On Mon, 2012-08-06 at 16:33 -0400, Matthew Barnes wrote:
> On Mon, 2012-08-06 at 13:31 -0600, Brian A Anderson wrote:
> > If I follow what you said below,  Does this mean that I have to convert
> > 2.24.5 data to 2.32 data and then to 3.4 or is it just one conversion?
> > Is there a utility that someone has that might do this job?
> 
> You don't HAVE to, but I think what Andre meant is the bigger the jump
> the bigger the risk.
> 
> Migration routines are written to convert data from the previous major
> release to the upcoming major release at the time of writing.  So if I
> wrote a new routine today it would convert something from the way it's
> represented in 3.4 to the way it will be represented in 3.6, and that's
> about as much testing as it receives prior to release.
> 
> The theory goes, as the routines execute chronological order, the data
> undergoes one or possibly multiple conversions but should end up in the
> currently supported representation.  But as time passes, old migration
> routines may bit rot and silently break.
> 
> Case in point: I believe the mbox-to-Maildir conversion itself still
> works in 3.4 but the detection for when the conversion needs to run is
> currently broken because it relies on some subtle aspect of the startup
> sequence that has changed since 3.0.
> 

In my experience, one of the better ways to do data conversion is to
write an independent program to do it. Name it something like
evolution-convert-3.2-3.4. After enough time, you would have a
conversion program for each version and they could be run in order. They
could also run without requiring a full working Gnome/Evolution
environment so they could be used to convert offline data. On a backup
server for example.

-- 
Knowledge Is Power
Power Corrupts
Study Hard
Be Evil

Attachment: signature.asc
Description: This is a digitally signed message part

_______________________________________________
evolution-list mailing list
[email protected]
To change your list options or unsubscribe, visit ...
https://mail.gnome.org/mailman/listinfo/evolution-list

Reply via email to