Re: lstream problems in 21.5 GnuPG signing a file with MailCrypt.

I just received this from Ben... ,---- | ugh. | | i'd have to hazard a guess that there's a problem in the internal md5 | code. i know this isn't well debugged since i redid all the coding | system stuff as part of my big mule patch. i bet it's something | related to the detect-coding part of the code in md5.c, since that | code is the only place that enters file-coding the way it does. | clearly it's a buffering problem, that's why you're seeing it only | bigger than a certain number of bytes. beyond this i can't say, | though, without sitting down at the code -- you might forward this | message to someone who can step through the c code in question and see | if they can identify what's going on. (granted, that code is a bit | tricky because there are lots of nested levels ...) | | ben `---- Does that help at all? |--==> "MS" == Michael Sperber <sperber(a)informatik.uni-tuebingen.de&gt; writes: SY> Invalid operation: SIGPIPE raised on process; closed it, "*GPG*" MS> No, SIGPIPE means the external process has died, and XEmacs isn't Yes, because I killed the external process. SY> Let me know if I can do anything to help get to the bottom of this. MS> Set a break point on the invalid_operation call in unix_send_process MS> (around line 1569). Reproduce the error. Mail in the backtrace, if MS> possible both C and Lisp. But the invalid_operation isn't the problem, that was being called because I killed the external process. My first thoughts were the same as Ben's, a buffering problem. -- |---<Steve Youngs>---------------<GnuPG KeyID: 10D5C9C5>---| | XEmacs - It's not just an editor. | | It's a way of life. | |------------------------------------&lt;youngs(a)xemacs.org&gt;---|