PS, yes, we've considered lots of different channel name options trying to preserve both the homage to the historical system and keep to three characters.
H1:OMC-A_{channel}_{details} and H1:OMC-B_{channel}_{details},
but it doesn't feel the same and gets ugly when you get into the {channel} {details}, like H1:OMC-A_ASC_QPD_A_PIT.
H1:OMC-OMCA_{channel}_{details} and H1:OMC-OMCB_{channel}_{details}
feels even worse.
And while we'd love to go back in time and make the IMC and OMC less special by changing to
H1:LSC-OMCA_{channel}_{details}, H1:LSC-OMCB_{channel}_{details} and H1:ASC-OMCA_{channel}_{details}, H1:ASC-OMCB_{channel}_{details},
we're of the opinion that too much code that relies on those channels to make the change even for O5, since we hope to still be able to use the BHD system while only controlling the IFO with one OMC in DC readout.
As the IFO gets more complex, and we're running low on three-letter-acronyms that make sense, we'd like the RCG team to consider allowing for more than 3 characters for "subsystems" in channel names. For example, in A+ O5, it makes a lot of sense to convert
H1:OMC-{channel}_{details}
to
H1:OMCA-{channel}_{details} and H1:OMCB-{channel}_{details}
where the channel structure will remain
{IFO}:{SUBSYSTEM}_<channel>_<details>
In discussing this with @ezekiel.dohmen and @david.barker, they suggest, ideally, the RCG would be revamped to build channel names and infrastructure not based on the first 5 characters of the .mdl file name, but instead the {IFO} be parsed from the CDS parameter block, and the {SUBSYSTEM} be parsed from the top_names of the blocks on the top level of the .mdl. The existing "top_names" feature does override that "last three of the first five define the channel subsystem" rule, but as it stands, I think those top_names still need to be three characters.
No urgent time-line on this request, but having this "by the end of O4" would probably be good so we can get a jump on the O5 controls infrastructure. All I'd need at this point is a promise that "yup, we'll definitely do that" soon, since we're in the final design phase of A+ O5 ISC and naming sensors, so we'd like to know that we can handle that.
Erik von Reis (4d3c7ffd) at 28 Feb 22:12
Update changelog for 6.7.2-1+deb11 release
Erik von Reis (e59d8c0e) at 28 Feb 22:12
Update changelog for 6.7.2-1+deb10 release
Erik von Reis (6f5d440f) at 28 Feb 22:12
Update changelog for 6.7.2-1+deb11 release
... and 15 more commits
Erik von Reis (16798697) at 28 Feb 22:08
Update changelog for 6.7.2-1+deb10 release
... and 12 more commits
Erik von Reis (8c9ab2f4) at 28 Feb 22:05
Update changelog for 6.7.2-1+deb12 release
Erik von Reis (149284f8) at 28 Feb 22:05
Be explicit about package directories.
Erik von Reis (54d1def6) at 28 Feb 22:05
Update changelog for 6.7.2-1+deb12 release
... and 12 more commits
Erik von Reis (339ef83a) at 28 Feb 21:56
Add Contec Binary DO 32L and new LIGO DAC to pcid
Erik von Reis (3fb194a9) at 23 Feb 01:36
Switched to debian12 version
Erik von Reis (ed1b64d2) at 23 Feb 01:36
Switched to debian12 version
Erik von Reis (041d6f34) at 23 Feb 01:35
Update changelog for 6.7.1-1+deb11 release
... and 5 more commits
Erik von Reis (13da9816) at 23 Feb 01:33
Update changelog for 6.7.1-1+deb10 release
... and 120 more commits
Ezekiel Dohmen (56538752) at 21 Jul 00:21
Ezekiel Dohmen (a44b2486) at 21 Jul 00:21
Ezekiel Dohmen (3e3076b7) at 21 Jul 00:21