Mon, 27 Feb 2012 10:03:31 +0100 [web] Exposes the undo feature to user through a undo-history view (closes #893940)
Anthony Truchet <anthony.truchet@logilab.fr> [Mon, 27 Feb 2012 10:03:31 +0100] rev 8267
[web] Exposes the undo feature to user through a undo-history view (closes #893940)
Mon, 27 Feb 2012 09:45:48 +0100 [undo, transactions] Adds a property in the transaction API. (closes #2202953)
Anthony Truchet <anthony.truchet@logilab.fr> [Mon, 27 Feb 2012 09:45:48 +0100] rev 8266
[undo, transactions] Adds a property in the transaction API. (closes #2202953) Required by forthcoming changes in undo_web.diff
Mon, 27 Feb 2012 09:43:15 +0100 [repo, undo] Finish repository-side implementation of the undo feature (closes #893940)
Anthony Truchet <anthony.truchet@logilab.fr> [Mon, 27 Feb 2012 09:43:15 +0100] rev 8265
[repo, undo] Finish repository-side implementation of the undo feature (closes #893940) 1) Changes API for undo_transaction in order to raise an exception: the previous API returned a sequence of error, the user of the API needed to be careful to check explicitly. We change this for it to raise an UndoException with the list of errors as an attribute. 2) The server-side support for undoing update actions is finished 3) The configuration for `undo-support` is now a simple boolean Beware of the costs, both in DB space and in number of requests on the RDBMS backend, once this feture is activated. 4) Adds a txuuid param/attribute to the existing NoSuchTransaction exception.
Fri, 24 Feb 2012 12:35:24 +0100 [rql rewrite] when a subquery has to be introduced, properly return the snippet expression so that further expressions are properly ored. Closes #2207180 stable
Sylvain Thénault <sylvain.thenault@logilab.fr> [Fri, 24 Feb 2012 12:35:24 +0100] rev 8264
[rql rewrite] when a subquery has to be introduced, properly return the snippet expression so that further expressions are properly ored. Closes #2207180
Fri, 24 Feb 2012 12:44:28 +0100 [migration/3.14.4] use the right API else we break sqlserver instances (closes #2211248) stable
Aurelien Campeas <aurelien.campeas@logilab.fr> [Fri, 24 Feb 2012 12:44:28 +0100] rev 8263
[migration/3.14.4] use the right API else we break sqlserver instances (closes #2211248)
Thu, 23 Feb 2012 19:06:14 +0100 backport stable
Sylvain Thénault <sylvain.thenault@logilab.fr> [Thu, 23 Feb 2012 19:06:14 +0100] rev 8262
backport stable
Thu, 23 Feb 2012 19:05:36 +0100 [devtools] make a copy of the xvfb-run script so we can kill it properly stable
Julien Cristau <julien.cristau@logilab.fr> [Thu, 23 Feb 2012 19:05:36 +0100] rev 8261
[devtools] make a copy of the xvfb-run script so we can kill it properly Since changeset a3d3bdd46463 we don't run firefox directly but through a wrapper script. This meant that on stopping the test, we would kill that wrapper script but leave the Xvfb server and firefox processes running. So make a local copy of xvfb-run, and make it kill Xvfb when it goes down, not just when its client exits.
Thu, 23 Feb 2012 12:15:52 +0100 [ajax] use AJAX_BASE_URL instead of JSON_BASE_URL
Adrien Di Mascio <Adrien.DiMascio@logilab.fr> [Thu, 23 Feb 2012 12:15:52 +0100] rev 8260
[ajax] use AJAX_BASE_URL instead of JSON_BASE_URL JSON_BASE_URL / 'json' should be replaced by AJAX_BASE_URL / 'ajax'
(0) -3000 -1000 -300 -100 -30 -10 -8 +8 +10 +30 +100 +300 +1000 +3000 tip