Loving XBRL Software

The XBRL Software Blog is dedicated to mainly technical evolutions in XBRL (including iXBRL, XBRL Specifications, etc.) Editor: Dennis Knochenwefel (Working for 28msec Inc., but expressing his own opinion)

[xbrl-software]

28 shooting #makingof #xbrl #fintech #xbrlife #database #zurich #friends #28msec http://ift.tt/1S7n4s7

[xbrl-software]

What a beautiful day for a business trip to Amsterdam #xbrl #fintech #xbrlife #database #lovethisbusiness #sunrise #beautifulday http://ift.tt/1JLt0yy

[xbrl-software]

Great #fintech event in #zurich at #interxion #banking #xbrl #28msec http://ift.tt/1iv228A

[xbrl-software]

2 Simple Reasons Why .xbrlx Would Be Better Than .xbrl, .xml, or .xsd

What is .xbrlx?

It is simply a zip archive containing XBRL files belonging together. For example, in the SEC case it could contain all of the following files:

Why .xbrlx?

Firstly, to group XBRL files that belong together into a sensible unit. Secondly, to prevent people looking at the bare file contents.

1 Grouping XBRL Files Together

In most cases, a single XBRL Instance Document doesn’t make sense. In addition, you need the schema file and the linkbases. Only if you have all of them, they make sense as a whole.

Accordingly, the SEC Instance Documents contain relative URLs, e.g.: 

As a consequence, the schema file won’t be found if you just send the XBRL Instance Document around. The same applies for the linkbase files referenced in the .xsd Schema file. Hence, it only makes sense to send all of them around - not just one of them.

For this reason, zipping them all together is the most reasonable thing to do. To put some lipstick on, I suggest renaming the .zip file to .xbrlx. With this, everybody has an idea what the file might contain.

2 Hiding the Raw XBRL Files

XBRL is not a simple language, given that it comprises more than 20 technical specifications. On this account alone, it doesn’t make any sense for a non-technical user to look at the actual file contents. Nevertheless, a lot of non-technical users do exactly that. Then, - surprise surprise - they complain about XBRL being too complicated.

In comparison, people would never unzip an Excel (.xlsx) file and look into the raw XML files underneath:

If non-technical users did look into these XML files they would definitely complain about excel being too complicated.

In summary, XBRL files should be hidden within an archive file (for example a .zip file) just like excel hides all the XML files. This zip archive should be renamed to end with “.xblrx”. By doing so, the archive file will indicate what it actually contains. And, non-technical users will not try to unzip it, but rather look for the right XBRL Software to work with it.

Current Developments

I’m noting with appreciation, that the Taxonomy Packages Specification is addressing this issue and it is on its way. Yet, it should rather use the file ending .xbrlx for package files instead of .zip. Overall, it sounds trivial, but - In my opinion - makes a lot of a difference.