locked
Project Crescent RRS feed

  • Question

  • I have downloaded the CTP 3 and have done some basic report building using the powerpivot functionality. While it seems like a great product I have the following questions / queries / suggestions :

    1    if I have a table / matrix on the page along with a chart, and if i click on one element of the chart, it filters the data of the table/matrix. but vice versa does not work. If i click on one element of the table, the chart should get accordingly filtered. Is this a bug or am I missing something ?

    2    i vaguely remember having read that i can 'bookmark' a particular view. What i mean is that for putting accross a point lets say i have put some filters in place etc.. and then i want to 'save' this view...is there a way of doing this ? in fact i should be able to save multiple views.

     

    3   in the matrix, i feel it is extremely important to provide for expand/collapse functionality as well as the drag and drop functionality.

    4  in the filter window, there should be a visual mechanism to identify on which field a filter has been placed. there should be a mechanism of removing filter of a field, or removing all filters from all fields.

    5  in the tile mechanism also there should be a mechanism of selecting all tiles, or identifying the active tile separately.

    6 is there a mechanism by which the window will cover my full screen instead of just a rectangular block in the middle of my screen, as i have a high res monitor ?

    Thats for now, but as i experiment, i would post more of such comments.

    Cheers to MS for a good product !

     

    Sanjay Shah

    sanjay@prosysinfotech.com

     

    Wednesday, September 7, 2011 12:34 PM

All replies

  • I have downloaded the CTP 3 and have done some basic report building using the powerpivot functionality. While it seems like a great product I have the following questions / queries / suggestions :

    1    if I have a table / matrix on the page along with a chart, and if i click on one element of the chart, it filters the data of the table/matrix. but vice versa does not work. If i click on one element of the table, the chart should get accordingly filtered. Is this a bug or am I missing something ?

    This is currently by design.  There is no selection enabled within the table or matrices.

    2    i vaguely remember having read that i can 'bookmark' a particular view. What i mean is that for putting accross a point lets say i have put some filters in place etc.. and then i want to 'save' this view...is there a way of doing this ? in fact i should be able to save multiple views.

    You cannot save "multiple" views.  I'm not sure what you are referring to.  You can save the report in it's current state.  We don't save the entire interactive state with the report, but some of it we do.  There may be some additional work we do in this area before we release.

    3   in the matrix, i feel it is extremely important to provide for expand/collapse functionality as well as the drag and drop functionality.

    Expand/collapse is not in CTP3 and will not be in the first version.  Drag and drop is another feature we are looking at.  We can't comment yet on whether it's in or not.

    4  in the filter window, there should be a visual mechanism to identify on which field a filter has been placed. there should be a mechanism of removing filter of a field, or removing all filters from all fields.

    There are some additional tweaks we will be making in this area.  Again, we can't necessarily comment on all of them at this point, but the feedback is noted and appreciated.

    5  in the tile mechanism also there should be a mechanism of selecting all tiles, or identifying the active tile separately.

    Can you clarify what you mean by identifying the active tile separately?  We do give the active tile a different visual representation with the navigation strip at the top (tabs).   As far as the "Select All" feature is concerned, it will not make the first release.

    6 is there a mechanism by which the window will cover my full screen instead of just a rectangular block in the middle of my screen, as i have a high res monitor ?

    Again, we can comment on post CTP3 functionality at this time.

    Thats for now, but as i experiment, i would post more of such comments.

    Cheers to MS for a good product !

     

    Sanjay Shah

    sanjay@prosysinfotech.com

     



    Program Manager, SQL Server Reporting Services This posting is provided "AS IS" with no warranties, and confers no rights. http://blogs.msdn.com/seanboon http://www.twitter.com/ssrs
    Thursday, September 8, 2011 12:25 AM
  • Hi Sean Boon, Robert Bruckner :

     

    Thanks for your feedback. Following are my comments / further observations :

     

    1  selection within table matrix : I understand that currently this is a limitation. but the lack of it makes the product look half baked. You would of course know that in an 'industrial strength' application of this we would be analysing areas like Inventory Analysis, Accounts Receivable Analysis, Project Profitability...where there are hundreds and thousands of items. so a cross connection between all the objects of the dashboard is very important, especially when the real purpose of crescent is 'adhoc' analysis directly by the user.

    2  while tiling and other features of selection appear attractive, in real life it is unlikely to cut much ice with professional business analysts like me. the reason is that there are too many discrete values and such selection features may actually block a lot of real estate. Pardon me for saying this, but i feel there are important analytical features that must be considered in version 1 to make it a very big sucess. one important feature which is lacking is the ability to search. since the full data is available in the background, i am sure that providing a search button which will quickly traverse all the fact/dimension values and show up matching results should not be difficult, considering the huge experience MS already has in this area.

    3  your answer re. saving multiple views is "you cannot save multiple views". I do appreciate this but please note that when users actually use such reports in a collaborative way, there is a very big requirement of sharing each others' analytical views. When i am doing my analysis, i may filter here, filter there and reach an 'aha' conclusion. now i want to share it with my boss. how do i do this ? Since the product is already server based, i think it is important that the 'view' is saved and shared. how else would you achieve 'collaboration' ? if this is feature is not provided, it would continue remaining a 'personal' tool albeit on the sharepoint. Request you to reconsider your stand pl.

    4 expand collapse in the matrix : why are is there such a big missout ? what would be the utility of a matrix with such an important feature. it would be impossible for me to wade through hundereds of rows to arrive at my area of interest. Again request you to reconsider your stand pl.

    5  Filter fields : when i do my analysis and do a lot of filtering etc.. at some point in time one tends to forget the applied filters etc..  Even in the tile trip, while the active tile is highlighted, if i scroll many times, the active tile will be hidden, and i would again have to do a lot of fwd/backwd scrolling to remember which tile i had clicked. would it not be easily possible to keep a stack of all the filters clicked by a user and show them in a window ? it would be such a use ful user interface. the user then could intereact with such a window to select/deselect what he wants and thereby see different views of the data ? Again request you to reconsider your stand pl. it is very important for users.

     

    6  Drag/Drop in Matrix : i am glad you are considering the same. This is very very important. we have been sadly missing this ever since owa was phased out. i was very eagely looking at this in Excel Services also, but here too it is missing. it is very important for the user to be able to do impromptu analysis himself.

    one point i forgot to mention in my initial email : would you be considering some very important alternate ways of showing measures in a matrix/table. For example you have already given count, sum etc... But it is very important for us to see % to total and such other fields which are now practically taken for granted in any matrix report environment thanks to the popularity of pivot table.

     

     

    6 New Points :

    6.1  Would a report developed in crescent share paramaters with a PPS dashboard ? I would design a corporate BI architecture as starting with a PPS dashboard and attached analytical charts/reports. At this point the users realise they have a problem for which they need to do a deepdive. so attached to this page would be a report designed in cresecent . the user would open the same. he would get the same view of data as the parameters would be shared. then using crescent he would quickly do a deep dive to the root cause of the problem. i would then share my concerns with my team members by storing the views and asking them to select a particular view from their pc's. In this way we would all be seeing at the same 'view'. i would see crescent as a part of an overall BI reporting strategy rather than as a stand alone tool. pl correct me if i am wrong.

     

    6.2   change axis :  it is many a times very important to dynamically change the axis. for example if i am do a 'sales analysis', i may have dimensions like product category, geography (continents), application Category, customer vertical etc.. The measure for all remaining the same viz Sales Amount. now i want to see the sames by product category, or by application category or by customer vertical. This changing is at present not possible at run time. i feel that in any graph you should provide a functionality of changing the axis, otherwise too many graphs have to be created and they eat up the screen real estate.

    6.3  drill down graphs : i should be able to drill down from my graphs, rather than use the graph just as a filter point. for example if i am seeing continents in my graph, i shold be able to click on the same which activates a filter, and then the user should be able to select the next drill down level viz country etc.. this enables me to quickly locate problem areas in my analysis

    6.4  discretisation : one very important analytical method, for which we normally have to do loads of work on the backend, is discretisation or 'bucketing' of data as we call it. to make sense out of thousands and millions of rows, discretisation is just too critical. for exampl if i have sold at different 'gross' margins , i should be able to discretise my sales into buckets like 0-10%, 10-20% etc.. these buckets should be user definable. in a typical sales analysis report i do this 'bucketing' at the back end, and i see my sales by these buckets and i also see the sales mix. this gives me a very powerful view of sales. As 'adhocism' is the key word for crescent, i feel this analytical method is very important. kindly consider this.

    6.5  comparison : another very important analytical method is comparing two values of the same dimension, for example compare sales of 2011 with sales of 2010 at a (say) customer  level. it is not sufficient to just place the two values side by side, which can be done in a matrix. but i would like to create the difference between the two, and then discretise the difference. what this fetches me are important business info like as compared to last year which customer has given me more business, which customer has given me less business, which are my new customers, and which customers i have lost. i feel such analytical methods should be built on the technical platform. this will make the tool more usable by business analysis like me.

     

    i feel that even if it takes a little more time, when Crescent finally comes out it should match or exceed the capaibilities of products like tabluaeu, qlikview etc.. In its current avatar there is a lot to be done.

    hope you take my suggestions/criticism in a positive spirit. as i delve deepr into crescent, i shall post more such observations.

     

    kind regards

     

    sanjay shah

    sanjay@prosysinfotech.com

     

     

     

     

     

     

     

    Thursday, September 8, 2011 5:34 AM
  •  

    Hi Sanjay,

    Couple of high level points.

    1. We do appreciate the feedback!  We absolutely take them in a positive spirit.  Your scenarios are certainly valid, and these types of in depth conversations around functionality are great.
    2. Crescent is not designed to be a 'high end' analytical application.  There is no perfect line between 'high end' and not 'high end', and certainly some features that you catalog below don't have to fall under that category.
    3. Crescent is a Version 1 product.  We simply cannot produce every feature that everyone wants and still ship the product.  That will actually be true of every release.  We have requests from customers for all aspects of the Reporting Services product line and every release we have to make trade-offs about what we can plan and build for the release.  What you see in Crescent for V1, is a very much the outcome of those types of decisions.
    4. Please do log your suggestions on Connect. The product team has direct visibility into these suggestions and we can certainly analyze them better versus them existing as free text in the format of forum threads.  We review Connect items on a continual basis.  

    Hi Sean Boon, Robert Bruckner :

     

    Thanks for your feedback. Following are my comments / further observations :

     

    1  selection within table matrix : I understand that currently this is a limitation. but the lack of it makes the product look half baked. You would of course know that in an 'industrial strength' application of this we would be analysing areas like Inventory Analysis, Accounts Receivable Analysis, Project Profitability...where there are hundreds and thousands of items. so a cross connection between all the objects of the dashboard is very important, especially when the real purpose of crescent is 'adhoc' analysis directly by the user.

    [Sean] - Yes, we do understand the value of being able to have table selection and filtering.  There is definitely more coming as it relates to filtering and we'll discuss those features when we are ready, but actual cell level selection driving filtering operations in other data regions is not something that is currently planned for the first release.

    2  while tiling and other features of selection appear attractive, in real life it is unlikely to cut much ice with professional business analysts like me. the reason is that there are too many discrete values and such selection features may actually block a lot of real estate. Pardon me for saying this, but i feel there are important analytical features that must be considered in version 1 to make it a very big sucess. one important feature which is lacking is the ability to search. since the full data is available in the background, i am sure that providing a search button which will quickly traverse all the fact/dimension values and show up matching results should not be difficult, considering the huge experience MS already has in this area.

    [Sean] - Are you referring to searching the tile container (the field that is defined in the Tile by field)?  This is another request that we are certainly aware of.  Again, we do have some additional features coming beyond CTP3 that may address the scenarios you are looking at.  I'm certainly interested to understand what your business question or goal is in your current work with the tiles.  

    3  your answer re. saving multiple views is "you cannot save multiple views". I do appreciate this but please note that when users actually use such reports in a collaborative way, there is a very big requirement of sharing each others' analytical views. When i am doing my analysis, i may filter here, filter there and reach an 'aha' conclusion. now i want to share it with my boss. how do i do this ? Since the product is already server based, i think it is important that the 'view' is saved and shared. how else would you achieve 'collaboration' ? if this is feature is not provided, it would continue remaining a 'personal' tool albeit on the sharepoint. Request you to reconsider your stand pl.

    [Sean] - You will certainly be able to save your report and view it in presentation mode.  That is certainly a scenario we are targeting.  If the user viewing the report would like to make modifications, they can jump into edit mode and start editing.  They may or may not have the right to save over the top of the existing report.  What we don't have is the notion of a "per user" version of the report that you get to persist.  

    4 expand collapse in the matrix : why are is there such a big missout ? what would be the utility of a matrix with such an important feature. it would be impossible for me to wade through hundereds of rows to arrive at my area of interest. Again request you to reconsider your stand pl.

    [Sean] There are ways to filter the contents of the matrix, such as selection on column/bar charts in other data regions.  There other ways to filter the contents of a matrix to a more manageable set of records that will be forthcoming.  As far as pure expand\collapse goes in the matrix itself, I would encourage you to log that request on Connect.  My guess is it is already there :), and if so, you can just add your vote to it.  We are certainly not discounting the scenario. 

    5  Filter fields : when i do my analysis and do a lot of filtering etc.. at some point in time one tends to forget the applied filters etc..  Even in the tile trip, while the active tile is highlighted, if i scroll many times, the active tile will be hidden, and i would again have to do a lot of fwd/backwd scrolling to remember which tile i had clicked. would it not be easily possible to keep a stack of all the filters clicked by a user and show them in a window ? it would be such a use ful user interface. the user then could intereact with such a window to select/deselect what he wants and thereby see different views of the data ? Again request you to reconsider your stand pl. it is very important for users.

    [Sean] - There may be more in this area as we get closer to release.  I would encourage you to log an item in Connect for this functionality as well.  

    6  Drag/Drop in Matrix : i am glad you are considering the same. This is very very important. we have been sadly missing this ever since owa was phased out. i was very eagely looking at this in Excel Services also, but here too it is missing. it is very important for the user to be able to do impromptu analysis himself.

    [Sean] - We aren't going to allow you to modify the data regions at run-time.  That said, you can always jump into edit mode and make the necessary edits (add and remove fields).  CTP3 already supports this.  If you don't want to save those changes, you don't need to.

    one point i forgot to mention in my initial email : would you be considering some very important alternate ways of showing measures in a matrix/table. For example you have already given count, sum etc... But it is very important for us to see % to total and such other fields which are now practically taken for granted in any matrix report environment thanks to the popularity of pivot table.

    [Sean] -  Some of the calculation scenarios are going to require you add the calcs to the model in PowerPivot or in the AS Tabular model.  % of Total will likely be one of those.  Again, an entry in Connect here makes sense.

    6 New Points :

    6.1  Would a report developed in crescent share paramaters with a PPS dashboard ? I would design a corporate BI architecture as starting with a PPS dashboard and attached analytical charts/reports. At this point the users realise they have a problem for which they need to do a deepdive. so attached to this page would be a report designed in cresecent . the user would open the same. he would get the same view of data as the parameters would be shared. then using crescent he would quickly do a deep dive to the root cause of the problem. i would then share my concerns with my team members by storing the views and asking them to select a particular view from their pc's. In this way we would all be seeing at the same 'view'. i would see crescent as a part of an overall BI reporting strategy rather than as a stand alone tool. pl correct me if i am wrong.

    [Sean] - As Crescent does surface in Sharepoint, we will be actively looking at ensuring that when used in the context of all of our BI products, that they work together well.  

    6.2   change axis :  it is many a times very important to dynamically change the axis. for example if i am do a 'sales analysis', i may have dimensions like product category, geography (continents), application Category, customer vertical etc.. The measure for all remaining the same viz Sales Amount. now i want to see the sames by product category, or by application category or by customer vertical. This changing is at present not possible at run time. i feel that in any graph you should provide a functionality of changing the axis, otherwise too many graphs have to be created and they eat up the screen real estate.

    [Sean] - Right now you can jump into edit mode and make such modifications. The line between editing and run-time in Crescent is pretty blurry.  yes, there is a single artificant that gets saved, but we don't block people from being able get into edit mode to make such changes.  they may not be able to save on top of the existing report though.

    6.3  drill down graphs : i should be able to drill down from my graphs, rather than use the graph just as a filter point. for example if i am seeing continents in my graph, i shold be able to click on the same which activates a filter, and then the user should be able to select the next drill down level viz country etc.. this enables me to quickly locate problem areas in my analysis

    [Sean] - Great feature.  Certainly agree.  I would recommend logging this in Connect as well.

    6.4  discretisation : one very important analytical method, for which we normally have to do loads of work on the backend, is discretisation or 'bucketing' of data as we call it. to make sense out of thousands and millions of rows, discretisation is just too critical. for exampl if i have sold at different 'gross' margins , i should be able to discretise my sales into buckets like 0-10%, 10-20% etc.. these buckets should be user definable. in a typical sales analysis report i do this 'bucketing' at the back end, and i see my sales by these buckets and i also see the sales mix. this gives me a very powerful view of sales. As 'adhocism' is the key word for crescent, i feel this analytical method is very important. kindly consider this.

    [Sean] - DAX enables these types of scenarios. They need to be defined in the model or PowerPivot workbook and they can take the current context of what is selected, etc. in order to recalculate.  If you want the user to be able to define such buckets in Crescent, I would file an item in Connect.  I definitely agree on the value of the scenario.

    6.5  comparison : another very important analytical method is comparing two values of the same dimension, for example compare sales of 2011 with sales of 2010 at a (say) customer  level. it is not sufficient to just place the two values side by side, which can be done in a matrix. but i would like to create the difference between the two, and then discretise the difference. what this fetches me are important business info like as compared to last year which customer has given me more business, which customer has given me less business, which are my new customers, and which customers i have lost. i feel such analytical methods should be built on the technical platform. this will make the tool more usable by business analysis like me.

    [Sean] - DAX allows you to define calculations like this for the model.  You can certainly build a this year - last year type of calculation in that way.  Again, you need to do this in the PowerPivot or AS Tabular model.  If you want to be able to define these in Crescent directly 

    i feel that even if it takes a little more time, when Crescent finally comes out it should match or exceed the capaibilities of products like tabluaeu, qlikview etc.. In its current avatar there is a lot to be done.

    hope you take my suggestions/criticism in a positive spirit. as i delve deepr into crescent, i shall post more such observations.

     

    kind regards

     

    sanjay shah

    sanjay@prosysinfotech.com

     

     

     

     

     

     

     



    Program Manager, SQL Server Reporting Services This posting is provided "AS IS" with no warranties, and confers no rights. http://blogs.msdn.com/seanboon http://www.twitter.com/ssrs
    Thursday, September 8, 2011 9:59 PM
  • Hi Sean :

     

    Thanks for your detailed reply. I shall certainly post my points on Connect. However following are a few more comments / suggestions :

     

    I appreciate the frankness with which you have replied. I also understand and appreciate that there are certain limitations under which the MS teams work. However as a MS-BI partner (my company Prosys has executed over 100 projects using MS-BI) and as a business analyst and a developer myself, I am using this platform to communicate to MS the important points which we feel are missing in the MS-BI platform.

     

    1   Response on the above points :

         1.1  While it is true that the 'edit' button is easily accessible in crescent, but IMHO however simple the development platform is made, there will always be developers and there will always be users.  Users prefer a lot of flexibility in the 'front end'. The example I gave above of enabling the user to change the axis on the graphs on the front end itself is a case in the point. You mentioned that the user can easily click the edit button and make the change. But generally the users will never do that. In fact the 'edit' button should be accessible based on AD permissions which should be definable somewhere in the crescent administration.

    Now what should be given in the front end and what should be given in the development platform is always a judgemental decision, but I would suggest you also take a look at competing products to arrive at this decision. one product i recently reviewed is Qlikview. When i worked with Excel Pivot graphs i always used to feel that why is it not possible for me to change the axis without having to go into the design mode. so when the qlikview consultant came to show me qlikview i asked him whether this was possible in QV, and hey presto, it was possible and very easily done ! there are a lot of such small features in QV which make the users life extremely easy.

    i would also like to share with you here a small addin we made at our company. In Excel Pivot table (front end) there is no way in which the user can be prevented from dropping a field at the column or row level. Sometimes users make mistakes like dropping a field like Product Code in the column area. As there are a huge number of products, it makes the matrix unreadable and messy. This property is available in the backend. we created an addin and brought in this property to the front end. You should have seen the smile on the face of our users as he could now fine tune his reports and make them more 'foolproof' in their usage.

     

    Ditto with the drag/drop in the grid.

     

     

         1.2  Search Functionality : My comment on the Search functionality is not wrt tile. it is a generic comment. As has been mentioned in the published literature of crescent, since at the backend there is an 'association' available so it is easy to click on one element of a graph and get all other elements filtered out. using this feature would it be possible for you to work on a 'search' functionality. How i would imagine it to be is that there is a textbox where the user types in a search text e.g 'Apple'.  The system will do a narrowing down search on all the field labels irrespective of dimensions and give him the matching values to select. the user selects the appropriate text and on doing that the system applies this as a filter.  the advantage of this is the user need not know in which dimensions the search text exists. he just searches intuitively...

     

     

    2.0    Some new observations :

           2.1   I have mentioned to you about the feature of saving a 'view' and then sharing it with others.  A related functionality here is to be able to save a 'comment'. How is the end user using the reports which we develop. For example CEO ( on seeing the sales dashboard for for April 2011) would say 'Hey SBU - A is dropping in sales, i need to ask the reasons..'. At that point he should be able to open a window and store a comment. the head of SBU A should be able to see the comments when he sees the report for April. Maybe in the comment window there would be a button to email the comment to SBU A. a very limited functionality is available in PPS dashboards, but it leaves a lot to be desired.

     

           2.2   Top nn : the way top nn is handled in many of MS tools is very good and very user friendly. But an important miss out is the ability to present the items not appearing in the top nn. For example if I want to see top 5% of my customers i should be able to see this, and the other customers should get automatically grouped into 'Others'. we were able to achieve this in SSRS using some workaround. But in QV it is available right on the front end. This is a very important feature for business analysts like me.

     

    3.0   Some general observations :

    Two of the biggest problems we face in the field when we work with MS-BI are :

    -   BI Front End

    -   Connectivity to SAP

    Let me explain :

     

    BI Front End

    As far as web-BI is concerned, there is PPS, SSRS, Excel services and now Crescent. As far as thick client is concerned there is Excel Pivot Table/Graphs and Excel PowerPivot. Each one has its pros and cons. The moot point is that why are there so many clients ? For example why does not PPS itself morph into Crescent ? I do understand that there are competing product teams inside MS, but if you put yourself in the shoes of a consulting partner, the result is quite confusing. we dont really know what is going where. Should there not be one single integrated platform for BI frontending ?

    Version 1 of Power Pivot was very good, but we could not use it for enterprise BI because it did not have security, hierarchies etc.. Now these are being built in which is really great. The biggest benefit i got out of PowerPivot was it enabled 'portable BI' for me. I have some customers who absolutely wanted to carry their BI with them. In india network access is not very good from all the nooks and corners of the country. i have a customer who had officess accross the length and breadth of the country and their salesmen wanted to access BI wherever they went. the normal excel pivot tables become too large in size, and so PowerPivot fitted the bill perfectly.  However PowerPivot and Excel graphs are not interactive or support drill downs.

    so i always keep wondering why is it not possible to make a BI front end which contains all the best features of all the tools and make this tool available seamlessly either on a thick client or on the web ? I am very positive that MS has all the experience, all the people and all the resources to make this happen quickly.

     

    Connectivity to SAP

    i dont know whether this is the right forum for this, but in the field connectivity to SAP is a very big bone of contention. Since SAP is the biggest enterprise software in the market, our biggest potential BI market are the SAP users. The first question we are asked is how will you get data out of SAP ? Many other softwares esp like QV have 'connectors' to SAP. MS does not have one out of the box (correct me if I am wrong).

     

    Thank you for a patient hearing. i shall post all the points in connect as well.

    however i do look fwd to hearing from you.

     

    Kind regards

     

    Sanjay shah

    sanjay@prosysinfotech.com

     

     

     

     

     

     

     

     

     

     

     

     

     

     

     

     

    Monday, September 12, 2011 5:59 AM
  • Hi Sean

     

    I logged on to connect.microsoft.com. It asked me whether i wish to report suggestions or bugs. In either of the cases Project Crescent was not listed.

    Am I missing something ? Would it be possible for you to send me a link pl.

     

    Sanjay

    sanjay@prosysinfotech.com

    Wednesday, September 14, 2011 10:59 AM
  • Crescent is part of SQL Server / SQL Server "Denali".

     

    HTH,
    Robert


    Robert Bruckner   http://blogs.msdn.com/robertbruckner
    This posting is provided "AS IS" with no warranties, and confers no rights.
    Thursday, September 15, 2011 3:44 AM
  • FYI, please check out the list of new capabilities in Denali RC0 here.

    Thanks,
    Robert


    Robert Bruckner   http://blogs.msdn.com/robertbruckner
    This posting is provided "AS IS" with no warranties, and confers no rights.
    Saturday, November 19, 2011 4:07 AM
  • Thanks, Robert.

    I shall certainly install RC0 and post my comments on the same.

     

    Kind Regards

    Sanjay

    Monday, November 21, 2011 11:22 AM
  • Hi Robert, following are my observations :

    1                     Overall improvements are very good

    a.       scrolling of the graph when the number of elements are more than what the size can accommodate is a great feature

    b.      automatic formatting of the series legends etc is very good.

    2                     table

    a.       go top, go bottom etc required

    b.      when a table row is selected, mechanism of filtering other graphs based on that

    c.       should be possible to select top/bottom nn or nn% esp on value columns.       

    3                     multiple graphs

    a.       when we click on an element the graphs filter out. But we then don’t know on which graph we had set the filter.

    b.      Some mechanism to know where we had applied the filter.

    c.       Mechanism of removing all filters is a bit confusing. Only if we click on the plot area the filters get removed.

    4                     Filter window

    a.       Mechanism of removing all filters from all windows

    b.      Delete filter should not be next to clear filter. The names are confusing.

    c.       There is some confusion on the filter window. Sometimes fields appear to go away when the filter icon on individual graphs is clicked.

    d.      I think when filter icon of individual graph is clicked there should be a separate window for the same, instead of sharing the same space.

    5                     Views

    a.       Naming of a view

    b.      It should be possible to save views in the normal mode (i.e. not in edit mode). The end user should be able to save a view and then email the link to some users.

    c.       Annotate a view. There should be a mechanism of commenting on views by each person. Since every view will be unique this should be possible.

    6                     Search in filter area

    a.       Search should be independent of the field. For example if I search ‘abc’ it should search in all fields and show me wherever abc appears.

    b.      That way I don’t have to remember which field I need to search. Becomes more like a browser search

    7                     Formatting

    a.       Formatting of background colors, selection of colors for the graphs etc.. does not seem to be possible.

    8                     Graphs

    a.       Animation should be available on all graphs. It is a good  visual treat to see the graph slowly coming to life.

    b.      Without going into the design mode I should be able to change the series. For example if I have one measure say ‘fp_ratio’ and several dimensions like year, month, factory, material_group etc.. I will create two / two important graphs say of month and factory. But if I want to check this over other dimensions, I will have to keep changing the report in edit mode. While all that is really required is only to change the series in the graph. That way we can make the charting much more dynamic.

    c.       Top nn : sometimes there are too many series and they are not very material. We should be able to filter out the top nn (either by value or by % etc). the ones which get left out should be grouped as ‘others’ so that the control remains the same. To give an example, if I am seeing all product wise sales, I should be able to select the top 10 products, and the balance should get grouped into others. This should be available on all elements of the view.

    d.      Other graphing and visual components like gauges etc should be available

    e.      Coloring of graphs should be controllable by the users.

    9                     Reports

    a.       It should be possible to add user based comments on each report and each view.

    b.      There should be a shortcut to send the link of a report on email.

    10                 Help

    a.       A generic help on how to navigate should pop up and a help icon should be available

    11                 SharePoint

    a.       Is there a web part in which this will be stored ? or will this be a link in the document library itself.

    b.      Will it support common parameters ?

    12                 Canvas

    a.       The reporting canvas should be bigger and scrollable

    b.      It should be possible to have multiple canvas in a tabbed mode

    c.       It should be possible to share parameters across multiple canvas.

    d.      It should be possible to link a powerpivot report to a powerview report via parameters

    13                 Linkages to other reporting components

    a.       Pps, powerpivot, powerview etc. should talk to each other

    Friday, January 6, 2012 11:08 AM