Identify tool does not highlight features consistently

[GoogleCodeExporter]

What steps will reproduce the problem?
1. Starting with a fresh MORIS page, use the Identify tool to draw a box around 
some features. The layers selected are listed in the Query Results window. 
Select a layer in the At a Glance section of Query Results. The features 
selected within that layer are highlighted blue.

2. Choose a different layer in At a Glance. The features that fall within the 
Identify box are highlighted a dark shade of orange instead of blue. This 
happens regardless of whether you use the Clear Identified Features tool before 
selecting a second layer from At a Glance, or whether you close the Query 
Results window and start the Identify tool from scratch. The only way to 
"reset" this tool for blue highlights is to reload the MORIS page. 

What is the expected output?
For the Identify tool to work properly, any time I select a layer from At a 
Glance the follow should/could happen:
1. Its features that fall within the Identify box should turn blue, whether 
it's the first layer selected or a layer selected thereafter.
2. If I hover/mouseover any of the highlighted features, a pointing hand cursor 
should appear.
3. The pointing hand cursor should allow me to click on a blue highlighted 
feature.
4. When I click on the blue highlighted feature, that feature should be 
highlighted in yellow, while the other features highlighted in blue should 
remain blue.  Also, the row describing the yellow highlighted feature in 
Feature Details should be highlighted.
5. Highlighting a feature within a selection (i.e., highlighting in yellow) 
should work from the map window (by using the pointing hand cursor) and by 
selecting a row, or multiple rows, from Feature Details.  Results should be 
reflected in both the map window and Feature Details section of the Query 
Results window.

What version of the product are you using? On what operating system?
MORIS v.0.35, though this bug has been present for many versions.

Please provide any additional information below.

Original issue reported on code.google.com by marc.car...@state.ma.us on 18 Feb 2011 at 7:10

Attachments:

• identify_inconsistency_screenshots.doc

[GoogleCodeExporter]

Original comment by marc.car...@state.ma.us on 18 Feb 2011 at 8:57

Attachments:

• identify_inconsistency1.jpg
• identify_inconsistency2.jpg
• identify_inconsistency3.jpg

[GoogleCodeExporter]

Original comment by Aleda.Fr...@state.ma.us on 24 Feb 2011 at 8:58

• Added labels: Identify

[GoogleCodeExporter]

This one was tricky.  The main thing that caused such confusion w/ the current 
mapper was that once an at-a-glance layer was selected, the rows in the feature 
details table WEREN'T selected while the features on the map WERE.  That was a 
complete disconnect.  So when you were clicking on the blue things, they were 
actually unselecting themselves.

So please review the attachment for an alternate workflow.  We actually have a 
few layers of selection going on, and I think putting them into different 
colored buckets makes most sense.  And I think that NOT selecting them by 
default keeps things understandable.

By 'few layers of selection' I mean that the bbox you make is the 1st layer.  
Clicking on a layer name in at-a-glance is another.  Then clicking on a feature 
detail row or rows is the 3rd.

I'm tempted to go ahead and commit my changes and ask for forgiveness later.  
These changes can easily be rolled back.

Original comment by cpl...@gmail.com on 19 Aug 2011 at 7:24

Attachments:

• query.docx

[GoogleCodeExporter]

Issue 13 folds into this issue.

Original comment by cpl...@gmail.com on 22 Aug 2011 at 4:57

[GoogleCodeExporter]

Issue 13 has been merged into this issue.

Original comment by cpl...@gmail.com on 22 Aug 2011 at 4:58

[GoogleCodeExporter]

I like the way this works now, CZM see if you agree.

Original comment by Aleda.Fr...@state.ma.us on 24 Aug 2011 at 6:26

[GoogleCodeExporter]

Original comment by Aleda.Fr...@state.ma.us on 24 Aug 2011 at 6:26

• Changed state: Fixed

[GoogleCodeExporter]

I like this a lot too.

The spelling of "high light" needs to be consistent in feature details.  Let's 
with "highlight" and "unhighlight"; hi-lite is a trademark.

Original comment by daniel.s...@state.ma.us on 25 Aug 2011 at 3:38

[GoogleCodeExporter]

Where is hi-lite used?  Can you provide a screenshot? 

Original comment by Aleda.Fr...@state.ma.us on 25 Aug 2011 at 3:50

[GoogleCodeExporter]

In v36, the Query results have two spellings.

Original comment by daniel.s...@state.ma.us on 25 Aug 2011 at 4:09

Attachments:

• highlight.png

[GoogleCodeExporter]

Sorry to be a lummox here, but on what page are you asking me to change a 
spelling from what and what to?

Original comment by cpl...@gmail.com on 25 Aug 2011 at 4:40

[GoogleCodeExporter]

Charlton, on the screenshot in #10, we have "highlight all" and "unhilight 
all."  The "gh" is missing.

Original comment by daniel.s...@state.ma.us on 25 Aug 2011 at 5:15

[GoogleCodeExporter]

Thanks for hiliting that.

Committed revision 80.

Original comment by cpl...@gmail.com on 25 Aug 2011 at 5:30

[GoogleCodeExporter]

I like the way this works, but I think there's still a bug with the pointing 
hand cursor that appears when I mouseover a highlighted feature. The pointing 
hand cursor only appears over layers highlighted in blue, and then when I click 
the feature with the hand cursor, I end up identifying with a new bounding box. 
Here are my steps:

1. Use the Identify tool with the MIMIC 2008 Monitoring Sites data layer.
2. Select the layer name in the At a glance table.
3. The selected sites are now highlighted orange on the map. If I mouseover one 
of the orange highlighted points, I don't get the pointing hand cursor. In 
MORIS v0.35, you do see the hand cursor (at least the first time you identify 
and features are highlighted in blue). Even though I don't get a pointing hand 
cursor, I can click on one of the orange highlighted features and it will turn 
blue and be highlighted in the Feature details table. I think it'd be nice if 
we saw the hand cursor when we hover over an orange feature. 
4. Close the Query results window and identify a second time on the monitoring 
sites data layer.
5. I can no longer click on one of the orange highlighted feature and have it 
turn blue. This should be fixed so that it is consistent with the way it works 
the first time a user identifies.
6. Select one of the rows in the Feature details table. The feature is 
highlighted in blue on the map. When you mouseover a blue highlighted feature, 
the pointing hand cursor appears. When I click the feature though, I end up 
identifying with a new bounding box. This should be fixed so that I can select 
the feature with the cursor rather than start a new identify.

I agree with Aleda regarding issue 13. We were looking for a red outline with 
no fill for the identify box. 

Original comment by emily.hu...@state.ma.us on 23 Sep 2011 at 4:11

[GoogleCodeExporter]

Thanks for the detailed directions.  It's been fixed.  Or at least the 
inconsistency between the behavior when 1st launched vs. the 2nd has been 
eliminated.

> 3. The selected sites are now highlighted orange on the map. If I mouseover 
one of the orange highlighted points, I don't get the pointing hand cursor. In 
MORIS v0.35, you do see the hand cursor (at least the first time you identify 
and features are highlighted in blue).

The way it behaved in 0.35 was actually incorrect.  The mouse cursor changed 
because it was actually already selected.  Which was bad.  There was a 
disconnect between the map and the table, so now everything is NOT selected by 
default.  Orange things are responsive to a click but they cannot trigger a 
different mouse cursor.

I have changed the box / poly hilite styling.

Original comment by cpl...@gmail.com on 23 Sep 2011 at 5:47

[GoogleCodeExporter]

Charlton, what's the thinking behind having the pointing hand cursor only 
appear over blue highlighted features? I would think we would want the cursor 
over both orange and blue highlighted features so that users know they can 
click them. Because I see the identify cursor over orange features, I would 
expect to start a new identify if I clicked an orange feature.

Original comment by emily.hu...@state.ma.us on 26 Sep 2011 at 1:58

[GoogleCodeExporter]

Comment 14 #5 seems to be fixed for me. 

Original comment by Aleda.Fr...@state.ma.us on 26 Sep 2011 at 4:23

[GoogleCodeExporter]

Emily, that's the default behavior of how the GeoExt library and OpenLayers 
interact.  I'll poke around to see if that's easily overwritten.

Original comment by cpl...@gmail.com on 26 Sep 2011 at 4:53

[GoogleCodeExporter]

I added another layer that highlights things in red on mouseover.  See what you 
think.  If you don't like the way it turns something temporarily to red when 
it's already blue, there will likely be nothing I can do about that.

Also, you pointed out a bug when querying points w/ a bbox and clicking on 
them.  Then returning to draw a new bbox and then clicking only to find out 
that a brand new query was fired.  I don't recall where you laid out the 
directions, but it should really be fixed this time.

Original comment by cpl...@gmail.com on 26 Sep 2011 at 7:11

[GoogleCodeExporter]

Features still are not highlighted consistently if a user uses the Clear 
identified features tool. Please see Issue 71 for details.

Original comment by emily.hu...@state.ma.us on 3 Oct 2011 at 3:13

[GoogleCodeExporter]

Can you reactivate this issue and roll issue 71 into this one?  I think keeping 
it here makes sense to me.

Original comment by cpl...@gmail.com on 3 Oct 2011 at 6:54

[GoogleCodeExporter]

Issue 71 has been merged into this issue.

Original comment by emily.hu...@state.ma.us on 3 Oct 2011 at 6:59

[GoogleCodeExporter]

The clear identify tool is buggy and unnecessary:

1. Identify a feature (it will turn orange); select the identified feature by 
clicking on it (it will turn blue); then unselect it by clicking on it again. 
It will turn back to orange. This works as it should (including the mouse-over 
effects).
2. Click the clear identify tool.
3. Repeat step 1. Now clicking on the highlighted feature (blue) initiates a 
new identify query instead of unselecting it (turning it back to orange) as it 
did in step 1. This only occurs after the clear identify tool is clicked.
The tool is unnecessary since users can initiate a new identify query by simply 
drawing a new box/poly (whether or not the query results window remains open). 
Please remove the clear identify tool from the toolbar, as space there is 
limited. Doing so will also solve this bug.

Original comment by emily.hu...@state.ma.us on 3 Oct 2011 at 7:01

[GoogleCodeExporter]

This button has been removed.

Original comment by cpl...@gmail.com on 6 Oct 2011 at 1:51

[GoogleCodeExporter]

It looks like there are still some bugs with the identify tools:

1. The Clear identified features button has not been removed.

2. A polygon drawn with the Identify by Poly tool does not clear from the map 
once the query results windows is closed. A bbox drawn with the Identify tool 
is cleared from the map once the query window is closed, so the drawn polygon 
should also be cleared to be consistent.

3. The Identify by Poly tool is still selected when a user switches from the 
Identify by Poly tool to the Measure by length or area tool. For example, if I 
use the Identify by Poly tool and then switch to the Measure by area tool, my 
cursor will still be the identify cursor and I'll draw two polygons - a red 
polygon for identify and a blue one for measuring area (see attached 
screenshot).

4. When I switch from using the Identify to the Identify by Poly tool (or vice 
versa), I can no longer click on orange highlighted features without starting a 
new query. For example: 1)Identify on the New England Mask using a bbox; 
2)Select the New England Mask layer in the At a glance table in the query 
results; 3)Click on one of the orange highlighted features on the map; 4)Close 
the query results; 5)Identify on the New England Mask by drawing a polygon; 
6)Select the New England Mask layer in the At a glance table in the query 
results; 7)Click on one of the orange highlighted features on the map. When I 
click on the map, I start a new query instead of highlighting the feature in 
blue.

5. If I use the Identify by Poly tool two times in a row, it sometimes 
highlights features on the map in orange as I'm drawing the polygon. For 
example: 1)Identify by Poly on the New England Mask; 2)Select the New England 
Mask layer in the At a glance table in the query results; 3)Close the query 
results window; 4)Draw another identify polygon with vertices that intersect 
some of the New England Mask features. Those features will be highlighted in 
orange as you draw the polygon.

Original comment by emily.hu...@state.ma.us on 18 Oct 2011 at 5:58

Attachments:

• measure_screenshot.png

[GoogleCodeExporter]

> 1. The Clear identified features button has not been removed.

Ah.  I had only removed the clear button from the popup window.  The main tbar 
has been updated now.

> 2. A polygon drawn with the Identify by Poly tool
> 5. If I use the Identify by Poly tool two times in a row

Those should both be fixed now.

#3 and #4 are related.  I hadn't imagined that you'd want to be able to switch 
between poly and point selection for the same set of query results.  That 
should not be allowed.  Similarly, if you have features selected and 
highlighted, I can't guarantee that they should still be available once you 
enter the measuring mode.  I'd like to turn off all highlighted features 
whenever the query mode / measurement is changed.

v. 0.63
Committed revision 173.

Original comment by cpl...@gmail.com on 18 Oct 2011 at 6:22

[GoogleCodeExporter]

Here are our comments from testing the latest round of updates:

> 1. The Clear identified features button has not been removed.

Fixed. Can the "Unselect and unhighlight all" please be added back to the 
pop-up window?

> 2. A polygon drawn with the Identify by Poly tool does not clear.

Fixed.

> 3. The Identify by Poly tool is still selected when a user switches from the 
Identify by Poly tool to the Measure by length or area tool.

Not fixed. When I switch from Identify by Poly to Measure by area or length, I 
can't stop drawing a polygon. The double-click doesn't stop the polygon 
creation. When I switch from Identify by bbox to Measure by area or length, I 
will identify as I draw the polygon or line.

> 4. When I switch from using the Identify to the Identify by Poly tool (or 
vice versa), I can no longer click on orange highlighted features without 
starting a new query.

Not fixed.

> 5. If I use the Identify by Poly tool two times in a row, it sometimes 
highlights features on the map in orange as I'm drawing the polygon.

Not fixed. FYI, the features that turn orange are only those that overlap with 
my identify poly vertices.

Original comment by emily.hu...@state.ma.us on 20 Oct 2011 at 3:08

[GoogleCodeExporter]

Re. 3-5, I didn't address those because I wanted to make sure that everyone was 
OK w/ losing the query highlights when switching out of the query mode.  E.g. 
if you entered in w/ a bbox select and made your way to having a few things 
highlighted, if you were to switch to poly query and/or measure something, I 
need to lose the query highlights.

Original comment by cpl...@gmail.com on 20 Oct 2011 at 4:23

[GoogleCodeExporter]

Oh, I suppose I did try and address #5!  It still doesn't work?  OK, I'll check 
again.

Original comment by cpl...@gmail.com on 20 Oct 2011 at 4:24

[GoogleCodeExporter]

For #5 Emily showed me how she was able to produce the problem and then I was 
able to produce it as well in FF 7.

Original comment by Aleda.Fr...@state.ma.us on 20 Oct 2011 at 4:32

[GoogleCodeExporter]

OK, one at a time.

#3 should be fixed.

v. 0.66
Committed revision 176.

Original comment by cpl...@gmail.com on 20 Oct 2011 at 4:54

[GoogleCodeExporter]

#1 should be fixed.  (Button restored.)

v. 0.67
Committed revision 177.

Original comment by cpl...@gmail.com on 20 Oct 2011 at 5:12

[GoogleCodeExporter]

#4.  I can't allow you to switch between the two modes and maintain your query 
hits.  Suggestions?

#5.  I can't reproduce that here.  I've tried!  Please advise.

Original comment by cpl...@gmail.com on 20 Oct 2011 at 5:20

[GoogleCodeExporter]

We all agreed that for #3-5 we do not expect to still see the query results on 
the map once the Query results window has been closed. 

Original comment by Aleda.Fr...@state.ma.us on 20 Oct 2011 at 5:49

[GoogleCodeExporter]

I'm concerned more w/ making sure that you're ok w/ losing the query results 
when switching modes.  So if you were t leave the window open and switch from 
bbox to poly, you need to expect the results to be requeried.

Original comment by cpl...@gmail.com on 20 Oct 2011 at 6:04

[GoogleCodeExporter]

Emily is going to walk me through #5 tomorrow morning.

To make sure that we're all running in the same direction . . . w/ my latest 
changes, has #4 been taken care of?

Original comment by cpl...@gmail.com on 20 Oct 2011 at 6:49

[GoogleCodeExporter]

#1. Fixed.

#3. This is fixed, but it looks like we no longer have the measuring cursor. 
When I switch from identify to measure, I still have the identify cursor. Would 
it be possible to go back to having the plus sign cursor like we have in 
http://maps.massgis.state.ma.us/map_ol/moris.php?

#4. This still isn't fixed, but I don't think I've done a very good job 
explaining it. I'll walk you through #4 and #5 tomorrow morning and hopefully 
that will help clear things up.

Original comment by emily.hu...@state.ma.us on 20 Oct 2011 at 7:39

[GoogleCodeExporter]

Emily, I think I have taken care of #4.

v. 0.70
Committed revision 180.

Original comment by cpl...@gmail.com on 20 Oct 2011 at 9:41

[GoogleCodeExporter]

#3's cursor issue should be fixed.

v. 0.71
Committed revision 181.

Original comment by cpl...@gmail.com on 21 Oct 2011 at 2:02

[GoogleCodeExporter]

#3. Fixed.  Measure tool has a crosshair cursor.

#4. Not fixed. Still can't click on orange highlighted features without 
starting a new query.

#5. Fixed.

Original comment by emily.hu...@state.ma.us on 24 Oct 2011 at 1:15

[GoogleCodeExporter]

I have spent a very long time trying to address this issue.  And I think it 
boils down to a but in GeoExt.  It turns out that you lose the disconnect 
between the query results table and orange dots whenever you change from a 
query mode to any other mode (like from bbox to poly or from query to measure). 
 I rarely admit defeat, so I am going to stop one step shy of that and simply 
say that it has to be a GeoExt issue.  I may post a bug to the listserve, but 
they aren't very responsive, so I won't hold my breath there.

What I suggest is this.  I have created an alternate universe here where 
changing from query to anything else does NOT goof things up like you currently 
experience, i.e. it doesn't launch a new query when you try and click on a dot. 
 BUT what you give up here is the ability to click on the dots and have the 
corresponding row in the table light up.  That is where the original issue lies 
(I believe), and there is no way for me to get around that.  You may still 
click on the table row and have the corresponding dot (or polygon) highlight, 
but NOT the other way around.

How do you feel about this?

Original comment by cpl...@gmail.com on 16 Nov 2011 at 4:49

[GoogleCodeExporter]

A 'but' in GeoExt.  That's about right!  The gnat's but(t).

Original comment by cpl...@gmail.com on 16 Nov 2011 at 5:01

[GoogleCodeExporter]

WAY TO GO, ME!  I fixed it.  I'll wrap up a release in a bit.

Original comment by cpl...@gmail.com on 16 Nov 2011 at 6:25

[GoogleCodeExporter]

v. 0.72
Committed revision 184.

I made a bunch of other changes in a local copy of map.js that I have here, but 
I reverted my changes to only include the TWO NECESSARY LINES OF CODE to fix 
the specific issue we affectionately call #4.

So . . . I'll keep my local fistful-of-changes map.js here while I wait to see 
if the TWO NECESSARY LINES OF CODE takes care of the whole enchilada.

Let's not divide the number of HOURS by TWO LINES OF CODE to determine my 
efficiency rate, OK?

Am I bitter?  Heck yes!  Can I get on w/ my life now?  We'll see . . . It's all 
up to YOU, Emily, to kick this into Verified territory.  No pressure.

Original comment by cpl...@gmail.com on 16 Nov 2011 at 6:31

[GoogleCodeExporter]

Good news, Charlton! #4 is fixed and you can get on with your life now!

Original comment by emily.hu...@state.ma.us on 16 Nov 2011 at 8:25

• Changed state: Verified