Best Of

Hi Mateo DE LOS RIOS

I can't find the reference at the moment but you might like to know that the functions of the Tilda database are being incorporated into a BOARD 11.x release.  This is good as it means that everybody now gets the benefits of centralised, advanced security control.  It is not so good if you've really gone to town with the Tilda and used it in everything you do as you will now have to re-do everything in an entirely new way which may, or may not, work as you are used to working (Guess which one my money is on?).

Personally, I welcome the incorporation into BOARD any of the original bolt-ons, like the Meta-tool (Worst kept BOARD secret super tool created by Jan Kaufmann), Log-analyser, Randomiser, Tilda, Transporter.  I also welcome the the replacement of BOARD features such as the MXC cubes. 

What is slightly irksome is the way these features and replacements are rolled out without adequate support to the few users that took the original tools and made them work, weaving them into their best practice to then have to learn an entirely new way and have to amend all existing models in order to be able to even use a new platform.

...Anyway , hope this helps

Hello Anastasia,

Many thanks for taking the time to respond and for explaining the technical issues currently being dealt with regarding the website.  Perhaps a BOARD notice to all community users would help in keeping BOARD developers aware of issues and cultivate support from them.

Thank you also for your comments regarding feedback. My aim in mentioning several points was to demonstrate that there are now a growing number of issues (a trend?) compounding to make the BOARD experience currently less enjoyable than it was.

As I'm sure you will know, each of the points I have outlined in my post has either been formally submitted as ideas or are current topics of discussion on the community website and I suspect the product team are already well aware of them.  On that note, I would welcome any opportunity to be involved in any follow up action.

Kind regards to you Anastasia,

Paul

Thanks Samuele Marchetto,

I like the nexel approach - I keep meaning to find reasons to use it but end up testing in it only to replace the solution with a dataflow or datareader process later. 

I was playing with the datareader date functions that you mentioned and that is another way and a method that will come in handy later and for another process I will be incorporating.

As a result of playing with your suggestions and so thinking about my needs more deeply, I was able to devise a method that uses a dataview to calculate the different date parts with algorithms.  As there is nothing, that I have found, on this matter and as the method is dynamic, I have submitted an Inspiration which outlines and discusses the method I chose to ensure it is more easily found by all.

However, for completeness and in summary, I created a dataview with several algorithms like this:

The cubes are populated with the numeric value of each of the dates so that calculations can be carried out.

Again, thanks for your response and so quickly too.

Using References, it is not possible to replace the member title with a custom one.  So long as I know how many columns I have 5, 10, [n] and that the number is fixed, I can create a custom entity/cube structure to sit above my dataview.  So, if I had 10 fixed columns, I would do this:

• Create an entity [Fake Title]
• Add the following members as follows

     1 - [blank]

     2 - [blank]

     3 - [blank]

     ...

     10 - [blank]

• Enter no of cilumns in MAX ITEM NR
• Create a cube [Fake Title cube]
• Add the entity [Fake title]
• Create an Admin Screen
• Create a dataview
• Add Cube [Fake Title cube]
• Add the only entity [Fake Title] to the column axis
• Select dataentry [x]
• You should now have a dataview with a single blank but yellow row.
• Type the custom titles into the yellow numbered columns - if no numbers, switch entity from default to code.
• Save the entries.
• Copy the dataview
• Return to your working screen and screen Hide the header
• Reduce the far left column to minimum = invisible on web
• Now set column width to that of target dataview and position on top - you may also need to set it to the back so as not to interfere with operation of source dataview.

hide the headers of the fake title dataview.

I think that's the correct sequence - was this any good to you?  Again, it is only useful for non-dynamic dataviews where the number of columns remains fixed and the data in them matches the fake title you have set.

Hi Bob, based on your question and diagram, you can achieve this through the block formatter as Helmut has said.  Below is the formatting, in red boxes, I set up to show a custom block, remembering to enable the custom block editor first.  Highlight the required level of cell Value, Total, Grand Total or all of them as I have.  Once highlighted, you can specify how it should look in options and Font boxes. 

Below is how I formatted a random dataview to meet your specifications (NB: I centred the % values to emphasise that the symbols before and after appear at the extreme end of the cell - this is not always desirable)

A good feature to use is the Copy and Paste which allows you to copy the completed format and paste it to the blocks that you want without having to format each one individually - useful for when you realise that you forgot to copy 'drag' the block with formatting, to create the subsequent blocks, at the beginning of the dataview creation - doh!

Was this what you're looking for?

Thank you all for your replies.  It's great to see several solutions to one problem as this really opens up so many more possibilities to different problems.  To summerise, there are 4 possible solutions:

  1.   CSV - save to text and use file to populate anything - creates independent artefacts which could be lost and must be tracked on promotion trough Dev > UAT & Production.

  2.   BOARD Replicated [R] Entities - How to use them - Exciting as I've not used this functionality yet but I do heed the warning from Bob Gill.  I will now be playing with [R] Entities on a dev model so thank you very much for pointing it out to me.

  3.   ROLAP - Really clean, simple and fits fits with BI Best Practice.

  4.   Mapping Cube - This is what I had originally set out to do but encountered issues.

I will go with the ROLAP approach - Thanks Björn Reuber.  However, Etienne CAUSSE, would you have a model I could look at (cpsx + hbmp) as I want to understand this approach and learn where I was going wrong?

Thanks all for your excellent advice.