Re: cl-launch and openmcl
Gary Byers <[email protected]> Sat, 7 Oct 2006 10:34:33 -0600 (MDT)
| Newsgroups | gmane.lisp.openmcl.bugs |
|---|---|
| Message-ID | <[email protected]> |
This message is in MIME format. The first part should be readable text, while the remaining parts are likely unreadable without MIME-aware tools. --0-869405748-1160238873=:57104 Content-Type: TEXT/PLAIN; charset=X-UNKNOWN; format=flowed Content-Transfer-Encoding: QUOTED-PRINTABLE There are some bugs/instabilities introduced in the 060826 snapshots (new "basic" stream are missing some support functions/methods). That'll be fix= ed Any Day Now. (What happens is that INPUT-STREAM-P returns false for someth= ing that is in fact an input stream in some cases; this happens when the standa= rd=20 streams are being initialized, and leads to an attempt to write an error me= ssage on an uninitialized stream.) On Fri, 6 Oct 2006, [ISO-8859-1] Far=E9 wrote: > I tried with the latest tarball from 060826, and got the same error. > I reproduce it, grab the latest cl-launch, and try > cl-launch -l openmcl -B tests 112 114 > > [ Fran=E7ois-Ren=E9 =D0VB Rideau | Reflection&Cybernethics | http://fare.= tunes.org=20 > ] > Reporter: Mr Gandhi, what do you think of Western Civilization? > Gandhi: I think it would be a good idea. > > > On 06/10/06, Far=E9 <[email protected]> wrote: >> Thanks! It works much better now. I've just released cl-launch 1.89 >> with improved openmcl support, including image dump and direct >> command-line arguments. >> http://www.cliki.net/cl-launch >>=20 >> Testing, however, reveals unstability in openmcl, at least the tarball >> I'm using. >> cl-launch -l openmcl -B tests 114 114 >> goes into the debugger with a >> Read error between positions 10 and 7932 in /local/fare/tsts/clt-out.sh >> while trying to compile-file clt-out.sh using the 1.1-pre-060608 >> (Beta: LinuxX8664) tarball. The second line in clt-out.sh begins with a "#|"; there is no terminating "|#" in the file, hence the end-of-file error. It's not clear to mw how/where the (partial) clt-out.sh file is written; if= =20 it's written from lisp and the last bufferful of output isn't getting flushed, that may be an OpenMCL bug. So, where and how does this file get written, and how might that process be isolated so that the bug can be identified ? --0-869405748-1160238873=:57104 Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ Bug-openmcl mailing list [email protected] http://clozure.com/mailman/listinfo/bug-openmcl --0-869405748-1160238873=:57104--