[GRASS-dev] Principle G3D library question

Soeren Gebbert soerengebbert at googlemail.com
Sat Jun 18 15:41:57 EDT 2011


Hello Hamish,
many thanks for code reviewing.

2011/6/18 Hamish <hamish_b at yahoo.com>:
> [r3.in.ascii]
>
>
> readHeaderString()
>   while (fgetc(fp) != '\n');
>
> probably it should use G_getl2(), which also accepts DOS and MacOS9 newlines?

Done in svn.

>
>> version:
>> order:
>
> once formally decided, we should modify the 6.x versions to deal with the
> extra header instructions, and continue on (with a message?) if it matches
> grass6's format, otherwise fatal_error out with a reasonable error message.

I don't think it is necessary to backport the header changes to 6.x.
The grass6 version works fine, maybe a manual update is meaningful?
For backward compatibility a flag was added to r3.out.ascii which is
IMHO not overkill. :)

>
> ASCII is an acronym of a named entity, and should remain uppercased.

Fixed in svn.

>
>
>> Note that unit tests for r3.in.ascii are implemented in the
>> <em>test.r3.out.ascii.sh</em> script located in the r3.out.ascii directory.
>
> if many/most modules will eventually get test scripts, is it worth
> mentioning that in the help page? maybe <!-- commented out --> so only
> developers see it?

Good test scripts provide valuable information about the module
capabilities, data handling and possible bugs. Test scripts should be
designed so that grass user can easily read and write them ...  at
least this is the goal we have. My hope is that user will write test
scripts in case they found a bug or in case test scripts are still
missing.
Therefor in my humble opinion the test scripts should be added
automatically to the end of the modules HTML manual pages. In case
they are missing in the module directory (r3.in.ascii) the manual
should explain why.

>
> I think is it enough to describe in the grass7 help page how
> to make a grass6-compatible G3D output file, probably with an
> example. A flag for that could be overkill.
>
>
> btw, I believe the "AV" responsible for parts of the G3D code is Alfonso Vitti.

Thanks, i was now able to find the g3d related mails in the grass5
developer mail archive. Very valuable information.

Best regards
Soeren


More information about the grass-dev mailing list