Skip Menu |

This queue is for tickets about the IO-Socket-Socks CPAN distribution.

Report information
The Basics
Id: 92511
Status: resolved
Priority: 0/
Queue: IO-Socket-Socks
People
Owner: Nobody in particular
Requestors: leonerd-cpan [...] leonerd.org.uk
Cc:
AdminCc:
Bug Information
Severity: Wishlist
Broken in: 0.63
Fixed in: (no value)

History Show all quoted text
  Mon Jan 27 13:34:16 2014 leonerd-cpan [...] leonerd.org.uk - Ticket created
Subject: Transparent IPv6 support for transport by 'use base IO::Socket::IP'
Being a subclass of IO::Socket::INET, the transport layer is fixed at IPv4. You say in the TODO notes, "IPv6 support". This can be achieved by using IO::Socket::IP instead. Small caveat currently though is that both IO::Socket::Socks and ::IP want to use a method called _configure, for their own purposes. So one of these would have to be renamed. (see also https://rt.cpan.org/Ticket/Display.html?id=92484) -- Paul Evans
  Tue Jan 28 12:28:54 2014 oleg [...] cpan.org - Correspondence added
Пнд Янв 27 13:34:16 2014, PEVANS писал: Show quoted text
> Being a subclass of IO::Socket::INET, the transport layer is fixed at > IPv4. > > You say in the TODO notes, "IPv6 support". This can be achieved by > using IO::Socket::IP instead. > > Small caveat currently though is that both IO::Socket::Socks and ::IP > want to use a method called _configure, for their own purposes. So one > of these would have to be renamed. > > (see also https://rt.cpan.org/Ticket/Display.html?id=92484)
Unfortunately I still have no IPV6 support on my network. So I don't need / can't test such thing. Also IPV6 support for IO::Socket::Socks requires not only support for IPV6 on socket level, but additionally should be implemented in SOCKS protocol itself. So this needs some work (in which I not interested for now). For now I can rename _configure if it will help somehow.
  Tue Jan 28 12:28:54 2014 The RT System itself - Status changed from 'new' to 'open'
  Sat Jul 26 14:20:46 2014 sri [...] cpan.org - Correspondence added
+1 IPv6 support gets more important every day. -- sebastian
  Sun Jul 27 01:02:35 2014 oleg [...] cpan.org - Correspondence added
On Sat Jul 26 14:20:46 2014, SRI wrote: Show quoted text
> +1 IPv6 support gets more important every day. > > -- > sebastian
Patches are welcome. Also I may try to implement it myself if somebody will provide me IPV6 environment
  Sun Jul 27 07:15:13 2014 leonerd-cpan [...] leonerd.org.uk - Correspondence added
On Sun Jul 27 01:02:35 2014, OLEG wrote: Show quoted text
> Also I may try to implement it myself if somebody will provide me IPV6 > environment
You don't need an IPv6 setup to test it (also every box will always support ::1 as localhost and the various fe80:... linklocal addresses). Simply use IO::Socket::IP instead of ::INET as the base, and trust that it would handle IPv6 if required. -- Paul Evans
  Sun Jul 27 13:24:33 2014 sri [...] cpan.org - Correspondence added
Show quoted text
> You don't need an IPv6 setup to test it (also every box will always > support ::1 as localhost and the various fe80:... linklocal > addresses).
This is what i go with when i need to add IPv6 support to something, optional ::1 tests. -- sebastian
  Tue Jul 29 10:58:03 2014 oleg [...] cpan.org - Correspondence added
On Mon Jan 27 13:34:16 2014, PEVANS wrote: Show quoted text
> Being a subclass of IO::Socket::INET, the transport layer is fixed at > IPv4. > > You say in the TODO notes, "IPv6 support". This can be achieved by > using IO::Socket::IP instead. > > Small caveat currently though is that both IO::Socket::Socks and ::IP > want to use a method called _configure, for their own purposes. So one > of these would have to be renamed. > > (see also https://rt.cpan.org/Ticket/Display.html?id=92484)
Paul, which version of IO::Socket::IP would you recommend as "minimum required" for this task?
  Tue Jul 29 11:34:47 2014 leonerd-cpan [...] leonerd.org.uk - Correspondence added
On Tue Jul 29 10:58:03 2014, OLEG wrote: Show quoted text
> On Mon Jan 27 13:34:16 2014, PEVANS wrote:
> > Small caveat currently though is that both IO::Socket::Socks and ::IP > > want to use a method called _configure, for their own purposes. So > > one of these would have to be renamed.
> > Paul, which version of IO::Socket::IP would you recommend as "minimum > required" for this task?
I would suggest taking 0.28 mininum, because of the _configure method: https://metacpan.org/source/PEVANS/IO-Socket-IP-0.31/Changes#L22 -- Paul Evans
  Tue Jul 29 11:43:12 2014 oleg [...] cpan.org - Correspondence added
On Tue Jul 29 11:34:47 2014, PEVANS wrote: Show quoted text
> On Tue Jul 29 10:58:03 2014, OLEG wrote:
> > On Mon Jan 27 13:34:16 2014, PEVANS wrote:
> > > Small caveat currently though is that both IO::Socket::Socks and ::IP > > > want to use a method called _configure, for their own purposes. So > > > one of these would have to be renamed.
> > > > Paul, which version of IO::Socket::IP would you recommend as "minimum > > required" for this task?
> > I would suggest taking 0.28 mininum, because of the _configure method: > > https://metacpan.org/source/PEVANS/IO-Socket-IP-0.31/Changes#L22
I plan to rename _configure(). This is not a problem. In Debian stable current version is 0.16. Is it good enough? As I see IO::Socket::SSL requires 0.20.
  Fri Aug 29 00:47:44 2014 oleg [...] cpan.org - Correspondence added
Looks like IO::Socket::IP has annoying bug - no support for connect timeout: https://rt.cpan.org/m/ticket/show?id=92075 And can't be used as IO::Socket::INET replacement. And in fact most distributions using IO::Socket::IP as socket class (LWP, HTTP::Tiny) has same bug. I'll give a try to IO::Socket::INET6
  Mon Jan 12 08:37:20 2015 oleg [...] cpan.org - Correspondence added
RT-Send-CC: sri [...] cpan.org
Птн Авг 29 00:47:44 2014, OLEG писал: Show quoted text
> Looks like IO::Socket::IP has annoying bug - no support for connect > timeout: https://rt.cpan.org/m/ticket/show?id=92075 > And can't be used as IO::Socket::INET replacement. And in fact most > distributions using IO::Socket::IP as socket class (LWP, HTTP::Tiny) > has same bug. > I'll give a try to IO::Socket::INET6
IO::Socket::Socks 0.66_2 introduces IPv6 support with help of IO::Socket::IP 0.36+ Everybody are welcome to test it
  Wed Jan 14 04:53:06 2015 oleg [...] cpan.org - Correspondence added
Пнд Янв 12 08:37:20 2015, OLEG писал: Show quoted text
> Птн Авг 29 00:47:44 2014, OLEG писал:
> > Looks like IO::Socket::IP has annoying bug - no support for connect > > timeout: https://rt.cpan.org/m/ticket/show?id=92075 > > And can't be used as IO::Socket::INET replacement. And in fact most > > distributions using IO::Socket::IP as socket class (LWP, HTTP::Tiny) > > has same bug. > > I'll give a try to IO::Socket::INET6
> > IO::Socket::Socks 0.66_2 introduces IPv6 support with help of > IO::Socket::IP 0.36+ > Everybody are welcome to test it
Fixed in 0.66
  Wed Jan 14 04:53:07 2015 oleg [...] cpan.org - Status changed from 'open' to 'resolved'