VOTable error response is DALI misaligned

[molinaro-m]

While WD-ConeSearch-1.1-20200828 tried to align with DALI also on error responses, REC-ConeSearch-1.03 still speaks of

• ID="Error" name="Error"

in INFO element for VOTable error response (w.r.t. name="QUERY_STATUS" value="ERROR").

Server/client implementations need be checked to see whether this change can be applied as minor revision to 1.03.

[molinaro-m]

Personally, I think that this change, even if implemented server side and accepted client side, could be considered minor but will be happy to be proven wrong.

[msdemlei]

On Fri, Aug 04, 2023 at 08:35:24AM +0000, Marco Molinaro wrote:

Personally, I think that this change, even if implemented server side and accepted client side, could be considered minor but will be happy to be proven wrong.

You'd have my blessing, but fankly, in implementation SCS's funky error messaging has been a relatively minor pain (for me; I'll admit I did wonder why SIAP and SSAP chose to do it differently, though, and hence we'd look better if we managed to align things). If the DALIfication of SCS diagnostics turns out to be trouble in any way (let me know and I'll try a few clients I can get my hand on) I think it's fine if we said "SCS predates DALI by a decade, and fixing this would break clients, so we chose to be DALI-incompatible in SCS 1".

[molinaro-m]

On Fri, Aug 04, 2023 at 08:35:24AM +0000, Marco Molinaro wrote: Personally, I think that this change, even if implemented server side and accepted client side, could be considered minor but will be happy to be proven wrong. You'd have my blessing, but fankly, in implementation SCS's funky error messaging has been a relatively minor pain (for me; I'll admit I did wonder why SIAP and SSAP chose to do it differently, though, and hence we'd look better if we managed to align things). If the DALIfication of SCS diagnostics turns out to be trouble in any way (let me know and I'll try a few clients I can get my hand on) I think it's fine if we said "SCS predates DALI by a decade, and fixing this would break clients, so we chose to be DALI-incompatible in SCS 1".

Ops... you know what? I wrote minor but I thought major i.e. I agree with you. Sorry for the mess!

[gmantele]

I have not developped any DAL client, but my impression is that many DAL clients are able to deal with different kinds of error, and especially with errors raised in a DALI way. So, I agree: I consider this change as minor. And as @msdemlei said, if it breaks some clients we can just add a sentence explaining the reason of this change: to make it DALI compatible. This sounds as a good reason to me.

[mbtaylor]

I've lost track a bit of who is agreeing/disagreeing with what here, but interestingly I notice that topcat does flag DALI-style errors in Cone Search output but does not (at least didn't, until 10 minutes ago) notice these SCS-1.03-style errors. So for topcat at least, such a modernisation would be beneficial rather than disruptive.

[gmantele]

Just to be more explicit on my agreement: I agree to this modernization as a minor change in SCS-1.03. If @molinaro-m wants to add a sentence (as suggested by @msdemlei ) explaining the possible break of clients, it is also fine with me.