2001-10-23 <NotZed@Ximian.com>
* mail-mt.c (mail_user_message):
(mail_get_password): Removed, all functionality moved to
mail-session.
* mail-config.c (mail_config_write_on_exit): Check
threaded/preview hash is null before using it, its setup on demand
so itmight nto be initialised here.
* mail-session.c (request_password): Remove password_current
stuff.
(alert_user): Redont, similar to get_pass. Do things as async as
possible, and dont even wait for a response if we're not asking
for the cancel button (this may or may not be right behaviour -
need to check). mail_user_message() code replaced from the stuff
in mail-mt.c
(MailSession): Added a lock field.
(init): Setup lock.
(finalise): fRee lock.
(register_timeout): Redone. We now allocate our own 'timeoutid's,
and <> to the real things asynchronously. Use async_event's so we
can make sure we have no outstanding ones after shutdown.
(mail_session_enable_interaction): If interaction has been
disabled, and we have either a message-box open, or a password
request open and/or any pending message boxes/passwords, blow 'em
away.
(main_register_timeout): If we have pending remove of this same
timeout, dont do anything.
(timeout_timeout): Properly honour the result, remove the timout
if it returns false.
svn path=/trunk/; revision=13952
2001-10-23 Christopher James Lahey <clahey@ximian.com>
* e-table-field-chooser-item.c: Got rid of the gc argument to
e_table_header_draw_button.
(etfci_font_load): Just use the style font here.
* e-table-group-container.c (e_table_group_container_construct):
Changed this to just use the font from the style. Fixes Ximian
bug #11882.
* e-table-header-item.c: Got rid of the gc argument to
e_table_header_draw_button.
(ethi_font_load): Changed this to just fallback on the font from
the style if the fontname is NULL or doesn't load to a font.
Fixes Ximian bug #11882.
* e-table-header-utils.c, e-table-header-utils.h
(e_table_header_draw_button): Got rid of the gc parameter here and
changed to use the gc from the style of a button created for this
purpose, but not shown. Fixes Ximian bug #13251.
svn path=/trunk/; revision=13949
2001-10-23 JP Rosevear <jpr@ximian.com>
* gui/e-meeting-model.c (build_etable): no longer set the
scrollbar policy here
* gui/e-meeting-time-sel.c
(e_meeting_time_selector_update_main_canvas_scroll_region): add an
extra row to the height so the click to add row can be properly
seen
(e_meeting_time_selector_construct): set the scrollbar policy for
the etable scrolled
svn path=/trunk/; revision=13948
2001-10-23 JP Rosevear <jpr@ximian.com>
* cal-util/timeutil.c (icaltimetype_to_tm): convert an
icaltimetype to a tm
(tm_to_icaltimetype): vice versa
* cal-util/timeutil.h: new protos
* conduits/calendar/calendar-conduit.c: replace all mktime and
localtime calls (except for debugging calls)
* conduits/todo/todo-conduit.c: ditto
(comp_from_remote_record): make sure the completed time is in UTC
svn path=/trunk/; revision=13946
2001-10-23 <NotZed@Ximian.com>
* mail-mt.c: Added missing errno.h
(mail_msg_new): Fix the logic a bit, dont try to open the log file
unless logging is actually requested.
svn path=/trunk/; revision=13944
* mail-mt.c (mail_msg_new): If the fopen() fails (eg, because
evolution-mail was started by oafd with PWD=/), don't try to use
the NULL file handle.
svn path=/trunk/; revision=13943
* camel-mime-utils.c (mail_list_magic): Remove an extra * in one
of the regexps that glibc apparently doesn't mind but bsd does.
svn path=/trunk/; revision=13942
2001-10-23 Rodrigo Moya <rodrigo@ximian.com>
* cal-client/cal-query.c (cal_query_construct) set priv->corba_query
to CORBA_OBJECT_NIL if there was an error
svn path=/trunk/; revision=13939
2001-10-23 Christopher James Lahey <clahey@ximian.com>
* gal/e-text/e-text.c (e_text_draw): Use the widget's style's
text_gc for drawing text if either draw_background or draw_button
is on.
svn path=/trunk/; revision=13937
2001-10-23 Christopher James Lahey <clahey@ximian.com>
* e-table-header-item.c, e-table-header-item.h: Got rid of the gc
field here and just use the fg_gc from the style for drawing the
buttons.
svn path=/trunk/; revision=13936
2001-10-23 Christopher James Lahey <clahey@ximian.com>
* configure.in: Bumped required GAL to 0.15.99.3 so EEntry's
"draw_button" argument can be used in
addressbook/gui/component/select-names/e-select-names.c.
From addressbook/ChangeLog:
2001-10-23 Christopher James Lahey <clahey@ximian.com>
* gui/component/select-names/e-select-names.c
(e_select_names_add_section): Set "draw_button" argument to TRUE
on this EEntry.
svn path=/trunk/; revision=13935
2001-10-23 Christopher James Lahey <clahey@ximian.com>
* configure.in: Bumped version to 0.15.99.3 for EText's
"draw_button" argument.
* gal/e-text/e-entry.c: Forward the "draw_button" argument to the
contained EText.
* gal/e-text/e-text.c, gal/e-text/e-text.h (e_text_draw): Added
new "draw_button" argument. If this is turned on and the first
widget that's a parent of the containing canvas with GTK_NO_WINDOW
turned off above the current one is a button, EText draws a button
underneath itself to match the containing button.
From gal/shortcut-bar/ChangeLog:
2001-10-23 Christopher James Lahey <clahey@ximian.com>
* e-shortcut-bar.c (e_shortcut_bar_add_group): Turn on
"draw_button" argument to EEntry.
svn path=/trunk/; revision=13934
2001-10-23 <NotZed@Ximian.com>
* mail-session.c (get_password): Completely re-done. We now hae a
completely async dialogue when requested from antoehr thread, and
dont use gtk_main() if we can avoid it (which is normally the
case). This stuff is only partially finished, and will mena the
removal of the same from mail-mt.c, and the mail_user_message()
code will be moved here and changed to work in a similar way.
* mail-callbacks.c (empty_trash): Dont try and connect to remote
stores just to get the trash. Also, always run empty trash async,
and make sure we unref the trash.
(empty_trash): Hmm, dont unref the trash, causes a problem on
exit, i suspect something else is doing funky unrefs on it.
* mail-tools.c (mail_tool_get_trash): Pass a 'connect' arg, tell
it whether it should tryand connect or not to the parent service,
if it isn't already connected.
* component-factory.c (owner_unset_cb): Dont try wait_all here,
could potentially deadlock.
(idle_quit): Keep returning TRUE if we have outstanding
processing. Note that this may busy-wait during exit processign
with busy tasks :(
(idle_quit): Keep calling ourselves till we no longer get called
(i.e. gtk_main really quits).
* mail-mt.c (mail_msg_active): New function, returns TRUE if
events are still active/outstanding.
(do_op_status): @$@$#@@!#@!! didn't unlock the mail_msg_lock if
data->activity was NULL and we had no global_shell_client anymore!
Also shortcut processing if this is going to be the case.
(mail_msg_init): Setup a temporary other gui_port for redoing with
new semantics password, user message and progress reporting.
(mail_get_password): #ifdef'd out all this code temporarily, till
it gets fully moved to mail-session.c
2001-10-22 <NotZed@Ximian.com>
* component-factory.c (owner_set_cb): Dont call enable_interaction
here.
(interactive_cb): But here instead, let the shell tell us when its
ok to go interactive.
svn path=/trunk/; revision=13933
2001-10-22 Christopher James Lahey <clahey@ximian.com>
* e-table-item.c (eti_idle_maybe_show_cursor): Make this check
immediately for whether to show the cursor, instead of in the idle
handler.
* e-tree-table-adapter.c (e_tree_table_adapter_node_set_expanded):
Added a pre_change here since we send two changes.
svn path=/trunk/; revision=13931
2001-10-22 Christopher James Lahey <clahey@ximian.com>
* e-tree-memory.c (e_tree_memory_sort_node): Call
e_tree_model_pre_change here where appropriate.
svn path=/trunk/; revision=13929
2001-10-22 Christopher James Lahey <clahey@ximian.com>
* e-table-item.c (eti_cancel_drag_due_to_model_change): Cancel the
drag or maybe drag if you have the mouse button down when the
pre_change signal comes in. Fixes Ximian bug #11309.
* e-table-without.c (add_row, remove_row,
e_table_without_show_all): Call e_table_model_pre_change where
appropriate here.
svn path=/trunk/; revision=13927
2001-10-22 Anna Marie Dirks <anna@ximian.com>
* evolution-message-composer.xml: Added an accelerator (H_TML) to the Format->Html
menu item.
svn path=/trunk/; revision=13926
2001-10-22 Jeffrey Stedfast <fejj@ximian.com>
* mail-display.h (mail_html_write_string): New fun macro.
* mail-format.c (write_one_text_plain_chunk): Use gtk_html_write
instead of mail_html_write so we avoid strdup'ing the text.
(handle_text_plain_flowed): Use the mail_html_write_string macro.
(handle_message_rfc822): Here too.
(mail_format_raw_message): We can do it here also...
(attachment_header): We can do it a little bit here...
(write_address): A microscopic bit here...
(write_hr): same.
svn path=/trunk/; revision=13925
2001-10-22 Damon Chaplin <damon@ximian.com>
* e-summary-tasks.c (generate_html): set the default timezone on the
server.
* e-summary-calendar.c (e_cal_comp_util_compare_event_timezones):
update to check for DATE values.
(uids_to_array): pass the default timezone
to cal_recur_generate_instances.
(generate_html): set the default timezone on the server.
svn path=/trunk/; revision=13921
2001-10-22 Damon Chaplin <damon@ximian.com>
* idl/evolution-calendar.idl: added setDefaultTimezone() method.
* pcs/cal-backend.c (cal_backend_get_default_timezone):
(cal_backend_set_default_timezone): new functions to call class
methods.
* pcs/cal-backend-file.c: lots of changes to handle the default
timezone and use it.
* pcs/query.c: use the default timezone.
* gui/dialogs/task-details-page.c (date_changed_cb): initialized
completed_tt.
* gui/dialogs/event-page.c: changed it to handle DATE values. The
'All Day Event' checkbox is only set now when the DTSTART and DTEND
are DATE values.
* gui/dialogs/comp-editor-util.c (comp_editor_free_dates): free the
CalComponentDateTime structs as well.
* gui/e-tasks.c: set the default timezone on the server.
* gui/tag-calendar.c:
* gui/gnome-cal.c:
* gui/e-week-view.c:
* gui/e-day-view.c: updates to handle DATE values.
* gui/e-calendar-table.c (date_compare_cb): updated to use the new
ECellDateEditValue values, so it now works.
(percent_compare_cb): updated to use GPOINTER_TO_INT values.
(e_calendar_table_init): use an ECellPercent for the percent field
and an ECellDateEditText for the date fields.
* gui/comp-util.c (cal_comp_util_compare_event_timezones): return TRUE
if the DTSTART or DTEND is a DATE value. We don't want to show the
timezone icons for DATE values.
* gui/comp-editor-factory.c (resolve_pending_requests): set the default
timezone on the server.
* gui/calendar-model.c: major changes to support sorting properly.
For date and percent fields we now use subclasses of ECellText, so
we don't use a char* as the model value. For the percent field we now
use a GINT_TO_POINTER. For the date fields we now use a
ECellDateEditValue* as the value.
* gui/calendar-config.c (calendar_config_configure_e_cell_date_edit):
set the timezone and use_24_hour flags of the new ECellDateEditText.
* conduits/todo/todo-conduit.c (pre_sync):
* conduits/calendar/calendar-conduit.c (pre_sync): set the default
timezone on the server.
* cal-util/timeutil.c (time_days_in_month): removed debug message.
* cal-util/test-recur.c: try to handle timezones in the iCalendar
file properly, and updated to pass default timezone.
* cal-util/cal-util.c (cal_util_generate_alarms_for_comp):
(cal_util_generate_alarms_for_list): added default timezone argument.
* cal-util/cal-recur.c: changed many of the functions to take a default
timezone, to use to resolve DATE and floating DATE-TIME values.
* cal-client/cal-client.c (cal_client_set_default_timezone): new
function to set the default timezone.
(cal_client_ensure_timezone_on_server): new function to ensure that
a given timezone is on the server.
* gui/e-cell-date-edit-text.c: new subclass of ECellText to display
and edit a date value.
* cal-util/cal-recur.c (cal_obj_byday_expand_monthly): changed week_num
to -week_num when calculating the weeks to go back from the end of the
month for things like BYDAY=-2WE. Fixes bug #11525.
(cal_recur_generate_instances_of_rule): only go up to MAX_YEAR (2037).
We can't really handle anything past that anyway.
(cal_recur_ensure_rule_end_date): initialize cb_date.end_date to 0,
so if the RULE doesn't generate COUNT instances we save 0 as the
time_t.
svn path=/trunk/; revision=13920
* e-shell-view-menu.c (command_open_folder_in_new_window): No need
to gtk_widget_show() the view anymore.
* e-shell-view.c (activate_shortcut_cb): No need to
gtk_widget_show() the view anymore.
* e-shell.c (impl_Shell_createNewView): Don't explicitly show the
new view.
(e_shell_create_view): gtk_widget_show() the new view and flush
the GTK+ event loop before sending the ::interactive notification.
(create_view): New helper function.
(e_shell_create_view): Use it.
(e_shell_create_view_from_settings): New.
(e_shell_restore_from_settings): Use
`e_shell_create_view_from_settings()'.
svn path=/trunk/; revision=13918
2001-10-22 Federico Mena Quintero <federico@ximian.com>
* gui/tasks-control-factory.c (tasks_control_factory_fn): Put up a
warning dialog box if we failed to create the tasks control.
Fixes bug #13033.
svn path=/trunk/; revision=13917
2001-10-22 Damon Chaplin <damon@ximian.com>
* e-cell-date-edit.c (e_cell_date_edit_set_popup_values):
(e_cell_date_edit_update_cell): use the new ECellText methods to get
the text and set the value.
* e-cell-percent.c: new subclass of ECellText to display & edit a
percentage value.
svn path=/trunk/; revision=13916
2001-10-22 Anna Marie Dirks <anna@ximian.com>
* mail-config.glade: Added several new accelerators to the edit account
dialog, and fixed the focus targets on all existing accelerators. I
was hoping that since the widgets used in the account editor are also
used in the first-time start-up druid and them mail-config druid,
that this would fix all accelerator problems for those druids as well,
but there is something wrong with Bonobo which is preventing
the druids from responding to their accelerators. Sigh. Anyway,
most accelerators in the edit account dialog work correctly
now, though there are few (such as _Full Name) which fail to
do anything. I don't know why.
svn path=/trunk/; revision=13914
2001-10-22 Damon Chaplin <damon@ximian.com>
* zoneinfo/*: updated all files again, placing current RDATEs first,
so Outlook uses those. Also fixed a few bugs in vzic which resulted in
a few changes.
* zoneinfo/Makefile.am (DIRS): added America/North_Dakota.
2001-10-22 Damon Chaplin <damon@ximian.com>
* src/libical/icaltimezone.c (icaltimezone_get_utc_offset):
(icaltimezone_get_utc_offset_of_utc_time): if we go off the start of
the changes array, return the TZOFFSETFROM of the first change.
Also changed the maximum year to 2037.
* src/libical/icaltime.c (icaltime_day_of_week):
(icaltime_start_doy_of_week):
(icaltime_week_number): init tm_hour to 12. Sometimes mktime()
adjusts the time, if that local time doesn't actually exist, leading
to the wrong day being returned. It is unlikely to adjust by 12 hours.
(icaltime_as_timet_with_zone):
(icaltime_from_timet_with_zone): change it back so it does convert
DATE values to/from the timezone. time_t values don't really go well
with DATE values, so be very careful when using them. We now assume
that the time_t points to the start of the day in the given timezone.
(We used to assume it pointed to the start of the day in UTC, but
that meant it was actually incorrect wrt the displayed calendar.)
* src/libical/icalrecur.c (expand_year_days): for FREQ=YEARLY with no
modifiers, we add one day, using the month and day from DTSTART.
(next_year): make sure we never go past 2037.
* scripts/mkderivedproperties.pl: Updated to allow DTSTART, DTEND, DUE
and RECURRENCE-ID to be set with DATE values. I think it now handles
all properties which can take DATE values, except RDATE which uses
DATE-TIME-PERIOD.
svn path=/trunk/; revision=13912
2001-10-22 Jeffrey Stedfast <fejj@ximian.com>
* mail-stream-gtkhtml.c (mail_stream_gtkhtml_new): New class that
wraps writing to a GtkHTML stream so that we don't have to write
to an intermediate GByteArray.
* mail-display.c (on_url_requested): Use the new Camel->GtkHTML
stream - this means we don't have to chew up nearly as much
memory...yay!
svn path=/trunk/; revision=13911
2001-10-22 <NotZed@Ximian.com>
* providers/local/camel-spool-summary.c (summary_rebuild): No, use
Storing, as the other code does :p
(spool_summary_check): Check for consistency.
svn path=/trunk/; revision=13904
