SW2007 For what it is worth

I went to the 2007 Rollout yesterday and am really impressed. I
haven't used the software yet but from what I saw it looks like SW has
really been listening to us after all. I probalby have between 30 and
50 requests for enhancements etc. For the most part my requests are
fairly "standard" and should be fairly obvious to users and SW. And I
would estimate that over 50% of thes issues have been addressed. Not
the least of which are a number of improvements within the drawing
system.The SWIFT technology is really fantastic.

It has taken a little time for SW to get around to addressing some of
these "smaller" issues but it is going to be great to finally get these
minor interruptions corrected. I think that the best news of all is
that SW is actually listening to the user base.

Now, all I have to do is to wait for SP1...

Ed

 

What in particular do you find fantastic about SWIFT?

 

Try SWIFT on a real part and tell us how it goes.
I'm not trying to be snarky... I'm actually curios. Does it work?

-Ed

 

Since I probably won't be getting or using 2007 until SP1, I can only
go on the demo's... Sorry..But, all four of the SWIFT features are a
step in the right direction.

As far as what looked goo to me, (more to address the previous reply):

1) Features: I have times when I have had to shift thing in the
feature tree so that fillets etc. would desplay properly. Even if
SWIFT isn't perfect in this area, it is a very good step forward.

2) Dimensions: I have not used the current AutoDimensioning because it
seems like there is more work then just adding the dimensions manually.
The SWIFT Dimension looks a lot more straight forward.

3) Sketch: From my perspective this is an extremely and long awaited
feature. How many times have I added a dimension to get the message
that there is a conflict. The solution for this has been to delete the
last dimension, remove the potential conflict and then to add the last
dimension back again... what a lot of fooling around and wasting of
time! To add insult, both SW and IV knows something about where the
conflict exists but until now the program would keep what the problem
is a secreat.... Uggg.

4) Mates: It is also extremely fruatrating how clugy constraints are,
especially when there is a conflict. IV was way more troublesome to
find a problem then SW but I have spent way too much time looking for
some "hidden" conflict. And I have always thought that if SW knew that
there was a problem that it should tell us give us a little better idea
of where to start looking. I believe that the SWIFT feature will help
with this.

Until we all get to use SWIFT, we may find that it is not perfect.
But, it is at least a light at the end of the tunnel for making SW more
effecient for users. And there is always the hope that this "trend" of
SW addressing some of our day to day effeciencies will continue.

Ed

 

Well Ed,

I just tried it on a "real" part. This is an intake manifold that, because
of draft requirements I could not previously filet in a couple of areas.
Just as a side note, SWIFT does not show up anywhere in the help!
Interesting, if you don't know that this feature is called FiletXpert then
looking for it in the help produces nothing.
Anyway, it does work! Well sort of. it gave me some funky webs on the
outside of some of the runners, but not a huge deal, and it did all in one
step.
On the inside of the runners I had to do the filleting in two steps using
FiletXpert, but again it did work.
One of the nicest things is that the Cancel button appears to work as
expected. I don't say that as a joke (well maybe I do) but it does erk me
that when dealing with something that takes a long time sometimes you just
want to cancle out and try something else.

Muggs

 

You've got to be kidding?!. The one place I would use something like
this is on a casting. And castings invariably have configurations.
Forging dies is another one that needs this and typically will have
configs.

 

Translation: Project managers in India really blew it.


It sure looks like SW2007 won't stable or usable to SP5 this time.

 

Yeah, that is true. The example that I gave of the intake manifold I had to
save as with no configs for it to work.

Muggs

 

The project mgrs for this were not in india.

What do you base this statement on? Based on a feature that is 50%
implemented? I have not used a more stable sp0. The worst thing I have
to say about this release is that several of the new additions and
changes (such as this one) don't go far enough, but it doesn't break any
existing functionality that I use, and it is no less stable than other
recent versions.

 

Not having VS2007 to play with myself, I really appreciate the
feedback. I would have never expected that the fillet feature would
not work on configurations. Have any of you looked at the other 3
SWIFT features?

Thanks,

Ed

 

I was directly involved with them on this particular feature, so yes, I
can be sure of this.

Yeah, well, a lot of us have been saying this for a long time. They've
done some things that were meant to help, with varying success. I think
the initial release is stable, but it isn't without new flaws.

Easily, that's pretty conservative. I'd bet early October or sooner.

By "stable", I take it you mean "crash free". Anyone who can read knows
what Salvador thinks. Still, he's not the only one here who has been
using the software, and I don't base my opinion on what he says.

Anyway, I don't argue that there are a bunch of half-baked features, but
this is not news. Still, this release is by most accounts more stable
than other releases. I have had and seen several crash problems which
have been more due to file specific issues than just general random
crashes, but overall crash problems are down.

 

Fair enough.
But the obvious also comes to mind.
The Beta program is managed poorly and doesn't last long enough.

That problem falls to the people at SW that decided on an unrealistic
and untenable
release schedule.

 

Tim Markoski wrote:
....

If you throw enough shit something will stick eventually.

I agree that beta did not last long enough this release, but what there
was of it was very well managed. What is it (other than Paul Salvador's
posts) that you base your information on?

 

My personal experience as a SW User.

 

Its interesting that you folks chose to engage on this. If they are
people in India or people in Concord, it still come down to the quality
of the people.
Frankly I don't care if my software is written by a group of pygmies
living in a trailer in antarctica as long as they are really freaking
talented, in-touch pygmies.
I suppose that one could theorize that the people in Concord would be
better than developers in India or those pygmies in a trailer, but
there is no real reason to think that's true (as I can see) The proof
is in the product, not where it originates from

When I hear 'stable' I don't think of crash-free. Because I don't even
register SWx crashes anymore - they happen, sure, but I have been
writing them off mentally as the price of doing business. When I
crash, I just shrug, start over and keep moving on. I have no idea how
many crashes I had last week. I know there was one, but the
week-long-tally was ten?? Twenty? Two? No clue. This is not a good
thing by any measure, but it's not worth getting wound up about or
really even thinking about anymore, or my life expectency would be
about 45.
----------------------------------------------

When I hear 'stable', I take it to mean 'does it work'. Perhaps Tim
can illuminate his intent, but for this post I am going to run with
'does it work'.

Usually, with a new release, there is all sorts of new stuff (and some
old stuff) that just doesn't work. Like the rendering of lightweight
assemblies that I posted about a few days ago - its new in 2006, its
promising, and you can actually render lightweight assemblies, but (in
my tests) it uses more memory and resources so you just can't use it
in the real world.

And that's why I was curious about SWIFT. It's new, it sounds cool and
promising, but does it work ( is it stable enough to actually risk
dollars on)?

When I first saw a demo around December of last year I noticed that -
on their demo part, which is always questionable because it has been
worked over a few dozen times so of-course-it-will-work - when they did
the SWIFT for fillets a wall of their part moved.

A wall moved.
Thats a big deal.

The reason the wall moved was because, at that time, SWIFT would work
through a bunch of iterations and re-order features until it found the
right combination to make the fille work. The downside is that when
your 'automaton' messes with feature order other things can change.
To be fair, this automaton has one goal - make fillet work (like my
favorite line from the terminator - its what he does... its all he
does!).
But as is typical when something is automated, it's limited goal does
not necessarily jibe with the goals of the meat-based-entity that
originally made the part. Sure (in this case) the automaton can
automatically place a fillet before a critical draft to make the fillet
work, but does the re-ordering of features that moves that wall make
the part work (and generate money?), which, as a meat-entity, is all I
care about?

So through the early part of this year I sent them about half a dozen
of parts to test this new technology out on. I said if it works on
these, it 'works' and I can embrace it. If not, don't release it. I
even followed up a few times, and to this date have heard no response
(which, is not an indictment - they have a ton of stuff going on and
i'm just this one guy in Illinois).

Based on what I saw, I will NOT use SWIFT unless I have a backup of the
part so I can use SWx utilities to compare the before and after
geometry and be sure no critical design-intent (part function, not the
'design for change' definition that matt says he favors) was violated.
I strongly urge anyone interested in money to do the same. Sure, use
SWIFT, but use it with caution and check it a lot.

It likely is 'news' to the newbies so its worth addresing.
SWx works hard to make features that will help us. If you ever go
there, you will see folks working harder and longer than just about
every company I have set foot in. They are impressive, and they care.
They also come up with great ideas (like 3D curve on surface, for
instance), and work to get it into the product. But, as with most
design work, at some point they hit a wall - it works, but it does not
work in every case. To quote matt, its half baked.
So the dilemma is - do we release it because we know it works a little
and can help some people, or do we hold off until we can make it work
for more people? Tough question, especially when you have an
enthusiatic, talented group focused on making their customers happier
that have dedicated six months+ of their lives to the task.
I think its obvious which way they lean - and why I deal with every new
feature/function like I have a bomb in my hand.
My company is really a baker of SWx, but we will probably not use 2007
for generating revenue until March or April of next year. Its just the
way it works.

Ed

 

Guys
I read your comments with interest. As a Solidworks administror of 10
years I would never install SP0, once bitten twice shy and all that.
Unfortunatly due to work commitments i wasnt able to take part in the
BETA program this year so
the other big thing I thank you for is that 2007 is not released in the
UK yet and you guys get to do the bug finding for me, so maybe there
are some advantages to living in the old country.
thanks
ken

 

This just begs a reference to "...from which all code can obviously and
trivially be derived by one-armed unloved monkeys from Elbonia". For the
uninformed here (are there any???) it's a reference to software design at
Dilbert's company. I wish I could find one of the specific cartoons that
reference it - perfect for the descension of this discussion.

WT

 

Well, I thrashed it around for a while. Here is what seems to be
happening. Creating a configuration will prevent the FeatureXpert from
running. In other words a config shuts down SWIFT. But, SWIFT can be
used prior to the creation of a configuration to get fillets to work.

So the work around seems to be to use SWIFT to create all needed fillet
combinations by creating a fillet, then suppressing and creating a
fillet in what will be the next configuration's geometry. When all
these fillets have been created then create a config and unsupress the
needed geometry in that configuration. Seems like a lot of work, but if
SWIFT can make the fillet quicker than you can it might be worth it.

This amounts to a kind of horizontal modeling and is not easy to
visualize in SW current feature tree. And it still isn't guaranteed to
work.