<div dir="ltr">Splinting a patch into separate check it's not a problem. I have organized commits in that way.<div><br></div><div>About a 'author' in name, <span style="font-size:12.8px">organization is also a author, so clocky is a 'one man </span><span style="font-size:12.8px">organization' :D.</span></div><div><span style="font-size:12.8px"><br></span></div><div><span style="font-size:12.8px">It's much more simple, I have no benefits in 'anonymous' contribution, in that case I would rather go into separate project.</span></div><div><span style="font-size:12.8px">From a user-point of view, i'm sure that no-one would like to see 'clocky', 'google', 'llvm' in a compiler output, and I agree with that.</span></div><div><span style="font-size:12.8px">But on file level it's much more easier to organize checks in that way, as I't keep author more responsible for own checkers.</span></div><div><span style="font-size:12.8px"><br></span></div><div><span style="font-size:12.8px">For future my proposal could be:</span></div><div><span style="font-size:12.8px">- leave files/folders name as they are</span></div><div><span style="font-size:12.8px">- add a check 'tag' engine:</span></div><div><span style="font-size:12.8px"><br></span></div><div><span style="font-size:12.8px">So output from --list-checks-ex could be:</span></div><div>Module Check-name Tags</div><div><span style="font-size:12.8px">google explicit-constructor [reliability, security]</span></div><div><span style="font-size:12.8px">clocky inefficient-container-insert [performance, c++11]</span></div><div><span style="font-size:12.8px"><br></span></div><div><span style="font-size:12.8px">In this way, user would be able to run check in few ways: clocky-</span><span style="font-size:12.8px">inefficient-container-insert, performance-</span><span style="font-size:12.8px">inefficient-container-insert, c++11-</span><span style="font-size:12.8px">inefficient-container-insert, </span><span style="font-size:12.8px">c++11-performance-</span><span style="font-size:12.8px">inefficient-container-insert, </span><span style="font-size:12.8px">inefficient-container-insert</span><span style="font-size:12.8px">, in short any combination of tags would be accepted. But it would select a check that have all tags.</span></div><div><span style="font-size:12.8px">And a name that user used to enable check would show up in warning, a default would be a first tag name performance-inefficient-container-insert, if not available then module name.</span></div><div><span style="font-size:12.8px">Also user would be able to enable all check from specific tag like --checks= "-*, performance"., or pair of tags --checks="performance-c++11" to enable all checks that are c++11 performance checks.</span></div><div>A output from --list-checks would be a list of default names.</div><div><br></div><div>In that way no new 'modules' would need to be created to tag a checks (<span style="font-size:12.8px">cert-*).</span></div><div><span style="font-size:12.8px">Tests still would use a module name as prefix and in file names.</span></div><div>Module could be also printed as a tag, instead of separate column.</div><div>Checks names would need to be unique.</div><div><br></div><div>Right now documentation of clang-tidy checks is poor, some template should be used.</div><div>Now I see that some of checks are duplicated in both google and misc module, that also would need to be resolved out.</div><div><span style="font-size:12.8px"><br></span></div><div>Thats, just my ideas...</div></div><div class="gmail_extra"><br><div class="gmail_quote">2015-10-06 17:50 GMT+02:00 Aaron Ballman <span dir="ltr"><<a href="mailto:aaron.ballman@gmail.com" target="_blank">aaron.ballman@gmail.com</a>></span>:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class="">On Tue, Oct 6, 2015 at 8:56 AM, Piotr Zegar <<a href="mailto:piotrzsl@gmail.com">piotrzsl@gmail.com</a>> wrote:<br>
> ClockMan abandoned this revision.<br>
> ClockMan added a comment.<br>
><br>
> As a 'corporation' in which I work has doubts that checks developed by my after work, but tested on copyright protected code should be released to public under my 'name'. I will drop "push request", until everything will clarify.<br>
<br>
</span>I look forward to seeing a new patch at some point. I agree with<br>
Eugene that any future patch should be split out into distinct patches<br>
that focus on just one check at a time.<br>
<span class=""><br>
> As for author name in module: For me is a: must-have, as I plan to extend these checks and develop more.<br>
> As a author name in check name: I'm thinking about some "tags", so check could be registered under multiple names: clocky-object-copy-in-range-for and performance-object-copy-in-range-for.<br>
<br>
</span>I would be opposed to using an author name in anything that is<br>
user-facing. We group the checkers either by functionality<br>
("modernize-*" and "readability-*"), by organizational name<br>
("google-*" and "llvm-*"), or by publication name ("cert-* and<br>
"cppcoreguidelines-*"). We actively discourage use of author names<br>
even in our code comments.<br>
<span class="HOEnZb"><font color="#888888"><br>
~Aaron<br>
</font></span><div class="HOEnZb"><div class="h5"><br>
><br>
> Best regards,<br>
> Clocky<br>
><br>
><br>
> Repository:<br>
> rL LLVM<br>
><br>
> <a href="http://reviews.llvm.org/D13444" rel="noreferrer" target="_blank">http://reviews.llvm.org/D13444</a><br>
><br>
><br>
><br>
</div></div></blockquote></div><br><br clear="all"><div><br></div>-- <br><div class="gmail_signature">inż. Piotr Zegar<br><a href="mailto:me@piotrzegar.pl" target="_blank">me@piotrzegar.pl</a><br><a href="http://www.piotrzegar.pl" target="_blank">http://www.piotrzegar.pl</a></div>
</div>