Vathix wrote:
> I guess this one could be considered socket.d 1.0. I've tested it pretty well, even made a test server that accepts multiple connections, listener.d included. It works good for windows and linux.
> 
> Still needed:
> - Host name resolving
> - Socket stream class
> 
> 

Good work! i have tested in linux and writen some code for connecting to listener.d and run fine.

thanks for the work...

regards

Julio Jiménez

On Sun, 29 Feb 2004 18:38:35 -0500, Vathix <vathix@dprogramming.com> wrote:

> I guess this one could be considered socket.d 1.0. I've tested it pretty
> well, even made a test server that accepts multiple connections,
> listener.d included. It works good for windows and linux.
>
> Still needed:
> - Host name resolving
> - Socket stream class

Will this code be added to the standard distribution? Robert

Robert M. Münch wrote:

> On Sun, 29 Feb 2004 18:38:35 -0500, Vathix <vathix@dprogramming.com> wrote:
> 
>> I guess this one could be considered socket.d 1.0. I've tested it pretty
>> well, even made a test server that accepts multiple connections,
>> listener.d included. It works good for windows and linux.
>>
>> Still needed:
>> - Host name resolving
>> - Socket stream class
> 
> 
> Will this code be added to the standard distribution? Robert

I hope so.

-- 
Christopher E. Miller

Vathix wrote:
> I guess this one could be considered socket.d 1.0. I've tested it pretty well, even made a test server that accepts multiple connections, listener.d included. It works good for windows and linux.

Just gave it a quick scan.  It's not bad so far.  A few things could probably be simplified (like toAddrString could be a single line calling inet_ntoa instead of all that string stuff), and you might want to consider using the newer functions that support IPv6, or at least making sure your interface will work seamlessly with the new connection methods.  I say let people use it and push for Phobos inclusion once it's been used a bit and you've gotten some feedback.


Sean

Sean Kelly wrote:

> Vathix wrote:
> 
>> I guess this one could be considered socket.d 1.0. I've tested it pretty well, even made a test server that accepts multiple connections, listener.d included. It works good for windows and linux.
> 
> 
> Just gave it a quick scan.  It's not bad so far.  A few things could probably be simplified (like toAddrString could be a single line calling inet_ntoa instead of all that string stuff), and you might want to consider using the newer functions that support IPv6, or at least making sure your interface will work seamlessly with the new connection methods.  I say let people use it and push for Phobos inclusion once it's been used a bit and you've gotten some feedback.
> 
> 
> Sean

I have to say that I don't know much about IPv6. I would think all you need to do is set the AddressFamily to INET6 in Socket's constructor and make a new Address class such as IPv6Address that uses C's struct sockaddr differently; but I don't know so I'll look into it. Thanks.

-- 
Christopher E. Miller