On Fri, 10 Dec 2004 09:24:42 +0100 (CET), Peter Svensson <[EMAIL PROTECTED]> wrote: > We have done all the authentication logic etc in perl. It is a lot easier, > even more so since we have a database back end.
How are you doing this? Are you using an $AGI->exec() to Read in the information, or are you doing the entire thing in Perl? Would it be possible to see a snippet of how you are accomplishing this? > When you move to perl you can also check for the existance of the file > directly. Good idea. I'll do that. > You may want to verify the room number here. It is less secure (you get > feedback if the room is valid earlier) but a lot easier on the callers. I > have found that callers sometimes make a mistake or a digit may be missed. > With the above solution you only get one shot at entering the room > number, and no feedback that it is invalid. Instead, the pin code will be > asked for three times. Actually it goes back to s,3 which is the conference room question, so when it is invalid it asks for the conference room _and_ pin again so that user doesn't have to worry about a single shot. If the conference room/pin pair is incorrect, you get another chance to enter the pair again. This seems a bit more secure since you don't get information about whether the conference room or pin is invalid, but just that that pair is wrong. > These are used by the agi below to make the announcement, right? Yes sir. Basically what happens is it asks the user to record the prompt, then the AGI script does the work of joining the conference and playing back the wave file. Once that is done, the AGI execution is complete and the caller is attached to the conference. Once the user is entered into the conference, the delay between the callfile being generated and the wave file is played back, the user is joined and hears their recording and then they can start talking. The user isn't muted so could probably talk over top of the announcement. Anytime I've used this type of feature though I've been able to hear my announcement being played (it also lets me know that I'm in the conference). Not sure if those implementations take into account to mute the channel while the announcement is played though. > You need to make sure that the temporary file is stored on the same device > as the actual spool file. Otherwise the 'mv' below will not be atomic. > /var/spool is not unheard of to have on it's own partition. Just a heads > up. I don't think I quite understand. Even if it is on a different partition, the path should still be the same should it not? /var/spool should still work whether it is on /dev/hda2 or /dev/hda5 ? > You probably want to close first and then move to make sure the output > buffers are flushed before the move. Yep... you're right :) Thanks, Leif Madsen. http://www.leifmadsen.com _______________________________________________ Asterisk-Users mailing list [EMAIL PROTECTED] http://lists.digium.com/mailman/listinfo/asterisk-users To UNSUBSCRIBE or update options visit: http://lists.digium.com/mailman/listinfo/asterisk-users
