Open timroland opened 4 months ago
Thanks for your review, I will make the changes.
Concerning the topics, actually there will be a separate topic for each instrument. We will always call it by: the instrument, followed by the name of the schema (trigger or refined) and potentially another keyword for more specific notices (slewing/not-slewing, catalog, etc.). Do you find it ok ?
I just made some simplification of the schema and examples.
Actually, the previous slewing
and not-slewing
topics transmit the same notices (with a different slew status). So they are merged in the same slew
topic
Thank you @timroland.
This rebase is touching other files, either address the conflicts or create a new PR for further feedbacks.
Yes, sorry, I messed up a bit my rebase. I think I fixed it
Yes, sorry, I messed up a bit my rebase. I think I fixed it
Thank you. Is this ok to review again or you are in the process of adding content?
No it is ok to review it, thanks ! (in parallel with the mission page)
I'm a bit confused about the number of schemas and examples provided. There are three schemas: svom.trigger
, svom.refined
, and svom.mxt.localization
. These schemas should correspond to the number of Kafka topics requested to us.
I cannot find the detector
property in the schema, especially since you are using examples from each instrument (eclairs, eclgrm, and grm). It would be helpful to provide the instrument names as an enum if you plan to use svom.trigger
and svom.refined
as parent schemas.
In the core schema, we have alert_type,
which can be utilized to consolidate the trigger
and refined
topics into one, using initial and updated fields, respectively.
Thanks for the review.
Actually I thought that the name of the schemas did not have to be the same as the topic names. But just need to appear in the topic name...
Actually the trigger
and refined
schemas are quite different because they contain different type of information. The first one contain information about the trigger and the second one contain information about the light curve analysis of the trigger (which comes a bit later in the sequence). Plus, both ECLAIRs and GRM can issued these 2 types of notices with exactly similar fields. That is why I made only 2 schemas instead of 4. But I can make dedicated schemas for each instrument if it is necessary for your system or just if it is cleaner.
Thanks for the review. Actually I thought that the name of the schemas did not have to be the same as the topic names. But just need to appear in the topic name... Actually the
trigger
andrefined
schemas are quite different because they contain different type of information. The first one contain information about the trigger and the second one contain information about the light curve analysis of the trigger (which comes a bit later in the sequence). Plus, both ECLAIRs and GRM can issued these 2 types of notices with exactly similar fields. That is why I made only 2 schemas instead of 4. But I can make dedicated schemas for each instrument if it is necessary for your system or just if it is cleaner.
I think it's good to have one topic for each instrument, with one schema for each. It will be easier for community to subscribe each topic corresponding to instrument of theirs interest. If you wish we can plan a meeting, for suggestions.
I added the 2nd set of notices for ECLAIRs, GRM and ECLGRM (joint analysis). These are defined in the
refined.schema.json
I also added the schema and an example for the 1st set of MXT notices in the dedicated folder.PS: All these notices are refered to level 2 notices in the SVOM nomenclature