Surely just re-compiling from new source (and not changing much else), would be a start. Where upstream bugs have been fixed, it would be an easy win, even where they are not, at least bugs reported on modern versions can then be addressed.
That document says little, it is a good summary of the situation, but it does not mention any actual bugs outstanding. I counted maybe 4 people offering help, was that followed up ?
Issues with packaging such as version dependencies or poor choice of versions for "stable" packages, notwithstanding. Maybe the answer is to leave "stable" versions of client versions asis, and build a (mostly) self contained server version, with no legacy library deps.. System tools can stay dependent on the "stable" versions, the server version can include newer libraries & tools.. no need to invent artificial barriers to progress..
Not to beat a dead horse and not to bash on Debian (personally Debian is the only distro I use), but to further help other people make a decision as to which version they should/may want to install: the slapd package included with Debian or the latest version from source:
The Debian community is fully aware of the numerous issues with their OpenLDAP package and acknowledges that it needs work, they have also been asking for help with OpenLDAP for some time (1878 days according to their "work-needing packages" list):
openldap (#512360), requested 1878 days ago
Description: OpenLDAP server, libraries, and utilities
Reverse Depends: 389-admin 389-ds-base 389-ds-base-dev
389-ds-base-libs 389-dsgw adcli alpine am-utils aolserver4-nsldap
apache2-bin (200 more omitted)
Installations reported by Popcon: 163954As the entry specifies bug #512360 in the BTS gives additional information about what work is needed: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=512360
Thanks,
Josh