You can subscribe to this list here.
| 2007 |
Jan
|
Feb
|
Mar
|
Apr
(1) |
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
(1) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2008 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
(1) |
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2009 |
Jan
|
Feb
(2) |
Mar
(12) |
Apr
|
May
|
Jun
|
Jul
(12) |
Aug
(1) |
Sep
(4) |
Oct
(4) |
Nov
(1) |
Dec
|
| 2010 |
Jan
|
Feb
(15) |
Mar
|
Apr
|
May
(22) |
Jun
|
Jul
(13) |
Aug
(2) |
Sep
(6) |
Oct
|
Nov
|
Dec
(5) |
| 2011 |
Jan
|
Feb
|
Mar
|
Apr
(51) |
May
(7) |
Jun
(2) |
Jul
|
Aug
(2) |
Sep
|
Oct
|
Nov
(2) |
Dec
|
| 2012 |
Jan
|
Feb
(11) |
Mar
|
Apr
(5) |
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
(2) |
Dec
|
| 2014 |
Jan
|
Feb
|
Mar
(1) |
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| S | M | T | W | T | F | S |
|---|---|---|---|---|---|---|
|
1
|
2
|
3
(2) |
4
|
5
|
6
|
7
|
|
8
|
9
|
10
|
11
|
12
|
13
|
14
|
|
15
|
16
|
17
|
18
|
19
|
20
|
21
|
|
22
|
23
|
24
|
25
|
26
|
27
|
28
|
|
29
|
30
|
31
|
|
|
|
|
|
From: Paul P. <pa...@im...> - 2010-08-03 21:38:08
|
On Tue, Aug 03, 2010 at 05:42:16PM +0200, Hannes Wallnoefer wrote:
> I can shed some light on the OSX backspace issue. Apparently there is
> a "Delete sends Ctrl-H" option in the terminal prefs, and if that is
> deselected, my patched jline works correctly. The user who came up
> with this said it was working for him before and suspected this
> setting to have changed during a recent upgrade to Snow Leopard.
I left this out of my summary but I had tweaked that option on and off
without having much luck orienting on a deterministic outcome: but it
had become one of those infinitely tedious combinatorial problems, and I
wasn't sure how far one had to back out before the option change was
recognized. (Start new tab? Restart Terminal? Log out and back in?)
In any case, finding the right combo for all platforms sounds like a
fascinating job for the right person.
News of possible interest to jline users: I have full bash event history
syntax working in the scala repl, with "magic space" invocation, so you
can do something like
scala> val x = !54<space>
and the text of that event will be filled in. Also !?, !!, !-num, and
!#. However, I wrote it all in scala so someone would have to port it
to java for general usage, and I am not personally motivated in that
direction.
Another feature I implemented is multi-line history entries, so if you
enter something like
scala> object Foo {
| def bar = 5
| }
then it is only a single entry in the history. Since there was no
history storage format whatsoever, I had to invent something to have any
chance of correctly delimiting events. Has anyone else done this? Is
there any kind of standard? Any ideas? I generated start/end tokens on
lines by themselves when there was more than one line.
It basically works with a giant "except", and that is the same problem
which haunts the current ctrl-R implementations: when mutiple lines are
displayed via history, most of them remain on the screen when moving to
the next entry. Not so terrible at 2 lines but with 10 or 20 line
entries it's pretty uncool.
I don't suppose anyone has a good idea of how to address it in jline?
Looking at the backspace code, it seems to make some pretty broken
assumptions, like buffer.size / termwidth is the number of lines one is
displaying, i.e. there's no such thing as a newline. At least, this
looks like the relevant code to me:
private final int backspace(final int num) throws IOException {
if (buf.cursor == 0) {
return 0;
}
int count = 0;
int termwidth = getTermwidth();
int lines = getCursorPosition() / termwidth;
count = moveCursor(-1 * num) * -1;
debug ("Deleting from " + buf.cursor + " for " + count);
buf.buffer.delete(buf.cursor, buf.cursor + count);
if (getCursorPosition() / termwidth != lines) {
if (terminal.isANSISupported()) {
debug("doing backspace redraw: " + getCursorPosition() + " on " + termwidth + ": " + lines);
printANSISequence("J");
flushConsole();
}
}
drawBuffer(count);
return count;
}
--
Paul Phillips | All men are frauds. The only difference between
In Theory | them is that some admit it. I myself deny it.
Empiricist | -- H. L. Mencken
i pull his palp! |----------* http://www.improving.org/paulp/ *----------
|
|
From: Hannes W. <ha...@he...> - 2010-08-03 15:42:23
|
2010/7/21 Paul Phillips <pa...@im...>: > On Wed, Jul 21, 2010 at 10:02:15AM +0200, Hannes Wallnoefer wrote: >> I don't own a Mac, so the commit was based on reports of others. But I >> just asked another Mac user and he confirmed backspace is working for >> him with my fork. > > There's quite a rabbit hole with this. I logged all the keycodes being > emitted by jline. The deal is that before your patch, both codes were > mapped to backward delete, as is apparently necessary because the stock > US keyboard settings in OSX terminal have that backward-deleting key in > the upper right (labelled "delete") emitting ^H but ctrl-H and fn-delete > emit ^?. I can shed some light on the OSX backspace issue. Apparently there is a "Delete sends Ctrl-H" option in the terminal prefs, and if that is deselected, my patched jline works correctly. The user who came up with this said it was working for him before and suspected this setting to have changed during a recent upgrade to Snow Leopard. Hannes > So as it was there is no forward delete. All three erase backward. > With your patch delete and fn-delete do the right thing, but ctrl-H goes > the wrong way since it emits the fn-delete code. > > I do not think there's any sane simple configuration which will do the > right thing for everyone. > > -- > Paul Phillips | Eschew mastication. > Caged Spirit | > Empiricist | > all hip pupils! |----------* http://www.improving.org/paulp/ *---------- > |