repeat
If market_is_mess then
begin
writeln('Market needs regulation');
Create_Institution
end;
if assigned(institution) and (institution=bureacratic) then//(second clause is always true given enough time
begin
Writeln('Cry wolf over institution, say market can regulate itself');
Writeln(stderr,'Communism failed');
Kill_institution;
end; until hell=frozen;
Classic Mac OS goes a long way. Application folders can be dragged and dropped, and dragging the OS folder to a newly formatted (external) HD effectively clones the system.
I didn't use the OS that much, so there could be downsides, but it was quite impressive
Problem is that reviews much be "hot", iow on the site quite quick after the release to be newsworthy.
A good review would cover quite a while, and e.g. describe how the distribution would hold up the first upgrade round or so, or how many times it was routed.
This is often where the non commericial *nix distributions (Debian, FreeBSD) shine.
>One language is not generically faster than another language
Of course it is. GC comes at a price, abstract datatypes come at a price, since they usually require runtime helpers to figure them out.
Of course compilers can soften this a bit for built-ins, and even a bit more for datatypes based on generic programming (templates, generics), but the fact remains.
>Leaving aside that you are now comparing a set of compilers to a language
A language is as fast as the fastest compiler I can acquire and use for my problem.
Yes, otherwise he has to ensure there is the right version Java machine correctly configured.
I take my changes with a P-IV optimized version that runs on anything till 486.
Since the majority of the cycles will be brought in by the P-IV and Athlon series, there is no problem, since Athlon runs non SSE P-IV code quite decently
And VS is not targeted at numeric work, but should be able to outrun Java.
Of course, if your work is one or two loops inside one method, or one method with e.g. possible tailrecursion, language doesn't matter much, only compiler optimization. I think that happens to your mandelbrot, but that is not typical for scientific calculation.
I don't know Eiffel that well, and can't judge the fitness of the language itself, but:
- Compiles to language X != As fast as X. Runtime helpers, higher level constructs etc. Eifel might be fast, but compiling to a language considered fast doesn't prove that. - The language is relatively unknown. This was another advantage of C/C++ and things like Java and Delphi: everybody knows it.
I'm not going to learn a language for Gnome, one for KDE etc etc.
All HDs have spare sectors to use when during life sectors die. (though I didn't know it were that much, I'd guess 5-10%). I assume they managed to format them.
Note that all drives have become about 1 1/2 the size, except the ultrasuperduper new one.
That is apparntly a current topmodel, and those expensive models with borderline technology are apparantly more likely to loose sectors (either technology, or the higher density itself is fragile)
Well, there is a good reason to avoid 4.4, if it really touches commonly used client libs. It is a hassle and sets a precedent.
I don't want to administer an OSS project that has to make dozens of these lines to each every stupid tool, and check them for each release or import.
It's like those freeware postcard licenses, the originators aren't evil, but in practice it simply puts too much of a break on usability the of the projects source.
If you knew something about databases, you would answer that yourself.
All databases have their niche. mysql can serve read-only data pretty fast. typical use: dumb website backend
postgresql is a bit better all round, while keeping some performance when it gets simple.
firebird (or oracle) can handle really complex loads with complex transactions decently.
Something like the domains is a specialistic task, and though I doubt they pick the domains database on a pure ulitarian basis, easy to use application connenctions and complex transactions are not really an item there:_)
Nobody are the amateurs, and the Slashdot wannabee ha4ors. They also don't get that this is the only commercial grade db system available for free. (and no, mysql doesn't live up to that).
Professionals would already know Interbase (Firebirds previous name). It's the main database technology behind Delphi, which is still the best selling bulk development tool after Microsoft's stuff. And the only major commercial development tool with a Linux version.
Slashdot Mirror
You can teach this to an infant in a few days
repeat
If market_is_mess then
begin
writeln('Market needs regulation');
Create_Institution
end;
if assigned(institution) and (institution=bureacratic) then
begin
Writeln('Cry wolf over institution, say market can regulate itself');
Writeln(stderr,'Communism failed');
Kill_institution;
end;
until hell=frozen;
Classic Mac OS goes a long way. Application folders can be dragged and dropped, and dragging the OS folder to a newly formatted (external) HD effectively clones the system.
I didn't use the OS that much, so there could be downsides, but it was quite impressive
Isn't one supposed to research demands first before
introducing new technologies?
Or did Microsoft buy the judge and the rest of the justice system too?
Problem is that reviews much be "hot", iow on the site quite quick after the release to be newsworthy.
A good review would cover quite a while, and e.g. describe how the distribution would hold up the first upgrade round or so, or how many times it was routed.
This is often where the non commericial *nix distributions (Debian, FreeBSD) shine.
How easily laymans get fooled and scared
There is a very nice substance, called chlorine, while maybe not a full steriliser, it is cheap, available in nearly any house, and cheap.
Some heavy chlorine in bucket, curtain in bucket, let it stand for a while, presto
>One language is not generically faster than another language
Of course it is. GC comes at a price, abstract datatypes come at a price, since they usually require runtime helpers to figure them out.
Of course compilers can soften this a bit for built-ins, and even a bit more for datatypes based on generic programming (templates, generics), but
the fact remains.
>Leaving aside that you are now comparing a set of compilers to a language
A language is as fast as the fastest compiler I can acquire and use for my problem.
HotSpot is great. But that means it is able to get somewhat closer to normal compiled languages than the old bytecode interpreter with its GC storms,.
Not that it is suddenly the apex of performance (except compared to the old bytecode interpreter)
Yes, otherwise he has to ensure there is the right version Java machine correctly configured.
I take my changes with a P-IV optimized version that runs on anything till 486.
Since the majority of the cycles will be brought in by the P-IV and Athlon series, there is no problem, since Athlon runs non SSE P-IV code quite decently
Sure, if take C's worst case, and HotSpot's best case, then of course I can make HotSpot look good.
In practice however, C/C++ code runs circles around nearly all Java code. Period
C++ should be able to run circles around Java.
And VS is not targeted at numeric work, but should
be able to outrun Java.
Of course, if your work is one or two loops inside one method, or one method with e.g. possible tailrecursion, language doesn't matter much, only compiler optimization. I think that happens to your mandelbrot, but that is not typical for scientific calculation.
Wonder why this got modded up Funny.
:-)
It is simply true, and so obvious it struck me immediately too.
Java for intense computation? Months? Go away
Nothing against Java, but it isn't suitable for this.
I don't know Eiffel that well, and can't judge the fitness of the language itself, but:
- Compiles to language X != As fast as X. Runtime helpers, higher level constructs etc. Eifel might be fast, but compiling to a language considered fast doesn't prove that.
- The language is relatively unknown. This was another advantage of C/C++ and things like Java and Delphi: everybody knows it.
I'm not going to learn a language for Gnome, one for KDE etc etc.
The FreeBSD project builds and ships ISOs. Then they are responsable for license problems in them
And when in doubt, read Orwell's 1984 for more pointers.
All HDs have spare sectors to use when during life sectors die. (though I didn't know it were that much, I'd guess 5-10%). I assume they managed to format them.
Note that all drives have become about 1 1/2 the size, except the ultrasuperduper new one.
That is apparntly a current topmodel, and those expensive models with borderline technology are apparantly more likely to loose sectors (either technology, or the higher density itself is fragile)
Well, there is a good reason to avoid 4.4, if it really touches commonly used client libs. It is a hassle and sets a precedent.
I don't want to administer an OSS project that has to make dozens of these lines to each every stupid tool, and check them for each release or import.
It's like those freeware postcard licenses, the originators aren't evil, but in practice it simply puts too much of a break on usability the of the projects source.
I saw several comments on the freebsd-ports list that the FreeBSD troops see no problem in the adoption either.
The reason was also the same, clientside libs seem to go free.
Agree,
and when not using an open source compiler: binary project files.
> if someone installs your work from disc 3 of some > Linux distro, they couldn't care less who you are.
The experience got me a job. It made me stood out from my classmates that lifted their way through college, copying the assignments etc.
I got an _interesting_ job easier, and I got my first promotion a lot easier.
Many of the others ones are doing helpdesk/smalltime sysadmining now, with their CS bachelor.
I'm not, and while my job pays slightly better (20-30%), it is much more varied and interesting
It's like saying a professional caterer will loose his job if he cooks once a week at the Salvation army.
stupid.
If you knew something about databases, you would answer that yourself.
All databases have their niche. mysql can serve read-only data pretty fast. typical use: dumb website backend
postgresql is a bit better all round, while keeping some performance when it gets simple.
firebird (or oracle) can handle really complex loads with complex transactions decently.
Something like the domains is a specialistic task,
and though I doubt they pick the domains database on a pure ulitarian basis, easy to use application connenctions and complex transactions are not really an item there
Nobody are the amateurs, and the Slashdot wannabee ha4ors. They also don't get that this is the only commercial grade db system available for free. (and no, mysql doesn't live up to that).
Professionals would already know Interbase (Firebirds previous name). It's the main database technology behind Delphi, which is still the best selling bulk development tool after Microsoft's stuff. And the only major commercial development tool with a Linux version.
I can do all that with ports (as local user) by adding some vars.
I can postpone to the actual install as user too. (but have to store my zipfiles outside the common archive)