On Apr 26, 2018, at 20:28, Noam Postavsky <npostavs@gmail.com> wrote:


I'm not sure I understand the question. This format call produces
messages that are semantically equivalent to the original concat calls
but the concat calls were not localizable so I changed them to something
that we'll eventually be able to properly extract and localize.

I meant just the outer (format "%s%s%s%s" ...) vs (concat ...).  Is that
also needed to make strings localizable?  If yes, I think adding a
comment would be helpful, because it looks a bit unnatural to me.

Oh, you mean (format "%s%s%s%s" ...) vs an eventual (concat "%s%s%s%s" ...) ?

I checked the lisp ref to find something that would make it a clear choice but did not find anything really conclusive.

What I did is I try to keep concat for simple strings. And anything that involved fixing complex native strings was either moved to format or something else.

There is nothing that makes concat of format etc more or less appropriate for l10n in my understanding. I guess I was influenced by the way the original authors abused concat to put native strings together...

+                (message "Operation finished. Packages that are no longer needed: %d. Type `%s' to remove them"

This line is getting pretty long, perhaps it should be broken up?

I guess you mean the line of code?

Yes.

It there a length limit in emacs code ?

Jean-Christophe Helary
-----------------------------------------------
http://mac4translators.blogspot.com @brandelune