Text error messages

[Markus Demleitner]

Hi again,

On Tue, Jun 30, 2026 at 03:41:48PM +0100, Dave Morris via dal wrote:
> On 2026-06-30 13:58, Markus Demleitner via dal wrote:
> > So... unless someone wants to seriously champion RFC 9457, I'd
> > hope for Pat's DALI-text spec and move on ...
>
> Challenge accepted :-)

Ok, so... what would become possible or noticeably simpler if we did
adopt RFC 9457 into DALI as compared to adopting the SODA errors
<https://ivoa.net/documents/SODA/20170517/REC-SODA-1.0.html#tth_sEc5.2>
into DALI?

And is that benefit really proportional to incurring the load of an
extra library?  Or the effort of parsing the JSON and XML oneself?

> In the spirit of the P3T team. If there is an existing RFC with several
> implementations behind it, we should at least evaluate it before we invent
> our wheel.

The trouble is: We already have (at least) two wheels (VOTable
QUERY_STATUS and SODA/Datalink plain text).  Pulling in another two
wheels (RFC 9457 in XML and JSON) and telling people: "Oh, the other
two wheels suck, and they'll go away in less than 20 years" is
possible but needs a more than a gut feeling to justify.

> In addition to the Python framework that James cited, it looks like there is
> a growing ecosystem around creating and consuming RFC-9457 messages.

"Growing" is a word that usually gives me the creeps when I think
about tech that's supposed to last 20 years.  Few things survive 20
years of growth.

> I think it is worth a look.

Sure.  But after a (superficial, ok) look I've not come up with a
credible scenario in which the extra complication over SODA-like
text/plain messages is worth it.

Thanks,

            Markus