view nextstep/AUTHORS @ 108750:3339da3cfeb3

Redesign bidi-aware edge positions of glyph rows, fix bug #6036. dispextern.h (struct glyph_row): New members minpos and maxpos. (MATRIX_ROW_START_CHARPOS, MATRIX_ROW_START_BYTEPOS) (MATRIX_ROW_END_CHARPOS, MATRIX_ROW_END_BYTEPOS): Reference minpos and maxpos members instead of start.pos and end.pos, respectively. xdisp.c (display_line): Compare IT_CHARPOS with the position in row->start.pos, rather than with MATRIX_ROW_START_CHARPOS. (cursor_row_p): Use row->end.pos rather than MATRIX_ROW_END_CHARPOS. (try_window_reusing_current_matrix, try_window_id): Use ROW->minpos rather than ROW->start.pos. (init_from_display_pos, init_iterator): Use EMACS_INT for character and byte positions. (find_row_edges): Renamed from find_row_end. Accept additional arguments for minimum and maximum buffer positions seen by display_line for this row. Don't use iterator to find the position following the maximum one; instead, increment the position found by display_line directly. Fix logic; eol_pos should be tested before the rest. Handle the case of characters delivered from display vector (bug#6036). Fix tests related to it->method. Handle the truncated_on_right_p rows. (RECORD_MAX_MIN_POS): New macro. (display_line): Use it to record the minimum and maximum buffer positions for glyphs in the row being assembled. Record the position of the newline that terminates the line. If word wrap is in effect, restore minimum and maximum positions seen up to the wrap point, when iterator returns to it. (try_window_reusing_current_matrix): Give up if in bidi-reordered row and cursor not already at point. Restore original pre-bidi code for unidirectional buffers. dispnew.c (increment_row_positions, check_matrix_invariants): Increment and check row->start.pos and row->end.pos, in addition to MATRIX_ROW_START_CHARPOS and MATRIX_ROW_END_CHARPOS. .gdbinit (prowlims): Display row->minpos and row->maxpos. Display truncated_on_left_p and truncated_on_right_p flags. Formatting fixes. (pmtxrows): Display the ordinal number of each row. Don't display rows beyond the last one. bidi.c (bidi_cache_iterator_state): Don't zero out new_paragraph: it is not copied by bidi_copy_it.

In addition to the folks listed in ../AUTHORS responsible for GNU Emacs itself,
the NeXTstep port owes to the following people:

Carl Edman
    original author and maintainer, mainly UI
Michael Brouwer
    heavy contributor, input handling and other areas
Christian Limpach
    help / maintenance on NeXTstep
Scott Bender
    OpenStep, Rhapsody ports
Christophe de Dinechin
    MacOS X port
Adrian Robert
    GNUstep port, update Emacs 20 -> 21+

Joe Reiss
    popup menu, dialog boxes; icons
Andrew Athan
    font panel integration
Scott Byer
    improved rendering code
Scott Hess
    keyboard handling suggestions

Rahul Abrol
    "hide others" patch
Adam Ratcliffe
    preferences panel documentation
Peter Dyballa
    assistance with non-ASCII rendering and keyboard handling
David M. Cooke
    fix to XPM crash bug
Carsten Bormann
    initial patch and assistance getting dired working for non-ASCII filenames
Andrew Moore
    assistance on ns-mark-nav extension

The GNUstep port was made possible through the assistance of Adam
Fedor, Fred Kiefer, M. Uli Klusterer, Alexander Malmberg, Jonas
Matton, and Riccardo Mottola.  Leigh Smith maintained the SourceForge
project for a period.

Suggestions from Darcy Brockbank, Timothy Bissell, Scott Byer, David
Griffiths, Scott Hess, Eberhard Mandler, John C. Randolph, and Bradley
Taylor all helped things along at one point or another.  Axel Seibert
<seiberta@@informatik.tu-muenchen.de> and Paul J. Sanchez
<paul@@whimsy.umsl.edu> offered their time and machines to make a
binary release possible.

We would also like to thank a number of people who kept up the
constant supply of bug reports, suggested features and praise: Hardy
Mayer, Gisli Ottarsson, Anthony Heading, David Bau, Jamie Zawinski,
Martin Moncrieffe, Simson L. Garfinkel, Richard Stallman, Stephen
Anderson, Ivo Welch, Magnus Nordborg, Tom Epperly, Andreas Koenig,
Yves Arrouye, Anil Somayaji, Gregor Hoffleit; and the few hundred
other people on the mailing list from whom we didn't hear much, but
the presence of which assured us that maybe this project was actually
worth doing.