Python 3 sucks

The third version of FIOC ( Python3000 for you cheese lovers ) is kind of lame, even lamer than classic FIOC. Minor changes suck and bigger ones weren't necessary.

PYTHONFAG

No shit.

I got 99 problems but python aint one

I'm still sticking to 2.6

>>1
Better Unicode support? nonlocal? Iterators everywhere? Some legacy crap out? print as a function?

What's bad of all that?

>>6
ONE WORD THE FORCED INDENTATION OF CODE THREAD OVER

>>6
enjoy your castrated lambda.

>>7
Didn't we have whole threads on how it is not forced, but merely strongly encouraged.

>>9
Whole threads that ended by the second post.

Better Unicode support?
LOL NO

0/10

ONE WORD THE FORCED INDENTATION OF CODE THREAD OVER ONE WORD THE FORCED INDENTATION OF CODE THREAD OVER ONE WORD THE FORCED INDENTATION OF CODE THREAD OVER ONE WORD THE FORCED INDENTATION OF CODE THREAD OVER ONE WORD THE FORCED INDENTATION OF CODE THREAD OVER ONE WORD THE FORCED INDENTATION OF CODE THREAD OVER ONE WORD THE FORCED INDENTATION OF CODE THREAD OVER ONE WORD THE FORCED INDENTATION OF CODE THREAD OVER ONE WORD THE FORCED INDENTATION OF CODE THREAD OVER ONE WORD THE FORCED INDENTATION OF CODE THREAD OVER ONE WORD THE FORCED INDENTATION OF CODE THREAD OVER ONE WORD THE FORCED INDENTATION OF CODE THREAD OVER ONE WORD THE FORCED INDENTATION OF CODE THREAD OVER ONE WORD THE FORCED INDENTATION OF CODE THREAD OVER ONE WORD THE FORCED INDENTATION OF CODE THREAD OVER ONE WORD THE FORCED INDENTATION OF CODE THREAD OVER


THE
    FORCED
        INDENTATION
    OF
CODE

☣ Please try to ignore troll threads! ☣

http://dis.4chan.org/read/prog/1231209853/10

* lambda still sucks dicks.
* ONE WORD, THE FORCED IMPORT OF REDUCE, GODDAMN GUIDO
* 0o12345. What the fuck?
* You can't open files with names that aren't valid unicode -- which means no more using Python to fix filenames after, e.g., unzipping a file that came from a system that used a different encoding -- which happens VERY OFTEN for some people, myself included. Phenomenally stupid.
* Actually after that last point, I'm not even going to continue this, because it's just not worth it. It's dead to me. Python used to be ok aside from some occasional gayness, but coding in Python 3 is like being forced to wear hot pants while dancing to YMCA.

On the subject, what other languages out there have *good* unicode support, and aren't written by retards who live in a happy-land where filesystem encodings are always perfect?

>>16
lol'd hard.

no more using Python to fix filenames after, e.g., unzipping a file that came from a system that used a different encoding
if your unzip utility is broken, you should fix it once instead of fixing the filenames everytime you unzip something.

>>18
http://datadriven.com.au/2008/12/08/zip-files-and-encoding-i-hate-you/
ergo your wrong bitch

>>19
the zip spec says only cp437 and utf-8 are supported.
any unzipping utility should do something like this:
if all filenames in archive are valid utf8
    encoding = utf8
else
    encoding = cp437

if you have a broken zip file, you should tell the person you got it from that it's broken so they can give you a non-broken one.

>>11
WTF?

>>16
lambda still sucks dicks.
True. But that doesn't make Python 3 suck any more than Python 2.

ONE WORD, THE FORCED IMPORT OF REDUCE, GODDAMN GUIDO
Guido hates functional programming. Anyways, I always import my additional FP tools, so reduce will come from there now.

0o12345. What the fuck?
That's actually a good idea. Consistency at last. 0x, 0o, 0b, I like it. 0123 for octal is error-prone and an unjustified shortcut for a seldom used numeric base.

You can't open files with names that aren't valid unicode -- which means no more using Python to fix filenames after, e.g., unzipping a file that came from a system that used a different encoding -- which happens VERY OFTEN for some people, myself included. Phenomenally stupid.
Not a massive failure of Python; just trying to support Unicode properly. Many others (like Java) exhibit this behaviour, or worse. Partly a failure of UNIX, but it's also a matter of design.

On the subject, what other languages out there have *good* unicode support, and aren't written by retards who live in a happy-land where filesystem encodings are always perfect?
NONE THAT I KNOW OF

>>16
On the subject, what other languages out there have *good* unicode support, and aren't written by retards who live in a happy-land where filesystem encodings are always perfect?
Perl, of course.

>>22
")Q#,=%(=Q/#.+\pig/disgusting(=#&?!!%T*.

>>23
VALID PERL CODE

>>22
Perl doesn't readily distinguish between "byte data" and "character data". It all looks like strings, until you screw up and try to treat one like the other, and then all hell breaks loose because obviously somewhere you didn't convert something that you should have. Which completely fucking defeats the point that having a bunch of line noise in front of every variable is supposed to accomplish in the first place, that being a notation to indicate what kind of stuff the variable is supposed to contain. Then again, Perl 6 arbitrarily messed with that concept by changing $x[1] to @x[1] so you can't even look at the beginning of an expression and say "that's a scalar" anymore, much less "that's a unicode scalar".

At least Python has type() and isinstance(), and the repr() for string and unicode types (or, unicode-string and byte-string in 3k) distinguishes the two in an obvious manner. Much like the rest of Python. Perl, again, just tries to be fucking clever about it, and it ends up making code much harder to read and debug.

>>21
lambda still sucks dicks.True. But that doesn't make Python 3 suck any more than Python 2.
No, but lambda being one of the most oft-complained-about bits of Python, I was kind of holding out hope that something would be done with it -- either allow a parenthesized lambda to wrap multiple lines, or get rid of it altogether to end the discussion on it.

I really don't see anything objectionable about something like
x = somefunc(1, 2, 3, (lambda (arg1, arg2):
    n = do_stuff(arg1)
    do_more_stuff(arg2)
    n), 'cocks', 'balls')
As long as the full lambda is parenthesized, and each following line uses the same indentation as the first.

The fact that it   contains a int   X ex X.

Lain.

Bump for still sucking.

>>26
Perl has isa and char/ord. But I don't know why I'm responding, because you're probably dead by now.

>>26
Didn't Guido say just to name the function? If he doesn't realise the entire point of lambda why would he fix it?

>>31
I'd love to see you explain what you think the point of lambda functions is.

>>32
It's a way to write functions without giving them a name, or am I missing something?

>>33
The only real point is to provide a nice, concise way to define inline functions, for which Python's lambda functions are basically as good as it gets.
The only possible objection to Python's lambda functions isn't an objection to the lambda functions themselves, but to the fact that there is a distinction between statements and expressions, which is fair enough.

>>33
It's a way to write functions as values. You can give them a name by variable assignment or binding or whatever it's called in your language.

>>35
functions were already first class in python, I was taking it as implied.

>>36
Oh, then - insert "anonymous" in front of "values".

>>33-35
You're (all) missing something. IHIHBT.

>>38
Fill us in, then, instead of pretending there are vast and mysterious hidden depths to something as basic as lambda functions.

>>39
NIKIHBT. Thanks, that was a load off my mind.