X-Authentication-Warning:	delorie.com: mail set sender to geda-user-bounces using -f
X-Recipient:	geda-user AT delorie DOT com
X-Envelope-From:	paubert AT iram DOT es
Date:	Fri, 6 Jul 2012 11:06:36 +0200
From:	Gabriel Paubert <paubert AT iram DOT es>
To:	geda-user AT delorie DOT com
Subject:	Re: [RFC, v2] [geda-user] [PATCH] Allow to create metric Gerber
	and drill files.
Message-ID:	<20120706090636.GA22475@visitor2.iram.es>
References:	<20120703140236 DOT GA12646 AT visitor2 DOT iram DOT es>
	<CAKakQcdTgG6E2h+UgWoh-ujP6vFDH=voY-PBJR3OFG2awqL0_A AT mail DOT gmail DOT com>
	<20120705101614 DOT GA19974 AT visitor2 DOT iram DOT es>
	<20120705173423 DOT GC12804 AT malakian DOT lan>
MIME-Version:	1.0
In-Reply-To:	<20120705173423.GC12804@malakian.lan>
User-Agent:	Mutt/1.5.20 (2009-06-14)
X-SPF-Received:	2
X-Spamina-Bogosity:	Unsure
X-Spam-Score:	-4.4 (----)
X-Spam-Report:	Content analysis details: (-4.4 points)
	pts rule name description
	---- ---------------------- --------------------------------------------------
	-1.8 ALL_TRUSTED Passed through trusted hosts only via SMTP
	-2.6 BAYES_00 BODY: Bayesian spam probability is 0 to 1%
	[score: 0.0000]
Reply-To:	geda-user AT delorie DOT com
Errors-To:	nobody AT delorie DOT com
X-Mailing-List:	geda-user AT delorie DOT com
X-Unsubscribes-To:	listserv AT delorie DOT com

On Thu, Jul 05, 2012 at 10:34:23AM -0700, Andrew Poelstra wrote:
> On Thu, Jul 05, 2012 at 12:16:14PM +0200, Gabriel Paubert wrote:
> > 
> > I have appended a second version of the patch; the only difference 
> > with the first one is that now the line in the header that starts 
> > with "G04 PCB-dimensions" now specifies the dimensions either in
> > metric or in imperial dimensions depending on the metric option.
> > 
> > I was expecting that the developers with git write access
> > would be a bit more reactive and tell me whether something
> > looked wrong with the patch. I must admit I'm not completely
> > satisfied with the names of some functions/variables, but I'm 
> > not too worried since they are local to the gerber.c file. For
> > global names, I'm much fussier.
> >
> 
> The patch is as well-done as it could be. I considered doing
> it a while back, but was turned off by all the if-statements
> necessary. Thanks for finally biting the bullet!
> 
> Unfortunately, I don't recall enough about the Gerber spec
> to review your patch, nor to address your other concerns. I
> might push it through when I get a free moment anyway, assuming
> it doesn't break the old behavior.
> 
> > From my limited tests, the patch does not affect the Gerber
> > code in imperial mode (diff output only lists the obvious:
> > timestamp in the header, and the board dimensions in the
> > second version of the patch, in any case only the comments
> > are affected).
> >
> 
> Does tests/run_tests.sh still pass?

It seems to, I have tested on two machines, which give the same 
results:the first 4 tests fail (the bom ones), but I don't see 
how they could even be remotely related to the patch. 

Note that with my current patch metric is on by default for Gerber
generation, so this means that the files compare equal after 
going through gerbv as a png exporter and comparing through.

I realise that setting metric by default may break existing 
scripts, so maybe we should stay with imperial Gerber and
drill files by default. Right now I have imperial in the
golden subdirectory and metric in the outputs. They do not 
compare at all equal from diff -r, but they do after conversion
to png. I believe that this is good enough for inclusion.

[Making a few tests]
Even if I stay with imperial, the differences between golden
and outputs are large: aperture definitions are not the same, etc...
If there is a bug, it is well hidden, except perhaps for octogonal
apertures, which I don't use and don't seem to be covered by the 
test suite.

	Gabriel

- Raw text -