> On 27 Jun 2017, at 20:44, Junio C Hamano <gitster@xxxxxxxxx> wrote: > > Lars Schneider <larsxschneider@xxxxxxxxx> writes: > >> "rot13-filter.pl" always writes "OUT <size>" to the debug log at the end >> of a response. >> >> This works without issues for the existing responses "abort", "error", >> and "success". A new response "delayed", that will be introduced in a >> subsequent patch, accepts the input without giving the filtered result >> right away. Since we actually have the data already available in our >> mock filter the debug log output would be wrong/misleading. Therefore, >> we do not write "OUT <size>" for "delayed" responses. > > I still do not get why you think it makes any difference that you > are hoarding the result in the mock program. If the filter needs to > read a prepared result from a file in t/t0021/ before responding to > a real request after it replies to "delayed", would that change the > argument above? From Git's and the t0021-conversion.sh test's point > of view, I do not think it makes an iota of difference---it's an > implementation detail of the mock program. > > I am totally lost. > > Isn't the point of removing the log output from response to "delayed" > that the filter does not give the output back to Git at that point, > hence generally the size would not be available in the real-world > use case (not in the mock program)? Correct! Sorry for the confusion. How about this? "rot13-filter.pl" always writes "OUT <size>" to the debug log at the end of a response. This works perfectly for the existing responses "abort", "error", and "success". A new response "delayed", that will be introduced in a subsequent patch, accepts the input without giving the filtered result right away. At this point we cannot know the size of the response. Therefore, we do not write "OUT <size>" for "delayed" responses. To simplify the code we do not write "OUT <size>" for "abort" and "error" responses either as their size is always zero. Thanks, Lars