[PEDA] Octagonal Pads

[Geoff Harland]

Dennis Saputelli wrote:
<snip>
> now if the purpose of an octagonal pad is to make room for a trace
> turning around it then altium (and i am not trying to be an apologist
> here) have solved that (AT LAST!) with the new rounded rectangle with
> defined corner radii, almost as good as the old razor blade on the light
> table to shave the edge of a pad to make a bit of clearance, but just
> today i wished i could have square corners at the bottom and radiused at
> the top without resorting to piled on primitives

I don't have a copy of AD6 (and thus can't confirm it for myself), but I
would have thought that you could select the 'Top-Middle-Bottom' value for a
pad's 'Padstack Mode' property, then select the 'Square' value for the Shape
property on the Middle and Bottom layers, and the 'Rounded Rectangular'
value instead on the Top layer.

Or have I "grabbed the wrong end of the stick" in this instance, and you
actually want just two of the four corners to be rounded instead? (And
perhaps on just one of the layers, such as the Top (copper) Layer.) That
would indeed require more effort to implement, and I would appreciate you
not wanting to do that unless there was a truly compelling reason for doing
so.

*If* AD (eventually) acquires all of the functionality which is currently
provided by P-CAD (2006), then you would in due course be able to define a
shape of your own choice on each layer, which would better cater for such
situations. (There is no guarantee that that will ever happen, but *unless*
that happens, P-CAD users will in due course lose some of the functionality
which is currently provided to them, as it is public knowledge that P-CAD
2006 will be the last major release of that application.)

> so i guess i don't really care if they have been remiss in fixing this
> problem and i am sure you are correct that it is fixable, i just think
> there are much bigger fish to fry

I would really be interested in hearing about which issues you think should
be rectified on an ASAP basis. (One area which I regard as needing a lot of
improvement involves DRC procedures, and especially various aspects
involving Internal Plane layers, non-"Simple" pads, and unplated pads.)

On the matter of octagonal pads though, you are definitely entitled to your
own POV, and I'm certainly not going to condemn you for it. And there are
doubtless many issues (perhaps even most issues) on which different users
would have different views (as far as the urgency in getting them rectified
is concerned). My POV on this particular issue though is that it involves
functionality of a core nature, and it can "catch out" users who are not in
the know (as indeed Steve Hendrix has reported), and that users should not
have to resort to the workaround of generating ODB++ files and then
exporting Gerber files from the CAMtastic file displaying those ODB++ files.

I regard the existance of that workaround as a "mildly" mitigating factor as
far as the urgency of rectifying the contents of "directly" generated Gerber
files is concerned, but it still doesn't change the fact that it is not
inherently obvious that "directly" generated Gerber files can be
problematic.

To deal with that aspect of the staus quo though, Altium's programmers could
write some code which would display a dialog box whenever users attempted to
"directly" generate Gerber files (and the PCB file concerned contained one
or more pads having an octagonal shape). That dialog box could alert users
that there would be issues with "directly" generated Gerber files, while
also notifying them of the workaround available, and also "polling" them as
to whether they still want to continue (with generating those files
"directly"). And the provision of that code would require less effort than
the effort required to rectify "directly" generated Gerber files *and* the
other simultaneous changes which I would regard as highly advisable (and
which I described in my previous message).

OTOH, if this issue was to be rectified eventually (and all of the other
changes recommended were either already implemented before then, or
otherwise on a simultaneous basis), then some programming time could be
saved by making all of those changes on an ASAP basis instead (and thus
avoiding the need to write any code for the dialog box just described). And
making those changes would "present" AD in a more professional manner, from
the POV of users, than the invocation of a dialog box advising them that
they would need to resort to a workaround to generate truly satisfactory
Gerber files.

OTGH (On The Gripping Hand, from "motie" aliens who have three arms, as
described within Niven and Pournelle's SF novel "The Mote in God's Eye"), it
would not take very much effort to write the source code for the dialog box
concerned, so the provision of that dialog box "for the time being" would at
least alert users that there is an issue in this regard, and thus
substantially reduce the liklihood that any PCBs which they ordered would be
"mis-manufactured". I would still regard it as highly preferable for this
issue to be rectified ASAP, but the provision of such a dialog box would at
least address its "gotcha" aspect, and also advise users of the available
workaround.

> i don't care a whit about gaining or losing altium's favor as i am a
> paying customer and as far as i am concerned they are working for me

I sometimes wonder whether Altium's management are working for anyone other
than themselves, as some of their actions could be regarded as being hostile
to the interests of their customers, their shareholders, and their
employees. If I was to take a charitable attitude though, they probably
really don't appreciate just how and why they have let all of those
stakeholders down on various occasions.

> and yes they should have fixed this octagonal pad issue or killed it
> long ago, i don't think it is even offered as a possibility anymore, is it
?

If support for pads having an octagonal shape was ever withdrawn, it
probably wouldn't be regarded as a loss by anyone who has never used such
pads. OTOH, people who have used such pads (and I have, on some occasions
(in conjunction with some extra steps and precautions)) could be expected to
have a very different view though.

I am definitely not hostile to the provision of the relatively new "Rounded
Rectangular" shape for pads, but what I still do find objectionable is
Altium's willingness to provide new features or functionality while failing
to rectify serious defects associated with existing features and
functionality. And in a number of cases, new features have resulted in
regression in previously provided features, and typically because the new
features have not been properly thought through. And for the same reason,
and/or manifestly inadequate testing, newly provided features have
themselves often left a lot to be desired as well (and regardless of/"over
and above" their impact upon previously provided features).

While making changes to how new functionality and features are developed and
implemented could potentially result in them being provided at a slower rate
than has been the case to date, I still think that there are far too many
cases of defects which should never have been "shipped" to users in the
first instance, and which have all too frequently still not been rectified
in a timely manner after users subsequently discover them (and sometimes
only after being "bitten" by such defects). And while there are probably a
number of users who wouldn't regard the issue involving pads with an
octagonal shape as being a "top priority" issue, and another complicating
factor is that it would be highly desirable (if not essential) to make some
other changes at the same time (or else beforehand) as well, I am still of
the view that this particular issue should still be rectified ASAP. Failing
that, the dialog box which I described previously should be implemented on
an ASAP basis instead, in order to at least mitigate the impact of that
defect.

Regards,
Geoff Harland.