9:07 p.m.
I found these little bugs using 1.6.2 under windows:
1- When trying to select several variables to display summary
statistics, I hold the ctrl button and start selecting individual
variables. Now, while pressing the left mouse button (or while having
it pressed) on a new variable, if I move the mouse pointer even by a
pixel, the whole previous selection is gone. The only way to select
multiple (not adjacent) variables is to make sure the mouse pointer
does not move while the left mouse button is pressed.
2- From the sample menu, selecting restrict based on criterion and
entering, for example, VALUE="" crashes the program (happens with "")
3- Pressing q (not Q) while renaming a model or graph closes the icon view.
4- Selecting "arrange icons" deletes the custom names given to models
returning back to Model1, Model2 etc. Similarly, model names are not
saved after saving a session so that I see Model1, Model2 etc.
However, double clicking a model, I can see the custom name in the
first row (so it was not completely lost).
Cheers,
Talha
--
"Money won't buy happiness, but it will pay the salaries of a large
research staff to study the problem." - Bill Vaughan
--
11:59 p.m.
Not at all. Also, when I create a model table, all my variables with
*** significance show up with ** next to them. At the bottom, only the
explanations for * and ** show up (nothing saying "*** indicates
significance at the 1 percent level". This also happens with the pdf
file created from the TeX->view menu. Finally, individual estimation
result screens look alright.
Cheers,
Talha
On 4/3/07, Allin Cottrell <cottrell(a)wfu.edu> wrote:
On Tue, 3 Apr 2007, Talha Yalta wrote:
> I found these little bugs using 1.6.2 under windows...
Thanks, Talha. Most of these sound like straightforward and
easily fixable bugs, with the possible exception of number 1,
the "Ctrl-click" issue.
Allin
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
--
"Money won't buy happiness, but it will pay the salaries of a large
research staff to study the problem." - Bill Vaughan
--
12:54 p.m.
On Tue, 3 Apr 2007, Talha Yalta wrote:
Also, when I create a model table, all my variables with ***
significance show up with ** next to them. At the bottom, only
the explanations for * and ** show up (nothing saying "***
indicates significance at the 1 percent level".
That's not a bug as such, it's a design decision (a debatable one,
perhaps). Given the premium on horizontal space in the model
table (particularly if you have a lot of models) I decided not to
include a distinct "***" marker for P-values < .01. Of course,
coefficients with P-values < .01 also have P-values < .05 ;-)
This is different from the presentation in the individual model
windows, but it's not that uncommon.
Allin.
1:17 p.m.
On Tue, 3 Apr 2007, Talha Yalta wrote:
1- When trying to select several variables to display summary
statistics, I hold the ctrl button and start selecting individual
variables. Now, while pressing the left mouse button (or while having
it pressed) on a new variable, if I move the mouse pointer even by a
pixel, the whole previous selection is gone. The only way to select
multiple (not adjacent) variables is to make sure the mouse pointer
does not move while the left mouse button is pressed.
This is tricky. I have noticed myself that you have to be very
careful doing Ctrl-click for multiple selection. I'll take
another look at it, but this may be the price we pay for a
relatively unusual feature that I find useful, namely the ability
to define a (contiguous) multiple selection simply by dragging or
"swiping" the mouse across the variables. The problem is that
movement of the mouse is taken as the start of a drag selection.
There may be a way to fine-tune this.
2- From the sample menu, selecting restrict based on criterion and
entering, for example, VALUE="" crashes the program (happens with "")
Now fixed.
3- Pressing q (not Q) while renaming a model or graph closes the icon
view.
Fixed.
4- Selecting "arrange icons" deletes the custom names given to models
returning back to Model1, Model2 etc. Similarly, model names are not
saved after saving a session so that I see Model1, Model2 etc.
However, double clicking a model, I can see the custom name in the
first row (so it was not completely lost).
Fixed.
Allin.
12:53 a.m.
Here are a few more:
6- When I have more than 41 tables I cannot add the 42nd and above to
the model table by dragging and dropping since the model table icon is
invisible and the icon window doesn't scroll while the mouse button is
pressed (due to dragging)
7- Sessions are not saved as a single compressed file. I have a, say,
session1.gretl file plus a session1 directory with the data, an xml
file, and all the model files.
8- What is worse, when I "save as" a new session, say session2, I get
a session2.gretl file, however, the old session1 directory is now
replaced with a session2 directory with all the files lost.
9- Suggestion: Being able to create a model table with more models
(more than 6, maybe 8) would be nice to have.
10- Suggestion: Whoever I show gretl to, they are simply blown away by
the icon view concept. How models and graphs can be saved as icons for
future reference, how they can easily be combined and exported to
LaTeX, the little utilities... This is some serious innovation IMO.
However, the default view doesn't do the program justice.
IMHO the UI can be set up such that there is one main gretl window
which will serve as a workspace. Both, the current gretl window (which
can be renamed as the analysis window) and the icon view can be opened
by default as separate windows within this work space. Furthermore,
some of the menus of the analysis window such as file, view, tools,
and help can be migrated to the new main window while data, add,
sample, variable, model menus are retained in the the analysis window.
This will allow for a good foundation to add new features for the
future versions including things like matrix toolbars or various
specialized palettes.
Finally, gretl is already much further away in terms of the UI and the
ease of use compared to any statistical package out there including
eviews. Still, IMHO Adobe's Photoshop can serve as a good model on how
to improve it. I think Photoshop is a great example to show how a very
large number of complex operations with lots of individual options can
be done using a clever graphical user interface.
Anyway, thanks for the already great program...
Cheers
Talha
On 4/4/07, Allin Cottrell <cottrell(a)wfu.edu> wrote:
On Tue, 3 Apr 2007, Talha Yalta wrote:
> 1- When trying to select several variables to display summary
> statistics, I hold the ctrl button and start selecting individual
> variables. Now, while pressing the left mouse button (or while having
> it pressed) on a new variable, if I move the mouse pointer even by a
> pixel, the whole previous selection is gone. The only way to select
> multiple (not adjacent) variables is to make sure the mouse pointer
> does not move while the left mouse button is pressed.
This is tricky. I have noticed myself that you have to be very
careful doing Ctrl-click for multiple selection. I'll take
another look at it, but this may be the price we pay for a
relatively unusual feature that I find useful, namely the ability
to define a (contiguous) multiple selection simply by dragging or
"swiping" the mouse across the variables. The problem is that
movement of the mouse is taken as the start of a drag selection.
There may be a way to fine-tune this.
> 2- From the sample menu, selecting restrict based on criterion and
> entering, for example, VALUE="" crashes the program (happens with
"")
Now fixed.
> 3- Pressing q (not Q) while renaming a model or graph closes the icon view.
Fixed.
>
> 4- Selecting "arrange icons" deletes the custom names given to models
> returning back to Model1, Model2 etc. Similarly, model names are not
> saved after saving a session so that I see Model1, Model2 etc.
> However, double clicking a model, I can see the custom name in the
> first row (so it was not completely lost).
Fixed.
Allin.
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
--
"Money won't buy happiness, but it will pay the salaries of a large
research staff to study the problem." - Bill Vaughan
--
1:44 a.m.
On Wed, 4 Apr 2007, Talha Yalta wrote:
6- When I have more than 41 tables...
Do you mean more than 41 models?
I cannot add the 42nd and above to the model table by dragging
and dropping since the model table icon is invisible and the
icon window doesn't scroll while the mouse button is pressed
(due to dragging)
I would say that more than a couple of dozen models per GUI
session is going beyond the design limits: with that number of
models you want a script, or set of scripts. But I'll take a
look.
7- Sessions are not saved as a single compressed file. I have a,
say, session1.gretl file plus a session1 directory with the
data, an xml file, and all the model files.
Sessions are in fact saved as a single compressed file. Open
sessions are also represented by a directory containing the files
you mention: the "save" file is a compressed archive of the
directory in question. (I guess I should make these directories
hidden so that people won't get confused.)
8- What is worse, when I "save as" a new session, say
session2,
I get a session2.gretl file, however, the old session1 directory
is now replaced with a session2 directory with all the files
lost.
Wait a minute: Are you saying that if you save as "session2" the
session won't open again? (Never mind what's going on in the
background -- just concentrate on things at the visible level not
working as they are clearly supposed to work.)
9- Suggestion: Being able to create a model table with more
models (more than 6, maybe 8) would be nice to have.
I guess this might be feasible for PDF formatting on a landscape
page. I judged that 6 was about the limit for on-screen
legibility.
10- Suggestion: Whoever I show gretl to, they are simply blown
away by the icon view concept. How models and graphs can be
saved as icons for future reference, how they can easily be
combined and exported to LaTeX, the little utilities... This is
some serious innovation IMO. However, the default view doesn't
do the program justice...
I agree that the session concept is not fully exploited at
present, and that is something we will come back to and
re-examine. Thanks for the suggestions. Right now we're focused
on particular econometric functionality (the bootstrap and VECMs)
but the UI does need reconsideration.
--
Allin Cottrell
Department of Economics
Wake Forest University, NC
1:01 p.m.
Do you mean more than 41 models?
Actually having more than 33
new icons (models or graphs) will not fit
inside my icon window even if it is maximized. This number would be
even less if I had a lower screen resolution, say 1024x768. I think
the problem comes from the fact that there are always 4 icons in each
row and enlarging the window only increases the horizontal spacing
between icons instead of reorganizing them to use the extra space
efficiently.
Wait a minute: Are you saying that if you save as
"session2" the
session won't open again? (Never mind what's going on in the
background -- just concentrate on things at the visible level not
working as they are clearly supposed to work.)
OK, I see now. The directory I am
referring to in the working
directory is a temporary one and it is deleted when the session is
closed. I also see that the individual .gretl files contain
everything. So it is OK. Still, this is confusing. I think it would be
better if the system temp directory was used for these temporary
folders and files.
I guess this might be feasible for PDF formatting on a landscape
page. I judged that 6 was about the limit for on-screen
legibility.
Even 4 might not fit in the page depending on the way user names each
model. I can easily fit 8 when I just assign (1), (2), ... for the
model names. Of course more can be fitted on landscape. IMHO 8 is a
good number (power of 2 = nice :-) ). The user should do the further
tweaking (by renaming the models, variables etc) if s/he has problem
fitting everything on the page.
11- Here is a new werid one:
a)-I open gretl, I open a session (the icon view appears automatically).
b)-I close the icon view by clicking on the x icon on the top right corner.
c)-I close gretl by clicking on the x icon on the top right corner in
the main screen.
d)-The program crashes saying gretlw32.exe has encountered a problem
and needs to close.
Cheers,
Talha
On 4/4/07, Allin Cottrell <cottrell(a)wfu.edu> wrote:
> On Wed, 4 Apr 2007, Talha Yalta wrote:
>
> > 6- When I have more than 41 tables...
>
Do you mean more than 41 models?
>
> > I cannot add the 42nd and above to the model table by dragging
> > and dropping since the model table icon is invisible and the
> > icon window doesn't scroll while the mouse button is pressed
> > (due to dragging)
>
> I would say that more than a couple of dozen models per GUI
> session is going beyond the design limits: with that number of
> models you want a script, or set of scripts. But I'll take a
> look.
>
> > 7- Sessions are not saved as a single compressed file. I have a,
> > say, session1.gretl file plus a session1 directory with the
> > data, an xml file, and all the model files.
>
> Sessions are in fact saved as a single compressed file. Open
> sessions are also represented by a directory containing the files
> you mention: the "save" file is a compressed archive of the
> directory in question. (I guess I should make these directories
> hidden so that people won't get confused.)
>
> > 8- What is worse, when I "save as" a new session, say session2,
> > I get a session2.gretl file, however, the old session1 directory
> > is now replaced with a session2 directory with all the files
> > lost.
>
Wait a minute: Are you saying that if you save as
"session2" the
session won't open again? (Never mind what's going on in the
background -- just concentrate on things at the visible level not
working as they are clearly supposed to work.)
>
> > 9- Suggestion: Being able to create a model table with more
> > models (more than 6, maybe 8) would be nice to have.
>
I guess this might be feasible for PDF formatting on a landscape
page. I judged that 6 was about the limit for on-screen
legibility.
>
> > 10- Suggestion: Whoever I show gretl to, they are simply blown
> > away by the icon view concept. How models and graphs can be
> > saved as icons for future reference, how they can easily be
> > combined and exported to LaTeX, the little utilities... This is
> > some serious innovation IMO. However, the default view doesn't
> > do the program justice...
>
> I agree that the session concept is not fully exploited at
> present, and that is something we will come back to and
> re-examine. Thanks for the suggestions. Right now we're focused
> on particular econometric functionality (the bootstrap and VECMs)
> but the UI does need reconsideration.
>
> --
> Allin Cottrell
> Department of Economics
> Wake Forest University, NC
> _______________________________________________
> Gretl-users mailing list
> Gretl-users(a)lists.wfu.edu
> http://lists.wfu.edu/mailman/listinfo/gretl-users
>
--
"Money won't buy happiness, but it will pay the salaries of a large
research staff to study the problem." - Bill Vaughan
--
7:55 p.m.
On Thu, 5 Apr 2007, Talha Yalta wrote:
Actually having more than 33 new icons (models or graphs) will
not fit inside my icon window even if it is maximized... I think
the problem comes from the fact that there are always 4 icons in
each row and enlarging the window only increases the horizontal
spacing between icons instead of reorganizing them to use the
extra space efficiently.
I've now made it so the number of icons per row adjusts if you
widen the session window.
OK, I see now. The directory I am referring to in the working
directory is a temporary one and it is deleted when the session
is closed. I also see that the individual .gretl files contain
everything. So it is OK. Still, this is confusing. I think it
would be better if the system temp directory was used for these
temporary folders and files.
I've now made these files hidden.
11- Here is a new werid one:
a)-I open gretl, I open a session (the icon view appears automatically).
b)-I close the icon view by clicking on the x icon on the top right corner.
c)-I close gretl by clicking on the x icon on the top right corner in
the main screen.
d)-The program crashes saying gretlw32.exe has encountered a problem
and needs to close.
I'll need to boot Windows to test that; I'm not seeing it on
Linux. If you have time, could you see if that's still happens
with today's Windows snapshot? Thanks.
Allin
12:27 p.m.
I've now made it so the number of icons per row adjusts if you
widen the session window.
Looks very nice. thanks.
I'll need to boot Windows to test that; I'm not seeing it on
Linux. If you have time, could you see if that's still happens
with today's Windows snapshot? Thanks.
I'm afraid it is still there.
Cheers
Talha
--
"Money won't buy happiness, but it will pay the salaries of a large
research staff to study the problem." - Bill Vaughan
--
12:57 p.m.
On Fri, 6 Apr 2007, Talha Yalta wrote:
> I'll need to boot Windows to test that; I'm not seeing
it on
> Linux. If you have time, could you see if that's still happens
> with today's Windows snapshot? Thanks.
I'm afraid it is still there.
Can I check I have this right? You said, open gretl; open a
session file; close the icon window via its window manager "X"
button; close gretl via the main window's "X" button -> crash.
Is that right?
I tried this on Windows and didn't get any ill effects. Could you
maybe send me a copy of the session file in question? I suppose
this might have something to do with the specifics of that file
(though right now I'm not sure how that could be).
Allin.
1:54 p.m.
I tried this on Windows and didn't get any ill effects. Could
you
maybe send me a copy of the session file in question? I suppose
this might have something to do with the specifics of that file
(though right now I'm not sure how that could be).
I just sent 2 of my session
files to your email (cottrell(a)wfu.edu). I
hope they are helpful.
Cheers
Talha
On 4/6/07, Allin Cottrell <cottrell(a)wfu.edu> wrote:
> On Fri, 6 Apr 2007, Talha Yalta wrote:
>
> > > I'll need to boot Windows to test that; I'm not seeing it on
> > > Linux. If you have time, could you see if that's still happens
> > > with today's Windows snapshot? Thanks.
> > I'm afraid it is still there.
>
> Can I check I have this right? You said, open gretl; open a
> session file; close the icon window via its window manager "X"
> button; close gretl via the main window's "X" button -> crash.
> Is that right?
>
I tried this on Windows and didn't get any ill effects. Could
you
maybe send me a copy of the session file in question? I suppose
this might have something to do with the specifics of that file
(though right now I'm not sure how that could be).
>
> Allin.
> _______________________________________________
> Gretl-users mailing list
> Gretl-users(a)lists.wfu.edu
> http://lists.wfu.edu/mailman/listinfo/gretl-users
>
--
"Money won't buy happiness, but it will pay the salaries of a large
research staff to study the problem." - Bill Vaughan
--
3:11 p.m.
On Fri, 6 Apr 2007, Talha Yalta wrote:
I just sent 2 of my session files to your email (cottrell(a)wfu.edu).
I
hope they are helpful.
Thanks. But here they open and close fine, so I'm still puzzled.
Could you tell me the path to these files on your system?
Conceivably it's something to do with path names.
Allin.
3:18 p.m.
They were in D:\Docs\Dissertation\GRETL. Can it be something to do
with what is installed on my computer etc?
On 4/6/07, Allin Cottrell <cottrell(a)wfu.edu> wrote:
On Fri, 6 Apr 2007, Talha Yalta wrote:
> I just sent 2 of my session files to your email (cottrell(a)wfu.edu). I
> hope they are helpful.
Thanks. But here they open and close fine, so I'm still puzzled.
Could you tell me the path to these files on your system?
Conceivably it's something to do with path names.
Allin.
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
--
"Money won't buy happiness, but it will pay the salaries of a large
research staff to study the problem." - Bill Vaughan
--
4:47 p.m.
Yes, it is a standard single user setup.
On 4/6/07, Allin Cottrell <cottrell(a)wfu.edu> wrote:
On Fri, 6 Apr 2007, Talha Yalta wrote:
> They were in D:\Docs\Dissertation\GRETL. Can it be something to do
> with what is installed on my computer etc?
I don't think that's likely. I presume you have permission to
write to that location OK?
Allin.
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
--
"Money won't buy happiness, but it will pay the salaries of a large
research staff to study the problem." - Bill Vaughan
--
5:21 p.m.
Some more ideas (shots in the dark actually) for bug replication:
- Do you guys have the same preferences settings (working dir etc.)?
- locale-specific? e.g. decimal separator setting, try running in English
- what exact windows version are you using?
A crash is not nice, so if you like I could try to reproduce it, too.
-sven
Talha Yalta wrote:
Yes, it is a standard single user setup.
On 4/6/07, Allin Cottrell <cottrell(a)wfu.edu> wrote:
> On Fri, 6 Apr 2007, Talha Yalta wrote:
>
> > They were in D:\Docs\Dissertation\GRETL. Can it be something to do
> > with what is installed on my computer etc?
>
> I don't think that's likely. I presume you have permission to
> write to that location OK?
>
> Allin.
> _______________________________________________
> Gretl-users mailing list
> Gretl-users(a)lists.wfu.edu
> http://lists.wfu.edu/mailman/listinfo/gretl-users
>
6:45 p.m.
On Fri, 6 Apr 2007, Sven Schreiber wrote:
Some more ideas (shots in the dark actually) for bug replication:
- Do you guys have the same preferences settings (working dir etc.)?
- locale-specific? e.g. decimal separator setting, try running in English
- what exact windows version are you using?
A crash is not nice, so if you like I could try to reproduce it,
too.
Quesion for Talha that might narrow it down: if you do as you
described, but exit gretl via /File/Exit, do you get this problem?
Or is it only if you close gretl via the window manager?
Allin.
9:22 p.m.
Well, I am using an English Windows XP so it can't be locale specific.
And yes, I get it when I exit via the window manager. One thing I
noticed is that it doesn't happen all the time but keep running gretl
and closing, it is bound to happen within a few times. Sometimes it
happens consistently for each try. I am not sure but might be that it
happens more with the tables-liquid.gretl file. In fact I created a
new session file and it never happened with that one.
Then I also created a 3rd session file (using one of the sample data)
and here something else happens (again not always but most of the
time):
I run gretl
I open the crash-test.gretl session file (attached)
I close gretl using the window decorator
and I get error messages such as:
The instruction at "0x7c91103c" referenced memory at "0x010c9294".
The
memory could not be "written".
or
The instruction at "0x7c91103c" referenced memory at "0x010c98f4".
The
memory could not be "written".
I hope this is helpful
Cheers
Talha
On 4/6/07, Allin Cottrell <cottrell(a)wfu.edu> wrote:
On Fri, 6 Apr 2007, Sven Schreiber wrote:
> Some more ideas (shots in the dark actually) for bug replication:
>
> - Do you guys have the same preferences settings (working dir etc.)?
> - locale-specific? e.g. decimal separator setting, try running in English
> - what exact windows version are you using?
>
> A crash is not nice, so if you like I could try to reproduce it,
> too.
Quesion for Talha that might narrow it down: if you do as you
described, but exit gretl via /File/Exit, do you get this problem?
Or is it only if you close gretl via the window manager?
Allin.
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
--
"Money won't buy happiness, but it will pay the salaries of a large
research staff to study the problem." - Bill Vaughan
--
12:01 a.m.
On Sat, 7 Apr 2007, Talha Yalta wrote:
Well, I am using an English Windows XP so it can't be locale
specific. And yes, I get it when I exit via the window manager.
By "via the window manager" I mean, by clicking the "X" or close
button on the window titlebar, which is what you indicated at
first. My question was, do you get the same unstable behavior if
you close gretl using the /File/Exit menu item?
Allin.
12:56 a.m.
No it never crashed when I used the /File/Exit menu item.
I experimented quite a bit with various combinations. It only happens
when clicking on the "x" on the window titlebar and not always but
often.
The crashtest.gretl is the one that usually produces the "memory could
not be "written" error when I click x on gretl window (without
clicking x on the icon view window first).
On the other hand, tables-liquid.gretl produces the "gretlw32.exe has
encountered a problem" crash after clicking on x on icon view and
gretl main window in that order.
On 4/7/07, Allin Cottrell <cottrell(a)wfu.edu> wrote:
On Sat, 7 Apr 2007, Talha Yalta wrote:
> Well, I am using an English Windows XP so it can't be locale
> specific. And yes, I get it when I exit via the window manager.
By "via the window manager" I mean, by clicking the "X" or close
button on the window titlebar, which is what you indicated at
first. My question was, do you get the same unstable behavior if
you close gretl using the /File/Exit menu item?
Allin.
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
--
"Money won't buy happiness, but it will pay the salaries of a large
research staff to study the problem." - Bill Vaughan
--
6473
days inactive
6478
days old
20 comments
3 participants
participants (3)
-
Allin Cottrell -
Sven Schreiber -
Talha Yalta