10:54 p.m.
Hi Allin,
Some entries from Model Window->Save are missing translation. No entry in
POT for the following function:
At line 5722
./gui2/library.c: blurb = g_strdup_printf("Statistic from model %d\n"
Hope that it was caught on time.
----
Attached is the file that I am preparing to have in gretl CVS root.
What do you think?
Thanks,
Hélio
3:24 p.m.
On Fri, 23 Sep 2011, Hélio Guilherme wrote:
Some entries from Model Window->Save are missing translation. No
entry in
POT for the following function:
At line 5722
./gui2/library.c: blurb = g_strdup_printf("Statistic from model %d\n"
Hope that it was caught on time.
OK, I've marked that in CVS.
Attached is the file that I am preparing to have in gretl CVS root.
What do you think?
Can you check your guide against the material in Appendix C2
of the Gretl User's Guide, please? I don't think zlibc is a
gretl dependency. Also, the simple suggestion in the manual
seems a good one:
apt-get build-dep gretl
Allin
11:33 p.m.
Hi,
On Sat, Sep 24, 2011 at 4:24 PM, Allin Cottrell <cottrell(a)wfu.edu> wrote:
Attached is the file that I am preparing to have in gretl CVS root.
> What do you think?
>
Can you check your guide against the material in Appendix C2 of the Gretl
User's Guide, please? I don't think zlibc is a gretl dependency. Also, the
simple suggestion in the manual seems a good one:
apt-get build-dep gretl
Allin
I did a step-by-step installation of the packages that ./configure reported
as necessary. In the search for zlib (with synaptic) the only package that I
saw to fit was that one. Since then ./configure stopped complaining, I
identified it has the required one. I will double check what are the zlib
packages in my system.
I also think that ./configure reported GMP as required. How can we confirm
this?
Now, in that README.packages file I will add the reference to Appendix C2.
The purpose of the README.packages file, is kind of a shortcut from
documentation. A user can see in gretl webpages how to get CVS tree, and
then can just copy the commands to get the packages.
Furthermore, I plan to have the commands for OpenSuSe and Fedora/RedHat or
any other Linux. This is because sometimes the packages do not have the same
names (and packages managers, like "zypper" and "yum").
---
Since you actually made me read Appendix C2 (I was not aware that
documentation had improved so much since I read it long time ago ;)), here i
a typo that I detected:
(...)
apt-get build-dep gretl
will download and install all the necessary packages for building the
version of gretl that is currently present in your APT sources. Techincally,
---
Thanks,
Hélio
12:37 a.m.
Well, the model description is not being translated.
You initialize them with N_(), shouldn't it be with _()?
(...)
case M_HQC:
descrip = N_("Hannan-Quinn Information Criterion");
val = pmod->criterion[C_HQC];
statname = "$hqc";
break;
default:
dummy_call();
return;
}
sprintf(vname, "%s_%d", statname + 1, pmod->ID);
blurb = g_strdup_printf(_("Statistic from model %d\n"
"%s (value = %g)\n"
"Name (max. 15 characters):"),
pmod->ID, descrip, val);
3:58 a.m.
On Sun, 25 Sep 2011, Hélio Guilherme wrote:
Well, the model description is not being translated.
You initialize them with N_(), shouldn't it be with _()?
You have correctly identified a problem with the translatable
strings, and that would be one way of fixing it, but I prefer
another way.
Could I request, please don't modify C files in the CVS repository
without getting approval first? It's not that I want to maintain a
monopoly, but it can get confusing if the C sources are changed
without notice. The tacit arrangement at present is that only Jack
Lucchetti and myself make such changes, although we can surely add
others to that list as seems appropriate.
Allin Cottrell
4:04 a.m.
On Sun, 25 Sep 2011, Hélio Guilherme wrote:
On Sat, Sep 24, 2011 at 4:24 PM, Allin Cottrell
<cottrell(a)wfu.edu> wrote:
>
> Can you check your guide against the material in Appendix C2 of the Gretl
> User's Guide, please? I don't think zlibc is a gretl dependency. Also, the
> simple suggestion in the manual seems a good one:
>
> apt-get build-dep gretl
>
I did a step-by-step installation of the packages that ./configure reported
as necessary. In the search for zlib (with synaptic) the only package that I
saw to fit was that one. Since then ./configure stopped complaining, I
identified it has the required one. I will double check what are the zlib
packages in my system.
I think you'll find that zlib1g is the required package; zlibc is a
rather strange package that is not needed by gretl.
I also think that ./configure reported GMP as required. How can we
confirm
this?
It is required, yes. You can see that in the ChangeLog entry for
gretl 1.9.2, at http://gretl.sourceforge.net/ChangeLog.html .
Now, in that README.packages file I will add the reference to
Appendix C2.
The purpose of the README.packages file, is kind of a shortcut from
documentation. A user can see in gretl webpages how to get CVS tree, and
then can just copy the commands to get the packages.
Furthermore, I plan to have the commands for OpenSuSe and Fedora/RedHat or
any other Linux. This is because sometimes the packages do not have the same
names (and packages managers, like "zypper" and "yum").
Good idea.
Since you actually made me read Appendix C2 (I was not aware that
documentation had improved so much since I read it long time ago ;)), here i
a typo that I detected:
(...)
apt-get build-dep gretl
will download and install all the necessary packages for building the
version of gretl that is currently present in your APT sources. Techincally,
OK, perhaps that needs to be clarified.
Allin
2:36 p.m.
On Sun, Sep 25, 2011 at 4:58 AM, Allin Cottrell <cottrell(a)wfu.edu> wrote:
On Sun, 25 Sep 2011, Hélio Guilherme wrote:
Well, the model description is not being translated.
> You initialize them with N_(), shouldn't it be with _()?
>
You have correctly identified a problem with the translatable strings, and
that would be one way of fixing it, but I prefer another way.
Could I request, please don't modify C files in the CVS repository without
getting approval first? It's not that I want to maintain a monopoly, but it
can get confusing if the C sources are changed without notice. The tacit
arrangement at present is that only Jack Lucchetti and myself make such
changes, although we can surely add others to that list as seems
appropriate.
Sorry for that. This is the second time you warn me about this.
I'll keep my hand off it.
Hélio
2:58 p.m.
On Sun, 25 Sep 2011, Hélio Guilherme wrote:
Sorry for that. This is the second time you warn me about this.
I'll
keep my hand off it.
No, please don't. Your contribution is most welcome. The only thing is,
gretl has become a huge project and it's very easy to introduce changes
which are good at first sight but in fact clash with something else which
lives in some other remote part of the code you know little about.
It still happens to me all the time. True, RC systems allow to revert
changes, but having to revert a commit still makes things messy. Allin and
I have found a way to deal with this quite satisfactorily: if I'm not 100%
sure of what I'm doing, I send a diff file to Allin instead of committing
my changes, so he can see what I'm trying to do and how. In most cases,
Allin edits my code, makes it more readable, more efficient and more
consistent with the overall structure and then commits it himself.
This system has worked pretty well over the years.
Riccardo (Jack) Lucchetti
Dipartimento di Economia
Università Politecnica delle Marche
r.lucchetti(a)univpm.it
http://www.econ.univpm.it/lucchetti
3:09 p.m.
On Sun, 25 Sep 2011, Riccardo (Jack) Lucchetti wrote:
On Sun, 25 Sep 2011, Hélio Guilherme wrote:
> Sorry for that. This is the second time you warn me about this. I'll keep
> my hand off it.
No, please don't. Your contribution is most welcome. [...]
Agreed, and sorry if I sounded tetchy.
Allin
10:23 p.m.
On Sun, 25 Sep 2011, Allin Cottrell wrote:
On Sun, 25 Sep 2011, Hélio Guilherme wrote:
> Since you actually made me read Appendix C2 (I was not aware that
> documentation had improved so much since I read it long time ago ;)), here
> i
> a typo that I detected:
> (...)
> apt-get build-dep gretl
> will download and install all the necessary packages for building the
> version of gretl that is currently present in your APT sources.
> Techincally,
OK, perhaps that needs to be clarified.
Hmm, I'm not sure about that. On my Ubuntu (lucid) system at home, I
just tried
sudo apt-get build-dep gretl --dry-run
to see what would happen. I have current CVS gretl installed under
/opt, but I don't have any standard Ubuntu/Debian gretl files on the
system. The command ran fine, and gave the following:
<shell>
allin@myrtle:~$ sudo apt-get build-dep gretl --dry-run
Reading package lists... Done
Building dependency tree
Reading state information... Done
The following packages will be REMOVED:
libreadline5-dev
The following NEW packages will be installed:
gnuplot libdb-dev libdb4.8-dev libreadline-dev libreadline6-dev
0 upgraded, 5 newly installed, 1 to remove and 0 not upgraded.
Remv libreadline5-dev [5.2-7build1]
Inst gnuplot (4.2.6-1 Ubuntu:10.04/lucid)
Inst libreadline6-dev (6.1-1 Ubuntu:10.04/lucid)
Inst libreadline-dev (6.1-1 Ubuntu:10.04/lucid)
Inst libdb4.8-dev (4.8.24-1ubuntu1 Ubuntu:10.04/lucid)
Inst libdb-dev (4.8 Ubuntu:10.04/lucid)
Conf gnuplot (4.2.6-1 Ubuntu:10.04/lucid)
Conf libreadline6-dev (6.1-1 Ubuntu:10.04/lucid)
Conf libreadline-dev (6.1-1 Ubuntu:10.04/lucid)
Conf libdb4.8-dev (4.8.24-1ubuntu1 Ubuntu:10.04/lucid)
Conf libdb-dev (4.8 Ubuntu:10.04/lucid)
</shell>
I'm not sure about libdb -- it's not a direct gretl dependency but
maybe is an indirect requirement? But anyway, the main point is that
apt-get was able to give me a dependency list even though no gretl
sources that apt knows about are installed on this computer.
Nonetheless I'm sure this is right: if the dependencies of current
CVS gretl were to differ from those of the latest gretl version
registered by Ubuntu/Debian, "apt-get build-dep gretl" would give
the latter, not the former, which might not be what the user needs.
But most of the time it shouldn't be far off.
Allin
4598
days inactive
4600
days old
9 comments
3 participants
participants (3)
-
Allin Cottrell -
Hélio Guilherme -
Riccardo (Jack) Lucchetti