< Back to Blog

Banded Reports are Evil


There are a large number of articles about how Pie Charts are Evil. Now evil is overstating it and there are times pie charts are appropriate. But the vast majority of the time when pie charts are used – they shouldn't be. (I think this is mostly due to people assuming pie charts are the standard way to display percentages.)

The same holds true for banded reports. There are cases where their use is appropriate, but that is a very limited sub-set of where they are used.

So what are banded reports? This is the approach used by Crystal Reports, SSRS, Pentaho, Jasper Reports, Actuate, etc. A report by definition is composed of a report header, a table, and a footer. The table component can have multiple levels of detail, with aggregation at those levels (this is where the term banded comes in).

When this fits the report you want, a banded report is a great solution. If the data comes from a single stored procedure, all you want to display is the data in the single table, and the detail levels and aggregation desired are easily set in the report designer – a banded report designer works well.

But when your needs are different, or more, then this locked in approach becomes a significant problem. Fundamentally banded reports are about limitations. Limitations that can be broached only with great effort and significant compromises.

So when should you look at an alternative approach?

  1. You want to display data outside the single table. Something as simple as placing the total of an invoice in the header rather than at the bottom of the table can be difficult to impossible.
  2. Multiple datasources used in a single report are difficult to impossible, because the table is tied to a specific dataset. For example, if you want a report showing a table of each salesperson's sales that month, followed by charts showing their telephone & email stats, then the same for the next sales rep. This is distinct data from two different databases laid out independent from each other.
  3. The banded table does not provide sub-detail and/or aggregation on the criteria you need. If the designer does not provide the functionality within the table – that's it. No way to get it. The limitations you hit in this approach tend to be absolute.
  4. Report details are broken by row (or column). You cannot iterate over the rows of data returned in any other way. For example if you can have multiple contact names and want to list one per line all within the cell for the contact name – sorry, no.
  5. Most banded report designers can import sub-reports – but only as a distinct completed report. They cannot be imported to become part of the main report being generated, using the live data populating the main report. This is a fundamental architectural limitation because of the concept of 1 table tied directly to a dataset.
  6. Most banded report designers do not support the concept of if or switch statements in the reports (and those that do have it for very limited use). If you wish to include formatted text at the bottom of an invoice based on the country, that can be very difficult.
  7. For some reason a lot of the banded report systems have trouble with paged output (i.e. to PDF). There's no fundamental reason why based on the concept of banding, but it appears that most place everything in the final report using absolute positioning as they build up the report, and that makes it difficult to then reformat for page breaks. Regardless, it's tends to be a problem.
  8. Banded reports strongly encourage you (for good reason) to create a stored procedure for the table in each report. This means any change in the report requiring additional data requires not just a change in the report template, but also in the underlying stored procedure. In many live systems changing a report template requires just minor acceptance testing but a change to the database, even if just a stored procedure, has a much larger effort to validate it before taking it live.
  9. The number of detail levels can be a problem, but most banded report designers allow 7 – 9 levels so this in practice is rarely an issue.

The beauty of free-form (as opposed to banded) report design is you can do anything you want. You select the data you need within the report, you iterate over any data, in any way, across any part of the template. You insert sub-reports as desired and they become part of the template populated with the data you are iterating over. You include charts, gauges, bitmaps, hyperlinks, etc. wherever under any criteria you wish.

With free-form a detail sub-set can be any set of data you wish (even from another database) broken out over any part of the template you wish (not just rows). There is no such thing as a report header, table, and footer, there is just the template where you can place anything anywhere. With free-form, you can structure your report any way you can imagine.

For the cases where the strict structure of a banded report designer is difficult and/or limiting, please take a look at Windward Reports. With Windward you design your reports in Microsoft Word, Excel, or PowerPoint. It provides you a totally freeform report designer, both for layout and formatting. Wicked fast, extremely easy, and very powerful.

So next time you want to tear your hair out over the limitations or difficulity in your banded report designer, take a look at a freeform alternative. It will turn something difficult/impossible into something very easy.

Tags Start & End

Tags Can Start & End Anywhere

Appendix B

.NET code for multi-page image output

Appendix A

Java code for multi-page image output

Data Bin Search

The Data Bin can now be searched to find a table, column, node or other piece of data without scrolling through it all.

Shrink to Fit

This will shrink the contents of a cell until it fits the defined cell size.

Time Zone Conversion

A new Windward macro has been added to help with converting dates and times from UTC time to the local time zone.

Image Output Format

New image output formats added.

PostScript Output Format

PostScript, commonly used with printers and printing companies, has been added as an additional output format.

New and Improved Datasets (Designer, Java Engine, .NET Engine)

Datasets have been re-written from scratch to be more powerful and easier to use.

Stored Procedure Wizard (Designer)

This works for all tag types that are connected to a SQL-based data source (Microsoft SQL Server, Oracle, MySQL, or DB2).

Boolean Conditional Wizard (Designer)

Before, conditional statements could only be written manually. Now they can also be built using our intuitive Wizard interface.

Reorganized Ribbon

The ribbon menus have been re-organized and consolidated to improve the report design workflow.

XPath 2.0 as Data Source

Adds various capabilities such as inequalities,descending sort, joins, and other functions.

SQL Select Debugger

SQL Select  Debugger

  • The look and feel was improved
  • Stored Procedure Wizard
  • Improved Exceptions pane

Tag Editor/Tag Selector

Added a Query tab as a field for typing or pasting in a select statement

  • Color Coding of Keywords
  • TypeAhead
  • Evaluate is now "Preview"

Rename a Datasource

All tags using that Data source will be automatically updated with that name.

Connecting to a Data Source

New single interface to replace 2 separate dialog boxes

Tag Tree

Displays of all the tags in the template, structured as they are placed in the template. This provides a simple & intuitive way to see the structure of your template. Also provides the capability to go to any tag and/or see the properties of any tag.

Added Javelin into the RESTful Engine

Support for Google Application Engine Integration

The ability to integrate the Windward Engine into Google’s cloud computing platform for developing and hosting web applications dubbed Google Applications Engine (GAE).

Additional Refinement for HTML Output

  • Improved indentation for ordered and unordered lists
  • Better handling of template header and footer images
  • Better handling for background images and colors

Redesigned PDF Output Support

This new  integration will allow for processing of complex scripts and bi-directional  text such as Arabic.  Your PDF output  will be much tighter and more closely match your template, and we’ll be able  to respond rapidly to PDF requests and fixes.

PowerPoint Support

Includes support for new ForEach and slide break handling, table header row repeat across slide breaks, and native Microsoft support for charts and images.

Tags are Color Coded

Tags are color coded in the template by type, making it easy to visually identify them.

Increased Performance

Version 13’s core code has been reworked and optimized to offer a reduced memory footprint, faster PDF generation and full documentation of supported features and limitations in the specifications for DOCX, XLSX and PPTX.

Advanced Image Properties

Documents can include advanced Word image properties such as shadows, borders, and styles.

Improved HTML Output

Windward has updated HTML output to reflect changing HTML standards.

Version 13 New Data Sources

Windward now works with a slew of new datasources: MongoDB, JSON, Cassandra, OData, Salesforce.com

Generate Code

The Generate Code tool in the designer allows you to open an existing template and, with a click of a button, automatically create a window with the code needed to run your current template with all data sources and variables. Simply copy this code and paste into your application's code in the appropriate place. You now have Windward integrated into your application.

You only need to do this once. You do not do this for each template. Instead, where it has explicit files for the template and output, change that to parameters you pass to this code. Same for the parameters passed to Windward. This example uses explicit values to show you what to substitute in where.

Pivot Tables Adjusted in Output

Any pivot tables in an XLSX template are carried over to the XLSX output. The ranges in the pivot ranges are adjusted to match the generated output. So your final XLSX will have pivot tables set as expected in the generated file.

This makes creating an XLSX workbook with pivot tables trivial.

Imported Template Can be Set to Match the Parent Styles

In an imported sub-template, if its properties for a style (ex. Normal) differ from the parent template's properties for the style, the use in the sub-template can be set to either use the properties in the sub-template, or the properties in the parent.

You set to retain when you don't want the child template's styling to change when imported. You set to use the parent when you want the styling of the imported template to match the styling in the parent.

Any explicit styling is always retained. This only impacts styling set by styles.

Tags can be Placed in Text Boxes

Tags can be placed in text boxes. Including linked text boxes. This gives you the ability to set the text in a textbox from your data.

Tags can be Placed in Shapes & Smart Art

Tags can be placed in shapes & smart art. This gives you the ability to set the text in a shape from your data.

HTML Output Supports Embedded Images

When generating HTML output, the engine can either write bitmaps as distinct files the generate HTML references, or it can embed the images in the HTML providing a single file for the output.

Footnotes & Endnotes can Have Tags

You can place tags in pretty much any part of a template, including in footnotes & endnotes.

Document Locking Supported in DOCX & XLSX

Any parts of a DOCX or XLSX (PowerPoint does not support this) file that are locked in the template, will be locked the same in the output.

Specify Font Substitution

If a font used in the template does not exist on the server generating a report, the font to substitute can be specified.
In addition, if a glyph to be rendered does not exist in the font specified, you can specify the replacement font. This can be set distinctly for European, Bi-Directional, and Far East fonts.

Process Multiple Datasources Simultaneously

If you need this - it's essential. And if you don't need it, it's irrelevant.

Windward enables you to build a document by applying multiple datasources to the template simultaneously. When Windward is merging the data into a template, it processes the template by handling each tag in order, and each tag pulls from different datasources. (As opposed to processing all of one datasource, then processing the next.)

This allows the select tag to use data from another datasource in its select. For example, if you are pulling customer information from one data source, you can then pull data from the sales datasource using the customer ID of the customer presently processing to pull the sales information for that customer. If you're interested in patching together your data from multiple datasources, read this post on our blog.

David Thielen

President/CEO at Windward Studios

From his early years as a Senior Developer at Microsoft, to legendary designer of the popular Enemy Nations strategy game, to reporting and document generation guru, Dave has never lost his passion for building superb software and teams.

Written by:_
David Thielen
President/CEO at Windward Studios
Windward © 2020 All Rights Reserved.


Got questions about reporting and document generation? We've got answers—let's connect!
Send a note
messaging, phone, or email contact optionsclose out button