May 2006 - flexi-streams-devel@common-lisp.net

[flexi-streams-devel] New version 0.5.5
by Edi Weitz 24 May '06

24 May '06

ChangeLog: Version 0.5.5 2006-05-24 Some small fixes for LW Download: http://weitz.de/files/flexi-streams.tar.gz

1 0

[flexi-streams-devel] New version 0.5.4 (Was: flexi-streams and CMUCL)
by Edi Weitz 18 May '06

18 May '06

[Please use the mailing list.] Hi Satyaki! On Sun, 14 May 2006 14:46:40 -0700, "Satyaki Das" <satyakid(a)gmail.com> wrote: > Thank you for writing flexi-streams. I needed the attached patch to > get in-memory-streams to work with CMUCL-19c (on x86 linux). > Basically the problem was that open-stream-p on a vector-stream was > always returning nil. Thanks for the report. I've uploaded a new version which should fix this. I've conditionalized the patches on CMUCL because other Lisps don't seem to need it. I also changed your patches a bit. I'd recommend asking on the CMUCL mailing list if this is considered a bug in their Gray streams implementation. The SBCL I have on my server (0.9.7.1) behaves like all the other implementations. Cheers, Edi.

1 0

[flexi-streams-devel] Re: flexi-streams feature idea
by Edi Weitz 13 May '06

13 May '06

[Please use the mailing list. See Cc.] Hi! On Sat, 13 May 2006 13:19:22 +0200, "Attila Lendvai" <attila.lendvai(a)gmail.com> wrote: > i know that you probably have tons of ideas yourself, but the > in-memory streams could support file-position, especially the > vector-streams. > > we are trying to serve a generated document made by cl-typesetting, > but it wants to seek in the bivalent output stream it demands, so a > direct network stream is a no-go. but using a vector-output-stream > would be fine if it supported file-position. > > if you don't have time for this i'll look into this eventually > myself, but for now /tmp is fine for us. I don't think you can support FILE-POSITION, at least I'm not aware of a way to do that with Gray streams. However, one could invent an accessor like, say, IN-MEMORY-STREAM-POSITION. But one would need to define the exact semantics of this operator. What would you propose? Cheers, Edi.

1 0