>I suppose the dialog between the two of you has answered this question, but I >can't figure out what the answer was. If the answer depends on decisions you >have made to change nmh defaults, assume that that has happened. > >How do I get mhshow to use my current browser for html parts? If there is no >way for nmh to know determine my current browser, how do I tell that to nmh?
The changes to mhshow were discussed before, but I don't think everyone quite understood the implications. Let me explain in greater detail. For 1.6, the default behavior for mhshow is: 1) text parts are to be processed with utilities that output the text on standard out. Before the processing of each part was free to do whatever it wanted, which in really old versions of MH did things like invoking an xterm. 2) Everything is run under one pager. Previously each part was responsible for invoking the pager. I would like to emphasize these are the defaults. You can change them back to the previous (IMHO, sucky) behavior). So, you want to change 1). That's easy enough; just arrange for an appropriate entry for your your mhshow-show-text/html entry in your .mh_profile. What's an "appropriate entry" look like? No idea! Well, I mean, it will probably look something like: mhshow-show-text/html: launch-my-browser %F The man page for mhshow has more details. I don't know what launch-my-browser is supposed to be for your system; for MacOS X it would be a command called "open", but someone else can give you advice there. Now, there is a wrinkle. mhshow, by default, is expecting something to appear on stdout. It doesn't really NEED anything to appear on stdout, but less will wait until everything has closed stdout (usually by exiting). So if your "launch-my-browser" program is long-running, less will never complete because it's waiting for it to finish. You could fix that by doing something like: mhshow-show-text/html: launch-my-browser %F > /dev/null So less isn't waiting for it to finish. You could also use the -noconcat option, which reverts to the old behavior where the program used to process each part was responsible for handling the pager. --Ken _______________________________________________ Nmh-workers mailing list [email protected] https://lists.nongnu.org/mailman/listinfo/nmh-workers
