This commit reflects a move of the 'ben-mac-port' branch start-point
tag on a subset of the source tree, via CVS's ability to move a branch
tag after the fact. The changed files are all copied from trunk r190.
Ben Harris [Tue, 6 Apr 1999 23:18:50 +0000 (23:18 +0000)]
Mac networking ios still as shafted as ever, but I should probably commit what
I _have_ done. It does manage to open a connection to the correct port on the
correct host -- it just then bombs the client with a "CHK error". Pity.
Ben Harris [Sun, 4 Apr 1999 18:23:35 +0000 (18:23 +0000)]
Lots of MacTCP networking stuff. I think all the code's written, but it
doesn't actually work yet. Also the telnet backend is still single-session,
and I haven't even touched the ssh one. Oh, and the "Main" segment just
overflowed 32k, so we have extra linker incantations. I'm off to the pub
now.
Ben Harris [Tue, 30 Mar 1999 19:44:52 +0000 (19:44 +0000)]
Fonts where boldface is wider than usual are now beaten up so they fit.
Note that this doesn't happen for non-COlor QuickDraw yet, due to the lack of
CharExtra().
Ben Harris [Sat, 27 Mar 1999 15:39:46 +0000 (15:39 +0000)]
General cleanup because MrC is rather pickier than SC. Not that the PPC build
actually links, you understand, but it does most of the rest of the things it
should.
Ben Harris [Tue, 23 Mar 1999 21:00:05 +0000 (21:00 +0000)]
Ha! Nasty intermittent bug nailed (I think)!
"That word, resLocked: I do not think it means what you think it means."
HLock seems to be my friend though.
Ben Harris [Thu, 18 Mar 1999 00:04:34 +0000 (00:04 +0000)]
Backends now seem to work. Pasting still doesn't.
macterm.c now uses UPPs for a few things so that it should work if compiled
natively for the PowerPC. Not tested it yet though.
Ben Harris [Sun, 14 Mar 1999 17:43:35 +0000 (17:43 +0000)]
Attribute masking implemented. The variable attr_mask stored the current
set of "interesting" attributes, and do_paint only takes those ones into
account when deciding when to break style runs. Makes redraw quite a bit
faster on displays which don't do ANSI colour.
Ben Harris [Sun, 14 Mar 1999 15:48:35 +0000 (15:48 +0000)]
There's something going wrong here. Occasional builds have difficulty loading
the wordness resource. I've added extra diagnostics, but I haven't worked out
what's going on.
This commit reflects the start of the 'ben-mac-port' branch in CVS.
Due to CVS's ability to branch files independently, most of the tree
comes from the tip of 'ben-early-mac-port', but terminal.c is the
version from r70
Ben Harris [Tue, 9 Mar 1999 11:57:13 +0000 (11:57 +0000)]
Removed support for Line Feed/New Line Mode on the grounds that it doesn't
do the same as lfhascr -- it also changes the "return" key to send CR/LF, which
lfhascr doesn't do (and shouldn't do). Also, Line Feed/New Line mode isn't
meant to affect IND, which lfhascr currently does (but arguably shouldn't).
Ben Harris [Sun, 7 Mar 1999 23:19:02 +0000 (23:19 +0000)]
Lots of things, notably:
* Tidied state machine, with better handling of unknown escape and control
sequences.
* Support for automatic newline mode, with lfhascr now being a session
variable.
* #include <string.h> for mem*() and str*().
Ben Harris [Wed, 3 Mar 1999 22:03:54 +0000 (22:03 +0000)]
Resizing of terminal windows added. Seems to work, but term_size might benefit
from calling scroll_display. This is where my scroll-optimisations look less
useful than Simon's.
Ben Harris [Tue, 2 Mar 1999 23:19:20 +0000 (23:19 +0000)]
We now have saner handling of 4-bit modes (disable colour, get the Palette
Manager to give us bg, fg, bold fg and cursor).
I've also removed some traces of the old scroll-optimiser.
Ben Harris [Mon, 1 Mar 1999 23:43:29 +0000 (23:43 +0000)]
We now use DeviceLoop for redraws, which allows for depth-specific tweaks.
The first of these is that in 1-bit modes we ignore colour attributes entirely,
which has nice effects on some of Fragments efforts if nothing else.
We still seem to be a lot slower than BetterTelnet, which is a bit worrying.
Ben Harris [Mon, 1 Mar 1999 22:26:50 +0000 (22:26 +0000)]
Right, so that's a hook for scroll optimisations (when Simon gets round to
them) and a vaguely-working scrollbar. Next stop: faster and more sensible
redraw in 1-bit modes at least. Waiting five minutes for the test case
to scroll past is _tedious_.
Ben Harris [Sun, 28 Feb 1999 17:05:11 +0000 (17:05 +0000)]
Erm, major change at the moment is a new test case. You need a file called
'fragment' to build it, but its licence prevents me distributing it at present.
Now, about those scrolling optimisations...
Ben Harris [Sun, 28 Feb 1999 02:38:41 +0000 (02:38 +0000)]
Right, we now have:
* The ability to load the configuration from the current resource file (ie the
application for now).
* A scroll bar and size box that look right, even if they don't actually _do_
anything yet.
* A load of other minor tweaks.
Ben Harris [Fri, 19 Feb 1999 23:51:21 +0000 (23:51 +0000)]
We now have an event loop, albeit one which ignores everything going on
around it, then proclaims itself to be bored and leaves in a huff. I
suspect this coulbe be turned into a satirical comment on someone, but I
can't be bothered right now.