Re: preliminary patch for a post tag trigger

[Mike Sutton]

"Mark D. Baushke" wrote:
> 
> Hi Mike,
> 
> First, I like the idea of the posttaginfo feature.
> 
> I have not yet looked closely at the patches, but in the absense of
> documentation, I have a few questions:
> 
>   a) Do you have any patches for the documentation for this new feature?
>   b) Do you have any tests to be added to src/sanity.sh to verify that the
>      feature is working correctly?
 
Not yet.  As I mentioned this prelimary patch does not conform to the
"rules" outlined in the "HACKING" file.  I'm just looking for initial
reactions.  

>   c) During a tagging operation there is the possibility that a tag was
>      not actually moved as a part of the tag command. It might be
>      desirable to know the the files for which the tag operation failed.
>      In fact, in my opinion, that would be the biggest benefit of the
>      feature. A way to tell the user or some other system that the tag
>      operation had failed for some (but not all) files and what files
>      succeeded and what files failed.

Good question.  I can't answer this one.

> 
> I did notice one thing when looking at the patch, I see this:
> 
> +    "# A non-zero exit of the filter program will cause the tag to be 
> aborted.\n",

This isn't really applicable for the posttaginfo file.  It should be
removed.  I just copied the tagino header information to use with
posttaginfo.

> 
> however, this information seems ambiguous to me as the tag has been
> applied to at least some files in at least one directory or this
> posttaginfo script would not have been called.
> 
> I am guessing you are going to something like the loginfo script does
> which will abort commits to subsequent directories after having
> committed to the first directory. I will grant you that it has
> precedent. I am just not sure if it is desirable or not.

One of the basic goals was to create a mechanism that would allow a
single email to be sent after a tag operation. 

Regards,

Mike Sutton