kicad-developers team mailing list archive

[Seth Hillbrand <seth.hillbrand@xxxxxxxxx>]

I agree with your point that gerber viewers act as an important check.
Toward that end displaying a warning message should alert the user that
they have a problematic Gerber file that shouldn't go to a manufacturer.

Fixing sounds dangerous to me as it modifies a file that explicitly doesn't
follow a standard.  The resulting 'fixed' file will be standards-compliant
but could very easily be not what the user intended.

Viewing non-standard Gerber files would be personally useful to me as I
often receive gerber files from other engineers at different institutions
working on a wide range of EDA tools.  Some of them generate non-standard
gerbers and their users have no interest in switching their workflows.  At
the moment, I keep 4 different viewers installed, two on a virtual machine
to ensure that I can look at the files.  I would really like to expand the
viewing of non-standard files in Kicad.  I'm happy to submit this patch if
we're open to the idea.

-S

On Thu, Sep 28, 2017 at 10:53 AM, José Ignacio <jose.cyborg@xxxxxxxxx>
wrote:

> I don't know. If anything it would be the most useful to be able to try to
> repair broken files like that (maybe a script?). Displaying broken files
> "correctly" is dangerous. One of the main uses for a Gerber viewer is to do
> a pre-manufacturing check, and if your gerbers are broken and they work in
> the viewer anyway it could be a problem.
>
> On Thu, Sep 28, 2017 at 12:47 PM, Seth Hillbrand <seth.hillbrand@xxxxxxxxx
> > wrote:
>
>> Looking at gerbv right now, it appears to silently handle decimal places
>> if they exist.  However, in the absence of an explicit decimal place, it
>> treats %FSD as %FSL, which is probably why Clemens' file was correctly
>> displayed, as opposed to being oversized by a factor of 100.
>>
>> Personally, I would love to see Kicad following a robustness principle
>> that allows more files to be displayed but with a definite warning message
>> detailing the formatting error and cautioning that the file _may_ not be
>> correctly displayed because of the bad format.
>>
>> Best-
>> Seth
>>
>>
>>
>> On Thu, Sep 28, 2017 at 9:17 AM, jp charras <jp.charras@xxxxxxxxxx>
>> wrote:
>>
>>> Le 28/09/2017 à 17:58, Jon Evans a écrit :
>>> > Perhaps another route is to improve the messaging given to the user in
>>> these cases, so that it's
>>> > easy for them to correct the file / report an issue to their tool
>>> vendor?
>>>
>>> Yes.
>>>
>>> In fact, %FSD is already supported by Gerbview because (a long time ago)
>>> I found Gerber files in
>>> decimal format (not documented, because %FSD was never a official Gerber
>>> format statement).
>>>
>>> This is the reason no error was reported: coordinates were read as
>>> floating numbers (in mm) and valid.
>>>
>>>
>>> >
>>> > On Thu, Sep 28, 2017 at 11:53 AM, Wayne Stambaugh <
>>> stambaughw@xxxxxxxxx
>>> > <mailto:stambaughw@xxxxxxxxx>> wrote:
>>> >
>>> >     On 9/28/2017 10:32 AM, jp charras wrote:
>>> >     > Le 28/09/2017 à 16:13, Wayne Stambaugh a écrit :
>>> >     >> On 9/28/2017 9:45 AM, jp charras wrote:
>>> >     >>> Le 28/09/2017 à 01:27, Clemens Koller a écrit :
>>> >     >>>>
>>> >     >>>> On 2017-09-26 13:38, jp charras wrote:
>>> >     >>>>> The Gerber file is broken:
>>> >     >>>>> the line:
>>> >     >>>>> %FSDAX33Y33*%
>>> >     >>>>>
>>> >     >>>>> is incorrect
>>> >     >>>>
>>> >     >>>> Thank you!
>>> >     >>>>
>>> >     >>>> Since I cannot do anything about this proprietary non
>>> compliant EDA tool, would it be
>>> >     possible to support these wrong but obvious lines anyway (maybe
>>> after showing a warning) - so
>>> >     would you accept a patch to support the %FSD gerber code?
>>> >     >>>>
>>> >     >>>> Regards,
>>> >     >>>>
>>> >     >>>> Clemens
>>> >     >>>>
>>> >     >>>>
>>> >     >>>
>>> >     >>> A patch is possible, but the actual issue is:
>>> >     >>> What is the meaning of %FSD format?
>>> >     >>>
>>> >     >>> I saw some "Gerber" files using %FSD for a decimal format
>>> (coordinates in floating point
>>> >     notation),
>>> >     >>> that differs from your Gerber file ( that is in fact a %FSLA
>>> format, nothing else ).
>>> >     >>>
>>> >     >>
>>> >     >> Unless %FSD is an obsolete gerber command, I'm opposed to this
>>> idea on
>>> >     >> principle alone.  KiCad should not be in the business of
>>> supporting
>>> >     >> broken file formats created by other tools.  The gerber file
>>> format is a
>>> >     >> published standard and we should be following it as closely as
>>> possible.
>>> >     >>  You should file a bug report with the vendor of the program
>>> that
>>> >     >> created these gerber files.
>>> >     >>
>>> >     >> Cheers,
>>> >     >>
>>> >     >> Wayne
>>> >     >
>>> >     > In latest Gerber doc, %FSD appears in "Errors and Bad Practices"
>>> list and is clearly called
>>> >     Invalid
>>> >     > Format Statement in the "Error" section.
>>> >
>>> >     In this case we should not support %FSD.
>>> >
>>> >     >
>>> >     > only %FSLA and %FSTA exit.
>>> >     > %FSTA is now on the deprecated list (Kicad uses the %FSLA
>>> option).
>>> >     >
>>> >     >
>>> >
>>> >     We will have to continue to support these for legacy gerber files.
>>>
>>>
>>> --
>>> Jean-Pierre CHARRAS
>>>
>>> _______________________________________________
>>> Mailing list: https://launchpad.net/~kicad-developers
>>> Post to     : kicad-developers@xxxxxxxxxxxxxxxxxxx
>>> Unsubscribe : https://launchpad.net/~kicad-developers
>>> More help   : https://help.launchpad.net/ListHelp
>>>
>>
>>
>> _______________________________________________
>> Mailing list: https://launchpad.net/~kicad-developers
>> Post to     : kicad-developers@xxxxxxxxxxxxxxxxxxx
>> Unsubscribe : https://launchpad.net/~kicad-developers
>> More help   : https://help.launchpad.net/ListHelp
>>
>>
>