On Monday, January 20, 2003, at 10:27 PM, Mark Garringer wrote: > I can't figure out where the extra line is coming from. I'd imagine it > is > how the color codes are being processed (slightly?!?) differently in > 3.1, > but I'm not really familiar with comm.c code or where I might start > looking. > > yeah things are done a bit different in 3.1 when i was making the patches for the color string replacer functions i realized that the best places to place a function to filter color could lead to disaster due to the fact that the output processing expects to know how much of the buffer was written out and if you expanded the buffer below you'd have to do a bit of work to return the characters written out from the original string and not an expanded string. I ended sticking the proc_color functions in vwrite_to_output even though it returns the length written to the buffer since nothing currently using the function cares about the size being returned. The output paths could use a little reworking to allow for fairly low level filter functions that perform text manipulations as close to the socket as possible with being able to access the character structure but allow the code above this point to be oblivious to any changes that would be going on. Ron Fenner -- +---------------------------------------------------------------+ | FAQ: http://qsilver.queensu.ca/~fletchra/Circle/list-faq.html | | Archives: http://post.queensu.ca/listserv/wwwarch/circle.html | | Newbie List: http://groups.yahoo.com/group/circle-newbies/ | +---------------------------------------------------------------+
This archive was generated by hypermail 2b30 : 06/26/03 PDT