On Thursday, 21st January 2010, "Julian H. Stacey" wrote:
>Reason I imagined it would bet better with an option on send,
>& auto detect on receive:
> There's perhaps people out there running CTM to distribute stuff
> other than FreeBSD source (other *BSD src, other data, some binary
> systems may have limited or no access to upgrade binaries except
> at release upgrade.
> Such users might not be on this list, as this list is more for
> the FreeBSD patches than the programs as such. So ideally a CTM
> would have a format rev. no, & receivers would first be updated
> to dual capable auto detect of old & new format, then later senders
> would reduce length of CTM lines sent.
ctm_rmail already works with any line length that is a multiple of 4.
A change to 72 characters per line in ctm_smail does not require anyone
to update ctm_rmail. We should just do it. It's safe.
In principle a revision number is a good thing for every file and
transport format, but at this late stage I doubt any benefit would
be gained by adding one to the ctm email format.