Re: [gnugo-devel] auto generated regression tests

[Gunnar Farneback]

Dan wrote:
> You could say we have an established practice.

No, we don't. 

> Our functions status_to_string and safety_to_string produce upper
> case output,

In my opinion that's a mistake, and they are used far from
consistently. See for example the -w output:

H4 : W dragon size 3 (10.62), genus 1, half eyes 0, escape factor 4, status 
unknown, matcher status alive, moyo size 0 safety weak, owl status alive
... neighbors F5 B11 C4

> and ALIVE, DEAD, ... are used consistently in both the source and
> the documentation.

In the source we of course use that since those are macros. In the
documentation use of upper case and lower case are mixed, try e.g.
'grep -i dead *.texi' and see. In my opinion it makes sense to use
upper case when it refers to values of specific data fields, but not
when talking about the status of a dragon in general.

In summary I see no good reason to change the output of the gtp
command dragon_status and I do see reasons not to change.

/Gunnar