Introducing additional plugin descriptor

John Mazzitelli mazz at redhat.com
Fri Apr 4 12:52:29 UTC 2014


Why aren't we sticking with XML? What does "not fancy" even mean? Is that seriously the only "con" to using XML here?

IMO, we should use XML for this kind of thing unless XML cam't do what we want. "Not fancy" is not a reason to move away from it. "It is not the latest thing all the cool kids are using" shouldn't be a reason we should consider.

What was the reason for not using XML in the first place?

----- Original Message -----
> Hello,
> 
> last week I demoed Canned DynaGroup Expressions provided by plugin [1]. On
> IRC we discusses what file format should we use (currently I wrote it
> using yaml). It's possible that in future we're going to introduce another
> descriptor, so we may want some consistency in file formats.
> 
> We have basically 3 possibilities yaml, json and xml.
> 
> Short summary of pros and cons:
> 
> yaml:
>   + human readable
>   + can be parsed to objects (similar to jackson)
>   - no declarative "yaml-schema" validation possible
> 
> json:
>   + harder to read/write
>   + json-schema
>   - gaYak does not like it :-)
> 
> xml:
>   + consistency (we already use xml - rhq-plugin.xml)
>   + xml-schema
>   + easy to write in schema-aware IDE
>   - not fancy
> 
> I'd personally stick with yaml, because it's readable. I don't miss
> validation using schema, because structure I am using in
> dynagroup-expressions.yaml is simple enough, and I still need to
> additionally validate/evaluate (even if schema would have existed) the
> expression.
> 
> Thanks
> 
> [1] http://youtu.be/pB5Qr4Sys2w
> --
> Libor Zoubek
> _______________________________________________
> rhq-devel mailing list
> rhq-devel at lists.fedorahosted.org
> https://lists.fedorahosted.org/mailman/listinfo/rhq-devel
> 


More information about the rhq-devel mailing list