This is the mail archive of the
mailing list for the GCC project.
Re: Add checks for tree.value, tree.minval, and tree.maxval and others
- From: kenner at vlsi1 dot ultra dot nyu dot edu (Richard Kenner)
- To: rth at redhat dot com
- Cc: gcc-patches at gcc dot gnu dot org
- Date: Thu, 18 Mar 04 20:22:54 EST
- Subject: Re: Add checks for tree.value, tree.minval, and tree.maxval and others
> Of course. But I always list files just to be safe.
Well, clearly it isn't safe.
I think it far better to check in too *little* than two much. I'll never
feel confident, whatever the procedure, in just blindy checking in all
modified files. It's just my conservatism. It's too easy to edit a file and
not notice it. Yes, that's a violation of the testing protocol, since the
patch is supposed to be done as the only change to the tree, but that error
will almost always have much smaller effect than the opposite error:
inadvertly checking in that edited file. Errors when too little is checked
in are easy to fix and will be noticed on my next diff anyway.
"cvs commit" to check it all in. And you won't miss anything.
On active trees, that can often take so long that somebody else has done a
checkin and my files are out of date. Even aside from the issue above, I
find it more reliable to list all the files. For one thing, that way I can be
sure that ChangeLog agrees with my checkin.