At 10:27 AM 7/13/01 +0100, Steve Wiseman wrote:
[some response which was rooted in a misunderstanding of the full meanings 
of my statements. It appears that we both agree on how the program actually 
works and any differences were only semantic. It is moot who said it the 
most correctly.]

If the Advanced Tab is used to annotate ? parts with a suffix, it will 
duplicate already-existing designators within the annotation range. So if 
R1A already exists, and new R parts are added as R?, and then 
Annotate/Advance is run to annotate all ? parts, and the range is set as 
1-n with suffix A, another R1A will be assigned, so that there are now two. 
This is a nuisance if one wants to keep existing suffixed reference 
designators, which would be quite common on revisions, actually, in my 
experience, it is the norm.

If one does not use a suffix, then Annotate will not duplicate designators, 
so, in the case given, only with no suffix on Annotate, one will end up 
with an R1 and an R1A. Annotate is "getting it right" with respect to its 
instructions, but this is likewise not satisfactory, since the A suffix is 
being used to indicate that these parts are in the A section of a 
multi-section design.

I ended up agreeing that this was a bug, of sorts; the nature of the bug is 
that checking for existing designators was omitted from the process of 
Advanced Annotation. It should not duplicate designators, period. We do not 
use an Annotate tool in order to get duplicate designators. It appears that 
the programmer only considered the possibility that we were completely 
annotating a section for the first time, i.e., annotate was set to "all 
parts," in which case it would overwrite existing designators.

[and I wrote:]
> > Problem is that the existing behavior is not good enough to deal properly
> > with the subject situation. Yes, you can reset designators and that will
> > solve one problem, but, as in the present situation, it will create
> > another, the lack of correspondence of designators between the old
> > schematic and the new one. This can make quite a mess.

[and Mr. Wiseman responded:]

>I've never experienced this, and I've added components to many schematics
>without forcing a renumbering. Protel gets this right, unless you're using
>suffixes, which I'd not tried before.

Yes, if you do not use suffixes, Protel handles the annotation correctly. 
However, that is not what I was describing. I was describing the use of 
Reset Designators, which, by definition, forces a renumbering as if none 
had existed.

I am not confident that the Synchronizer will properly handle this. It 
might, but I'd want to make sure that everything was solidly backed up in a 
way that it could be recovered even if I thought everything was running 
fine. The old Netlist Load procedure would be completely bollixed, as would 
have been just about every CAD system out there. PCB designer's least 
favorite message from engineer: "I made some changes to the schematic and 
so that everything will be nicely numbered now, I re-annotated. Please add 
the new parts and make the changes on the PCB."

I had this happen to me with a 20-page schematic done in Tango. I managed 
to recover by writing a utility that analysed the symbol positions on the 
schematic pages and made a correspondence table, then I used another PCB 
was-is utility to renumber the PCB to the new numbers. But it was at least 
a day's work.

> > The synchronizer might handle it properly, but it would not fix any
> > accessory documentation, like the note that says "Do not stuff C101A."
>That information I keep in a separate file as a kindness to my
>assemblers. Since there's no C101A in the BOM or on the board, it's not
>terribly important information, it just helps reduce confusion.

Not in this case. I'm writing about the general case. There may or be such 
accessory documentation, either as notes on an assembly drawing or 
elsewhere. The designer might not even be aware of such notes if there is 
anyone else involved, such as an engineer checking the project or a 
purchasing agent. When I am working on a client project, a lot of 
information may be passed back and forth with specific instructions about 
"keep U6 close to U17," for example. If the design is renumbered, all those 
notes become next to meaningless. This is why I'm not a big fan of 
renumbering, but that is really a separate question. We want to be able to 
renumber, and we want to be able to intelligently add new designators 
without renumbering the existing ones.

> > And here I run into a brick wall. I have a vague memory that we have
> > discussed in the past a shortcoming of the global edit Copy Attribute
> > operation. I could find no way to use it to add a suffix. Perhaps someone
> > knows how to do this!
>Hmm. Not me...
>Is this the kind of thing I ought to be using macros / hand-crafted
>servers for? (obviously not for this particular problem, which can be
>solved elsewhere, but it's a nice, small, clearly defined problem...)

By the kindness of Mr. Williams and Qualecad, it appears we have such a server.

Abdulrahman Lomax
P.O. Box 690
El Verano, CA 95433

* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* To post a message: mailto:[EMAIL PROTECTED]
* To leave this list visit:
*                      - or email -
* mailto:[EMAIL PROTECTED]?body=leave%20proteledaforum
* Contact the list manager:
* Browse or Search previous postings:
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

Reply via email to