C++

Strings. They are not interoperable with platform strings, so you end up using std::vector half of the time. The copy policy (copy on write or deep copy) is not defined, so performance guarantees can not be given for straightforward syntax. Sometimes they rely on STL algorithms that are not very intuitive to use. Too many libraries roll their own which are unfortunately much more comfortable to use. Unless you have to combine them. Variety of string representations Now, this is a little bit of a platform problem - but I still hope it would have been better when a less obstinate standard string class would have been available earlier. The following string representations I use frequently: generic LPCTSTR, LPC(W)STR allocated by CoTaskMemAlloc, BSTR, _bstr _t (w)string, CString, std::vector a roll-my-own class (sigh) that adds range checking and basic operations to a (w)char * buffer of known length Build model. I am sick to death of all the time spent muddling around with who-includes-what, forward declarations, optimizing precompiled headers and includes to keep at least incremental build times bearable, etc. It was great in the eighties, but now? There are so many hurdles to packing up a piece of code so it can be reused that even moms dog gets bored listening to me. Hard to parse This makes external tools especially hard to write, and get right. And today, we C++ guys are lacking mostly in the tool chain. I love my C# reflection and delegates but I can live without them. Without great refactoring, I can't. Threading is too hard Language doesn't even recognize it (by now), and the freedoms of the compiler - while great - are to painful. Static and on-demand initialization Technically, I cheat here: this is another puzzle piece in the "wrap up code for reuse": It's a nightmare to get something initialized only when it is needed. The best solution to all other redist problems is throwing everything into headers, this problem says "neeener - you cannot".

诚然，其中许多内容超出了严格的语言范围，但在我看来，整个工具链都需要进行判断和发展。

Perl

Mixed use of sigils my @array = ( 1, 2, 3 ); my $array = [ 4, 5, 6 ]; my $one = $array[0]; # not @array[0], you would get the length instead my $four = $array->[0]; # definitely not $array[0] my( $two, $three ) = @array[1,2]; my( $five, $six ) = @$array[1,2]; # coerce to array first my $length_a = @array; my $length_s = @$array; my $ref_a = \@array; my $ref_s = $array; For example none of these are the same: $array[0] # First element of @array @array[0] # Slice of only the First element of @array %array[0] # Syntax error $array->[0] # First element of an array referenced by $array @array->[0] # Deprecated first element of @array %array->[0] # Invalid reference $array{0} # Element of %array referenced by string '0' @array{0} # Slice of only one element of %array referenced by string '0' %array{0} # Syntax error $array->{0} # Element of a hash referenced by $array @array->{0} # Invalid reference %array->{0} # Deprecated Element of %array referenced by string '0' In Perl6 it is written: my @array = ( 1, 2, 3 ); my $array = [ 4, 5, 6 ]; my $one = @array[0]; my $four = $array[0]; # $array.[0] my( $two, $three ) = @array[1,2]; my( $five, $six ) = $array[1,2]; my $length_a = @array.length; my $length_s = $array.length; my $ref_a = @array; my $ref_s = $array; Lack of true OO package my_object; # fake constructor sub new{ bless {}, $_[0] } # fake properties/attributes sub var_a{ my $self = shift @_; $self->{'var_a'} = $_[0] if @_; $self->{'var_a'} } In Perl6 it is written: class Dog is Mammal { has $.name = "fido"; has $.tail is rw; has @.legs; has $!brain; method doit ($a, $b, $c) { ... } ... } Poorly designed regex features /(?=regexp)/; # look ahead /(?<=fixed-regexp)/; # look behind /(?!regexp)/; # negative look ahead /(?<!fixed-regexp)/; # negative look behind /(?>regexp)/; # independent sub expression /(capture)/; # simple capture /(?:don't capture)/; # non-capturing group /(?<name>regexp)/; # named capture /[A-Z]/; # character class /[^A-Z]/; # inverted character class # '-' would have to be the first or last element in # the character class to include it in the match # without escaping it /(?(condition)yes-regexp)/; /(?(condition)yes-regexp|no-regexp)/; /\b\s*\b/; # almost matches Perl6's <ws> /(?{ print "hi\n" })/; # run perl code In Perl6 it is written: / <?before pattern> /; # lookahead / <?after pattern> /; # lookbehind / regexp :: pattern /; # backtracking control / ( capture ) /; # simple capture / $<name>=[ regexp ] /; # named capture / [ don't capture ] /; # non-capturing group / <[A..Z]> /; # character class / <-[A..Z]> /; # inverted character class # you don't generally use '.' in a character class anyway / <ws> /; # Smart whitespace match / { say 'hi' } /; # run perl code Lack of multiple dispatch sub f( int $i ){ ... } # err sub f( float $i ){ ... } # err sub f($){ ... } # occasionally useful In Perl6 it is written: multi sub f( int $i ){ ... } multi sub f( num $i ){ ... } multi sub f( $i where $i == 0 ){ ... } multi sub f( $i ){ ... } # everything else Poor Operator overloading package my_object; use overload '+' => \&add, ... ; In Perl6 it is written: multi sub infix:<+> (Us $us, Them $them) | (Them $them, Us $us) { ... }

Ruby:

太慢了 自负的群体 这不是闲谈 在nil上调用方法时出错，而不是返回nil à la Objective C 非线程

Python。

虽然前面提到了python处理作用域的奇怪方式，但我觉得最糟糕的结果是:

import random

def myFunction():

    if random.choice(True, False):
        myString = "blah blah blah"

    print myString

也就是说，if块内部的作用域与函数的其余部分相同，这意味着变量声明可以出现在条件分支内部，并且可以在条件分支外部访问。大多数语言要么阻止你这样做，要么至少为你提供某种严格的模式。

此函数有时会成功，但有时会抛出异常。虽然这是一个人为的例子，但这可能会导致一些微妙的问题。

JavaScript

Function object syntax: f = new Function( "foo", "bar", "return foo+bar;" ); (It takes n arguments, the first n-1 are arguments for the function, then nth is the actual function, in string form. Which is just silly.) Function arguments can be repeated. f = new Function( "foo", "foo", "return foo;" ); The last repetition is the only one ever used, though: f( "bye", "hi" ) // returns "hi" f( "hi" ) // returns undefined E4X should just die. My users are always complaining that it doesn't work the way they think it will. Let's face it, when you need a page and a half of psuedocode for a setter, it's time to rethink things. A standard notion of stdin/stdout/stderr (and files!) would be nice. null != undefined It's irritating to have to handle them both. Sometimes it's useful, but most languages manage to limp along fine with one.

Lua:

I understand the reasons, but seriously. Variables should be local by default, with a global keyword, not vice versa. I'm in general not a huge fan of the do/end style semantics. I much prefer C-style braces. Dynamic typing. I know, some of you go "Huh?" but I've been entirely spoiled by knowing exactly what type of data will be in a given variable. Constant if (type(var) == "string") then stuff() end is a pain. Variables need not be defined before they're used. I would much rather be explicit about what I'm trying to do than risk a typo causing what I like to call "wacky beans".

PHP:

同样，动态类型。 缺少闭包。你可以用$function($arg);但这不算。 同样，变量可以在定义之前使用。我有一个个人策略，总是在使用任何变量之前显式地将其初始化为已知值，并且我将其扩展到我可以控制的任何最佳实践文档。

C / C + +:

头疼=脖子疼。 不支持闭包。(我对c++ 0x很兴奋，因为c++ 0x有这些功能。) 静态类型。“等等，”你说。“你刚才说你不喜欢动态类型!”是的，我确实这么说过。但是静态类型也会让人头疼。(如果有选择的话，我仍然会选择静态类型。)最理想的情况是，我希望语言默认是静态类型的，但也支持动态类型。(我还想要一匹小马，500亿美元，还有整个世界。)