Am 27.12.2011, 16:11 Uhr, schrieb Andrei Alexandrescu
<seewebsiteforem...@erdani.org>:
On 12/27/11 4:51 AM, deadalnix wrote:
Le 27/12/2011 05:25, Andrei Alexandrescu a écrit :
https://github.com/D-Programming-Language/dmd/commit/675898721c04d0bf155a85abf986eae99c37c0dc
Andrei
Maybe I'll seem bitter, but I do not think this changement was really
that important. This is nice, ok, but we have some other really serious
flaw, like shared not doing what it is supposed to do.
Imagine how bitter I am that the string lambda syntax didn't catch on!
Andrei
I used it and I am not sure if it had draw backs, but looking at the code
to support them, it gave me the feeling that you were doing part of the
job, the compiler should be doing: test if the alias is a string, if it
can be compiled and assigning variable names to parameters. I once adapted
that style to write something similar to binaryFun and the algorithm
templates. It worked for me after a while, but I found it a bit
complicated. (Before D, I only used templates as far as generic types go.)
The source code didn't *look* exactly concise and going with the language
to me, if you know what I mean. It shows a nice capability of the language
- maybe the first time newbies stumble over mixin string, but a short
lambda syntax is as good (because the string mixin doesn't get us anything
beyond that) and saves the if-else to identify the kind of alias and the
support code. Honestly, I never really liked the idea of string mixins for
general use :D like saving on boilerplate code, multi-inheritance etc. It
may seem odd, but I prefer a dozen language features over one string mixin
feature that can cover them all.
That brings me to another point: I like to use IDEs and they must have
difficulties in understanding string mixins and giving hints. There are no
rules to what you can put in a string, but a lambda syntax defined in the
specification can be checked in the IDE.
