Re: [PATCH 06/15] sequencer: lib'ify read_populate_todo()

Johannes Schindelin <Johannes.Schindelin@gmx.de> writes:

I think you are referring to the patch part in this case.

As I was not making point-by-point comments on the proposed commit log message, quoting only that part and cutting the patch text would have been pointless. I could have cut the proposed log message and left the patch in, though, because the comments were not about what was in the proposed log message, but about what was not in it [*1*].

I left the patch part for other people's use, to make it easy for them to see what I was saying was correct and appropriate for what the patch does. Removing that would not have made much sense.

But that is only true in this case. I try to see if I can trim quote more aggressively, but I would still err on the side of over-quoting than under-quoting [*2*].

[Footnote]

*1* As to what was IN the proposed log message, I have one comment. I do not think "To be truly useful" adds any value, as there is nothing "truly" about what this series does. The original was "truly" useful for the purpose of the sequencer machinery and its use in the current callers, just like with this series it becomes "truly" useful for envisioned new callers that want to handle error conditions themselves. The change is making it "more useful" for one different use case. It may not be "truly" useful for other use cases that sequencer machinery could be used even with the "eradicate die and exit" change and other people may bring a new use case that wants it to be even "more useful". The fact that it may not be directly usable by a new use case without further change does not make the result of applying this proposed series less than "truly useful". The "truly" adjective implies absolute, but there is nothing absolute in incremental improvements. It is always relative to the context within which the machinery was designed to be used. "To make it usable for callers that want to handle errors themselves (instead of just dying and the calling process handle it), let's turn die's and exit's to returning negative values." is probably closer to what I would have expected.

*2* As I read the quoted part before sending my response out, erring on the side not to underquote tends to avoid a mistake that invites: "I think you misunderstood what I wrote in the part you snipped from your quote; perhaps you skimmed it without fully reading."