Re: [dumb] one you run ./configure then make

On Tue Dec 31 2019 Harry Putnam wrote:
> [email protected] (Dave Goldberg) writes:
> 
> > I went through this (finally) a couple years ago.  Yes, just put a
> > copy (good to keep a backup of the 2.X version for at least a while)
> > of your old .bbdb in ~/.  When you run M-x bbdb from emacs it will
> > detect the old version, convert it and offer to save.  It worked about
> > 90% perfectly for me.  I don't recall all the details but one issue
> > that came up is that some phone numbers, all of which were recorded as
> > strings in my 2.X file, were converted into an array.  bbdb itself
> > didn't seem to care which format was used - I only noticed because
> > some custom code I had choked on the records that were turned into
> > arrays.
> 
> Good to hear some personal experience on this... I wonder if arrays are
> less labor to retrieve as needed and less space intensive when
> storing?
> 
> When you said bbdb didn't seem to care about the format... do you mean
> it would display phone numbers in more than one format?  Whatever it
> was saved as?

With bbdb-phone-style bound to 'nanp ("North American Numbering
Plan", the default), BBDB tries to treat phone numbers like numbers
used in North America.  How this works is described in the docstring
of bbdb-parse-phone.

The NANP numbers are stored as a vector of numbers.  If BBDB cannot
treat a number according to NANP, it will store a string where it
preserves your input whatever it was.  (You may include any
characters if you like.)

If you use BBDB to let your computer dial numbers, the vector of
numbers are probably the better way to go.  You need to set
bbdb-dial-function for this.  (There is some protocol for this that
I am not familiar with.  By default, this goes via browse-url, so it
is not anything specific to BBDB.)

Beyond that, unless you like to put all numbers of your local white
pages into BBDB, I do not expect any significant difference between
the two formats for storing phone numbers (and I do not know which
one might turn out to be more efficient according to whatever
criteria).

All phone numbers are formatted in your *BBDB* buffer by calling
bbdb-phone-string.  This handles both NANP numbers and other
numbers.  So from that perspective you should not notice much of a
difference between the two formats.