CLIM mail archive


Re: 1.1 dissappearing text ...

    Date: Wed, 20 Oct 1993 17:24 EDT
    From: Daniel D Suthers <>

    hmm ..... 

    > It appears in a context where we are using both
    > filling-output and indenting-output

Hmm, very interesting.  I definitely need a losing test case.
FILLING-OUTPUT is one of the few formatted output facilities that is
implemented by encapsulating other streams; it is certainly possible
that I simply bungled one of the places that communicates with the
incremental redisplay protocol.

    I do use indenting-output and filling-output ... just tried ablation:
    - presence/absence of indenting-output has no effect
    - the problem goes away when I remove filling-output, but I don't 
      know if that's just because of redraws due to scrolling.
    Looks like one more reason to revert to my old hack that did
    filling-output by inserting newlines (the other being I want the text
    displayed as it is "filled" so the user doesn't have to wait for filling
    of a long document to start reading it.) 

    I think this problem showed up when we started attempting incremental
    redisplay, but I'm not sure exactly when, since we were focusing on the
    graphics and not the text for a while. Using Scott's suggestion I did
    determine that the text disappears when the display is redrawn inside a
    call to CLIM:REDISPLAY. (It is drawn correctly when trace shows only my
    displayers being called; it is drawn as whitespace when the identical
    sequence of calls to my displayers are invoked via clim:redisplay.) 

     Dan Suthers           | LRDC, room 505A     | 3939 O'Hara Street
     (412) 624-7036 office | University of Pittsburgh
     (412) 624-9149 fax    | Pittsburgh, PA 15260


Main Index | Thread Index