- From: <jose.kahan@w3.org>
- Date: Thu, 10 Feb 2000 14:49:30 +0100 (MET)
- To: Uwe F Mayer <mayer@tux.org>
- CC: www-amaya@w3.org
In our previous episode, Uwe F Mayer said: > > In fact, further investigation has shown that Amaya sees the outside world > if a numeric IP address is used, but does not connect to symbolic > addresses, even as named is running on the local machine, and netscape, > telnet, ssh, ftp all seem to have no problem resolving names. However, > Amaya is able to resolve "localhost". My installation of named is > several years old, and is configured as a caching/forwarding server. Any > ideas? Uwe, The name resolution is done thru gethostbyname(), a standard a Unix system call (to be more precise, the code is in the libwww/Library/src/HTDNS.c file). There must be some kind of incompatible with your named, as we haven't noticed this problem before, in a number of systems. What kind of system are you running (e.g., do a uname -a)? I can eventually give you a small program that makes the request using the libwww code and we can strace or truss it to find out where's the problem. If you're using an Amaya that you compiled yourself, it'll be even easier. >In fact, many of the files in Amaya/config contain spelling errors. I have > attached a diff-file that fixes those that I found. In particular I > thought is was funny, that according to one message in en-amayamsg > "These changes won't be effective until Amaya is retarted" Thanks! We'll merge the messages. It's a bit harder than doing a CVS diff because some of those files have evolved since the release. It'd be much easier to have a database oriented system to handle translations and then use it to generate the dialogue files... but we're not there yet :) Right now, we'll make Amaya much smarter by changing that particular message :) Thanks for all your help and tests! -Jose
Received on Thursday, 10 February 2000 08:49:40 UTC