Selective delete in windchill 10

Forum|Forum|13 years ago
November 6, 2012
27 replies
13195 views

Ok so this was possible in ILINK3.4 how do you do it in WC10. See picture below. I would like to delete 8010-120_mfg.asm version A.1 from the data base. Unfortunately I cannot I can delete A.4 through A.1. Purge is not the answer. I do not want to delete from A.1 through A.3 and leave only A.4 I would like to selectively delete these items from the database.

Ron

[cid:image001.png@01CDBC07.911FDD70]

Other

Show previous replies

DarrinHiebert

1-Visitor

I've been following this and I'm definitely in Don's camp on this.

Our current third party PDM (our solution is not really a PLM) does NOT
allow for ANY iteration deletion, and that IS the way that we want it.
Iterations in the middle are just history, and they are in there. They
don't hurt anything, except maybe making your database bigger. You do NOT
have to use them.

With all of that being said, I'm no expert on WC, and in fact we are
currently working towards our WC test bed setup.

But I have two comments in general.

1) PTC did this intentionally: I think that it's a distinct
possibility that PTC realized that they way that we WERE doing it in
Intralink by allowing iterative deletion was NOT a good idea. I have to
wonder if they weren't getting complaints from people (admins) that were
deleting things, and screwing things up, so they just took the
functionality away. As was stated earlier in this thread, you have to be
very very careful when you start deleting iterations, as it could cause
other things to fail. Even though you think that you are doing the right
thing, there could be something that you didn't account for. PDM IS all
inter-twined together, and it IS a huge jigsaw puzzle, and deleting one
seemingly insignificant piece could have dire consequences that you don't
see coming

2) Swiss army knife PLM: It's important to understand that WC is NOT
just PDM anymore, like Intralink, it's PLM. This is actually one of the
things that we LIKE about WC, and one of the reasons that we are
considering moving towards WC. For US ,and I realize that YMMV, we want
to have as much knowledge in ONE location as we can possibly get in said
location. Is that not the definition of PLM? It sure feels like this
should ultimately make us more efficient corporately, because we wouldn't
need to maintain several systems, and we would only need to train everyone
on ONE system, that being WC. But, as was stated earlier, you have to
realize that while it will do everything, it may not do all things the
best, and at the end of the day, that's OK, because the trade off is that
you get ALL information to a common point of entry / exit, and that far
outweighs, in our opinion, the maintenance involved with multiple systems.
For us, we have an ERP system. It sucks at some things, and some things
it's better at. For the things that it sucks at, we often have
work-arounds that are outside the system, and those work-arounds cause us
more headaches than they solve. They're complicated, and they are
somewhere else, so you have to leave ERP to find that answer. That is NOT
a total solution. The holy grail of PLM, I think, is to have ALL systems
tied together in one place that either share a common database, or at the
minimum can read from different databases if different functions are
controlled by different programs. If you've been to conference over the
last couple of years, I think that it's very very obvious that this is the
way that PTC is going with their PLM. They are trying to get as many of
these systems tied together as is possible, and I don't think that's a bad
thing. I also think that it's a matter of time before has, acquires, or
is ACQUIRED by an ERP vendor. At that point, think about how powerful WC
would be IF it was a one stop - do all ENTERPRISE wide system PRODUCT
LIFECYCLE MANAGEMENT system.

There was a post on here recently about which should drive which, PLM vs.
ERP, or ERP vs. PLM. The argument was made that PLM should drive
EVERYTHING, and I think for some that makes sense. I think that the rub
with WC, to Joel's point, is that this is the way for some companies, but
for some, especially smaller ones, this may not be the way that they want
to go, and I can definitely see where a smaller company could see WC as
way to much complication. So, while WC may be the Swiss army knife of
PLM's (or attempting to be), some users only want a pocket knife, and from
PTC, they don't have any options. That is a very valid argument as well.

DonSenchuk

12-Amethyst

I single iteration in the middle of a sequence has to be deleted for legal reasons? I'm not touching that with a 10-foot pole.

When you try to delete that iteration and windchill forbids it, well, there's your warning not to delete it.

In Reply to Michael Reece:

Whether you understand why or who cares is irrelevant. Whether you want to remove it for legal reasons of for other maintenance reasons, I should be able to remove that object.

If removing the object is doing something unreasonable to the dataset, Windchill should warn me, but if I have sufficient permissions, it should not stop me from removing C.3

jfrankovich

12-Amethyst

So, with all of the response to this I think it is clear that we all have very different ways that we need WC to work. I for one really did like the ability to delete selected iterations in 3.4 and that was a major pain point when moving to WC 9.1, but we had to learn to deal with it or work around it.

Deleting selected iterations is all well and good as long as you are always dealing with the LATEST versions of files. So on a typical day for most of us we want to use or edit the latest version of a part, assembly, or drawing. However, what about when we go back and look at files as they were stored or as they existed just before release? Deleting selected iterations would, just as it did in 3.4, indiscriminately remove iterations from the database and any transaction records. Not typically acceptable for a PLM as a system of record or a system that is intended to link to non-CAD data or tie in with other systems.

I personally would like to see that ability back on a limited basis in WC, BUT not until there is an acceptable way of dealing with the links that you are breaking. For those wanting WC as a simple PDM and willing to fix everything that breaks after the fact like in Intralink, simply allow unrestrained deletion of CAD files. (Would have loved that a few years ago.) But for those that need to be very cautious with the links and references to other files, iterations, and records, give us a way to handle the missing iterations such as automatic rollback to the previous iteration, selection of a replacement iteration, or removal/isolation of the object with a flag on any recorded references to that deleted iteration so that the object appears to be deleted but references to it indicate the deleted iteration. And record the deletion of the object in WC so that I can audit object deletion and see the affected items and records before and after the fact.

Just my 2,
John Frankovich

DarrinHiebert

1-Visitor

So, since I'm WC illiterate, can I offer an idea about how we would handle
this situation:

"However, what about when we go back and look at files as they were stored
or as they existed just before release?"

Our CURRENT PDM will not do this either, BUT we have a need to go back to
prior revs at points. What we do is go and get that prior rev the way
that we want it, verifying that it's correct, and saving it off somewhere
in a safe location. Then, we check out the LATEST version from our PDM,
and delete those files off of disk. We call up the PRIOR version that
have saved off somewhere, and check that back into our PDM. Now, we're
back a revision, AND we still have the history that we went forward, and
then decided that we would go back. Yes, this is somewhat of a pain, but
I guess that our take on it is that some day in the future, we may really
like to know that we changed the design but decided to go back. If we
were to delete the LAST revision, we would lose that history.

Iterations to us are history. It seems like a bad thing to delete
history.

SteveVinyard

1-Visitor

You have the following:

1- Massive Relational Database structure with links to everything (Change Models, Promotion Requests, Secondary Content, Multiple Visualization files for each object, IBAs, etc)

2- CAD Models with parametric information that is also managed, controlled and protected by the database/Windchill

3- All history is meticulously maintained automatically

4- All data integrity is maintained including rename operations etc.

5- All CAD Models are stored with an As-Stored configuration automatically (a baseline) that can be referenced at any point in time to see EXACTLY how a model was built, rev for rev, 15 years ago.

You don't have:

1- The ability to delete a historical revision very easily

[cid:image002.gif@01CDBDBB.E9DC5F00]

Steve Vinyard
Application Engineer

Anonymous

If the iteration has no links to anything else (highly unlikely), then
I'm pretty sure you can delete it (as admin). (Been a long time since
I've done so).

More directly/importantly though, when I've been asked this question in
the past, my response has always been to ask why you want to delete old
iterations?

Typically the answer comes down to wanting a "clean database". i.e. some
attempt to keep the house clean so guests don't see how messy life is
(can be). Or about file server space.

Having a clean house can be done with states, upon release, set all the
in work stuff to history (or similar) and only let admins or such see
history items.

File space should not really be an issue, the incremental cost for data
storage in a well configured datacenter is very inexpensive (less than
$1000/TB).

The only issue IMO is in database/metadata/table size. And here, Oracle
has solutions for extremely large databases (larger than I've ever had
to worry about). So may require more Oracle expertise than we normally
use in the WC world, but I'm confident it is manageable.

As for legal questions, I'm involved in two federal cases as an expert
witness, and if missing files (intermediate iterations) are not there,
it causes questions and problems in the legal world. Missing iterations
cause concerns from both legal parties, even if explained as clean-up
etc. Now, if the database contains files that were illegally obtained,
there is a lot more that needs to be done than just deleting the native
file and no application should allow such to be done through any user
interface.

~Dan

Dan Harlan
Mechanical Engineer / CAD Administrator
480.940.0036 x178 Office
480.940.0039 Facsimile

481 N. Dean Avenue
Chandler, AZ 85226
dharlan@aitint.com
www.aitint.com

kbrault1

1-Visitor

Good response Dan (I've known Dan since his days at the Nixon administration).

Kevin Brault

Sign up

Please use your PTC eSupport account.

Welcome to the PTC Community

Please use your PTC eSupport account.