On 7 December 2015 at 21:05, Sarah Graves <[log in to unmask]> wrote:
> Hi All,
>
> We (at EAO) were thinking it was time to have another Starlink Release.
> Unfortunately this still won't include HDF5, as we still haven't had time to
> test it.
>
> We will need to know which version of the Starjava software to include -- I
> see that TOPCAT/Stilts are in the middle of a release?
>
> We've started a very optimistically titled 2015B release notes page, which
> may well have to be changed to 2016A...
>
> http://starlink.eao.hawaii.edu/starlink/2015B
> http://www.oracdr.org/oracdr/2015BChangeLog
>
> so please do include notes on any changes you've made, or papers you've
> published on Starlink software.
>
> I included all the Starlink applications and libraries for which I could see
> git commits since the 2015A release, and for those with very few commits I
> copied them in or summarised them. I haven't put anything into ORAC-DR yet.
I've never been sure about these release notes. Presumably their
purpose is to give starlink users a summary of the main new
interesting features - interesting to a user that is. But invariably
we seem to fill them with technical details that are probably going to
be of no real interest to users (even if they understood what they
mean). People that want these sorts of details are surely going to
check out the individual package news files. Including them in the
release notes will probably just make most users give up reading at
about line three. I'm thinking of things like: "NDG: Check for
missing KeyMap elements" etc. Can I suggest that we make the release
notes shorter than usual and focus just on user-relevant highlights?
David
> This was unfortunately prompted by the recent discovery that the 2015A
> release did not correctly deal with the July 2015 leap second. As a result,
> it grids JCMT data taken after July wrongly by 15" in RA. We were going to
> try for a fast time frame on this release as a result. If that isn't
> possible (especially given Christmas coming up), I think we'll need to do an
> errata for 2015A instead, but I do think we should still push for a 2016A
> release early in the new year.
>
> (At the moment, the JCMT support scientists are directing affected users to
> use the rsynced versions pending a 2015B/2016A release or an updated 2015A.)
>
> Cheers,
>
> Sarah
>
>
> --
> Dr Sarah Graves
> Scientific Programmer
> East Asian Observatory/JCMT
> [log in to unmask]
|