Open chad-earthscope opened 4 years ago
I don't have strong feelings on 8 vs 10 chars, but I do think the network code should always be a 'code' or abbreviation, not a name. Too much length adds a burden on implementations and should be supported by a reason. Was any concrete reason given?
8 chars seems sufficient, allowing 4 chars for the code even when the full 4 digit year is included.
If there is a specific use case where 10 would make usage easier or better, then 10 is fine, but otherwise I feel 8 is enough.
The reasons for 10 characters, as I remember it, were:
SEIS1998
and
I agree that we don't want something too big. I could go either way.
My feel for how this will go is that for the forseeible future (years to a decade), all new temporary networks will want to be identified by a 2 char plus year identifier to allow interoperability with existing systems, so the SEISEX experiment in 2025 will probably want XQ2025 to be the network code instead of SEISEX2025.
At some later point, if/when systems that depend on 2 char network codes have been upgraded, then distinction between temp and permanent networks will just go away and the SEISEX2 experiment in 2035 may not even care about the start year part and will request the network code of just SEISEX2.
I kind of feel that 10 chars is kind of entering vanity license plate territory, and probably doesn't serve a real need. I do not see a huge benefit of 10, but don't see a huge downside either to be honest.
No strong feeling either, I would stick with 8 if there isn't a compelling reason for 10
A suggestion to allow larger network codes beyond 8 characters has been posed, such as 10 characters, was posed at the 2019 FDSN meetings. This would allow a full 4-digit year in temporary network codes and more space for description of the deployment.