Re: kern/49207

[christos%zoulas.com@localhost (Christos Zoulas)]

The following reply was made to PR kern/49207; it has been noted by GNATS.

From: christos%zoulas.com@localhost (Christos Zoulas)
To: "Kamil Rytarowski" <n54%gmx.com@localhost>, gnats-bugs%NetBSD.org@localhost
Cc: kern-bug-people%netbsd.org@localhost, gnats-admin%netbsd.org@localhost, 
	netbsd-bugs%netbsd.org@localhost
Subject: Re: kern/49207
Date: Sat, 27 Sep 2014 17:27:27 -0400

 On Sep 27, 10:55pm, n54%gmx.com@localhost ("Kamil Rytarowski") wrote:
 -- Subject: Re: kern/49207
 
 | Hello,
 | 
 | Quick questions to make it clear:
 | 1. Just to be sure. Is the extraction of these constants from clock_subr.h a good idea?
 
 I don't know; historically I've seen them defined in many places with
 different names. I guess it is an improvement to put them all in one place.
 
 | 2. After changing the macro names to longer forms. Is it a good idea to add compatibility defines in clock_subr.h? This way:
 | #define SECMIN SECONDS_PER_MINUTE
 
 I don't see why. It is not an API.
 
 | 3. Can I assume that with going back for the multiplication convention a user knows seconds per day (to be honest I never remembered it before :-) ). In other words, is the anti-16-bit-int campaign desired? Actually I believe that a good compiler will catch overflow immediately (clang, gcc).
 
 I personally would not worry about it. If you are using 16 bits for seconds,
 you should know you are going to get into trouble quickly :-)
 
 christos