TM1 Forum

Hi Nick,

I'm the dev lead for Cognos Insight at IBM (though these opinions are my own - not official..) Thanks for engaging with CI - we are very interested in your feedback, even if it does occasionally come from the wrong side of the bed.

Here is my 2 cents worth and I have got out from the wrong side of the bed so ...

We had a demo of TM1 Insight with an established user of QlikView last week and frankly these are some of the conclusions we came

1) Visualization , lack of quality charting, the charting is best described as average compared to QlikView or Tableau.

The initial release of Cognos Insight uses the same charting engine as our Business Intelligence platform, however doesn't offer as many settings to allow users to customize their charts. Going forward, IBM is getting really serious about visualization, and you can expect some coolness to be showing up in the days ahead.

2) TAB's are nice but I would like to have security based on TAB's in the CDD Files. I guess TM1 security will apply.

Interesting - I've heard this a few times, but not sure why you wouldn't build different workspaces for different user groups rather than secure individual tabs. Setting up security across many users and many functions can quickly become cumbersome, so I generally tend to try and keep these things as simple as I can.

3) TI's take approx 3 times the time they take via Architect or Perspectives.

To make it easier for uninitiated users, Cognos Insight import does three things by default that aren't available in Architect. First we qualify member names so that uninitiated users don't run into trouble with duplicate member names. Second, we handle blanks in the incoming data, and third we always create a top element for each dimension. If you turn these features off, the performance should be the same.

4) Performance i.e. took 15 seconds to move between tabs. This could be subjective so I would like members opinion on this.

It should not take 15 seconds to move between tabs. We do have to load tab contents on first access, so if there is a complex query driven by explore points or an interesting grid configuration there may be a small first time delay. We are continuing to add to our performance test suite as we encounter customer examples of data sets and queries - please do let us know through support if you have a data set we need to look at.

5) Pass CDD's around , Why can't we store them in Traditional Applications folder ?

Interesting - we currently support sharing through Cognos Connection - which allows sharing of any number of .cdd files, offers folder object security (read/write/traverse) to grant access to the right group of people, and supports provisioning of Cognos Insight for users who don't have it installed.

6) Promotion from Dev to Prod, I know there are "workarounds" but that's simply not good enough.

Agreed - though this isn't specific to Cognos Insight, data integration between TM1 servers is a use case that will get more and more attention as we make it easier to create and deploy TM1 applications.

7) Printing, You can't print these reports ! Can't export them to PDF.

You can't print the canvas, but you can right click on the canvas and "Copy to Clipboard". This makes it easy to put an image of a Cognos Insight tab into another document. I know that's not a good enough answer, but hope it helps in the meantime.

8) Spreading , if its a contribution tool then that's a must.

Cognos Insight does support spreading (though classic TM1 interfaces offer a few more options). Are there specific spreading options you are looking for?

9) How is user going to refersh the Tab's apart from logging off and logging back on ? Have I missed a point here ?

Not sure what the use case is here? Normally Cognos Insight is either used by an individual creating a workspace to be shared with others or consuming a workspace that has been authored and published by someone else. Maybe you are thinking of "Central-mode" where multiple users can be performing modelling actions at the same time?

10) Why is only one CDD file allowed to published via TM1 Applications ?

Sorry, don't understand this one. You can publish as many Cognos Insight applications as you want.

11) In Connected mode the performance even by IBM admission is not recommended.

Interesting - we run global "blitz tests" with users all around the world working in connected contribution mode every week. I'm thinking you should tell your IBM contact to give me a call..

The standard reply we got from IBM was "but it's a good Pre Sales tool" .. We ain't taking it to anybody ever again.

Ouch - that one stings a little. The idea here is to make it vastly easier to create and deploy Business Intelligence and Performance Management applications. Our competitors claim to deliver applications in hours and minutes - we can do it in seconds, plug into an enterprise BI infrastructure, and give you a glidepath to all of the rich functionality in our TM1 and BI platforms. As a consumption vehicle for PM & BI, I hope you'll find the rich layout capabilities useful to better explain the context in which the data lives. Distributed mode delivers higher scalability and responsiveness than you've seen before, and explore points give you new ways to explore TM1 data. Fundamentally CI is about agile performance management - if we haven't hit that mark we want to hear from you so we can work together to "make it so".

Cheerio

Thanks again - lets keep the conversation going - I won't be able to monitor the forum all the time so get to us through the normal channels as well, but glad to hear from real people trying to do real things with our software.

Thank you for your reply and I appreciate your feedback on my post. Since my post we have engaged the right people at IBM to provide more clarity and support around new CI features and potential enhancements that as a user community we would like to see.

I would also like to thank the Moderators for not deleting this post.

Cheerio

Great to have you here Floyd. I hope your bosses let you stay.

I think Cognos Insight shows good promise, but it's early days and one can't expect too much from the first release.

Could you have a word with your mates in marketing though? There's Cognos Insight and Cognos Business Insight, both quite different tools yet very similar names, especially given that "Business" doesn't actually mean anything in this context. That is confusing enough, but it'll get worse because the uninformed will call Business Insight "BI", pulling a third tool into the mix.

nick_leeson wrote:I would also like to thank the Moderators for not deleting this post.

One of the joys of the forum not being controlled by IBM.

Martin

Hi - thanks for the welcome! Yes we know the product naming thing needs work and a fix is coming "Real Soon Now". As for the early days part - its really important that we work on the right things next. Let us know what you need to be successful, else we'll end up like this:

http://dilbert.com/fast/2012-05-07/

Do send enhancements through the normal channels, but I welcome any discussion here and will try and keep track of the ideas raised.

Thanks!
Floyd

FloydKelly wrote:

2) TAB's are nice but I would like to have security based on TAB's in the CDD Files. I guess TM1 security will apply.

Interesting - I've heard this a few times, but not sure why you wouldn't build different workspaces for different user groups rather than secure individual tabs. Setting up security across many users and many functions can quickly become cumbersome, so I generally tend to try and keep these things as simple as I can.

Hi,

I (very real guy who is trying to do things with your software ) think I can give you a use case for that.

We're running a multi tenant environment, so there will be an application for every tenant (i.e. 75+ tenants per environment).
We have two types of users. The users who can do the whole sales planning process and the users who just can do some steps. I don´t want the limited users to see all tabs cause there are some tabs where they aren´t allowed to see anything.

So I have do build an application for those users what doubles up the amount of applications in the portal and increases the maintenance work. If there would be a security on tabs I could build just one application for every tenant.

(I already logged this through the official channel but maybe it´s interesting for you)

Best regrads
Froyo

Awesome - now I understand the use case and yes we have discussed it internally. (I didn't make the connection immediately)

Keep it coming...

FloydKelly wrote:

8) Spreading , if its a contribution tool then that's a must.

Cognos Insight does support spreading (though classic TM1 interfaces offer a few more options). Are there specific spreading options you are looking for?

I think this is also a very interesting point for us.
As we are coming from Enterprise Planning we are very used to some spreading functions which I don´t see in TM1/CI at the moment.

First example:
I have a time dimension but I don´t want the WholeYear item to be calculated as Sum all the time (which could be done with a hierachy). I also want to have Average Items or TimeLast Items ect., which is an Out-Of-The-Box feature in Enterprise Planning.



In TM1 I´m not able to write on the WholeYear item because I have to use a rule to realize this item.
At the moment I don´t see how to solve this problem.

Second example:
I have sales targets which are calculated. To keep it simple: a rate (Item1 = Item2/Item3). I realize Item1 by using a rule.



But now I´m not able to write on Item1, even if I´d like to. In EnterprisePlanning I can define this calculation AND write on the calculated items and it´s spreaded down.
UseCase: I want to lock Item2 and increase Item1 because I want to see how Item3 must decrease to reach my target.



This is a simple example and gets even more complex when there are more dimensions included which have to be spreaded for their own (organization structures maybe).

For such use cases it would be very neccessary to have the spreading possibilites which are in Enterprise Planning Contributor already implemented.


Best regards
Froyo

Ok got it - we've talked about some of these use cases internally - some will be easier to do than others. I'll pass these on to product management for prioritization, and keep an eye out for them architecturally as well.

Cheers,
Floyd

Maybe my idea is not new, but I think ability of groupping cubes, dims, processes must be extremely usefull. Now we have A LOT OF CUBES with LOT OF DIMENSIONS for different applications, so it's inconvenient to navigate through model. Objects can be sorted by us for diff purposes, for ex. system (backup, etc), planning, forecast...

ioscat wrote:Maybe my idea is not new, but I think ability of groupping cubes, dims, processes must be extremely usefull. Now we have A LOT OF CUBES with LOT OF DIMENSIONS for different applications, so it's inconvenient to navigate through model. Objects can be sorted by us for diff purposes, for ex. system (backup, etc), planning, forecast...

Yup, it's certainly something which quite a few would consider as an enhancement, as per this thread.

Its amazing I remember way back when, umm version 7.1 where so many questions like this were asked and still here we are so many years later and still, the same questions are being asked about things that would appear, on face value, should have been addressed many times over by Applix, Cognos and now IBM.

Why can't the TM1 Button allow users to rebuild a "workbook" ? its possible to rebuild sheet but not workbook.

Why?

event based scheduling would be good...

i.e. file is copied to Tm1 server, TM1 loads file...

yes there are workarounds but would be great if it was 'inbuilt'.

I am extremely new to TM1 as my company decided to move to it from EP about 4 months ago. My strength is more on the Cognos BI data architecture and reporting side. With that said, if I could ask the TM1 Development Team a question, it would be...can you please work more on how TM1 data interacts with BI? While I truly am impressed with the transfer of data security from TM1 to BI and the wysiwyg ease that causes report authors to not have to do a lot of work on manipulating data within queries, there have been some issues that are truly frustrating. The inability to manipulate data at the framework manager level was the first frustrating spot. The measure data being locked into a state of confusion as to its data type was the second frustrating spot. Please realize, developers, that businesses are going to put string and numeric data elements into dimensions that will later be identified as measures. This is not ideal, but it makes for a frustrating situation when a picklist-created string field that holds numbers is not allowed to be treated as a number when it appears in a report on the BI side. What is otherwise a beautiful report is marred by the fact that not only will this field not sort as a number, but it cannot be cast as an integer without mysteriously casting the entire measure dimension with it. The data model is not set up ideally, but in the past when dealing with EP, any situations such as this could be worked around within Framework Manager or within Report Studio using SQL functions.

Please keep up the good work on developing towards an otherwise good product. I am continually fascinated with learning how it works.

First of all, I love the CI product, and the new features in 10.2. It is a joy to work with.

11) In Connected mode the performance even by IBM admission is not recommended.



Interesting - we run global "blitz tests" with users all around the world working in connected contribution mode every week.

However in Connected Mode my CI crashes the TM1 server, just by opening the Explore pane.

Am I doing something wrong?

Regards,

I'd like to have feature 'deactivating objects' - marking them transparent or with special pic. System will interact like this objects are deleted, but they can easily be recovered without backing up and restarting model.

it will improve development when you have a lot of old unused cubes but not sure to delete them finally.

Please please please please stop or put some controls around subset names being available as weird psuedo objects in dimensions. :

Sorry thats not a question

Steve Rowe wrote:Please please please please stop or put some controls around subset names being available as weird psuedo objects in dimensions. :

Sorry thats not a question

I think it comes under the catch-all question of "undocumented 'features' which are probably more accurately described as 'unintended design consequences' (UDCs)". The undocumented UDC of being able to paste a single value into a single DBRW (or DBRA come to that), but not being able to do the same thing with a range of values (meaning that most users are unaware of this functionality, resulting in them sometimes deleting or changing data unintentionally) also comes under this category.

The question?

Imagine a TM1 user on his / her knees in a thunderstorm (the UDCs in this metaphor), looking up at the heavens (the TM1 design offices in this metaphor) with arms outstretched, and yelling:
"WHYYYYYYYYYYYYY????????"

THAT was the question.

Is it coincidence that the documented name for the feature I think Steve is describing (User Defined Consolidation) has the common acronym UDC? Or have I misunderstood Steve's request?

Duncan P wrote:Is it coincidence that the documented name for the feature I think Steve is describing (User Defined Consolidation) has the common acronym UDC? Or have I misunderstood Steve's request?

No, that one's actually documented. Kinda sorta. As I mentioned in some e-mail exchanges with user support the documentation isn't consistent, but it's there.

(Thus spake me:

In the Subset Editor, if you want to do a user-defined custom consolidation, what's the button that you click? It's captioned "Rollup". So if you're an admin, and you want to know how to enable/disable that functionality, what word are you likely to search the documentation for? Rollup, of course. But the only thing you'll find is a reference to the button in the subset editor in the user guide. If you're clued in on that day you may notice the term "User Defined Consolidation" in there and think that might be worth searching for, at which time you'll find the server parameter that you need (which is actually called UserDefinedCalculations rather than consolidations) in the Operations Guide.

)

However I believe that Steve is referring to the phenomenon described by Robin in this thread.