I do appreciate the reply, but it's not my TAN nor has it been an issue
for me. It's just funny (peculiar) that a long standing
counter-intuitive behavior has only now been considered for change. As I
wrote - as if they had a meeting and decided the worst outcome (for the
user) was the one to use.
Other, similar concepts in peculiar interface design - placing OK and
Cancel/Exit/Quit buttons as neighbors on menus. One missed pixel
positioning the pointer, a click, and down the tubes with the effort. Or
simplifying the interface by adding more ways to perform an action -
depending on the path taken: e.g. create a datum plane from the icon and
you get the new method. Create a datum plane as part of a cross-section
and you get the old method. Or hiding the means to re-orient the axis
extension lines on a drawing view until the axis is selected and then a
right-mouse-button is available. Is Edit/Rotate counter-intuitive?
Since Rev 13 I've wanted to read the meeting notes where some of the
more interesting decisions were made.
I asked at the conference who was in charge of interface for
Pro/Engineer. The answer was that all the development heads were
responsible. As if the dozen (or so) of them, so involved in their own
product and keeping the underlying data straight, are also going to have
a 12 way discussion on usability.
Before the oft held gridiron pile-on - Yes Pro/E is great. It was the
best choice at the time (Rev <11?) and it is still a pretty good one.
Yes, it has greater breadth than competitors. Yes, no software is
perfect -blah, blah, blah on imperfect software. Maybe there's only the
12 guys developing all the software and their to-do list is pretty long
and this drawing view item just bubbled up after so many (more than 10)
years have gone by.
The concept that matters is someone got asked - what happens when the
references are incomplete - and the answer was to reorient the view and
hose view related items so that some user and drawing checker had all
that work to do again.
Dave S.
Bjarne Frandsen wrote:
>
> David,
> if you create your 3D views without using any feature references by
> just using the current orientation, they will never fail. This could
> be done just after leaving the sketcher and the model is in a desired
> orientation. From there you can use the dynamic orient to turn the
> model in precise angles around principal axis to get other saved views.
> Here we have a J-Link application that will create all standard views,
> just be changing the view matrices in the model from the current
> orientation. This gives us views without feature references and
> concrete stability.
> Bjarne
>
>
> *David Schenken <->*
>
> 14-12-2009 15:43
> Please respond to
> David Schenken <->
>
>
>
> To
> -
> cc
>
> Subject
> [proecad] - TAN147281 - Drawing View Orientation
>
>
>
>
>
>
>
>
>
> When references are lost for drawing view orientation, the view is
> re-oriented. Somehow this was intended, like there was a meeting and
> they said - "We can leave the view orientation alone when the
> references go, or we can just ruin the appearance of the drawing.
> Votes for "ruin"? Votes against" The ayes have it."
>
> The TAN concludes " PTC development is considering an enhancement ..."
>
> My vote is that the drawing orientation not change, and not lose all
> the dimensions. There can be a selection to freeze the orientation. If
> the orientation is re-established and the orientation part of the
> transformation matrix is very similar to the frozen one, then the
> dimensions depicted in that view and related views are retained at
> their prior locations and with their prior adjustments.
>
> The current way to avoid this problem is to use datum planes or other
> datum features as view orientation references, especially those
> established in the model before any other features. Unless they are
> deleted these datum references are unlikely to change and therefore
> unlikely to fail in orienting views.
>
