Use more durable IDs for IMF_Beta and IMF_Beta3

[aboddie]

We would like to update our documentation from using PandaSDMX to using this package but would like to have a durable id for our source so we do not need to go back and update our documentation and code after launch. It would be more convenient if the fact that these end points are in beta was conveyed in the name of the source rather than the id.

There is no plan to retire SDMXcentral, so this existing IMF source will continue to be valid. The new data portal is just replacing the existing data portal. The existing IMF source should be better described to explain its role.

At this time user_agent needs to be passed in the header, this will change in a week or so. In the meantime the required header information should be part of the source declaration.

I will create a pull request shortly with proposed changes to reflect this.

[khaeru]

Hi @aboddie—thanks for this information and your PR!

From your use of “we” and “our” it seems you work for or with IMF. It's always super helpful and very welcome to have direct input from data providers on how the web services are intended to be used, instead of having to hunt for docs and guess at details. It's also great to hear that sdmx1 would be recommended by the first-party IMF documentation.

Tomorrow I will go through the PR and add any comments or questions that come up so that we can address those, merge, and release the changes.