On 2005-10-12 13:21:36 -0400, Derek Martin wrote: >> Reflowing to the width of the terminal or GUI windows is, what >> format=flowed is all about. Not doing it is certainly a bug. I'm sorry, but that argument goes too short. format=flowed tells us that the author is happy about having their text re-flowed. It also makes it easier to re-flow text. But why does that mean that we *have* to reflow it, instead of making the best use of that information to make messages as readable as possible? > I would have to agree. I would never use this personally, but it > rather does seem to be the point. If everyone wanted to read > lines at n < 80 columns, we would not have any need for this > method of formatting at all! Wrong. People read e-mail on PDAs and mobile phones (with 40 or less columns -- i.e., line lengths that are far below the kind of formatting that you would want senders to inflict on "ordinary" readers), and format=flowed comes in very handy there. >>>Currently, the format=flowed handler will try hard not to >>>output lines over 77 characters long. I regularly use mutt in >>>xterms well over 80 columns wide, because that's useful for >>>the index. But I would object to a change to mutt that would >>>reformat any format=flowed message to the full width of that >>>window just because we can. >> Then this should be made configurable as almost everything else >> in mutt > Certainly, mutt could have a variable such as > flowed_text_max_width to solve this. If 0, mutt uses the whole > width of the window. It could also do that if set to a value > greater than the width of the current terminal. Then, everyone > gets what they want. Well, no. There's a second decision in the code: To only re-flow things if that has to be because the line doesn't fit the display. It shouldn't be too difficult to change that, but still, I rather stubbornly believe that mutt's current code gets it right. -- Thomas Roessler · Personal soap box at <http://log.does-not-exist.org/>.
Attachment:
pgpGHiEegCrGq.pgp
Description: PGP signature