldbach's conjecture is a
Pi^0_1 statement of arithmetic (a measure of the intrinsic logical
complexity of a statement that includes the Gödel sentence), so as far
as we know, we might not know of a formalisation capable of proving it
to be true.
To reasonable people, these are reasons *not* to split up MS beyond necessary.
from using its unique position to leverage its monopolies, not to sate
our vindictiveness. Separating the OS from the applications
accomplishes this goal.
Remember being a monopoly is not the harm that antitrust law sets
out to correct. It is using that monopoly power in restraint of
competition. I'm in favour of the original DOJ submission (split MS
in two). I think the more drastic remedies would hurt current MS
consumers.
Obviously this renders the crypto export restrictions in the US
redundant: you can export anywhere from the US in two hops. I see
three main options for US policy makers (from least likely to most
likely): drop their own export restrictions, reimpose crypto
restrictions or pretend it is not happening.
It's a story about fighting injustice, and the idea that the good guys
might be the bandits in the forest, and that's better captured by the
`rob from the rich to feed the poor' than it is by your nationalist
take on it.
Ummm... Richard the Lionheart (the king to whom Robin Hood was loyal)
was actually a Norman.
The DMCA has specific provisions allowing reverse engineering; also
there are other laws ensuring the legality of reverse engineering.
interoperability. Since interoperability is one of the key points of
Kerberos, if that is so, then they are trying to derail an open
standard.
If that wasn't their aim, then why haven't they tried to defend
themselves on this point, instead of the lunatic `trade secret' route
they chose?
Well, like the joke says, that there's a sheep in the field is your
opinion. The only factual evidence you have is that there appears to
be something in the field that seems from one side to be a sheep. In
the light of Michael's post, I'm inclined he's in the right to say
what he does about TrustE. Drop the patronising `Please learn the
difference between a fact and an opinion'.
The game is $42, and that gives you both client and server software.
Most of the development was done on the linux platform. There's
a developer journal at Station Terminus that gives lots of this kind of
information.
They've gone to a lot of effort on this game. The idea has been
around for four years, it's been developed properly with a team of 12
for over two years, and the beta release was last Septemeber. It
looks like they've taken their time to get things right.
at information for this game. Here's the links...
And a couple of fan sites:
This games generated quite a bit of excitement, and deservedly so.
From the inteviews at the fan sites, this is an example of a game
where the developers decided to go it alone after poor initial
reception from publishers (eg. publishers wouldn't do a lionux port,
but insisted that noone else would be allowed to do one as a
precondition), and have ended up in the position of being able to
dictate terms to the publishers now.
Oh, and it isn't massively multi-player, and that's a *good* thing:
they've gone for allowing people to create their own servers, and for
a gameplay model that seamlessly moves from single player to
multiplayer, whilst retaining lots of depth. Very ambitious, and it
looks really successful. I'm impressed.
Wrong url...the preview is here
There's an in-depth preview at the Adrenaline
Vault, which was published last month. It gives a pretty good idea as
to why folks are excited about this one.
Counterexample to what? I don't believe they were copyrighted, I believe they were a military secret.
One of the reasons cited in the article for Boo's collapse is `overly
complex websites', that most customers could not read. Some meat to
the `Viewable with any browser' campaign methinks...
The point is that MS *used* to be very vocal in supporting standards
making (rightly) a lot of noise about Netscape's proprietary
extensions to HTML. But now they're the biggest browser...