Back [03:12:59]<SWPadnos__> SWPadnos__ is now known as SWPadnos [03:39:49]<jtr> jtr is now known as jtr_away [04:28:07]<SWPadnos> SWPadnos is now known as SWP_Away [21:17:45]<SWP_Away> SWP_Away is now known as SWPadnos [21:32:23]<SWPadnos> SWPadnos is now known as SWP_Away [22:20:27]<SWP_Away> SWP_Away is now known as SWPadnos [22:26:41]<rayh_> I've been trying to get halcmd to "open" in a tcl bidirectional pipe but it isn't having any. [22:27:22]<SWPadnos> how is this normally done in tcl? [22:27:24]<cradek> can you do that with any other program? [22:27:31]<SWPadnos> (or is it normally done in tcl? :) ) [22:27:38]<rayh_> I've got a model built upon bc [22:27:44]<rayh_> the unix calculator. [22:27:53]<SWPadnos> ok -and that works [22:27:56]<SWPadnos> ? [22:27:59]<rayh_> yes [22:28:13]<rayh_> send it 2+2 and it returns 4 [22:28:27]<SWPadnos> ok. what halcmd options are you using? [22:28:34]<rayh_> but bc has not prompt like halcmd: [22:28:47]<SWPadnos> use "-skf" [22:28:55]<SWPadnos> script mode has no prompt [22:28:56]<rayh_> okay. [22:29:04]<rayh_> I'll try. Thanks. [22:29:15]<SWPadnos> (k=keep going if there's an error, f=use a file) [22:29:15]<rayh_> I thought that was a bearing company. [22:29:20]<SWPadnos> so it is ;) [22:29:29]<rayh_> I was using -kf [22:30:09]<rayh_> * rayh_ brb testing. [22:30:11]<SWPadnos> ok - the -s will prevent the prompt - I put it in for just this situation [22:38:34]<rayh_> still dies. it does pass the first command like newsig MySig1 bit. [22:38:57]<rayh_> but hangs after that now. [22:40:31]<SWPadnos> hang as in more stuff in = nothing out? [22:53:14]<rayh_> I set the channel for writing "w" rather than "r+" and it seems to be working. [22:53:23]<SWPadnos> ah -good deal [22:53:31]<rayh_> I do get a response back to standard out [22:53:48]<rayh_> Just have to get it to where it needs to be. [22:53:57]<SWPadnos> errors only, or other stuff as well? [22:54:00]<rayh_> Thanks [22:54:13]<SWPadnos> does "rw" work? [23:00:01]<alex_joni> -rw :D [23:02:14]<SWPadnos> CD or DVD? [23:06:36]<rayh_> I'll try that. [23:08:23]<alex_joni> or blue ray [23:08:24]<rayh_> No. The possibles are "r" "r+" "w" and "w+" [23:08:25]<alex_joni> :D [23:08:36]<SWPadnos> hm [23:08:38]<alex_joni> how fit.. [23:08:59]<rayh_> w is the only one that works. [23:09:28]<rayh_> and that returns to the terminal running the program. [23:09:43]<rayh_> Seems like it is darn close. [23:09:58]<SWPadnos> there should be some way of opening two pipes then, one for command output and one for response input [23:11:58]<SWPadnos> are you using "flush"? [23:13:24]<SWPadnos> you can try using fconfigure to turn off buffering as well [23:13:46]<SWPadnos> (according to a tcl 8.5 help page at http://www.tcl.tk/man/tcl8.5/tutorial/Tcl26.html ) [23:32:47]<alex_joni> night all [23:32:53]<SWPadnos> good night [23:36:47]<rayh> that was fun! [23:37:02]<SWPadnos> I was just going to ask - that wasn't a crash was it? [23:37:27]<rayh> Nah that was a switch box that went astray. [23:37:33]<SWPadnos> ah - phew [23:38:01]<rayh> what was that tcl link again. [23:38:15]<SWPadnos>http://www.tcl.tk/man/tcl8.5/tutorial/Tcl26.html [23:38:20]<rayh> tnx [23:38:23]<SWPadnos> sure [23:38:32]<SWPadnos> the buffering looks like a likely candidate for problems [23:40:08]<rayh> Right. halcmd does NOT buffer? [23:40:25]<rayh> or does it buffer by line [23:40:32]<SWPadnos> no - it's in tcl file handling (or the kernel - not sure) [23:40:52]<SWPadnos> you can turn it off in tcl (8.5 at least) [23:42:27]<rayh> buffering may be the problem with the open xxx "r+" version. [23:42:41]<SWPadnos> it looks like you can issue something like "fconfigure pipe_name -buffering line" to put into line buffer mode (or use none for no buffering) [23:46:02]<rayh> phone [23:46:09]<SWPadnos> line buffering is probably best for this application, since there's always a newline after every command or response [23:46:11]<SWPadnos> ok