A number of software products that output IATI compliant XML files currently include a note containing the name of the product or application that created the file - usually in the format eg:
****** Generated by Aidstream *****************
Therefore I propose that this information which is very useful for any data user is formalised into the IATI standard perhaps by adding a 'generated-by' attribute to the iati-activities element?
It is unclear for me why this would be useful. Why does it matter which system produced the data? What is important in my opinion is that the data are conforming to the standard so that the data can be consumed, irrespecteively of the system producing this data. Adding this element, might encourage creating niches of IATI dialects which would be detrimental to the interoperability of the standard (e.g. this reporting application only works when you provide IATI data produced with system 'XYZ'). The standard should in my opnion be about aid and, not in any way be linked to the underlying technical systems.
I disagree Herman. Firstly it could be useful for consumers of the data to know how different systems render IATI xml. Secondly it is up to the standard in general to ensure that we don't allow niches of dialects.
(And as a pedantic aside, the standard should be about development and humanitarian activities with associated resource flows - not about "aid")
This proposal has been discussed amongst the IATI Technical Team following the end of the initial suggestion phase of the v2.02 upgrade.
The IATI Technical Team do not consider that there is a compelling case for this proposal at this time. However, we invite further discussion and encourage others who support this proposal to post accordingly. At this stage, and unless others come in and argue for this proposal to be included, it is likely that it will not be taken forward.
To add in my two-pennysworth, a generator tag does appear in other technical standards, with HTML being a notable example. I can see both sides of the argument and am personally on the fence about it, leaning towards objecting: It seems to be an excess payload with the effect of eating clock cycles, disk space and transmission packets.
Meta tags are only useful if searching, indexing or special use consumers use them - if anyone can think of an actual use case then I'm happy to be persuaded.
In any case - and as with HTML - if this is included, I would suggest adding in guidance that this value should not be used on pages not generated by software, e.g. pages whose markup was written by a user in a text editor.
My experience is that knowing which files have been published by common tools like CSV2IATI or Aidstream is sometimes useful for investigating problems with the files published with those tools. However, for this purpose, a consistent comment (like these tools already insert) is probably sufficient.
This proposal has again been discussed amongst the IATI Technical Team. As there are mixed views on the usefulness of this proposal, this will not be taken forward as part of the version 2.02 upgrade of the standard.
In the meantime, the addition of a consistent comment immediately after the <iati-activities> or <iati-organisations> elements is suggested, for example Aidstream use: <!-- Generated By AidStream -->
This proposal has been moved back to the 'Modifications, Additions, Improvements' forum, as this may become more relevant in the future if there is increased use of xml generation tools.
8 Comments