2:05 a.m.
Apologies to those who receive double notification -- and I won't
do this again, because I'll post right now the URL for the
gretl-announce list
http://ricardo.ecn.wfu.edu/gretl_announce.html
but, for anyone who's not on the announce list, gretl 1.6.4 is now
available at http://gretl.sourceforge.net/
This is mostly a quick fix for breakage in the Windows version of
gretl.
--
Allin Cottrell
Department of Economics
Wake Forest University, NC
8:21 a.m.
Allin Cottrell schrieb:
but, for anyone who's not on the announce list, gretl 1.6.4 is
now
available at http://gretl.sourceforge.net/
This is mostly a quick fix for breakage in the Windows version of
gretl.
Thanks for the quick reaction.
The old problems are gone, but there seems to be a new one:
Gretl does not persistently accept new preference settings for the user
directory. It is now set to "C:\Dokumente und
Einstellungen\<loginname>\Anwendungsdaten\gretl" (which BTW seems to
have changed?). When I try to set it to "C:\Dokumente und
Einstellungen\<loginname>\Eigene Dateien" (this is the My Files --or My
Data?-- equivalent in German), it seems to go ok, but reopening the
preference dialog shows the old unchanged setting.
This leads to gretl not finding my function packages, which are in the
other directory. (Of course it's not difficult to workaround...)
Sorry to be the bearer of bad news; I'm available for further testing
on Windows, of course.
cheers,
sven
2:38 p.m.
On Mon, 14 May 2007, Sven Schreiber wrote:
The old problems are gone, but there seems to be a new one:
Gretl does not persistently accept new preference settings for
the user directory...
OK, I see what's wrong. I put in place a fallback setting for the
user directory, using CSIDL_APPDATA. This is intended for the case
where
* somebody has installed gretl on a given machine, and then
* somebody else logs onto the machine and runs gretl
The second "somebody" will not have a userdir in the registry.
However, I forgot to make this code conditional on userdir being
an empty string! This is now fixed in CVS.
Should I put out 1.6.5?
And while we're at it:
The directory identified via CSIDL_APPDATA is described on msdn as
"The file system directory that serves as a common repository for
application-specific data. A typical path is C:\Documents and
Settings\username\Application Data."
We also have CSIDL_MYDOCUMENTS: "The virtual folder representing
the My Documents desktop item."
Also on msdn, under "How to Comply with Data and Settings
Management Requirements": "The My Documents folder is for
user-created data, but not for temporary storage or application
state data. Non-user data must be stored using the proper AppData
path..."
I've been assuming that CSIDL_APPDATA is the "correct" place to
put the gretl userdir, by default. But actually this directory
seems to be doing double duty. It might be better to split
"userdir", with temporary files going under CSIDL_APPDATA and
explicitly user-saved files under CSIDL_MYDOCUMENTS.
As you might imagine, it would probably take some time to get this
right, so I'm not about to do it right now. Moreover, it's not
totally obvious to me exactly how the breakdown between the two
locations should go. For example, which one should hold
X-12-ARIMA and/or Tramo-related files?
Allin
5:43 p.m.
Allin Cottrell schrieb:
However, I forgot to make this code conditional on userdir being
an empty string! This is now fixed in CVS.
Should I put out 1.6.5?
I don't know what the release policy is, but I would tend to answer yes;
the user dir problem should mostly affect existing gretl users, who will
be quite puzzled why they can't get their old settings to work.
In any case I personally would recommend that no new features are
introduced until 1.6.5. If you don't want two releases in such a short
time span, I can give you one or two more bugs that could be squashed in
the meantime... ;-)
...
I've been assuming that CSIDL_APPDATA is the "correct"
place to
put the gretl userdir, by default. But actually this directory
seems to be doing double duty. It might be better to split
"userdir", with temporary files going under CSIDL_APPDATA and
explicitly user-saved files under CSIDL_MYDOCUMENTS.
As you might imagine, it would probably take some time to get this
right, so I'm not about to do it right now. Moreover, it's not
totally obvious to me exactly how the breakdown between the two
locations should go. For example, which one should hold
X-12-ARIMA and/or Tramo-related files?
Hm, I'm not sure if it's worth the effort. Would you then do the
Mac-specifics also? Apart from the fact that new bugs will be introduced
by that switch... IMHO that would be a new feature, and it seems to me
that gretl could use some time to stabilize all the new stuff that has
been added in recent months.
just my 2 €-ct (can you read the Euro sign in your encoding?)
-sven
10:37 p.m.
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Sven Schreiber wrote:
Allin Cottrell schrieb:
> However, I forgot to make this code conditional on userdir being
> an empty string! This is now fixed in CVS.
>
> Should I put out 1.6.5?
I don't know what the release policy is, but I would tend to answer yes;
the user dir problem should mostly affect existing gretl users, who will
be quite puzzled why they can't get their old settings to work.
In any case I personally would recommend that no new features are
introduced until 1.6.5. If you don't want two releases in such a short
time span, I can give you one or two more bugs that could be squashed in
the meantime... ;-)
...
> I've been assuming that CSIDL_APPDATA is the "correct" place to
> put the gretl userdir, by default. But actually this directory
> seems to be doing double duty. It might be better to split
> "userdir", with temporary files going under CSIDL_APPDATA and
> explicitly user-saved files under CSIDL_MYDOCUMENTS.
>
> As you might imagine, it would probably take some time to get this
> right, so I'm not about to do it right now. Moreover, it's not
> totally obvious to me exactly how the breakdown between the two
> locations should go. For example, which one should hold
> X-12-ARIMA and/or Tramo-related files?
>
Hm, I'm not sure if it's worth the effort. Would you then do the
Mac-specifics also? Apart from the fact that new bugs will be introduced
by that switch... IMHO that would be a new feature, and it seems to me
that gretl could use some time to stabilize all the new stuff that has
been added in recent months.
just my 2 €-ct (can you read the Euro sign in your encoding?)
-sven
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
Hi all,
I just want to add my 2 cents also.
Once I did make a program that used the Windows Registry, and I found
the best location to save info for the user, to be in the key:
HKEY_CURRENT_USER\Software\MyProgram
This way when the program starts for the first time reads keys from that
location or creates them for the first run.
(sorry if am saying something you already know, I did not even look to
the gretlw32 code )
Hélio
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
iD8DBQFGSOSeHvesJJ5UqvkRAvMAAJ9UnNRfWJA350af5OK18l1xREIWIQCg7JXp
o9BXOiOGzNnREJRfHWSZNHs=
=vlSP
-----END PGP SIGNATURE-----
10:50 p.m.
On Mon, 14 May 2007, Hélio Guilherme wrote:
Once I did make a program that used the Windows Registry, and I
found
the best location to save info for the user, to be in the key:
HKEY_CURRENT_USER\Software\MyProgram
Yes, that makes sense, and that's what gretl uses for per-user
settings.
The question, though, is what we should use as the default
location for user files (e.g. saved data files, scripts, sessions
and so on)? I'm now inclined to think under "My Documents" may be
better than the current default, but I don't have time to test
that change so I'm not going to make it just now.
Allin.
10:41 p.m.
New subject: Please test! (was Re: gretl 1.6.4)
On Mon, 14 May 2007, Sven Schreiber wrote:
[ Should we release 1.6.5 real soon? ]
I don't know what the release policy is, but I would tend to
answer yes;
the user dir problem should mostly affect existing gretl users, who will
be quite puzzled why they can't get their old settings to work.
Yes, agreed. I've fixed the userdir issue (I think) and added
what Andreas requested (which was quite self-contained and
unlikely to break anything else). Since that involved a few new
strings I've also updated .pot and .po files.
Now: please beat on the Windows snapshot and see if it's OK!
Thanks.
http://ricardo.ecn.wfu.edu/pub/gretl/gretl_install.exe
If it's OK I'll swallow my pride and put yet another release.
Allin
2:52 p.m.
New subject: Please test! (was Re: gretl 1.6.4)
Allin Cottrell schrieb:
Yes, agreed. I've fixed the userdir issue (I think) and added
what Andreas requested (which was quite self-contained and
unlikely to break anything else). Since that involved a few new
strings I've also updated .pot and .po files.
so does that mean the translations should be updated as well? -- I guess so.
Now: please beat on the Windows snapshot and see if it's OK!
Thanks.
I only found trivial issues (so far;-):
1) the installer is 9MB, on the website it says 7MB
2) when I select the new menu item "nonparametric tests", I get the same
error dialog twice ("no suitable data...")
3) when I run a (py4gretl) function script, the session symbol window is
automatically closed -- but this is better than before, because no
spurious matrices appear anymore!
Some days ago I had the unpleasant experience that gretl "ate" some of
my variables, after I had created lags and done other stuff, but I can't
reproduce that anymore. I think an error in a function package also had
an influence. When it happens again, I'll try to track it down.
W.r.t. the discussion about storage locations: actually I'm wondering
about the use of the "current working directory" as it's currently
presented as an option in the preferences. You warned that it may not be
writable; so what's the use case for it anyway? (If it's abolished there
would be one preference pane less.)
Thanks for tackling the windows problems!
-Sven
3:31 p.m.
New subject: Please test! (was Re: gretl 1.6.4)
On Tue, 15 May 2007, Sven Schreiber wrote:
> Yes, agreed. I've fixed the userdir issue (I think) and
added
> what Andreas requested (which was quite self-contained and
> unlikely to break anything else). Since that involved a few new
> strings I've also updated .pot and .po files.
so does that mean the translations should be updated as well? --
I guess so.
Yup, if you have time.
> Now: please beat on the Windows snapshot and see if it's OK!
> Thanks.
I only found trivial issues (so far;-):
1) the installer is 9MB, on the website it says 7MB
OK, fixed.
2) when I select the new menu item "nonparametric tests", I
get
the same error dialog twice ("no suitable data...")
Fixed.
3) when I run a (py4gretl) function script, the session symbol
window is automatically closed -- but this is better than
before, because no spurious matrices appear anymore!
We'll have another go at that for the next "real" release (after
the upcoming fix release).
Some days ago I had the unpleasant experience that gretl
"ate"
some of my variables, after I had created lags and done other
stuff, but I can't reproduce that anymore. I think an error in a
function package also had an influence. When it happens again,
I'll try to track it down.
Thanks, please do.
W.r.t. the discussion about storage locations: actually I'm
wondering about the use of the "current working directory" as
it's currently presented as an option in the preferences. You
warned that it may not be writable; so what's the use case for
it anyway? (If it's abolished there would be one preference pane
less.)
This is a useful thing if you're launching gretl from an xterm
(which I usually am). However, maybe it should be suppressed in
the Preferences dialog on Windows? On the other hand, I could
_imagine_ somebody setting up a desktop shortcut to gretl with a
specific working directory and using this feature -- has anyone
actually done that?
Allin.
5:06 p.m.
New subject: Please test! (was Re: gretl 1.6.4)
On 15/05/07, Allin Cottrell <cottrell(a)wfu.edu> wrote:
In general for most programs I advise students to set up a directory
and to enter that directory in the start up item on the properties tab
of the shortcut. This means that all files relevant to an individual
project are kept in that directory. I think that some Gretl file
access and save default to the directory specified in the program
short-cut.
If an other user Gretl directory is specified in the General tab of
the gretl:options could this be used to duplicate the directory
structure of the main Gretl directory and have script, data etc.
subdirectories which are the property of the user and which he may use
in different projects. I suspect that this may appear obvious to a
Linux user but it is not how many people work in Windows. Section 4.8
of the user's guide might be extended to explain the recommended
practice .
One trivial point. The references to the Latex Companion on page 151
might be updated to refer to the second edition (2004)
@BOOK{Mittelbach2004,
title = {The \LaTeX Companion Second Edition},
publisher = {Addison-Wesley},
year = {2004},
author = {Frank Mittelbach and Michael Goessens },
timestamp = {2007.05.15}
}
Best Regards
John
This is a useful thing if you're launching gretl from an xterm
(which I usually am). However, maybe it should be suppressed in
the Preferences dialog on Windows? On the other hand, I could
_imagine_ somebody setting up a desktop shortcut to gretl with a
specific working directory and using this feature -- has anyone
actually done that?
Allin.
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
--
John C Frain
Trinity College Dublin
Dublin 2
Ireland
www.tcd.ie/Economics/staff/frainj/home.html
mailto:frainj@tcd.ie
mailto:frainj@gmail.com
2:37 p.m.
New subject: Please test! (was Re: gretl 1.6.4)
On Tue, 15 May 2007, Sven Schreiber wrote:
3) when I run a (py4gretl) function script, the session symbol window
is
automatically closed -- but this is better than before, because no
spurious matrices appear anymore!
Actually this does not sound right. Could you send me test cases
(scripts) for py4gretl, please? Then I can, hopefully, shake out
the remaining bugs in the function calling apparatus. Thanks.
Allin
1:03 p.m.
New subject: Please test! (was Re: gretl 1.6.4)
Allin Cottrell schrieb:
On Tue, 15 May 2007, Sven Schreiber wrote:
> 3) when I run a (py4gretl) function script, the session symbol window is
> automatically closed -- but this is better than before, because no
> spurious matrices appear anymore!
Actually this does not sound right. Could you send me test cases
(scripts) for py4gretl, please? Then I can, hopefully, shake out
the remaining bugs in the function calling apparatus. Thanks.
Sure, the following produces the effect for me (this should hopefully
also work with the public py4gretl_vecrestrict version on the server).
Of course you have to open the session symbol window before executing
the script.
BTW, there are still some bugs related to function packaging, gretl
still gets confused by functions with the same names (for example, from
an existing package that was run during the session, and then running a
new version for debugging and/or development purposes). So my rule for
the moment is to quit and restart gretl if I want to package a function.
But since that normally won't affect regular users, I think it's not
that urgent to fix it. However, that was the reason why I mentioned
gretl could use some settling-down time. It may also be that those bugs
sometimes affect the mere running of existing packages as well, I'm not
sure.
cheers,
-Sven
open denmark
list endo = LRM LRY
list rexo = null
list uexo = null
scalar n_y = nelem(endo)
scalar n1 = n_y + nelem(rexo) # not if determ_case = 2 or 4
scalar cirank = 1
matrix a_pat = ones(nelem(endo), cirank) * 99
matrix b_pat = 99*ones(n1-cirank, cirank) | I(cirank)
matrix test2 = py4gretl_vecrestrict(endo, 2, cirank, 3, 1, rexo, uexo,
a_pat, b_pat, 2000, 6)
3:52 p.m.
New subject: Please test! (was Re: gretl 1.6.4)
On Wed, 16 May 2007, Sven Schreiber wrote:
> Actually this does not sound right. Could you send me test
cases
> (scripts) for py4gretl, please? Then I can, hopefully, shake out
> the remaining bugs in the function calling apparatus. Thanks.
Sure, the following produces the effect for me (this should hopefully
also work with the public py4gretl_vecrestrict version on the server).
Of course you have to open the session symbol window before executing
the script.
Thanks. The closing of the session icon window was just due to
the fact that the calling script opens a new data file. If you
open denmark.gdt manually first, then run the test script with the
"open" line commented out, the session icon window is not closed,
and it displayed the temporary matrices. This is now fixed.
Other small fixes. (a) The Help button in the function call
window: the function help text is now word-wrapped rather than
showing very long lines. (b) If you fail to specify a required
argument in the GUI function call window, and press OK, you get an
error message. That's fine, but previously the call window closed
on this error. Now it stays open, waiting for correct input,
unless you click Cancel.
Allin.
2:10 p.m.
New subject: Please test! (was Re: gretl 1.6.4)
Allin Cottrell schrieb:
On Wed, 16 May 2007, Sven Schreiber wrote:
>> Actually this does not sound right. Could you send me test cases
>> (scripts) for py4gretl, please? Then I can, hopefully, shake out
>> the remaining bugs in the function calling apparatus. Thanks.
> Sure, the following produces the effect for me (this should hopefully
> also work with the public py4gretl_vecrestrict version on the server).
> Of course you have to open the session symbol window before executing
> the script.
Thanks. The closing of the session icon window was just due to
the fact that the calling script opens a new data file.
Ah ok, dummy me!
open denmark.gdt manually first, then run the test script with the
"open" line commented out, the session icon window is not closed,
and it displayed the temporary matrices. This is now fixed.
Thanks!
-sven
6188
days inactive
6190
days old
13 comments
4 participants
participants (4)
-
Allin Cottrell -
Hélio Guilherme -
John C Frain -
Sven Schreiber