User Agent: Mozilla/5.0 (Windows NT 6.1; rv:31.0) Gecko/20100101 Firefox/31.0 (Beta/Release) Build ID: 20140716183446 Steps to reproduce: I have configuration to support customized comment columns in LDAP search autocomplete list. Reference is: https://developer.mozilla.org/en-US/docs/Mozilla/Thunderbird/LDAP_Support And my config was: user_pref("mail.autoComplete.commentColumn", 2); user_pref("ldap_2.servers.<ldapsvraddr>.autoComplete.commentFormat", "[Title],[telephoneNumber],[managername],[departmentNumber]"); Actual results: When I upgraded to TB 31.0, I found in the auto-complete drop down list, the 2nd column is always the LDAP server address, and mine original customized configuration does not work any more. Is this feature be removed? I love the original feature! Expected results: Expected it to work as before - shown LDAP server address makes little sense for me because for a specific outgoing account, we use just one LDAP server.

Didn't verify it yet, but i suspect this is fallout from bug 754265. xref bug 1003856

I guessed wrong. hg log --removed mailnews/addrbook/public/nsIAbLDAPAutoCompFormatter.idl points to bug 452232 so actually it looks like it got removed. There's no comment about it there though... but assuming it's true, we should at least update the documentation at https://developer.mozilla.org/en-US/docs/Mozilla/Thunderbird/LDAP_Support

So, if an addon wanted to support this, they could replace the LDAP autocomplete component with a tweaked version that added extra columns to the attribute map and then formatted the result?

Magnus, are you working on a patch?

Sorry, not atm, feel free to take it.

Andreas, Do you know when will this bug be planned? I'd really love that functionality! Thank you so much! Regards, Steven

Andreas, One more question - what does this mean: tracking-thunderbird38? Thanks, Steven

tracking-thunderbird38? means that Andreas wants the core developers to give this issue some attention for our upcoming release, Thunderbird 38. Given though that this functionality did not exist in Thunderbird 31, it is unlikely to be a pre-release priority in Thunderbird 38. So I'm going to have to deny the tracking (since I'm the guy who stares at these tracked bugs). We would love to restore this functionality, so if someone work on a patch we would try to include it. But nobody has picked this up yet.

Andreas is advocating that this bug is important, but we will not have time to look at it prior to the Thunderbird 38 release. But let's look at it after that, hoping to be able to port the change to a dot release of Thunderbird 38.

Kent, Thank you so much for your explanation! Best regards, Steven

Hello Thunderbird 38.0.1 has the same issue. ldap_2.servers.EmployeeDirectory.autoComplete.commentFormat is ignored. This functions is really crucial for any for large organizations,colleges and universities. ldap_2.servers.EmployeeDirectory.autoComplete.commentFormat", "[Title] [Department] [telephonenumber]"); For users with the same or similar last name it helps if department and title showing in autocomplete field. Thank you Marina

I would like to repeat the sentiments of Marina, as our organization also heavily used the ldap department information to distinguish between users with similar names when composing a message. Would like to see this functionality return in a future dot release of Thunderbird!

Andreas, It seemed you were going to work on it, and we are looking forward to it's resolved. :D Not sure when we can happily see it's back? Thanks, Steven

This is the kind of feature that is mostly used by enterprises, yet the active developers are all individuals. It's hard to get motivated as an individual to fix, for free, an enterprise issue. Long-term we need some solution to that, which means figuring out some way to get enterprises to contribute back to the project so that there are people paid to specifically consider the enterprise issues. The most likely way for this to happen is some sort of enterprise-focused paid distribution of Thunderbird. Yet those who have tried this tell me it is unlikely to be successful. In the current system, the way to get this fixed is for one of the affected enterprises, possibly in combination with others, to offer to pay a developer to first investigate and then fix the issue. Knowing nothing about the issue myself, a wild guess of the cost of that would be $500 for the initial investigation, then $500 - $5000 for the actual fix.

> In the current system, the way to get this fixed is for one of the affected > enterprises, possibly in combination with others, to offer to pay a > developer to first investigate and then fix the issue. Knowing nothing about > the issue myself, a wild guess of the cost of that would be $500 for the > initial investigation, then $500 - $5000 for the actual fix. OK, we would chip in to have this fixed. Not sure we are good for $5000, but anybody else want to step in? Any developer want to do it for less than $5K?

(In reply to John Wade from comment #17) > OK, we would chip in to have this fixed. Not sure we are good for $5000, > but anybody else want to step in? Any developer want to do it for less > than $5K? What I was implying is that first someone needs to took at this to understand what the issue is, and my estimate of that was $500. Only after this is done could there be any reasonable estimate of the actual cost to fix.

(In reply to Kent James (:rkent) from comment #18) > > What I was implying is that first someone needs to took at this to > understand what the issue is, and my estimate of that was $500. Only after > this is done could there be any reasonable estimate of the actual cost to > fix. I understood this, I was just noting that paying someone $500 to analyze the scope of the required changes may be necessary, but if the cost to fix turns out to be $5,000, it is more that it is worth to us. FYI, if anyone would like to pay $5K to fix this, I would personally be happy to investigate for free and implement the fix, payment upon delivery of functional code.

Just the other day I stumbled across bug 452232, which might be causing this issue.