SolidWorks testing and "bugs"

If you've ever wondered about SolidWorks testing procedures or
complained about "bugs" you might find this article interesting.

http://designsmarter.typepad.com/rob_rodriguez/

 

Complete and utter BS.

 

Great that you think so... How about you elaborate a bit? I could say I
hate cars but that means nothing to anyone. If I told you why I hate
them and presented a argument then that would have some meaning. In the
mean time you have just pointed out nothing.

Want to fill us in as to why you think it is BS?

Ben

 

"Complete and utter BS" for the Questions or the Answers?

I think the Questions were a bit too softball, and the Answers read
like the SolidWorks QA manual & PR kit for management to use in
interviews.

I think (as I've stated before) that a 12 month cycle is too quick for
major releases, as the time required to get familiar with a release is
somewhat long, and then when you add bugs to it, that becomes very
tiring. Someone needs to corner SolidWorks execs on the actual Risk
vs Reward to end users of their 12 month upgrade cycle. I think the
only reason for a yearly cycle is the demands by the CFO for continual
high income from maintenance fees.

Sooner or later, it sure seems like to me that SolidWorks can't
continue doing a major relase every year, and I wonder about the long
term advisability of changing the file format every year, as that
seems intended to force users to upgrade as fast as possible.

SWks' "Kamesh" is the Vice President of Quality Engineering & relates
users into groups, which I think are a bit erroneous.

He describes "pragmatist" users being the 1/3 of users who adopt just
before mid-year as "solid citizens who do not like to take the risks
of pioneering...", when in fact I see those & later users as the ones
who realize that lost time is not only costly, but the time loss can
never be made up. That is just engineering logic & cost accounting.
I have one friend & his primary company customer in medical equipment
who are all still on SolidWorks 2005. My friend's emphasis is fast
output of new parts.

"Early Adopters" @ about 1/8 of the users are ones with "...vision to
adopt an emerging technology...", meaning being a user either with no
hair or one who wears a hat to keep from pulling his hair out. Early
Adopters are ones with a lot of free time on their hands, as a serious
designer can't afford to lose hours easily. Generally the older users
I've met get burned out on Beta Testing. It can take over your life
at times.

If the beta testing was better done over a longer time, I would adopt
right at SP 0, but that just doesn't make sense.

Bo

 

What he said

 

SW Corp has a release schedule that simply doesn't allow for proper testing.
The quality of each initial release is decreasing steadily.

SW makes changes that make ZERO sense and most certainly weren't requested
by any user.
Anyone who has used the new SW Explorer/PDMWorks combination knows what
a cluster @#$! it truly is.

PDMWorks is now Integrated into SW Explorer.
What a colossal ABORTION!
To start with, the installation on the server side doesn't take into
account your previous installation folder when you upgrade and manages
to recreate the
performance issues due to the MFC4.2 files.
One would think that at this stage SW could actually build a setup for
this that actually worked properly if they read their own technical
bulletins.
They integrate MS Desktop search for no logical reason at all.
It is a solution in search of a problem.

This wouldn't be so bad if SW actually provided a usable interface.
They actually spent time on a God awful skinned interface that looks
childish.
They didn't put any effort into functionality that would actually help
end users.
The API hasn't been updated at all as far as client functions are concerned.

One of the MAJOR screwups is that the local file view does NOT refresh when
a document is checked in and deleted from the local view.
You still can't check in more than one document at a time.
This makes absolutely no sense.

If you select a project in the vault view and then move to the local view,
PDMWorks doesn't remember the project you selected.

PDMWorks 2006 had it's problems on the client side but it worked.
We now have an "enhanced" application that is a HUGE step backwards.

A couple more gems from this newly "enhanced" version.

The upper right pane of SW Explorer does NOT allow ANY PDMWorks
functionality to be accessed via right-click. All you get is "Locate in
Vault Tree"
or "Locate in Local Tree" .
Yet another completely useless function.
You should be able to at least Check in or Check out documents from the
upper right pane.

eDrawings is now the default Viewer for PDMWorks.
The problem is that it defaults to the animated mode with no way to
change it.

 

This is a big change from the way SW used to do testing. These changes
were starting back in the 3 Amigo days and have since moved forward.
It used to be that the previous Quality guy could hold up a release if
it didn't come up to snuff. It used to be that individuals at SW were
responsible for particular areas of SW to make sure they worked. I
didn't hear Kamesh say that or whether he actually has ever done that.
They are continually testing a moving target and playing the odds that
enough people will be happy to continue buying with the level of bugs
they leave in. (see Breaking Vegas).

 

Where is "Breaking Vegas"?

 

I too find the choice of words used to describe the users to be most
telling. Those who quickly jump when SW says jump are "visionaries".
Those who choose to jump only after making sure that it is prudent to
do so are "laggards". Very, very telling. SW needs to understand
that most businesses cannot afford to use leading edge tools unless
they are proven to increase effiency or add value to the end product.
Don't give me a $3900 Swiss Army Knife when what I really need is a
screwdriver that always works exactly as it should. (My sincere
apologies to the Swiss).

 

In fact, if any company has a tight QUALITY SYSTEM for ISO or FDA GMP
compliance, they may be forced to even skip using a particular year's
version of SolidWorks if there are serious issues that don't get
fixed.

Bo

 

It is a documentary about the MIT group that "broke Vegas" in
Blackjack by using stastical methods. McEleny was featured in the
show. It can be seen on the History channel.

 

Hello to everyone,

In a nutshell. This article shows that Solidworks is trying. I also
shows that in fact they are overwhelmed with the task of
delivering more robust software because their release cycle is too
short. I am certain that 90% plus user would welcome
a more reliable software release. In the end, it comes down to a
simple conclusion. Does a new release enhance productivity
and guarantee savings to justify an upgrade? Only then it makes sense
to upgrade. And one more thing. The reason why
Solidworks can continue their marketing strategy is in fact also our
fault. If we would strive more to refine our skills
we would not need the fancy gadgets. As an example. I am sure that
users like Ed Eaton, Marc Biasotti etc. can
kick butt with SW 2001. I am certainly supporting innovation, but
nowadays it seem that the more Solidworks or any other
company provides new solutions the more we loose the ability to think
for ourself and sharpen our abiltiy to find solutions.

Elmo

 

As an example. I am sure that

Hopefully, Ed will chime in on this. My thinking is that Ed could not do a
lot of what he does nearly as well as he does with 2001. There are just too
many features missing.

As for me, I don't do models as fancy as Ed's, but absolutely cannot make
up for the shortcomings of 2001 with operator skill. So many little things,
like managing multiple mate groups and not being able to move hole wizard
holes to new (or lost) faces.

Perhaps operator skill could also be applied to work around the bugs? In
that case, give me new features that also save me time. They are most
definitely in there.

 

Hello Dayle,

With all do respect, the point I wanted to make was to emphasize the
problem, that
too many times poor technique & lack of knowledge is compensated with
a cry for
automation. I did not intend to proove that the ones that improve
their skills
past the ability to use new features can kick but with version 2001.
My point was to
instil a sense of self-responsibilty and stop blaming Solidworks for
all wrong doing,
without obviously excluding Solidworks. In the end Solidworks listens,
and if enough
of us jump on the new-version-band-wagon, then Solidworks has enough
reason to
maintain their course. And it goes without saying that the newer
release offer
functionality that goes beyond operater skill, you are totally right
their. But this is more
of a 10 foot few, rather than looking at the 30000 foot perspective...

Elmo

 

I've kind of settled on 2004 as having enough features, decent
performance and fairly stable.

 

You just said in one statement what should be the rule of thumb. If
you find a version that fits your needs and is stable then stick with
it.

And if you have to work with customers that have a future version and
can't work with dumb imported geometry then.... I'm sorry to hear
that. Your SOL. lol

Don