On Mon, Apr 25, 2016 at 12:10 PM, Eduard Moraru <[email protected]> wrote:
> On Fri, Apr 22, 2016 at 8:08 PM, Thomas Mortagne <[email protected]>
> wrote:
>
>> On Fri, Apr 22, 2016 at 5:59 PM, Clemens Klein-Robbenhaar
>> <[email protected]> wrote:
>> >
>> >> On Fri, Apr 22, 2016 at 5:34 PM, Thomas Mortagne
>> >> <[email protected]> wrote:
>> >>> Sounds good.
>> >>>
>> >>> What about upgrades, do you plan to provide a migrator ?
>> >>
>> >> (I would love to see my old calendars converted to new style ;))
>> >>
>> >
>> > I have to check out the
>> http://extensions.xwiki.org/xwiki/bin/view/Extension/Nested+Pages+Migrator+Application
>> > and see how it works / if it can be adapted to nested pages, etc.
>> >
>> > I guess this will be a separate application / helper etc in any case, or
>> is there a way to trigger the migration when installing a newer version?
>>
>> There is an ExtensionUpgradedEvent event that an extension can listen
>> to be called after it's upgraded.
>>
>
> Cool! Did not know about that. So we can have per-extension migration
> scripts, which would kind of superseed the XWiki migration framework (that
> work with the XWiki DB version) and we would only be still using that for
> core changes. Right now we use the XWiki DB version even for bundled app
> migrations.
>
> I guess we should document that somewhere and make it a best practice (if
> not already).

It's OK in Java but impossible to use that in a XAR extension that
might be installed without programming right (since wiki component
require PR).

>
> Thanks,
> Eduard
>
>>
>> >
>> > No deadline for this in my scedule yet, however ....
>> >
>> >
>> >>>
>> >>> On Fri, Apr 22, 2016 at 5:31 PM, Clemens Klein-Robbenhaar
>> >>> <[email protected]> wrote:
>> >>>>
>> >>>> I see there has been some discussion on the list, as far as I
>> understand there is no decision about it in general.
>> >>>>
>> >>>> However I would like to drop support for pre-nested pages for the
>> mocca calendar on the master branch right now.
>> >>>>
>> >>>> The pre-nested spaces structure is:
>> >>>>
>> >>>>   - Calendars are plain pages
>> >>>>   - Events are pages that have their calendar as parent page, but are
>> usually placed as siblings in the same space
>> >>>>     (necessarily w/o nested pages)
>> >>>>
>> >>>> It is more natural to have calendars as nestable, non-terminal pages,
>> and the corresponding events nested inside these calendar pages.
>> >>>>
>> >>>> Bugs that would be easy to solve after dropping support:
>> >>>>
>> >>>>  http://jira.xwiki.org/browse/MOCCACAL-91  "Cannot create two events
>> with the same title in the same calendar "
>> >>>>    This has been fixed in 2.5.1 for pre-nested pages, but that fix
>> does not work for nested pages.
>> >>>>    Admittedly it could also be fixed with an  if (nested space) do x
>> else y
>> >>>>    I have to admit that I am reluctant to implement the if-else fork
>> and fully test it (If I do not test both nested and pre-nested spaces it
>> will be broken with high probability)
>> >>>>    In some places there are already horrible if-else clauses to
>> support both colibri and flamingo, which I put there and the experience
>> makes me unwilling to try this again with nested spaces.
>> >>>>
>> >>>>  http://jira.xwiki.org/browse/MOCCACAL-93  "Preinstalled "Other
>> Events" Calendar should not be a terminal page"
>> >>>>    This actually can not be fixed in a pre-nested spaces compatible
>> way (I think)
>> >>
>> >> Another great feature of nested space based calendar: you can watch a
>> >> single calendar just by watching its page instead of having to watch
>> >> eveything.
>> >>
>> >>>>
>> >>>>
>> >>>> Proposal:
>> >>>>  a) there is already a "stable-2.5" branch.
>> >>>>     all fixes and improvements for pre-nested spaces will/should go
>> on this branch
>> >>>>  b) master will switch its xwiki platform dependency from 5.4 to 7.4.2
>> >>>>     (not 7.2 for me, because 7.4.2 is what I am willing to test
>> against ... any takers to test with 7.2 ?)
>> >>>>  c) for now, the old "parent-child" relationship will still be used
>> to determine which events belong to which calendar
>> >>>>     so no migration from pre-nested spaces is really necessary - at
>> least I think so. Of course this should be tested.
>> >>>>     (It would be nice to have such migration, of course. I am not
>> sure how to write this, however.)
>> >>>>
>> >>>>
>> >>>> Any opinions? Should I send an official [vote] request?
>> >>>>
>> >>>> Clemens
>> >>>> _______________________________________________
>> >>>> devs mailing list
>> >>>> [email protected]
>> >>>> http://lists.xwiki.org/mailman/listinfo/devs
>> >>>
>> >>>
>> >>>
>> >>> --
>> >>> Thomas Mortagne
>> >>
>> >>
>> >>
>> >
>> > mit freundlichen Grüßen
>> > Clemens Klein-Robbenhaar
>> >
>> > --
>> > Clemens Klein-Robbenhaar
>> > Software Development
>> > EsPresto AG
>> > Breite Str. 30-31
>> > 10178 Berlin/Germany
>> > Tel: +49.(0)30.90 226.763
>> > Fax: +49.(0)30.90 226.760
>> > [email protected]
>> > www.espresto.de
>> >
>> > HRB 77554 B - Berlin-Charlottenburg
>> > Vorstand: Maya Biersack, Peter Biersack
>> > Vorsitzender des Aufsichtsrats: Dipl.-Wirtsch.-Ing. Winfried Weber
>> > Zertifiziert nach ISO 9001:2008
>> > _______________________________________________
>> > devs mailing list
>> > [email protected]
>> > http://lists.xwiki.org/mailman/listinfo/devs
>>
>>
>>
>> --
>> Thomas Mortagne
>> _______________________________________________
>> devs mailing list
>> [email protected]
>> http://lists.xwiki.org/mailman/listinfo/devs
>>
> _______________________________________________
> devs mailing list
> [email protected]
> http://lists.xwiki.org/mailman/listinfo/devs



-- 
Thomas Mortagne
_______________________________________________
devs mailing list
[email protected]
http://lists.xwiki.org/mailman/listinfo/devs

Reply via email to