kicad-developers team mailing list archive

[Jon Evans <jon@xxxxxxxxxxxxx>]

Yeah, being able to ignore/accept ERC/DRC warnings is another feature that
probably is orthogonal to real-time operation.

What I was thinking is that there are some types of ERC errors that are
more valuable than others to run in real-time
A few examples:
- Duplicate refdes
- Conflicting labels on a net or bus (new check in my branch)
- Mismatch between hierarchical sheet and ports (a new check I will be
adding)

Some other errors I think would be annoying to run in real time but better
to run in batch, such as:
- Unconnected pins
- Pin type conflict (i.e. the matrix of input/ouput/etc pin types)

Those "online" errors could even be shown in a less obtrusive way than the
ERC markers that are created during batch ERC.
I will play with some options for this.

-Jon

On Mon, Mar 25, 2019 at 12:28 PM Christopher Buckley <
mallardproductions@xxxxxxxxx> wrote:

> A few ideas:
>
> Option 1 an error type selection menu, possibly with levels so that an
> error can be flagged / marked / "ignored" in successive automatic/manual
> rule checks until manually resolved.
>
> Option 2 might be classifying errors by significance (ie connecting a wire
> from +5 to gnd:== critical (duh-op!), open pin == notice, backwards
> transistor or electrolytic == important) (too much work!)
>
> Option 3 might be a setting to "flag" all potential errors (ie post it
> note), but ignore warnings until a specific rule check is performed
> manually.
>
> Option 4, manually mark the "flag" or "net point" as "accepted" so it
> never shows up again (possibly dangerous ;)
>
>
>
> --------------------------------------------
> On Mon, 3/25/19, Jon Evans <jon@xxxxxxxxxxxxx> wrote:
>
>  Subject: Re: [Kicad-developers] [RFC] online (real-time) ERC/DRC
>  To: "Wayne Stambaugh" <stambaughw@xxxxxxxxx>
>  Cc: "KiCad Developers" <kicad-developers@xxxxxxxxxxxxxxxxxxx>
>  Received: Monday, March 25, 2019, 9:59 AM
>
>  Understood. I will
>  work on a user experience proposal on the schematic side,
>  and I agree it makes sense to wait on the PCB side until we
>  are in a better position with DRC code (which I guess might
>  change dramatically as part of V6 development)
>  On Mon, Mar 25,
>  2019, 11:52 Wayne Stambaugh <stambaughw@xxxxxxxxx>
>  wrote:
>  Jon,
>
>
>
>  I would be open to a real time error ERC checking as long as
>  it met the
>
>  following criteria:
>
>
>
>  It can be disabled by the user.
>
>
>
>  A sane way to determine when it makes sense to actually
>  start and stop
>
>  the real time ERC testing can be designed.  I don't
>  want a bunch of ERC
>
>  indicators every time I add a couple of new symbols to a
>  design after I
>
>  have run the ERC and fixed the initial issues.  This would
>  get annoying
>
>  pretty quickly.
>
>
>
>  On 3/23/2019 5:57 PM, Jon Evans wrote:
>
>  > Hi all,
>
>  >
>
>  > I would like to start laying the groundwork for online
>  ERC/DRC.
>
>  >
>
>  > This means, for example:
>
>  > - Objects will know if they have any ERC/DRC errors
>
>  > - If objects are edited, associated error markers can
>  be cleared/updated
>
>  >
>
>  > Eventually, we could add a (optional) mode where new
>  ERC/DRC errors can
>
>  > be added in real-time as the user edits.  To begin,
>  I'd only be clearing
>
>  > or updating existing errors (generated with the offline
>  / batch ERC/DRC
>
>  > process we have now).
>
>  >
>
>  > For the schematic, I think all ERC checking will able
>  to be done in
>
>  > real-time once my new connectivity stuff is merged.
>  For the PCB, of
>
>  > course there are some DRC checks that are way too slow
>  to be done in
>
>  > real-time, but we could explore whether some checks are
>  fast enough.
>
>
>
>  Real time DRC testing is a different conversation.  From
>  what I've
>
>  observed with some of the more complex designs I've
>  tested, I don't
>
>  think we are anywhere near ready for this conversion without
>  making a
>
>  lot of performance improvements to our DRC code.
>
>
>
>  Wayne
>
>
>
>  >
>
>  > I would start with the schematic ERC since that's
>  where my head is at
>
>  > right now, but what I do should be able to carry over
>  to PcbNew in case
>
>  > it ends up working well.
>
>  >
>
>  > Any thoughts on this?
>
>  >
>
>  > -Jon
>
>  >
>
>  > _______________________________________________
>
>  > 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
>
>  _______________________________________________
>  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
>
>  -----Inline Attachment Follows-----
>
>
>