• lurker (5/5) Feb 14 2011 Given that large parts of Phobos are still in flux or planned for
• lurker (2/2) Feb 14 2011 Also, I wouldn't mind a small example of a server using std.socket.
• Jonathan M Davis (14/19) Feb 14 2011 I've never used std.socket and don't really know how mature it is, but i...
• Daniel Gibson (9/28) Feb 14 2011 When the new code is accepted std.socket should be really usable (it was...
• Andrej Mitrovic (5/5) Feb 14 2011 There's a broken example for D2 in the samples folder. It uses D1
• lurker (3/3) Feb 14 2011 Are you talking about listener.d? I managed to compile it with the
• Andrej Mitrovic (3/6) Feb 14 2011 You've fixed it from 2.051, right? The one released with 2.051 doesn't
• lurker (6/6) Feb 15 2011 Yes, 2.51:

lurker <lurker mailinator.com> writes:

Given that large parts of Phobos are still in flux or planned for
replacement I'd like to ask, how mature is std.socket?

I'm thinking of writing a small server app and would like to know if
there are serious problems awaiting the uninitiated?

Thanks

lurker <lurker mailinator.com> writes:

Also, I wouldn't mind a small example of a server using std.socket.

Thanks

Jonathan M Davis <jmdavisProg gmx.com> writes:

On Monday, February 14, 2011 17:16:33 lurker wrote:
 Given that large parts of Phobos are still in flux or planned for
 replacement I'd like to ask, how mature is std.socket?
 
 I'm thinking of writing a small server app and would like to know if
 there are serious problems awaiting the uninitiated?

I've never used std.socket and don't really know how mature it is, but it looks 
like the author of that module made some improvements to it recently which 
haven't made it into Phobos yet: 
http://d.puremagic.com/issues/show_bug.cgi?id=5401

So, I suspect that other than that, very little work has been done on it 
recently, though I don't know how much work it needs, if any. Presumably, the 
updates in that enhancement request deal with the most pressing issues.

Regardless, I'm unaware of any current issues for std.socket (nothing really 
stands out in the bugzilla from what I can see), and I'm unaware of any plans
to 
replace it. A socket type doesn't seem like the sort of thing that would really 
need do be overhauled though. Presumably, it's pretty low level. Bug again, I 
haven't really used it or looked at it. Maybe someone else would know more.

- Jonathan M Davis

Daniel Gibson <metalcaedes gmail.com> writes:

Am 15.02.2011 02:36, schrieb Jonathan M Davis:
 On Monday, February 14, 2011 17:16:33 lurker wrote:
 Given that large parts of Phobos are still in flux or planned for
 replacement I'd like to ask, how mature is std.socket?

 I'm thinking of writing a small server app and would like to know if
 there are serious problems awaiting the uninitiated?

 I've never used std.socket and don't really know how mature it is, but it looks
 like the author of that module made some improvements to it recently which
 haven't made it into Phobos yet:
 http://d.puremagic.com/issues/show_bug.cgi?id=5401

 So, I suspect that other than that, very little work has been done on it
 recently, though I don't know how much work it needs, if any. Presumably, the
 updates in that enhancement request deal with the most pressing issues.

 Regardless, I'm unaware of any current issues for std.socket (nothing really
 stands out in the bugzilla from what I can see), and I'm unaware of any plans
to
 replace it. A socket type doesn't seem like the sort of thing that would really
 need do be overhauled though. Presumably, it's pretty low level. Bug again, I
 haven't really used it or looked at it. Maybe someone else would know more.

 - Jonathan M Davis

When the new code is accepted std.socket should be really usable (it was 
usable before if you didn't want to do too fancy stuff like feeding your 
own socket_t into std.socket.Socket or deriving from the Socket class.. 
but this is fixed in the new code).
Better support for IPv6 (an Internet6Address class or something like 
that) and local connections (AF_UNIX) would be nice, though.

Cheers,
- Daniel

Andrej Mitrovic <andrej.mitrovich gmail.com> writes:

There's a broken example for D2 in the samples folder. It uses D1
code, so it needs to be translated properly to D2. char[] -> string,
string.find -> algorithm.find (+ algo's find is different) and some
other tidbits.

I'll see if I can try to fix it.

lurker <lurker mailinator.com> writes:

Are you talking about listener.d? I managed to compile it with the
latest DMD:

http://pastie.org/1565017

Andrej Mitrovic <andrej.mitrovich gmail.com> writes:

On 2/15/11, lurker <lurker mailinator.com> wrote:
 Are you talking about listener.d? I managed to compile it with the
 latest DMD:

 http://pastie.org/1565017

You've fixed it from 2.051, right? The one released with 2.051 doesn't
compile without modifications.

lurker <lurker mailinator.com> writes:

Yes, 2.51:

Digital Mars D Compiler v2.051
Copyright (c) 1999-2010 by Digital Mars written by Walter Bright
Documentation: http://www.digitalmars.com/d/2.0/index.html
Usage:
...