Mail Archives: geda-user/2014/08/13/09:18:41
On Aug 12, 2014, at 9:35 PM, Dave Curtis <davecurtis AT sonic DOT net> wrote:
> On 08/12/2014 04:19 PM, John Doty wrote:
>>
>> On Aug 12, 2014, at 4:57 PM, Dave Curtis <davecurtis AT sonic DOT net> wrote:
>>
>>> Because if "CR and LF have no effect", then why the admonition against CR/LF against places after a comma?
>>
>> Often, specs for readers say “be liberal in what you accept” while specs for writers say “be conservative in what you emit”. That seems to be the case here. Don’t make a big deal of it.
>>
>
> Yes, I totally understand that philosophy. If I was writing a reader, I would have no problem being liberal. If I was writing a writer, I would have no problem being conservative. But... I am writing a validation tool, which leaves zero margin for lack of precision.
Well, commonly you have “errors” and “warnings”. Issue an error for anything that a liberal reader cannot interpret, and a warning for anything that a conservative writer should not emit.
>
> I first did logic design on mainframe CPU's in 1981. I spent a large part of my life combing architecture documents for nits, making the lives of the spec authors quite miserable until they told me precisely what they meant. Logic design mistakes in CPUs can be expensive. At one point in my life I signed off on the Pentium processor errata sheet every month before it went out.....
>
>> John Doty Noqsi Aerospace, Ltd.
>> http://www.noqsi.com/
>> jpd AT noqsi DOT com
>>
>>
>>
>
>
>
John Doty Noqsi Aerospace, Ltd.
http://www.noqsi.com/
jpd AT noqsi DOT com
- Raw text -