cvs commit: ports/chinese/tin Makefile ports/chinese/tin/files
patch-attrib.c patch-cook.c patch-init.c patch-mail.c
sahil at tandon.net
Wed Jan 4 16:35:13 UTC 2012
On Jan 4, 2012, at 6:02 AM, Alexey Dokuchaev <danfe at FreeBSD.org> wrote:
> On Tue, Jan 03, 2012 at 10:31:37PM -0500, Sahil Tandon wrote:
>> On Tue, 2012-01-03 at 15:54:38 +0000, Alexey Dokuchaev wrote:
>>>> Revision Changes Path
>>>> 1.1 +11 -0 ports/chinese/tin/files/patch-tin.defaults (new)
>>>> 1.2 +0 -11 ports/chinese/tin/files/patch-tin_defaults (dead)
>>> This renaming of the patch file just because someone thought that dot
>>> looks better than underscore is gratuitous and should not have been
>>> done. (Not to mention that things that this patch tries to accomplish
>>> are usually done with one simple REINPLACE_CMD line.)
>> I suspect the motivation for the change was to canonicalize the filename
>> as described in the PH.
> I understand the motivation; for newly created files, PH rules obviously
> should be followed. For already existing patches, esp. when their contents
> stays the same, such blunt renames only cause unnecessary stress on the repo
> and taint the history. Ergo, should be avoided.
Sigh, please spare me the same old lecture; your logic is simple and no one is questioning that repo churn should be avoided. In your original mail you made reference to someone who thought one character "looking better" than another was perhaps the motivation for the name change. I pointed out a more likely rationale and ultimately agreed that such things are anyway best handled in Makefile rather than standalone patches. I also used an emoticon to convey that I was simply guessing the likely motivation for name change and not in some way defending it; however, you explicitly removed that and other relevant parts of my reply in your quoting above.
Over and out,
More information about the cvs-all