[GRASS-dev] Holiday GUI updates to GRASS

I just committed a number of GUI updates to GRASS which will hopefully make
some people happy for the holidays.

1. The title bar on the map layer window looks a bit nicer
2. You can now select panel options by clicking the icon/button in the map
layer window. Previously, you had to click the text.
3. I did a complete makeover to the EPSG projection setting GUI. There is
now a search engine that allows you to search for any text string, will
automatically grab the associated EPSG code, and put it into the EPSG code
entry window.
4. I put all the define new region code into the tcltk script (it was in a
separate bash script) so that it is now Windows-compatible (I hope). Some of
this code will go away with planned improvements to g.proj.
5. When you define a location using the EPSG module, it automatically
updates the locations list so that you don't have to restart GRASS to see
the new location.
6. I did a few cosmetic improvements to the main startup window (button
placement and consistent relief height).

Cheers
Michael
__________________________________________
Michael Barton, Professor of Anthropology
School of Human Evolution & Social Change
Center for Social Dynamics and Complexity
Arizona State University

phone: 480-965-6213
fax: 480-965-7671
www: http://www.public.asu.edu/~cmbarton

Hi,

Nice stuff Michael. The searchable EPSG code file is a great enhacement!

However, a glitch in gis.m is the long standing issue with single click
zoom not preserving the initial region aspect ratio. Have you thought
of fixing it before 6.2.2?

Maciek

Glad the EPSG changes are useful.

With regard to one-click zooming, it works exactly as it is designed to
work. I've said this a couple times on the list, but perhaps you've missed
it. No need to leave this as a bug report waiting to be fixed.

Michael

On 12/21/06 7:30 AM, "Maciej Sieczka" <tutey@o2.pl> wrote:

Hi,

Nice stuff Michael. The searchable EPSG code file is a great enhacement!

However, a glitch in gis.m is the long standing issue with single click
zoom not preserving the initial region aspect ratio. Have you thought
of fixing it before 6.2.2?

Maciek

__________________________________________
Michael Barton, Professor of Anthropology
School of Human Evolution & Social Change
Center for Social Dynamics & Complexity
Arizona State University

phone: 480-965-6213
fax: 480-965-7671
www: http://www.public.asu.edu/~cmbarton

Michael Barton wrote:

Glad the EPSG changes are useful.

Yes they are. Thanks.

With regard to one-click zooming, it works exactly as it is designed to
work.

Do you mean that the single click zoom is *not* supposed to preserve
the initial region aspect ratio in the *constrained* zoom mode?

I'd appreciate a short, straight answer.

I've said this a couple times on the list, but perhaps you've missed
it.

I'll take for a joke.

Cheers,
Maciek

On 12/22/06 12:52 PM, "Maciej Sieczka" <tutey@o2.pl> wrote:

With regard to one-click zooming, it works exactly as it is designed to
work.

Do you mean that the single click zoom is *not* supposed to preserve
the initial region aspect ratio in the *constrained* zoom mode?

I'd appreciate a short, straight answer.

I've said this a couple times on the list, but perhaps you've missed
it.

I'll take for a joke.

???

I thought the answer above was pretty short and straight. Here is my post of
27 November on this in its entirety. It's somewhat longer, but should
suffice.

From: Michael Barton <michael.barton@asu.edu>
Date: Fri, 17 Nov 2006 08:16:04 -0700
To: Markus Neteler <neteler@itc.it>
Cc: <grass-dev@grass.itc.it>, Hamish <hamish_nospam@yahoo.com>
Conversation: [GRASS-dev] fixes before 6.2.1 [was: ps.map: scale wrong and no
output]
Subject: Re: [GRASS-dev] fixes before 6.2.1 [was: ps.map: scale wrong and no
output]

Fixing the gis.m, so that it preserved the aspect ratio for
single-click zooming in the restricted mode, is also a critical issue.

This is not a bug. The single click zoom in (and out) uses the window geometry
to zoom in to fill the display as you get 'closer', while centering the view
(i.e., display region) on the point you click. You can set the aspect display
geometry with a box or with g.region--or you can simply make the display
window the aspect geometry you want and single-click away. It works like the
zoom in/zoom out button that GRASS lacks. IMHO, it's pretty slick. While some
might wish it were otherwise, it is this way with forethought and intention,
and works exactly as designed--so not a bug.

I've tried to notify you about all bug fixes that should be backported. And if
I remember correctly, you've told me that each one was backported. The only
possible outstanding bug is the slightly misleading error message generated
when the GUI fails to start up because gdal is not installed or incorrectly
installed. However, in discussions with Hamish, it is looking like this would
be better trapped before the GUI tries to launch.

Eric Patton's text/laton grid problem is a 6.3 issue that I've not yet been
able to replicate. The same goes for the in-progress nviz updates.

Let me know if there is anything else that comes to mind.

Michael

Michael
__________________________________________
Michael Barton, Professor of Anthropology
School of Human Evolution & Social Change
Center for Social Dynamics & Complexity
Arizona State University

phone: 480-965-6213
fax: 480-965-7671
www: http://www.public.asu.edu/~cmbarton

Michael Barton wrote:

On 12/22/06 12:52 PM, "Maciej Sieczka" <tutey@o2.pl> wrote:

With regard to one-click zooming, it works exactly as it is designed to
work.

Do you mean that the single click zoom is *not* supposed to preserve
the initial region aspect ratio in the *constrained* zoom mode?

I'd appreciate a short, straight answer.

I've said this a couple times on the list, but perhaps you've
missed it.

I'll take for a joke.

???

I thought the answer above was pretty short and straight. Here is my post of
27 November on this in its entirety. It's somewhat longer, but should
suffice.

Your message you quote below is not an answer to my question. I will
ask it again and, again, kindly ask you for a short, straight reply:

Do you mean that the single click zoom is *not* supposed to preserve
the initial region aspect ratio in the *constrained* zoom mode? Arguments?

A hint: if the single click zoom is *not* supposed to preserve the
initial region aspect ratio in the *constrained* zoom mode, then how
does it differ, in terms of the look of the map in the Map Display,
from the *unconstrained* mode?

See how d.zoom behaves. It allows for a single click zoom-out (middle
mouse key), and it *does* preserve the initial region aspect ratio
then. That's how gis.m's single click zoom tool *should* behave as
well. Isn't that logical for you? Why do you prefer it not preserving
the initial aspect ratio, like if it was the *unconstrained* mode???

Find more relpies below.

From: Michael Barton <michael.barton@asu.edu>
Date: Fri, 17 Nov 2006 08:16:04 -0700
To: Markus Neteler <neteler@itc.it>
Cc: <grass-dev@grass.itc.it>, Hamish <hamish_nospam@yahoo.com>
Conversation: [GRASS-dev] fixes before 6.2.1 [was: ps.map: scale wrong and no
output]
Subject: Re: [GRASS-dev] fixes before 6.2.1 [was: ps.map: scale wrong and no
output]

This is not a bug.

No, you only *think* this is not a bug, because you designed the tool
wrong and it performs according to your wrong design. The problem is
the design was wrong.

A similar situation would be if you wrote a calculator program, which
you designed to calculate 2+2 as 5. Obviously, the result would be
wrong in terms of truth and reasonable user's expectations, but your
argument would be that this is not a bug, because you designed it to
calculate 2+2 as 5 instead of 4, and it works as you designed it.

It is hard to argue with you because you look at complicated problems
only from your very own perspective, and hardly accept a reasonable
argument of someone else. Unless it is Glynn :). Let me remind you of
the issue with gis.m's box-zoom tool not maintaining the region
properly. I kept telling you about this problem for circa 3 months, and
you kept denying it. Until Glynn spoke up, after which you fixed the
tool according to his suggestion quite soon. When it was me, you just
kept treating me as an annoying pest. And to get rid of me you flooded
me with longish elaborates, which often didn't address the concerns and
looked like if you didn't understand the problem, or didn't want to
understand it.

Now you act the same. And you even repost your past replies, or make me
look for them in the archives myself, advertising them as an "answer".
It is driving me nuts that you are waisting my (and your) time. And all
I can do then is take it for a joke.

I'm sorry if I'm not making myself clear. I'm not a native English
speaker. Writing an understandable, detailed problem report is a
substantial effort for me. I'm writing them not to tease or piss off
people, but to help. Please respect this.

The single click zoom in (and out) uses the window geometry
to zoom in to fill the display as you get 'closer', while centering the view
(i.e., display region) on the point you click.

That's the problem. It shouldn't "fill the display as you get
'closer'". It should preserve the initial region aspect ratio instead,
because it is the *constrained* zoom mode, not the *unconstrained* one.

You can set the aspect display
geometry with a box or with g.region--or you can simply make the display
window the aspect geometry you want and single-click away.

This takes using more tools than needed, thus it is a sub-optimal
solution; a workaround, actually.

It works like the zoom in/zoom out button that GRASS lacks. IMHO, it's
pretty slick. While some might wish

Another bad practice, please don't do it. Don't understimate my
reasonable problem report as a "wish", a feature request.

it were otherwise, it is this way with forethought and intention,
and works exactly as designed--so not a bug.

See my notes above about the wrong desing as a bug.

Maciek

Michael Barton wrote:

I just committed a number of GUI updates to GRASS which will hopefully
make some people happy for the holidays.

1. The title bar on the map layer window looks a bit nicer
2. You can now select panel options by clicking the icon/button in the
map layer window. Previously, you had to click the text.
3. I did a complete makeover to the EPSG projection setting GUI. There
is now a search engine that allows you to search for any text string,
will automatically grab the associated EPSG code, and put it into the
EPSG code entry window.
4. I put all the define new region code into the tcltk script (it was
in a separate bash script) so that it is now Windows-compatible (I
hope). Some of this code will go away with planned improvements to
g.proj.
5. When you define a location using the EPSG module, it automatically
updates the locations list so that you don't have to restart GRASS to
see the new location.
6. I did a few cosmetic improvements to the main startup window
(button placement and consistent relief height).

brilliant on all counts.

Hamish