I must have missed some revision

classic Classic list List threaded Threaded
5 messages Options
Reply | Threaded
Open this post in threaded view
|

I must have missed some revision

Federico Bruni-2
Hi,

I'm updating the Learning Manual.
I've generated a diff of all the files to be translated with the
following command:

cd Documentation
make ISOLANG=it NO_COLOR=1 check-translation > ~/lily-it.diff

As I do not trust scripts too much, I did some double checks for every
file.  Also, I read through the online manual to find broken links and
some other errors.

And I've noticed that in the italian translation of LM 3.1.2 there are
two snippets saying "insert the whole score of a Wagner opera in here".
See:

git grep Wagner

(the first is not translated, the second is translated)

The second snippet has actually been removed in the original file.
But the diff file doesn't show this difference!
That's a big issue for me :(

How can I track the commit that made this change?
And how can I have missed that?

Thanks,
Federico



Reply | Threaded
Open this post in threaded view
|

Re: I must have missed some revision

Francisco Vila
2011/6/20 Federico Bruni <[hidden email]>:
> Hi,
>
> I'm updating the Learning Manual.
> I've generated a diff of all the files to be translated with the
> following command:
>
> cd Documentation
> make ISOLANG=it NO_COLOR=1 check-translation > ~/lily-it.diff

Keep in mind that this diff shows differences made to original files
from committish IDs in your files, be they correct or wrong.

> As I do not trust scripts too much, I did some double checks for every
> file.

In my experience, these scripts are trustworthy, what is not is our
use of them and specifically our committish IDs are wrong from time to
time.

>Also, I read through the online manual to find broken links and
> some other errors.
>
> And I've noticed that in the italian translation of LM 3.1.2 there are
> two snippets saying "insert the whole score of a Wagner opera in here".
> See:
>
> git grep Wagner
>
> (the first is not translated, the second is translated)

I obtain a confusing output from this command. One instance in
original English, one in es/fr/ja/nl , two in hu/ , four in it/

> The second snippet has actually been removed in the original file.
> But the diff file doesn't show this difference!
> That's a big issue for me :(

You added the string, then updated the Committish ID.  So the diff
doesn't show it because it thinks the file is updated.

See commit e2c733c0e3b8c1fc9110ee0797afeb792c536e78
Author: Federico Bruni <[hidden email]>
Date:   Sun Apr 11 23:01:18 2010 +0200

    Web-it: add italian translation

/Documentation/it/web/introduction.itexi
+of the Overture to Wagner's @emph{Tannh...

> How can I track the commit that made this change?

I do something naive but effective: git log -p [here you enter in a
less session] then search by means of the '/' key. Type "Wagner"
<Enter>, search forward with the 'n' key, backwards with 'shift-n' and
reset to to with 'p'.

> And how can I have missed that?

You probably translated an old version and marked the file as updated,
i.e. in sync with latest version of the original.  This is the commit
that changed the original:

commit bf982420290accec610083a8cfa7a6c7768355f7
Author: Colin Campbell <[hidden email]>
Date:   Sun Mar 14 19:22:31 2010 -0600

    DOC: minor fix in website Intro & examples

    Removed obsolete Liszt-Wagner example.
    Credited Schenker graph to Karl Shaffer
(...)
-of the Overture to Wagner's @emph{Tannhäuser}. The spacing


(I do not pretend the above information to be complete, just giving
some clues that could hopefully help)

--
Francisco Vila. Badajoz (Spain)
www.paconet.org , www.csmbadajoz.com


Reply | Threaded
Open this post in threaded view
|

Re: I must have missed some revision

Francisco Vila
2011/6/20 Francisco Vila <[hidden email]>:
>> How can I track the commit that made this change?
>
> I do something naive but effective: git log -p [here you enter in a
> less session] then search by means of the '/' key. Type "Wagner"
> <Enter>, search forward with the 'n' key, backwards with 'shift-n' and
> reset to to with 'p'.

You can also use

  gitk --all &

and use "search for [<<][>>] revision which adds/removes string: Wagner"
the use the [<<] button to go backwards in time.

--
Francisco Vila. Badajoz (Spain)
www.paconet.org , www.csmbadajoz.com


Reply | Threaded
Open this post in threaded view
|

Re: I must have missed some revision

Federico Bruni-2
Il giorno lun, 20/06/2011 alle 23.37 +0200, Francisco Vila ha scritto:
> You can also use
>
>   gitk --all &
>
> and use "search for [<<][>>] revision which adds/removes string:
> Wagner"
> the use the [<<] button to go backwards in time.

this helped me a lot, thanks
I've found that the change was made in August 2009, commit:
c287061603fba87f9987c79aea45a00eeeca0033

My first commit of it/learning/fundamental.itely was in December 2009,
so many months after this change.
What can be happened here?!?

The only problem I see is that the committish of my first translation
gives 'fatal: bad object':
a30fa41fac8fb2771a76f6d1edbec9714298ee36

But this does not explain why I missed that change.





Reply | Threaded
Open this post in threaded view
|

Re: I must have missed some revision

Francisco Vila
2011/6/21 Federico Bruni <[hidden email]>:

> Il giorno lun, 20/06/2011 alle 23.37 +0200, Francisco Vila ha scritto:
>> You can also use
>>
>>   gitk --all &
>>
>> and use "search for [<<][>>] revision which adds/removes string:
>> Wagner"
>> the use the [<<] button to go backwards in time.
>
> this helped me a lot, thanks
> I've found that the change was made in August 2009, commit:
> c287061603fba87f9987c79aea45a00eeeca0033
>
> My first commit of it/learning/fundamental.itely was in December 2009,
> so many months after this change.
> What can be happened here?!?

I am sorry but it's difficult for me to answer this unespecific
question.  References to the problem are now a bit sparse.  That
commit you are talking about is 9181df064925f1de66 Docs-it: initial
translation, from 2009-12-11.  There, you add

+Un blocco @code{\score} deve sempre contenere una sola espressione musicale,
+e questa deve trovarsi subito dopo il comando @code{\score}.
+Ricorda che un'espressione musicale può essere qualsiasi cosa, da una
+singola nota a una grande espressione composta come
+
+@example
+@{
+  \new StaffGroup <<
+    @var{...insert the whole score of a Wagner opera in here...}
+  >>
+@}
+@end example

i.e. you don't translate the @example.  Then,

> The only problem I see is that the committish of my first translation
> gives 'fatal: bad object':
> a30fa41fac8fb2771a76f6d1edbec9714298ee36

This is an unexistant version.  Where did you take it from? Anyway,
you can still guess wich version did you translate the file from, put
a correct commitish in your file and run the script again; that should
give you a correct diff of changes to the original.

> But this does not explain why I missed that change.

I think yes, but I could be wrong.  I could need a coffe just now.

--
Francisco Vila. Badajoz (Spain)
www.paconet.org , www.csmbadajoz.com