Ruby port and its socket library

Tabitha McNerney tabithamc at gmail.com
Mon May 26 23:58:43 PDT 2008


Hello everyone,

I'm just updating this email post that I made originally on the 8th of this
month because the problem continues and I realized that I'm not the only one
who noticed this -- as there have since been some tickets opened up about
this:

Here:

http://trac.macports.org/ticket/15299

and here:

http://trac.macports.org/ticket/15254

Thanks for bearing with me ... Ruby is used by a lot of people in the
MacPorts and non-MacPorts communities and I am perplexed as to what changed
on Leopard to cause these conflicts with MacPorts builds of Ruby?

T.M.

On 5/8/08, Tabitha McNerney <tabithamc at gmail.com> wrote:
>
> Hello,
>
> I accidentally uninstalled my previous installed version of the ruby
> MacPort (ruby 1.8.6 patch 110) and then I upgraded the port to the current
> ruby port (ruby 1.8.6 patch 114). The previous version of Ruby was made on
> Mac OS X Leopard 10.5.1 and it worked perfectly. I had since upgraded to
> 10.5.2 and now when I attempt to install the new portfile, I am getting
> errors regarding Ruby's attempt to build its socket library. As I do a
> verbose install (as in # port -v install ruby) here's the excerpt of where I
> am seeing a problem in the output:
>
>                           kconv.rb: m...........c.........
>>                              nkf.c: m...
>>                           socket.c: ccccccccccm.
>> No definition for bsock_do_not_rev_lookup
>> .
>> No definition for bsock_do_not_rev_lookup_set
>> .
>> No definition for bsock_s_for_fd
>> .
>> No definition for bsock_close_read
>> .
>> No definition for bsock_close_write
>> .
>> No definition for bsock_shutdown
>> ...
>> No definition for bsock_getsockname
>> .
>> No definition for bsock_getpeername
>> .
>> No definition for bsock_send
>> .
>> No definition for bsock_recv
>> ..
>> No definition for ip_addr
>> .
>> No definition for ip_peeraddr
>> .
>> No definition for ip_recvfrom
>> .
>> No definition for ip_s_getaddress
>> ...
>> No definition for socks_init
>> .
>> No definition for socks_s_close
>> .
>> No definition for tcp_accept
>> ..
>> No definition for tcp_sysaccept
>> .
>> No definition for tcp_svr_init
>> ..
>> No definition for udp_init
>> .
>> No definition for udp_connect
>> .
>> No definition for udp_bind
>> .
>> No definition for udp_send
>> ..
>> No definition for unix_init
>> .
>> No definition for unix_path
>> .
>> No definition for unix_addr
>> .
>> No definition for unix_peeraddr
>> .
>> No definition for unix_recvfrom
>> .
>> No definition for unix_send_io
>> .
>> No definition for unix_recv_io
>> .
>> No definition for unix_s_socketpair
>> .
>> No definition for unix_s_socketpair
>> .
>> No definition for unix_svr_init
>> .
>> No definition for unix_accept
>> ..
>> No definition for unix_sysaccept
>> ..
>> No definition for sock_initialize
>> ..........
>> No definition for sock_s_socketpair
>> .
>> No definition for sock_s_socketpair
>> .
>> No definition for sock_gethostname
>> ..
>> No definition for sock_s_gethostbyaddr
>> ...
>> No definition for sock_s_getnameinfo
>> .
>> No definition for sock_s_pack_sockaddr_in
>> .
>> No definition for sock_s_pack_sockaddr_in
>> .
>> No definition for sock_s_unpack_sockaddr_in
>> .
>> No definition for sock_s_pack_sockaddr_un
>> .
>> No definition for sock_s_pack_sockaddr_un
>> .
>> No definition for sock_s_unpack_sockaddr_un
>>
>>                         stringio.c:
>> c.....................................................
>>                          strscan.c:
>> cc.............................................
>>                             zlib.c:
>> mcccccccccccccccccc...................................................................................
>>
>
> This is the nuttiest thing and I can't figure out what the problem is yet
> (I'm starting to dig around in the Ruby source files to see what might be
> happening). Has anyone else noticed this? My hardware is kind of new too
> (its 64-bit EFI firmware, a Mac Pro early 2008 tower running 10.5.2  ...
> don't know if Apple might have done something low level which the sockets
> library for Ruby depends on)?
>
> Any suggestions would be highly appreciated.
>
> Regards,
>
> T.M.
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.macosforge.org/pipermail/macports-users/attachments/20080526/27aafd63/attachment.htm 


More information about the macports-users mailing list