FYI: new 2.00 whois server runs now on port 4444
Dear all, The new 2.0 whois server runs now on port 4444. People might want to give it a try before we put it in production. New features: - twice the speed for any classless lookups - the lookup algoritm is changed such that: Matyas Matyas can finally be found and also Daniel David hasn't problems anymore with finding himself in the database - unlimited number of matches allowed you can now find all davids in the RIPE database As always: please send bug reports directly to me and I will fix the problem. The service will be a little bit less reliable then our standard whois server since I am still working on the code. In general: I will try to keep the new server only 5 seconds behind the normal server between 9:00-19:00 local time. We will put the server in production when I have finished the code for the role object and auto NIC handle generation. I plan to finish that in about a week time. Kind regards, David K. --- How to try this out: Do the next command if your whois client supports choosing different ports then the default one (or edit the RIPE perl whois source code) $ whois -p 4444 -h whois.ripe.net Matyas Matyas Or do: $ telnet whois.ripe.net 4444 <return> Trying 193.0.0.194... Connected to dbase.ripe.net. Escape character is '^]'. Matyas Matyas <return> (The new database distribution will include a whois version that supports the selection of the port)
Dear all,
The new 2.0 whois server runs now on port 4444. People might want to give it a try before we put it in production.
(The new database distribution will include a whois version that supports the selection of the port)
It might be better if you were to get an actual assigned port number for your production code. That would slow down the port hopping that you are doing now. -- --bill
On Thu, 27 Jun 1996 10:26:34 -0700 (PDT) Bill Manning wrote:
The new 2.0 whois server runs now on port 4444. People might want to give it a try before we put it in production.
It might be better if you were to get an actual assigned port number for your production code. That would slow down the port hopping that you are doing now.
I might be missing something, but whois has been sitting on port 43 since the stone ages, i.e. there is a defined port. I think that it makes sense to use a different port for testing purposes. Asking for an 'officlal port' just for experimental code seems a bit off to me. Geert Jan
Hi Bill,
Bill Manning writes :
The new 2.0 whois server runs now on port 4444. People might want to give it a try before we put it in production.
(The new database distribution will include a whois version that supports the selection of the port)
It might be better if you were to get an actual assigned port number for your production code. That would slow down the port hopping that you are doing now.
I thought we already have one: port 43 for whois services. We will put the production code on port 43 as soon as the code looks stable for a while. Please correct me if I missed your point. David K.
I thought we already have one: port 43 for whois services.
We will put the production code on port 43 as soon as the code looks stable for a while.
Please correct me if I missed your point.
David K.
I am sorry that I was not clear. I beleive that, even with a superset of the whois protocol, this is no longer whois as defined and should therefor apply for a new well-known port. Local conversations with several others here at IETF do not support this view, so I expect that it (the view) will remain a minority opinion. Its something I thought worth mentioning in passing... :) Of course experimental stuff can use any port for testing. This, being a purely local matter can occur on any port the sysadmin chooses. -- --bill
In message <199606272053.AA03700@zephyr.isi.edu>, Bill Manning writes:
I thought we already have one: port 43 for whois services.
We will put the production code on port 43 as soon as the code looks stable for a while.
Please correct me if I missed your point.
David K.
I am sorry that I was not clear. I beleive that, even with a superset of the whois protocol, this is no longer whois as defined and should therefor apply for a new well-known port.
Local conversations with several others here at IETF do not support this view, so I expect that it (the view) will remain a minority opinion. Its something I thought worth mentioning in passing... :)
Of course experimental stuff can use any port for testing. This, being a purely local matter can occur on any port the sysadmin chooses.
-- --bill
Bill, As long as the is backward compatibility no protocol has had to apply for a new port number when it was extended. Curtis
Hi Bill,
Bill Manning writes :
I thought we already have one: port 43 for whois services.
We will put the production code on port 43 as soon as the code looks stable for a while.
Please correct me if I missed your point.
David K.
I am sorry that I was not clear. I beleive that, even with a superset of the whois protocol, this is no longer whois as defined and should therefor apply for a new well-known port.
Local conversations with several others here at IETF do not support this view, so I expect that it (the view) will remain a minority opinion. Its something I thought worth mentioning in passing... :)
Thanks for your explanation. As Curtis points out: I don't see a need for a new port number as long as we stay backward compatible. However, one might want to put a special 'for computer usage' server behind another port, to make a clear separation of functionality possible. But it's still not really needed since Gerald Winters claims that the RAWhois server is fully backwards compatible with the old scheme. Kind regards, David K.
participants (4)
-
bmanning＠ISI.EDU -
Curtis Villamizar -
David.Kessens＠ripe.net -
Geert Jan de Groot