[Kde-pim] Marketing blocker collection, DEADLINE: 2013-03-10
Martin Steigerwald
Martin at lichtvoll.de
Thu Mar 7 18:36:23 GMT 2013
Am Donnerstag, 7. März 2013 schrieb Andras Mantia:
> Lindsay Mathieson wrote:
> > Aside from that - just how are users going to fix a corrupted or out
> > sync "Cache but not really a cache". Because it has and will happen.
>
> I still feel confusion here. :) What gets out of sync of what?
> Akonadi db and the original storage? For that we have the syncing.
> Akonadi db and Akonadi config files? Well, we are just working on this
> issue, no?
>
> I guess I need to write the part2 of my blog:
>
> http://blogs.kde.org/2011/11/13/akonadi-misconception-1-where-my-data
>
> As I see the part1 was indeed not that complete regarding what exactly is
> in the database and how it acts (although it is valid what I wrote
> there).
I´d greatly appreciate it, cause reading the thread I get more and more
confused as to what is where.
Also I agree with Sven that robustness is absolutely critical. I think its
good to make it difficult for the user to destroy his data. As thus I also
think that Akonadi shall survive any deletion of the database without any
data loss on the mails - except for possibly not yet stored mails from the
write cache of it.
Any possible references from the text file bassed KMail configuration to the
database shall either break cleanly, so that filter rules just do nothing
anymore, or somehow be preserved during a database delete and replace
action.
I am not sure whether I´d cache the mails in the database at all. I always
thought the maildir like approach for caching mails in extra directory to be
quite nice. I´d cache the mails as files in the filesystem.
If all of this is not feasible and a deletion of the database cannot be
supported, then I think its crucial, to provide the users with easy
solutions to all the cases where they currently use a database removal and
recreate action to solve issues.
But still, even then I think its crucial to detect such a database
replacement action, put a BIG FAT warning window on next Akonadi start and
refuse anything that can do harm to the actual PIM data as mails, calendars
and such.
A PIM application suite stores *important* data for the users. And I think
the protection of the data has to go *first*. If the design does not allow
for it, I think its important to improve the design.
Loosing user data in not easy to understand circumstances for the users,
i.e. when the user itself did not delete it by her own hands, i.e. did not
rm -rf the directory with the mails itself or the mails stored on the mail
server, IMHO is a big fat absolute and top notch marketing blocker.
KMail 1 had some of these issues as well, like the no subject empty body and
header mails after some index breakage on some circumstances and upon faces
it I always felt the urge to use something else. Its just that I do have so
much stuff in KDEPIM that it would be quite some hassle to migrate.
Thus, the original PIM data is authoritative. KDEPIM / Akonadi may not
delete or destroy it in any way without explicit user consent.
--
Martin 'Helios' Steigerwald - http://www.Lichtvoll.de
GPG: 03B0 0D6C 0040 0710 4AFA B82F 991B EAAC A599 84C7
_______________________________________________
KDE PIM mailing list kde-pim at kde.org
https://mail.kde.org/mailman/listinfo/kde-pim
KDE PIM home page at http://pim.kde.org/
More information about the kde-pim
mailing list