Continuous trace from 2 days appear on 2 lines

[jschaeff]

Hello,

is this output norma ?

╰─ http "https://eida.gein.noa.gr/fdsnws/availability/1/query?net=HI&sta=ARG2&cha=HNE&start=2023-11-22&end=2023-11-23"
#Network Station Location Channel Quality SampleRate Earliest                    Latest                     
HI       ARG2             HNE     D       100.0      2023-11-22T00:00:00.000000Z 2023-11-23T00:00:00.000000Z
HI       ARG2             HNE     D       100.0      2023-11-23T00:00:00.000000Z 2023-11-23T05:34:57.000000Z

I would expect the 2 lines to appear as one continuous trace.

[petrrr]

+1

I also observed this behavior and was wondering how IRIS handles similar riquests:

• services.iris.edu/fdsnws/availability/1/query?network=IU&station=ANMO&channel=BHZ&start=2010-02-23T00:00:00&end=2010-03-24T00:00:00

#Network Station Location Channel Quality SampleRate Earliest Latest
IU ANMO 00 BHZ M 20.0 2010-02-23T00:00:00.000000Z 2010-03-25T00:00:00.000000Z
IU ANMO 10 BHZ M 40.0 2010-02-23T00:00:00.000000Z 2010-03-25T00:00:00.000000Z

Have not checked the specs and if this is explicitly adressed, but we should strive for uniformity.

BTW: Adding mergegaps=0.0 to the query does the job, but shouldn't this be default behaviour?

• https://eida.gein.noa.gr/fdsnws/availability/1/query?net=HI&sta=ARG2&cha=HNE&start=2023-11-22&end=2023-11-23&mergegaps=0.0

[petrrr]

I think issue #2 is actually relevant here as well.

[jschaeff]

FDSN spec does not force any behavior.

But it states: "This ‘text’ output format contains one timespan per line"

And :

So I think it's not consistent to have a different output with mergegaps=0.0 and without.

I don't see the point for the user in splitting continuous traces on different lines. I'm in favor of merging them.

[petrrr]

I would agree with @jschaeff.

So I think it's not consistent to have a different output with mergegaps=0.0 and without.

mergegaps=0.0 is supposed to be the default already, so there should not be any different in adding it or not. I just wanted to add the observation in order to point where the problem might be. It seems some piece of code is not executed w/o putting the mergegaps parameter explicitly.

[jbienkowski]

Hi guys,

Thanks for your investigation, I'm adding it to our todo list. We will come with a fix sometime in January, most probably.

Cheers,
Jarek