John Levon wrote:
On Tue, May 09, 2006 at 10:27:17PM +0200, Michael Gerz wrote:
case LFUN_CHANGE_REJECT: // what about these two
case LFUN_ALL_CHANGES_ACCEPT:
case LFUN_ALL_CHANGES_REJECT:
- flag.enabled(buffer_ && buffer_->params().tracking_changes);
+ flag.enabled(buffer_);
Will there be a plan to re-enable the test at some point? I.E. are you
going to make "there are unaccepted changes" an O(1) operation on the
buffer?
No. The test has been removed on purpose. In fact, the major motivation
of the CT code changes is the user requirement to switch CT off without
having to accept all changes before. I often switch off ct for simple
editoral changes in MS Word such that only major changes show up in the
documents that I share with others.
- if (features.isAvailable("dvipost") && params().tracking_changes
- && params().output_changes)
+ if (features.isAvailable("dvipost") && params().output_changes)
features.require("dvipost");
What does the UI for these two options look like?
Change tracking and change output can be enable/disabled. The current
status is indicated in the LyX menu.
// Are we in a DELETED change-tracking region?
- if (buf && buf->params().tracking_changes
- && lookupChangeType(cur, true) == Change::DELETED
+ if (buf && lookupChangeType(cur, true) == Change::DELETED
Presumably this is or is going to be a fast operation now?
I haven't cared for performance yet. AFAICS, CT will be at least as
"fast" as it is today, maybe even a bit faster because changes with type
UNCHANGED won't be stored anymore.
Michael