I've never cared about this or seen a rule for it. You might benefit by posting a rules page or something, to show the "logic" as the foundry (?) sees it.
I think it will go best for you, if the scale factors
are "round number" multiples / divisors. Can't do
"fractional pixels" so you end up "notchy" in any
conversion / resizing, unless integer multiples of
initial DBU?…
Yesterday I encountered by chance. the new-to-me
Brand X "Selection Protection" feature.
I al,so have discovered that the old "inmstance box"
scheme I so love, has been broken for some time (was
a big thing back around 2010, …
Yes, this sounds like on-theme (how to untangle multiple selectables).
This particular scenario IMO wants the "pop up preselection list". We could talk about what that does or doesn't do. Could have "class" filters (instance, PC…
If this is about choosing -which one- of multiple selectable instances, then that's another face of the other "locking" thread.
Sometimes you can "riffle through the deck"
by continually clicking and watching the lower lwft cor…
While my preference remains for the "instance
selectable extent modifier rectangle" (Brand X
use of instance/drawing layer object to supersede
the raw cell-extent), I could see something like use
of GDS properties to make a per-instance…
I don't believe so, although this is something I've also asked for in the past. Personally I prefer a "selectable extent" polygon approach (ala Brand X's use of the instance/dwg layer, although this is becoming "busted" in some…
One thing that might be handy, presuming deep access to PDK, is to create for each "slottable" layer, a LayerN_slot counterpart. This would be directly checkable for size as plain rect or polygon (for my 45s). Tapeout would use a Boolean A…
Tutorial is a big help for git-clueless like me.
A semi-smart script which would determine the
target-host "all needed packages" and maybe also
inquire about the build desired particulars (like all
the associated options, are they desir…
I see similarities here, to what Brand X does - sorta.
Brand X does not verify directly against layout. There
is an "extract" process which creates a new cellview
("extracted") and creates a flat (at least, in the versions
of ve…
Might I suggest a "watchdog timer" that will "shake the Etch-a-Sketch" if it ends up "stuck"? That's an approach in systems for making layers of "self-righting" (this being the back-backstop).
Then that ev…
The other way I have seen (may have mentioned) is to proliferate in-layer "purposes" and enforce a discipline that can be "codified" in DRC and LVS.
For example Met3 (33,0) might be attended by M3VDDA, M3VDDA, M3PGND, M3HV... …
Very interesting. So now I must ask another question -
is it possible to use the same scheme (!blahblah append)
to somehow "tag" a device by its name (recent discussion)
or any concerns with it?
and another, can !blahblah and !anotherthin…
I think there's a lot of opportunity to add a "meta"
layer using the properties but that seems like it
needs a lot of thought and definition, and risks
to back-compatibility to lesser GDS editors might
be lost?
However if the meta "…
In my (industrial) experience the voltage rating
runs with the device design specifics and the
unique device master should point to the rule.
The "vanilla CMOS" flows that have come out
for OSEDA tend to not exercise this (even old
MOSIS…
Let me relate what I think I understand about how
Brand X does (or used to do) it.
Just as a schematic may have a symbol or "stop on"
an alt view (like veriloga or spectre), for verification
I see "auLVS" reps on occasion. Beli…
Let me pose a "pathfinder" question-bundle.
Suppose that I made a transistor layout of fixed
geometry. Further suppose that I wanted to place
a text in that layout, which would "reflect" a GDS
"property" added (say, c…
Re the cutter experiment that didn't work out - might need to look
at the logic of the "source" terminal identification, may be that
things still get "over-merged" by some other layers' meddling?
Like connect statement if the m…
There should be prune rules somewhere? But sounds like
you want them applied "surgically"?
Perhaps the rule that does the pruning, could be modified
or enhanced to make a net that contains a pin-polygon (I
guess we don't do that over her…
Yes, I think we understand it the same. I'm thinking that smart
folks could come up with a simple architecture for using
properties (how many uses, really?), set a standard (API?)
and then enable more sophisticated yet less difficult checking
a…
This might be a use for a "cutter" feature, a path or polygon
that would ANDNOT with things coincident for extract logic
purposes only? Like for occasions when "connect" logic
connects things you didn't mean to and breaks device…
If properties are a thing, albeit unused, I'd like to see some
thought given to how they could be. Perhaps a few properties
like 'instanceName' would be big levers. I certainly do like
using instance names as first guess for correspondence.
And i…
You might make this easier by "baking in" features to the pad
layout cells or PCells. A placed-by-construction text is easy to
grab origin, of. If you made the text value inherited (is this
possible?) you might even get padName tied to P…
I've seen some tools which have select-direction modality
and I try to turn it off, first thing. Same with "strokes". Too
much conflict with "muscle memory" from simpler times.
Now I would be in favor of more modal options in s…
I don't know nuthin' from scripting but are you sure that
instances and layers are "valid"/ selectable, either as a
default on startup / layer-gen-from-readin or asserted?
Is "one" object selectable under same "preamble&qu…
I think I'd begin (presuming this is an ASCII techfile)
with finding and inspecting the flagged line for odd
qualities that the import might choke on.
For example Brand X is prone to be verbose and long-
winded, and the line might exceed a baked i…
While I'm wishing for installation ease, how about a script or
even a simple list of all dependencies? I've installed klayout
to fresh Linux installs a dozen or more times and always
end up chasing packages, fail-message after fail-message
until d…
I looked at an openaccess pdk and it sure seemed
full of proprietary data format stuff to me. Not sure
how "open" it really is. Other than branding.
Streamin to a Cadence database format or an "open"
format has nothing to do …