Status:  U
From: gene heskett <gheskett@wdtv.com>
To: coco@maltedmedia.com
Date: Wed, 25 May 2011 20:22:15 -0400
User-Agent: KMail/1.13.7 (Linux/2.6.38.7-pclos1.pae.bfs; KDE/4.6.3; i686; ; )
References: <20110524212513.GA6439@brawl.fslf.org>
	<9B6B5EC6-BC3F-4088-B92E-7A47ABA349BA@ocs.net>
	<20110525230414.GA14913@brawl.fslf.org>
In-Reply-To: <20110525230414.GA14913@brawl.fslf.org>
MIME-Version: 1.0
Message-Id: <201105252022.15919.gheskett@wdtv.com>
Subject: Re: [Coco] cprep19 __FILE__ fix?
Precedence: list
Reply-To: CoCoList for Color Computer Enthusiasts <coco@maltedmedia.com>
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Sender: coco-bounces@maltedmedia.com
Errors-To: coco-bounces@maltedmedia.com

On Wednesday, May 25, 2011 08:15:20 PM Willard Goosey did opine:

> On Wed, May 25, 2011 at 02:49:23PM -0700, Michael Furman wrote:
> > Willard, I looked into this a few months ago.  Off the top of my
> > haead:
> > 
> > 1) I have some source for the Krieder C library, but when I
> > 1) assembled it, it was not exactly the same as binary copies
> > 1) floating around.  This reduces my confidence that the resulting
> > 1) clib after hacking will actually work properly, and makes it
> > 1) difficult to generate a patch that will always work.  The date
> > 1) command returns the correct year.
> 
> Bleah.  I'm still trying to debug curses, I'm not about to start
> messing around with clib.
> 
> > 3) I took a look at Nitros9's date command and it has some Y2Kish
> > 3) stuff there but it's not clear if it works correctly.  The
> > 3) algorithm there seemed a bit convoluted or worked the problem
> > 3) backwards (from my perspective) for some unknown reason (to
> > 3) optimize it? for 8-bit math apparently?).  The comment in
> > 3) date.asm is from Grandpa Gene:
> > *   5      1996/09/25  Gene Heskett
> > * Made Y2K compliant.
> 
> I believe there was some trouble about making sure <2000 dates in file
> timestamps and things like that didn't break, but I don't really
> remember the discussion, I just used the results.
> 
> Gene?  Any comments?
> 
ISTR I worked on the date command to add an offset of 100 if it was before 
1980 or so.  It WILL eventually break, but likely long after they boil me 
done & burn the remains.

> > 4) It appears that Nitros9 keeps a 1-byte year,

This has been true since os9 version 1.00.  Every byte of the DP0 page was 
allocated from the gitgo, so there is no room to expand.

> > so this is what one
> > 4) would expect F$Time to return.  According to OS9Defs:
> > D.Time         equ       .                   System Time
> > D.Year         rmb       1

And ISTR the clock module constrains that byte to a $00 to $63 value.
 
> Heh, gonna be trouble in 2157, or maybe even 2028 if something tries
> to look at that as a signed number.

Either way there won't be a lot I can do about it by then...
 
> > I'll try to fill in the rest of details from my previous
> > investigation of this issue later this evening.
> 
> Good luck!
> 
> > The result was that to fix UUCPBB (Which reported the year as 19111
> > in some places and 111 in others) I basically changed any occurrence
> > of '19%d', year to '%d', 1900+year.  This was a hack that doesn't
> > address the real lower level problem.
> 
> Heh, yup.  That's the plan for cprep19. ;-)  Since the date string is
> of fixed length, I know which characters to overwrite.
> 
> Willard


-- 
Cheers, Gene
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
<http://tinyurl.com/ddg5bz>
<http://www.cantrip.org/gatto.html>
H. L. Mencken's Law:
	Those who can -- do.
	Those who can't -- teach.

Martin's Extension:
	Those who cannot teach -- administrate.

--
Coco mailing list
Coco@maltedmedia.com
http://five.pairlist.net/mailman/listinfo/coco