I am sorry for the long delay. Finally I have installed a bunch of
changes for BBDB. See the ChangeLog attached below for the complete
list of changes. Thanks to Sam Steingold for submitting several
patches. Also, thanks to Philip Hudson for a bunch of docstring
fixes.
BBDB is available at
http://savannah.nongnu.org/projects/bbdb/
To check it out, use
git clone git://git.savannah.nongnu.org/bbdb.git
I'll try to summarize important changes:
I renamed bbdb-no-duplicates to bbdb-allow-duplicates.
The problem was that this feature was broken in several ways: First,
the docstring of bbdb-no-duplicates did not match what the variable
name suggested. Second, the docstring recommended not to have
duplicates, but by default duplicates were allowed. Third, the old
code was broken that was supposed to detect duplicates.
For the renamed variable, I did not change its docstring. By default,
BBDB now throws an error if you try to use duplicate names, and all
this should work as promised.
Another old bug (going back to BBDB v2): bbdb-complete-mail did not
obey bbdb-completion-list if this was a list of symbols. This should
be fixed now.
On Mon Dec 26 2011 Sam Steingold wrote:
I want to avoid putting the email uid into the first name slot.
I.e., when creating a record for f...@bar.com, I want the names to
be void, not foo .
Set the new variable bbdb-message-mail-as-name to nil.
Also, you might want to customize the new var bbdb-message-clean-name-function.
On Tue Dec 27 2011 Sam Steingold wrote:
When visiting an article from foo b...@baz.com who is not in my db,
I am offered to add name foo and address b...@baz.com to the record
of First Last - foo, i.e., a person whose _organization_ is foo.
I think this is wrong: matching the name foo against organization
foo.
The related searching of the hash table should be more robust now
such that it uses only the relevant fields of the records.
On Wed Dec 28 2011 Sam Steingold wrote:
If I set bbdb-check-auto-save-file to t, modify the db and not
save it right away, I am asked all the time about recovering bbdb
from the auto save file. with v2, I was asked this only when bbdb
was first loaded, which was useful.
Now with bbdb-check-auto-save-file set to t, BBDB should check the
auto-save file only once when actually reading bbdb-file (though as
we discussed before, under normal cirumstances there should be no
need for bbdb-check-auto-save-file being set to t).
On Thu Feb 9 2012 Sam Steingold wrote:
Here is yet another patch:
it is very annoying that some messages are marked as coming from a known
sender but ':' does not show them because known senders are matched by
e-mail and name and the ':' searches just by email.
this patch makes them consistent.
Done.
On Fri Mar 16 2012 Sam Steingold wrote:
I get asked the above question for matching records with an email but
without a name:
Handling of records without a name should be more robust now, though
there could still be unexpected surprises. I consider such records
somewhat exceptional and expected behavior becomes a matter of
personal context (and taste).
On Tue Jan 3 2012 Sam Steingold wrote:
I have 3 more patches which, I think, are quite uncontroversial:
* 6abeb87..: Sam Steingold 2011-12-27 Use `symbol-value' instead of `eval' for
symbols.
Done.
* 77d6afa..: Sam Steingold 2011-12-27 ignore Makefiles and configure files;
autoloads-loaddefs
I could not find this patch.
* 2aa061a..: Sam Steingold 2011-12-27 * lisp/bbdb-mua.el
(bbdb-annotate-message): Do not offer to replace the name if it is already an
AKA.
Done.
BBDB internals:
bbdb-record-set-field is now the recommended low-level function to
set any field. Here the symbol `notes' denotes the most common note
field, whereas the symbol `Notes' denotes the element of the BBDB
record structure holding all note fields. To avoid name clashes,
`Notes' may not be used to denote a note field.
2012-06-24 Sam Steingold s...@gnu.org
* lisp/bbdb-mua.el (bbdb-update-records): Use bbdb-message-search
instead of bbdb-search when `update-p' is `search' so that all
senders are displayed.
* lisp/bbdb-mua.el (bbdb-annotate-message): Do not offer to
replace the name if it is already an AKA.
* lisp/bbdb.el (bbdb-label-completion-list): Use symbol-value
instead of eval for symbols.
* lisp/bbdb-mua.el (bbdb-update-records): Ditto.
2012-06-24 Philip Hudson phil.hud...@iname.com
* lisp/bbdb.el (bbdb-read-only, bbdb-initialize-hook)
(bbdb-mode-hook, bbdb-layout-alist, bbdb-case-fold-search)
(bbdb-message-caching, bbdb-complete-mail-allow-cycling)
(bbdb-after-save-hook, bbdb-completion-display-record)
(bbdb-update-records-address, bbdb-warn, bbdb-split, bbdb-concat)
(bbdb-read-string, bbdb-current-record, bbdb-debug)
(bbdb-timestamp, bbdb-creation-date, bbdb-gethash)
(bbdb-hash-record, bbdb-record-name, bbdb-record-name-lf)