[9fans] Shall we fix the use of Up/Dn arrows?

[9fans] Shall we fix the use of Up/Dn arrows?

[Yi DAI]

[-- Attachment #1: Type: text/plain, Size: 691 bytes --]

Hi,


I am new to Plan9. This request sounds quite simple/naive. Don't think that
I dislike the Plan9 ways. I enjoy it a lot. But it doesn't mean that it
needs no improvement/fix, right?

OK, here is my suggestion. We shall fix the use of Up/Dn arrows, I mean,
instead of using it to go one page up/dn, use it to move the cursor up/dn
one line. Here are two reasons for this:

1. We use Lf/Rt arrows to move the cursor, don't we? Then use Up/Dn to do
the same kind of task is symmetric, and thus more reasonable.
2. We already have PageUp/PageDown keys which do the task of paging up/dn,
and symmetrically, Home/End to move to the first and last page.


EOT



--
DAY

[-- Attachment #2: Type: text/html, Size: 885 bytes --]

Re: [9fans] Shall we fix the use of Up/Dn arrows?

[Akshat Kumar]

naah, let's use arrow keys to print arrows; there's nothing more literal,
and thus reasonable, than that

or, or, here's another rationalisation of some other obtuse usage...


nvm
ak


On Sat, Jan 23, 2010 at 1:15 AM, Yi DAI <plm.day@gmail.com> wrote:
> Hi,
>
> I am new to Plan9. This request sounds quite simple/naive. Don't think that
> I dislike the Plan9 ways. I enjoy it a lot. But it doesn't mean that it
> needs no improvement/fix, right?
> OK, here is my suggestion. We shall fix the use of Up/Dn arrows, I mean,
> instead of using it to go one page up/dn, use it to move the cursor up/dn
> one line. Here are two reasons for this:
> 1. We use Lf/Rt arrows to move the cursor, don't we? Then use Up/Dn to do
> the same kind of task is symmetric, and thus more reasonable.
> 2. We already have PageUp/PageDown keys which do the task of paging up/dn,
> and symmetrically, Home/End to move to the first and last page.
>
> EOT
>
>
> --
> DAY
>

Re: [9fans] Shall we fix the use of Up/Dn arrows?

[Bela Valek]

I agree, but only because PgDown and PgUp do the same as the Down and
Up arrows. Having two pairs of keys doing the same thing on the
keyboard is a waste. On the other hand, I found it easy to get used to
the Plan 9 approach. Its comfortable for me either way, I only think
that this modification would add a new functionality to the interface.

2010/1/23 Yi DAI <plm.day@gmail.com>:
> Hi,
>
> I am new to Plan9. This request sounds quite simple/naive. Don't think that
> I dislike the Plan9 ways. I enjoy it a lot. But it doesn't mean that it
> needs no improvement/fix, right?
> OK, here is my suggestion. We shall fix the use of Up/Dn arrows, I mean,
> instead of using it to go one page up/dn, use it to move the cursor up/dn
> one line. Here are two reasons for this:
> 1. We use Lf/Rt arrows to move the cursor, don't we? Then use Up/Dn to do
> the same kind of task is symmetric, and thus more reasonable.
> 2. We already have PageUp/PageDown keys which do the task of paging up/dn,
> and symmetrically, Home/End to move to the first and last page.
>
> EOT
>
>
> --
> DAY
>

Re: [9fans] Shall we fix the use of Up/Dn arrows?

[John Stalker]

> 2. We already have PageUp/PageDown keys which do the task of paging up/dn,
> and symmetrically, Home/End to move to the first and last page.

I don't have PageUp/PageDown, but I do have arrow keys.  Of course,
I'm typing this on a laptop.  I'm not saying this decision was made
for that reason, but it's something to consider.
--
John Stalker
School of Mathematics
Trinity College Dublin
tel +353 1 896 1983
fax +353 1 896 2282

Re: [9fans] Shall we fix the use of Up/Dn arrows?

[cinap_lenrek]

[-- Attachment #1: Type: text/plain, Size: 572 bytes --]

why not just hack rio and see how well it works out?  put the patch
somewhere for others to try. hacked a new piemenu patch that
allows sub menus (have put the hidden windows in a sub menu so that
the positions of the items in the pie doesnt change depending on the
hidden windows) and compacter layout algo.

so just do it...

i would prefer when the arrow keys start a new instance of
games/catclock btw. mutch more convinient :) for scrolling you
can use the mouse... this is mutch more consitant.... keyboard
should be for text and catclocks!

--
cinap

[-- Attachment #2: Type: message/rfc822, Size: 4839 bytes --]

[-- Attachment #2.1.1: Type: text/plain, Size: 691 bytes --]

Hi,


I am new to Plan9. This request sounds quite simple/naive. Don't think that
I dislike the Plan9 ways. I enjoy it a lot. But it doesn't mean that it
needs no improvement/fix, right?

OK, here is my suggestion. We shall fix the use of Up/Dn arrows, I mean,
instead of using it to go one page up/dn, use it to move the cursor up/dn
one line. Here are two reasons for this:

1. We use Lf/Rt arrows to move the cursor, don't we? Then use Up/Dn to do
the same kind of task is symmetric, and thus more reasonable.
2. We already have PageUp/PageDown keys which do the task of paging up/dn,
and symmetrically, Home/End to move to the first and last page.


EOT



--
DAY

[-- Attachment #2.1.2: Type: text/html, Size: 885 bytes --]

Re: [9fans] Shall we fix the use of Up/Dn arrows?

[Rudolf Sykora]

> 1. We use Lf/Rt arrows to move the cursor, don't we? Then use Up/Dn to do the same kind of task is symmetric, and thus more reasonable.

In case of being in an editor, editing some text, I'd say your idea
has a point. Also I find it disturbing when instead of moving up one
line the scrolling comes instead.
On the other hand, being in a rio window or an acme win window, moving
the cursor up or down using the arrows doesn't have much sense (in any
case you have to use a mouse to do something useful with a text
there).

What I'd appreciate, from the user's point of view, would be some kind
of history mechanism tied to the up/down keys in rio/win (where I
interact with a shell), up arrow bringing up the last command e.g.,
and a normal movement behaviour when editing a 'text' file (no direct
interaction with the shell, where 'enter' is just a newline).
On the other hand, implementing such a mechanism would probably lead
to ugly complications, which might be better be avoided. For now,
simplicity wins.

Btw., I know the " and "" scripts by R Cox, I use them. Without
something like them, it would be hell. Anyhow, hitting the up/down key
a few times is just quicker.

R

Re: [9fans] Shall we fix the use of Up/Dn arrows?

[Yi DAI]

[-- Attachment #1: Type: text/plain, Size: 319 bytes --]

On Sat, Jan 23, 2010 at 8:52 PM, Rudolf Sykora <rudolf.sykora@gmail.com>wrote:

> For now, simplicity wins.
>

> Btw., I know the " and "" scripts by R Cox, I use them. Without
> something like them, it would be hell. Anyhow, hitting the up/down key
> a few times is just quicker.
>
>
Agree, :)


--
DAY

[-- Attachment #2: Type: text/html, Size: 780 bytes --]

Re: [9fans] Shall we fix the use of Up/Dn arrows?

[Patrick Kelly]

> -----Original Message-----
> From: 9fans-bounces@9fans.net [mailto:9fans-bounces@9fans.net] On
> Behalf Of John Stalker
> Sent: Saturday, January 23, 2010 5:22 AM
> To: Fans of the OS Plan 9 from Bell Labs
> Subject: Re: [9fans] Shall we fix the use of Up/Dn arrows?
> 
> > 2. We already have PageUp/PageDown keys which do the task of paging
> > up/dn, and symmetrically, Home/End to move to the first and last page.
> 
> I don't have PageUp/PageDown, but I do have arrow keys.  Of course, I'm
> typing this on a laptop.  I'm not saying this decision was made for that reason,
> but it's something to consider.

What about Alt+'Direction' for those who don't have such keys?
This way everything works as expected (Law of least astonishment) and yet there is no functionality loss.

> --
> John Stalker
> School of Mathematics
> Trinity College Dublin
> tel +353 1 896 1983
> fax +353 1 896 2282

Re: [9fans] Shall we fix the use of Up/Dn arrows?

[blstuart]

> What I'd appreciate, from the user's point of view, would be some kind
> of history mechanism tied to the up/down keys in rio/win (where I
> interact with a shell), up arrow bringing up the last command e.g.,
> and a normal movement behaviour when editing a 'text' file (no direct
> interaction with the shell, where 'enter' is just a newline).

Please no!  I am constantly having to figure out how to turn off
that misfeature in this week's version of bash.  Maybe it only
happens once in 1000 times, but in that 0.1% of the time, my
finger coming down on the return key accidentally hits the up
arrow and the timing just so happens that I rerun the previous
command instead of the one I just typed, sometimes with very
unpleasant consqeuences.  I should point out that this of course
is primarily an issue with laptops and other small keyboard
designs.  Of course, those are the only ones I like to use...

If one really wants to give the shell some kind of previous
command shortcut, then stick with the EMACS ^p or the old
C-shell !! (the former, of course, allowing you to scroll through
the history if that is what is desired).

As for up-arrow moving up a line, this comes up from time to
time.  It's primary usefulness is in editors.  But in a proportional
font, what exactly does moving up a line mean?  Move to the
position that is the same number of characters from the beginning
of the line?  Move to the same geometric position?  What if that
position is in the middle of a character?  No one ever seems to
float a proposal that hits everyone as having "the ring of truth."
I have thought there could be some usefulness in this, but I
haven't thought of a policy that is so beautiful it draws me to
the source to code it.

As with any proposal like this, if it's what you want, then implement
it.  Put it out there and if enough others like it too, it may find
its way into general use.  If not, you still have the functionality
that you want.

BLS

Re: [9fans] Shall we fix the use of Up/Dn arrows?

[Rudolf Sykora]

> Maybe it only
> happens once in 1000 times, but in that 0.1% of the time, my
> finger coming down on the return key accidentally hits the up
> arrow and the timing just so happens that I rerun the previous
> command instead of the one I just typed, sometimes with very
> unpleasant consqeuences.

This is the 1st time I hear about such a misfortune... :)

> As for up-arrow moving up a line, this comes up from time to
> time.  It's primary usefulness is in editors.  But in a proportional
> font, what exactly does moving up a line mean?  Move to the
> position that is the same number of characters from the beginning
> of the line?  Move to the same geometric position?  What if that
> position is in the middle of a character?  No one ever seems to
> float a proposal that hits everyone as having "the ring of truth."
> I have thought there could be some usefulness in this, but I
> haven't thought of a policy that is so beautiful it draws me to
> the source to code it.

All usual editors that use proportional fonts have to cope with this somehow.
E.g. one of the most common in linux, openoffice's writer, seems to be
trying to move the cursor approximately to the same geometric
position.

R

Re: [9fans] Shall we fix the use of Up/Dn arrows?

[Russ Cox]

1. Many of us believe there is no problem to fix.  I am quite happy
with up and down for scrolling.

2. The Plan 9 editors have no built-in concept of a line.
(Going to line 10 in sam requires counting ten newlines from the beginning.)

3. Up/Down and PgUp/PgDown move different amounts: 1/3 and 2/3 of the
screen, respectively.

Russ

Re: [9fans] Shall we fix the use of Up/Dn arrows?

[Rudolf Sykora]

2010/1/23 Russ Cox <rsc@swtch.com>:
> 1. Many of us believe there is no problem to fix.  I am quite happy
> with up and down for scrolling.

Well, nobody did say there is anything to be fixed.
I myself am quite happy with it, too.

> 2. The Plan 9 editors have no built-in concept of a line.
> (Going to line 10 in sam requires counting ten newlines from the beginning.)

Not sure how the concept of a line delimited by newlines relates to
moving the cursor up one physical line on the screen.

Thanks
R

Re: [9fans] Shall we fix the use of Up/Dn arrows?

[Russ Cox]

>> 2. The Plan 9 editors have no built-in concept of a line.
>> (Going to line 10 in sam requires counting ten newlines from the beginning.)
>
> Not sure how the concept of a line delimited by newlines relates to
> moving the cursor up one physical line on the screen.

That's exactly the point.

Re: [9fans] Shall we fix the use of Up/Dn arrows?

[Frederik Caulier]

You're welcome to make changes and distribute those via contrib.

To get your own directory under /n/sources/contrib, mail
contrib@plan9.bell-labs.com.

You might also be interested in reading [0].

If you insist on using arrow keys to get from one line to another you
could also simply use the left/right arrows, although it might take a
bit longer to reach your destination. Using the mouse is definitely
faster.

Keep in mind the Plan 9 way of doing things; the mouse is used to
*control* stuff and the keyboard mainly for *text input*.

Best regards,
F. Caulier

[0] http://plan9.bell-labs.com/wiki/plan9/how_to_contribute/index.html

Re: [9fans] Shall we fix the use of Up/Dn arrows?

[Yi DAI]

[-- Attachment #1: Type: text/plain, Size: 853 bytes --]

Thanks Frederik, your and Russ' reply was quite informative.


On Sun, Jan 24, 2010 at 3:08 PM, Frederik Caulier <aediks@gmail.com> wrote:

> You're welcome to make changes and distribute those via contrib.
>
> To get your own directory under /n/sources/contrib, mail
> contrib@plan9.bell-labs.com.
>
> You might also be interested in reading [0].
>
> If you insist on using arrow keys to get from one line to another you
> could also simply use the left/right arrows, although it might take a
> bit longer to reach your destination. Using the mouse is definitely
> faster.
>
> Keep in mind the Plan 9 way of doing things; the mouse is used to
> *control* stuff and the keyboard mainly for *text input*.
>
> Best regards,
> F. Caulier
>
> [0] http://plan9.bell-labs.com/wiki/plan9/how_to_contribute/index.html
>
>


--
DAY

[-- Attachment #2: Type: text/html, Size: 1311 bytes --]

Re: [9fans] Shall we fix the use of Up/Dn arrows?

[Tim Newsham]

> 1. We use Lf/Rt arrows to move the cursor, don't we? Then use Up/Dn to do
> the same kind of task is symmetric, and thus more reasonable.

Wasn't this a concession? That slope is pretty slippery, isn't it?

> DAY

Tim Newsham | www.thenewsh.com/~newsham | thenewsh.blogspot.com

Re: [9fans] Shall we fix the use of Up/Dn arrows?

[John Floren]

On Sat, Jan 23, 2010 at 4:15 AM, Yi DAI <plm.day@gmail.com> wrote:
> 1. We use Lf/Rt arrows to move the cursor, don't we? Then use Up/Dn to do
> the same kind of task is symmetric, and thus more reasonable.

This is the behavior of Acme-SAC, as I recall, so one would think such
a change should be easy to do (since it's already been done in the
inferno version).

That said, I hate that behavior; it's probably easy to fix but it's
one of the reasons I don't use Acme-SAC much.

If you go back in the archives, you'll find post by myself and others
about very similar things. These days, the only thing I might ask
would be Ctrl-P to bring back previous commands, simply because it can
be annoying scrolling all the way back up above very verbose output
(i.e. I screw up some complex command, read the man page, get to the
bottom, then don't want to go back to the top to find my command again
so I can fix it).


John
--
"Object-oriented design is the roman numerals of computing" -- Rob Pike

Re: [9fans] Shall we fix the use of Up/Dn arrows?

[Charles Forsyth]

> 1. We use Lf/Rt arrows to move the cursor, don't we? Then use Up/Dn to do
> the same kind of task is symmetric, and thus more reasonable.

you're assuming the task is symmetric; perhaps left/right and up/down
aren't analogous, because x movement in a fixed y (left/right) is not the same
as y movement for varying x (up/down without fixed-length lines).

Re: [9fans] Shall we fix the use of Up/Dn arrows?

[Pavel Klinkovsky]

> What I'd appreciate, from the user's point of view, would be some kind
> of history mechanism tied to the up/down keys in rio/win (where I
> interact with a shell),

http://9fans.net/archive/2009/01/356

Pavel

Re: [9fans] Shall we fix the use of Up/Dn arrows?

[maht]

On 23/01/2010 19:01, Rudolf Sykora wrote:
>>   Maybe it only
>> happens once in 1000 times, but in that 0.1% of the time, my
>> finger coming down on the return key accidentally hits the up
>> arrow and the timing just so happens that I rerun the previous
>> command instead of the one I just typed, sometimes with very
>> unpleasant consqeuences.
>>
> This is the 1st time I hear about such a misfortune... :)
>
>
I've done the same, and it was something scary like rm -rf *

When people complain in irc has no history I say if you need to re-do
commands, use a script and /dev/text is your friend if you are half way
through.

The Scout Motto is: BE PREPARED which means you are always in a state of
readiness in mind and body to do your DUTY.
         * Be Prepared in Mind by having disciplined yourself to be
obedient to every order, and also by having thought out beforehand any
accident or situation that might occur, so that you know the right thing
to do at the right moment, and are willing to do it.
         * Be Prepared in Body by making yourself strong and active and
able to do the right thing at the right moment, and do it.

"To do the right thing at the right moment" can be extreme:

     "Where a man has gone so far as to attempt suicide, a Scout should
know what to do with him."
     "BE PREPARED to die for your country if need be, so that when the
moment arrives you may charge home with confidence, not caring whether
you are going to be killed or not"

>  Not sure how the concept of a line delimited by newlines relates to
moving the cursor up one physical line on the screen.

Working out where to move the cursor to

Re: [9fans] Shall we fix the use of Up/Dn arrows?

[Rudolf Sykora]

>>  Not sure how the concept of a line delimited by newlines relates to
>>moving the cursor up one physical line on the screen.
>
> Working out where to move the cursor to

Still I dare claim that moving a cursor up one visible line has
nothing to do with the concept of a line delimited by newlines. For
the movement, such delimited lines are completely irrelevant. For the
movement there is no difference whether there is a newline on the
previous visible line or the line was broken due to the width of the
window.

What I speak about is 'as if' clicking the mouse up one visible line.

R

Re: [9fans] Shall we fix the use of Up/Dn arrows?

[yy]

2010/1/26 Rudolf Sykora <rudolf.sykora@gmail.com>:
>>>  Not sure how the concept of a line delimited by newlines relates to
>>>moving the cursor up one physical line on the screen.
>>
>> Working out where to move the cursor to
>
> Still I dare claim that moving a cursor up one visible line has
> nothing to do with the concept of a line delimited by newlines. For
> the movement, such delimited lines are completely irrelevant. For the
> movement there is no difference whether there is a newline on the
> previous visible line or the line was broken due to the width of the
> window.
>
> What I speak about is 'as if' clicking the mouse up one visible line.
>
> R
>
>

what if the current selection is a block of text? anyway, I don't
think acme really knows where your pointer is (in x,y coords). I could
be wrong, though

-- 
- yiyus || JGL . 4l77.com

Re: [9fans] Shall we fix the use of Up/Dn arrows?

[Robert Raschke]

[-- Attachment #1: Type: text/plain, Size: 866 bytes --]

On Tue, Jan 26, 2010 at 11:50 AM, Rudolf Sykora <rudolf.sykora@gmail.com>wrote:

> >>  Not sure how the concept of a line delimited by newlines relates to
> >>moving the cursor up one physical line on the screen.
> >
> > Working out where to move the cursor to
>
> Still I dare claim that moving a cursor up one visible line has
> nothing to do with the concept of a line delimited by newlines. For
> the movement, such delimited lines are completely irrelevant. For the
> movement there is no difference whether there is a newline on the
> previous visible line or the line was broken due to the width of the
> window.
>
> What I speak about is 'as if' clicking the mouse up one visible line.
>
> R
>
>
How would you go about doing that?

There's no cursor based addressing available in the screens. And good
riddance too, methinks.

Robby

[-- Attachment #2: Type: text/html, Size: 1254 bytes --]

Re: [9fans] Shall we fix the use of Up/Dn arrows?

[erik quanstrom]

> > nothing to do with the concept of a line delimited by newlines. For
> > the movement, such delimited lines are completely irrelevant. For the
> > movement there is no difference whether there is a newline on the
> > previous visible line or the line was broken due to the width of the
> > window.
> >
> > What I speak about is 'as if' clicking the mouse up one visible line.
> >

> How would you go about doing that?

> There's no cursor based addressing available in the screens. And good
> riddance too, methinks.

and

> what if the current selection is a block of text? anyway, I don't
> think acme really knows where your pointer is (in x,y coords). I could
> be wrong, though

acme, sam, rio use frame(2).  frcharofpt(2) and frptofchar(2)
convert between characters and cartesian coordinates.
to find the closest character in the next line you're going to
have to start at the beginning of the line and add character
widths until the next character puts you past your target.
this requires a bit of work, since the text is stored in a different
data structure.

please use the up/down arrows in vim for a few minutes with
code and think about how many special cases there are.  for
example, you'll notice that vim stores an ideal x coordinate
for arrowing.

now if you still think this is such a great idea....
go forth and code!  or don't.  either way, let's pick a different
topic.

- erik