Multi-Object Spectrograph, and may
also enable some tests of on-the-fly
scheduling typical of what the LSST
will need once it becomes opera-
tional.
For proposal submission, obser-
vation preparation, and program
monitoring, we anticipate a set of
interconnected web-based tools to
replace the current large download-
able packages. We will absolutely not
simply take the existing tools and re-
implement them on the web; we in-
tend to take a full step back and have
a clear view of requirements and us-
ability before we even think about a
line of code.
Figure 8.
The new seqexec,
currently under
development, running in
a web browser.
As the Observing Database (ODB) lies at the
heart of it all, we made it an early candidate
for replacement; we are now progressing on
a modern Postgres SQL database design to
replace the current bespoke database, and
a new web-based Sequence Executor (se-
qexec) to go with it (Figure 8). Along with
these changes, we’re developing a new “se-
quence model” (which represents the de-
tailed observing sequence within the OT),
as the current model is overcomplicated and
the source of many maintenance headaches.
We plan to deploy the new database with
the new seqexec in early 2018; this will be
usable with FLAMINGOS-2 and the Gemini
68
GeminiFocus
Therefore, we will be starting a work-
ing group to develop the high-level
user requirements. The membership
of this group will include Gemini staff,
NGO representatives, and members
of the user community. The working
group will review all feedback we
have had on the existing tools, dis-
cuss possible fundamental changes
in approach, and make recommen-
dations for top-level needs and re-
quirements, with examples written
as user stories.
Note that the fundamental change in the un-
derlying infrastructure may also make pos-
sible some other changes, such as enabling
Principal Investigators to request physical
observing conditions (seeing, etc.) “on tar-
get” rather than by conditions percentiles.
What’s the timescale for all this? A little hard
to say given the scale of the work, but we
hope to switch off the old OCS by the end of
2019. Meanwhile, there will be incremental
releases of the various tools and facilities as
they develop.
January 2018 / 2017 Year in Review