guide to the Report manager and queries

[vmcdaniel]

As you have potentially figured out from my prior questions, I have embarked down the reporting trail. You all seem to have some great solutions for me thus far, and I thank you.
I would like to know how you gained your knowledge, is there some guide that can help me figure out things like, on a Change Request where are the comments stored?
Or is figuring that out some combination of applying knowledge gained from experience and trial and error, with a dash of tribal knowledge?

Thanks - Vaughn

[MikeLockwood]

Query builder reports are an absolutely gold mine (Report Manager).

But - It doesn't come with a very good treasure map. 🙂

Our best success to date has come from:

- Get somewhere close

- Export the qml and provide to tech support on a case

- Ask them for specific incremental help; they've usually been great on this

[jzupfer]

Hello,

I agree with Mike that both the community and PTC Technical Support are great resources--I learn a lot here! If you haven't already done so, you may also want to take a look at the Windchill reporting documentation, which includes a module related to Query Builder and a Query Build tutorial. You can find this information in the Windchill Help Center, beginning with this overview topic:

http://www.ptc.com/cs/help/windchill_hc/wc102_hc/index.jspx?id=ReportOview&action=show

I'd be very interested in your feedback.

Regards
Jane

Jane Zupfer
VP, Enterprise Products Publications

T
E -

[JeffZemsky]

Jane,

We could use some example based approaches like this old GS doc. Also a table of the more common joins would be VERY helpful in getting people started (see the last three pages)

[cid:image001.png@01CEBF97.340E4380]

[jzupfer]

Thanks, Jeff! We'll review the document you attached and follow up with you.

Best,

Jane

[SteveVinyard]

I'd be interesting this type of Doc using InfoEngine instead of Query Builder. Most of my Query Builder reports break during an Upgrade (9.1 to 10.x) but InfoEngine hardly has any issue. I'm told that it's much easier to grab data as well once you have some expertise. I've seen some of the basic docs on this but am interested in more, especially good examples with charts and basic metrics.

[cid:image002.gif@01CEBF81.B695F390]

Steve Vinyard
Senior Solution Architect

[jessh]

Be careful when using Info*Engine here...

By this, I don't mean Info*Engine is bad. Rather you can do a lot more
with Info*Engine and do things in different ways -- including doing
things in ways that are a really bad idea. With Query Builder
everything boils down to 1 (potentially quite complicated) relational
query pulling back just the columns you want. With Info*Engine you
*could* do something horrible like pulling back many more times the
objects you need and then filter them in memory.

In general whichever you use you really want to focus on pulling back
just what you need from the database in as few database queries as
possible and using the most efficient query criteria possible. I'd
argue that it's easier to focus on this with Query Builder than with
Info*Engine.

Of course there's lots Query Builder cannot do -- which is a perfect
time to use Info*Engine.

--
Jess Holle

[srector]

Personally, I learned to use the report manager by trial and error, and by disecting the many examplesgenerously shared by the other members of this forum.

The other tool I find useful is "Squirrel SQL Client". This Open-Source tool allows me to explore the table structure of Windchill using a graphical interface. A word of caution though. This tool is designed for database management, not just viewing. If you decide to use it, tread lightly.

Steve

[SteveVinyard]

Hi Jess, this is obviously very good to know. When you talk about filtering in memory, is that just done in the background (outside of Cognos) or is there someway to tell if this scenario is happening? I guess this boils down to, how do I know if I'm doing it wrong in InfoEngine?

[cid:image001.gif@01CEC018.E9CE1BB0]

Steve Vinyard
Senior Solution Architect

[jessh]

Filtering in memory would be done in the thread processing the
Info*Engine task request (unless you explicitly coded it otherwise).

As for whether you're doing it or not, the key is to understand what
your code (or webjects it invokes) *is* doing.

Anything (in any language) that looks like:

1. do something to get some Group/Collection/Result via a query
2. next, do something to find the things you really wanted (or remove
what you really didn't want) from the results of #1

is generally a problem.

If everything is the query is simply done via 1 webject, then it's
really a question as to how that webject is implemented...