Member
Last active 14 years ago
As I don't see your internals, I can only guess how you get the underlying data. My previous post (which I just corrected; there were a couple of syntax errors) was an attempt to help, clarify, and collaborate not tell you how to write your software. I hope it didn't come across that way. While I mostly develop in Java, wild SQL statements have always appealed to me.
Regards,
Ted
It could be done with a single SQL statement. If you assume a variable EXTEN has the extension of the FOP2 user:
(select * from fop2buttons where `group` in (select `group` from fop2buttons where exten="148") and type="extension" order by exten) union (select * from fop2buttons where `group` not in (select `group` from fop2buttons where exten="148") and type="extension" order by `group`,exten)
The first select gets the users in the same group and as there is only one group, it only has to sort by exten. The second selection gets all the extensions not in the same group as the current FOP2 and sorts them by the group name then the exten.
Ted
Currently groups are ordered by the lowest extension in the group. The SQL to simulate it is:
select `group` from fop2buttons group by `group` order by min(exten);
Is it possible to sort the groups alphabetically (with the exception of having the first group be the logged-in user's group as it is now)?
Tx,
Ted
Being able to collapse and expand Groups on the panel is a powerful way to find the right extension. It seems to have a built in delay to make it visually pleasing. While I am all for visual pleasure, that pleasure will be lost on an operator with a stack of calls to route if it takes three seconds for a group to expand and collapse. Maybe a configuration option could set the speed for an installation.
While thinking about that use case, I could also see "expand all" and "collapse all" buttons on the tool bar.
Thanks for your consideration,
Ted
I figured out the problem with calls being dropped when transferring to Park. For some reason I did not have your copy of extensions_override_freepbx.conf in /etc/asterisk. The version I had was missing the fop2-park context. I copied your file (from /usr/local/fop2) to /etc/asterisk and bingo it worked.
If anyone else has a similar problem, go to /etc/asterisk and enter the command
grep fop2-park *
If nothing is found, you have the same problem I did.
Ted
Nicolas,
Thanks for the response.
WRT direct transfer to voicemail, I believe I found a bug in autoconfig-buttons-freepbx.sh. It refers to "extenvoicemail". Your documentation and the panel code refers to "voicemailext". When I changed all occurrences in autoconfig-buttons-freepbx.sh (five I think) of "extenvoicemail" to "voicemailext" the transfer to voice mail button started to work.
And those pesky assisted transfers in Asterisk. I've spent a month developing a solution that allows the people being called to see the caller id information of the call currently on hold by the person transferring. Eventually I made the assumption that the last call put on hold is the call that will be transferred and store the channel name in a map (Java) keyed on the extension. When the call is first made (by reception) to the recipient my process looks up the callers extension (you pass it to me in checkdir.php) in that map and presents the underlying caller id information as part of the web call (to a tomcat server). It works nicely.
While doing that work I discovered that Asterisk 1.6 does have a Transfer Event that allows you to know what is happening at time of transfer. I suspect you could use this information to successfully identify what is in the parking lot. In my development I have been keeping the channel name around and when I see a hangup event I use that value to clear out the various maps and lists I have used to track the call to keep from leaking memory.
Agreed that a blind transfer to the parking lot does work, although I like that the person putting the call into park gets the message through the phone of the parking lot number. Without it the user must use FOP to know where it went.
Ted
I am working off Asterisk 1.6 (latest Elastix distro) and FOP 2.11.
From FOP2 I can:
From FOP2 I cannot:
The transfer to parking lot works from a SIP/soft phone; I don't know how to the transfer directly to a user's voicemail from the phone so I cannot confirm it that works.
Finally, when I park a call from a SIP or softphone, if I blind transfer it, the call is visible in the Parking Lot button, if I call with consultation (which I need to do to know where it is parked), the call appears in the Parking Lot while "consulting" with the parking lot attendant, but disappears as soon as I complete the transfer.
I've reviewed my configuration against Nicholas' documentation and the Elastix documentation.
Any ideas what is going on and how to fix these?
Tx,
Ted
Still not getting it.
I am using fop2 2.11 against the most recent Elastix. I downloaded and installed the fop2manager.
First I logged onto Elastix admin console and picked Unembedded FreePBX. On the FreePBX console I picked the Tools tab on the left and the Fop2 User option under the Third Party Addon separator. I added my test extensions, set their secrets to the same as their secrets added in Elastix and set the permissions to "all". After rebooting the server and reloading FOP2 on the client, I didn't get the action submenu. I changed the permissions to be the full list, including queuemanager and still no luck.
I gave up fop2manager and hand edited the /usr/local/fop2/fop2.cfg file, adding the lines:
user=123:12345:all
then
user=123:12345:all,queuemanager
More reboots/reloads and still no joy.
I did confirm that fop2manager did interact with the system by successfully changing one of the extension labels.
I have only used Firefox on the client, but under Linux, Win 7, and Win XP. The rest of the interface works without flaw.
Any ideas of what I should try next?
Thanks,
Ted
If custom_popup is calling an Asterisk connected service (I've been playing with tomcat and Asterisk-Java), that process can find out everything about the inbound call (e.g. if it is a transfer, what is the originating caller id). With the extensions one can infer the channel, but getting it explicitly would be great.
Tx,
Ted
Did this get resolved? I am having the same problems on Win 7, Win XP, and Ubuntu 10.4
Tx,
Ted