pkgsrc-Changes archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]
Re: CVS commit: pkgsrc/mail/roundcube-plugin-carddav
roundcube-plugin-carddav: Update to 4.4.2
pkgsrc changes:
---------------
* Remove PHP_VERSIONS_ACCEPTED. No PHP incompatibility found. The only requirement is PHP >= 7.1.
upstream changes:
-----------------
Version 4.4.2 (to 4.4.1)
* Revert to a single release tarball. The new approach (compared to 4.4.1) to
avoid the issue with conflicting dependencies between those coming with
roundcube and those coming with the RCMCardDAV release tarball is to append
the RCMCardDAV to the end of the autoloader list, so the roundcube
dependencies are always tried first. This means if a library used by
RCMCardDAV already comes with roundcube, RCMCardDAV will also use that
version of the library. There is still possible problems left (i.e. package
that only comes with roundcube might have a dependency for that an
incompatible version is already included with roundcube). In the end, I
don't think there is a clean solution to this issue. If you want to avoid
this mess, don't use the release tarball but install using composer.
* No changes to RCMCardDAV itself
Version 4.4.1 (to 4.4.0)
* Create release tarballs with a PHP version emulated to 7.1.0 (minimum
needed by RCMCardDAV)
* No changes to RCMCardDAV itself
Version 4.4.0 (to 4.3.0)
* MySQL/PostgreSQL: Increase maximum length limit for addressbook name (Fixes
#382)
* Fix: log messages could go to the wrong logger (carddav_http.log) for a
small part of the init code
* Support setting roundcube's collected senders/recipients to addressbooks
from preset (Fixes #383)
* Provide tarball releases in two variants: Guzzle v6 (roundcube 1.5) and v7
(roundcube 1.6) (Fixes #385)
Version 4.3.0 (to 4.2.2)
* New: For preset addressbooks, are re-discovery is performed upon every
login by default. This means that newly added addressbooks on the server
are discovered and added, whereas addressbooks that have been removed from
the server are also removed from roundcube.
o NOTE TO ADMINS: If you are using addressbook presets, please read the
documentation on the new preset setting rediscover_mode to decide if
re-discovery is desired or not. The new default is functionally safe,
but performance can be improved if the new behavior is not needed.
o For manually-added addressbooks, this will require changes to the
rcmcarddav data model, which is planned for version 5.
o Version 5 will also have a more elaborate version of re-discovery that
will allow to configure it such that it does not happen on every login.
* MySQL: Convert potentially used row format COMPACT (was default up to MySQL
5.7.8, Maria DB 10.2.1) to DYANMIC in migration 12, which would otherwise
fail (Fixes #362). It requires some other settings that have to be configured
in the MySQL server configuration additionally, all of which are also
defaults since MySQL 5.7.7 / Maria DB 10.2.2.
Version 4.2.2 (to 4.2.1)
* Fix: Detect login via OAuth and prevent usage of encrypted password scheme
in this case. The passwords encrypted cannot be decrypted anymore when the
access token changes. In case such passwords have already been stored to
the DB, the user must enter and save the password via the preferences
again. In case of an admin preset where the password is marked as a fixed
field, the password should be updated on next login of the user.
Version 4.2.1 (to 4.2.0)
* Updated French translation (#355)
* Updated German translation
* Fix: Display error message when the sync fails (instead of showing a
success message with a duration of -1 seconds)
* Fix: Display error message when a card cannot be updated because it changed
on the server since the last sync.
* Fix #356: Don't create group vcards with duplicate members, don't fail to
process them if we receive one from the server
* New: Action in the preferences to clear the local cache of an addressbook,
to allow a full sync from scratch. This is meant to fix errors in the local
state that cannot be repaired from incremental syncs, as might be the
result of issue #356
Home |
Main Index |
Thread Index |
Old Index