Run CLDF conversion and check

[LinguList]

@MuffinLinwist, what is the status on teh cldf conversion, are there any errors now?

[LinguList]

Should I check from my side?

[MuffinLinwist]

I have run the command and it gives me this error:
ValueError: tables must have distinct url properties

[MuffinLinwist]

Yes, please. I didn't get this error the other times I run it.

[LinguList]

Please check now, @MuffinLinwist, I made all necessary changes, and it runs now. Please let me know if there are questions. Basically, the strategy was not correct. It would have been better to check with actual cldf dataset conversions, where you can see the difference between add_component and add_column, but also the namespace was quite a problem here, as columns were re-defined, although these components, like a languagetable, come with predefined columns.

But this is all fixed now, and the code is also smaller ;)

[MuffinLinwist]

Thanks, @LinguList , for the modifications! The dataset is so much better now. I have one question:

• It is still not clear to me why adding something as a component or as a column. When I put: data_fnames={"ParameterTable": "parameters.csv", "LanguageTable": "languages.csv"}, and add the LanguageTable as a column, it yields the same result. Why choosing one or another? I thought that components are for data re-used across modules. If that's the case, then why the different coding between the two tables?

[LinguList]

It yields the same results, so you choose whatever you like.

[MuffinLinwist]

Great :)

[FredericBlum]

The CLDF conversion runs smoothly and we can close this issue, right? #10 is independent from this?