On 7/20/12 11:42 PM, sebb wrote: > On 20 July 2012 22:22, Dennis E. Hamilton <[email protected]> wrote: >> I've seen the comment about SCMs being sufficient for tracing the provenance >> of code and the changes that are made. That puzzles me. >> >> - History doesn't appear in source-code tarballs. >> - It requires the original SCM repository or a history-preserving port of >> the SCN to be available to interested parties. >> >> Basically, it is not a durable form of the information. > > Indeed, especially since SVN log messages are not versioned. > > I think log messages should only be used to inform the reader of the > commit message why the commit was done. > They should not be used for comments that are useful / necessary to > the reader of the code; those should be included as comments in the > code itself. > [Though of course such comments can go in the log message as well.]
Is it possible that you mix something here. We don't talk about useful comments that give some further information to specific code section or explain functions etc. Juergen > > However, does the end user of the source need to know provenance and history? > >> Just sayin' ... > > DItto. > >> - Dennis >> >> -----Original Message----- >> From: Pedro Giffuni [mailto:[email protected]] >> Sent: Friday, July 20, 2012 12:52 >> To: [email protected] >> Subject: Re: Coding guideline or common rules >> >> FWIW; >> >> >> ----- Original Message ----- >> ... >>> >>> Hi, >>> >>> I just stumbled over a commit message for the new UOF filter. >>> >>> I think we should agree on a common guideline for our code and how we >>> contribute changes and bring them in the code. >>> >>> SCM's manage the change sets and the information who made the change, >>> that means we don't need further comments like this >>> >>> ///Begin Added by wangyumin for uof2-filter from cs2c >>> ... >>> /// End Added by wangyumin on 2012-2-22 14:32:18 >>> >>> It is somewhat redundant and makes the code not really better readable. >>> Can we agree on the common understanding that we don't need such >>> comments and that we don't want them in the code. We should remove such >>> comments wherever we see or find them. >>> >> >> Indeed, I did mention in our local svn tutorial that those comments should be >> avoided. SVN does a wonderful job maintaining the origin information. >> >>> Any opinions? >> >> >> As a side note, I recently found similar prominent begin/end lines in another >> project and the culprit on that project was the GPLv2 section 2a: >> >> "You must cause the modified files to carry prominent notices stating >> that you changed the files and the date of any change." >> >> >> It's probable that old code from GPLd derivatives still carry such notes. >> >> Someone will have to clean them ;). >> >> Pedro. >>
