Hi, On mer, fév 22, 2006, Adeodato Simó wrote: > So how should, in your opinion, things work? I mean, to which other > field do you steal one space so that upon overflowing the assigned > length for %C, thread markers are not shifted?
(Sorry for getting back to this bug so late.) There are a lot of ways in which it could be handled, with drawbacks and advantages for each different algorithm (surprize). It seems to me the layout of columns is a per-line process, with no reference to things such as "longest text of this column". With the current system, I would have to force the index column to grow to 4 chars of width, that would eat space in all folders I browse, even if they have less than 10 or less than 100 messages. It would be enough to have an index column always as large as the largest entry in the column. I imagine how it would be implemented, but I fear it might have serious speed drawbacks to implement it with strlen() calls on all values of all columns. Also, the message index is a bit special in that it's easy to have a clue on it's width via the number of messages. It's also different of the sender in that it's desirable not to truncate the index field, while it's certain that the sender field will be overflowed for certain values of the sender and of the window width. So, perhaps this column with the width being the width of the widest value only makes sense for the index. Bye, -- Loïc Minier <[EMAIL PROTECTED]> Current Earth status: NOT DESTROYED
