Closed gmcninch closed 6 years ago
Oh dear. I had been subclassing an internal eieio class I wasn't supposed to, and in 0.5.4 changed the class structure, which meant renaming the "object-name" slot to "label". There's migration code to catch the old slot name and re-name it, but for some reason that code isn't doing its job in your case. I really don't know why.
Would you try this:
(ebdb-shutdown)
initialize-instance
method.If that doesn't fix the problem then I am well and truly stumped. If it does fix the problem, it means that there's something I don't understand about upgrading packages via the package manager (but implies that the problem is solvable).
Thanks!
Hi, thanks for the reply. Actually, the problem seems to have been my fault. I had written some "defmethods" for display of ebdb fields. I just commented that code out, restarted, and everything worked. I had "ruled that out" earlier, because I didn't think those methods would have been called. But I was evidently wrong...!
Looking more carefully, I had code for
(cl-defmethod ebdb-fmt-field ((_fmt ebdb-formatter) (field ebdb-field-address) (_style (eql collapse)) (_record ebdb-record)) ...
and
(cl-defmethod ebdb-fmt-field ((fmt ebdb-formatter-ebdb) (field ebdb-field-role) _style (record ebdb-record-person))
and I think that the point of my code was to include "multi-line" address information in the ebdb display.
I do note that at least the second of my defmethod's refers explicitly to
(slot-value field 'object-name)
Maybe it'll "just work" if I change that to 'label instead? I'll give it a shot...
Anyhow, sorry for the noise!
[edit: seems to work fine after replacing 'object-name with 'label ]
Excellent! Glad that wasn't my fault :) No guarantees about code structure stability until 1.0!
But I am glad that people are playing with this level of display code. It's an area that is still in flux. Do let me know if you have any requests/comments/confusions/ideas.
I just upgraded ebdb to 0.5.4 via emacs package manager. I then restarted emacs, and started gnus. Now if I try to compose a message (e.g. "m" from the group buffer), I get an error:
I'm not entirely sure why message/ebdb is looking at field-roles at this stage, but I note that I get the same sort of error if I enter a (partial) mail recipient in the "To:" field and press
<tab>
.This problem doesn't occur with 0.5.3 (but since no one else has reported this, it makes me wonder if there isn't something flaky about my contacts data?).
-george