Affiliation
American Association of Variable Star Observers (AAVSO)
Sat, 11/26/2016 - 03:49

I want to thank the membership for entrusting me with a Council seat for the upcoming biennium.  One of the planks in my campaign platform was attention to what I have dubbed the "Computing Infrastructure" (CI).  Towards this end, I wanted to begin by gathering comments about our basic data services: WebObs, Light Curve Generator, Variable Star Plotter, etc.: problems or deficiencies that you have identified in the functionality of these and related systems.  I don't want to go into performance issues here (that is yet another can of worms to be opened), and I am leaving out VPhot, VStar, Transform Generator and Applier, and Zapper.  Those applications have active support, and I want to start with the orphan services.  I'll begin with my own "punch list," built up over the past few years.

Tom

 

Variable Star Plotter

If you fire up VSP for the purpose of checking the photometry table for a known chart ID, you must generate the corresponding finder chart first.  The option to produce a photometry table instead of a finder chart is ignored when you specify a chart ID rather than a star.

 

Data Download Tool (ascii) 

The web page instructions refer to "question mark icons" that can be used to get help.  These icons don't exist.  The file returned does not contain observation type in the records (eg: DSLR, CCD). 

 

Light Curve Generator

The tool will allow the user to enter a one-sided magnitude range limit, but only implements such a limit on minimum magnitudes, not maximums.

 

WebObs Search

The tool will reject an observer ID that is not in upper case (no other tool does this).

 

WebObs PEPObs

Tool computes incorrect sidereal times.

 

Inconsistent Front-ends for Data Access

LCG, WebObs Search, and Data Download do not have coherent interfaces.  LCG permits filtering by band, but not by observation type or observer.  WebObs permits filtering by observation type and observer but not by band.  Data Download does not permit filtering by type, observer, or band.  Data download and LCG expect civil date filters to be entered as MM/DD/YYYY, while WebObs expects YYYY-MM-DD.  LCG and WebObs will accept a blank "end date" to include up-to-the-minute observations, but Data Download rejects it.  LCG's interpretation of end dates differs from that of Data Download and WebObs.  The latter applications consider the end date as inclusive, while the former considers it exclusive.  Meanwhile, the AAVSONet photometry interface permits filtering only by Julian date or V magnitude and, strictly speaking, does not allow specification by star name [Yes, you can use a variable star identifier, but that merely invokes a tight cone search around the star's coordinates.  Data for non-variable stars, of which there is plenty, cannot be explicitly retrieved at all].

Affiliation
American Association of Variable Star Observers (AAVSO)
Existing AAVSO Software

Hi Tom

Sounds like a good punch list for a crack programmer having access to the code available for the web interface. 

I got used to it's ideosychronies. If it's a quick fix, it would be good for new users.

I am keen on making the web site faster, but have no idea how to. Glad you are working on things.

Ray

 

Affiliation
American Association of Variable Star Observers (AAVSO)
VSP Web API

Hi, Tom:
Great list.

I've also run into problems with the web access API for VSP. If you use it to fetch photometry for a given Chart ID, you get back obviously incorrect data. (The API for fetching photometry using a star name works great, though!)

Thanks for pulling the list together.

- Mark

Affiliation
American Association of Variable Star Observers (AAVSO)
VSP and Internal Identifiers

I just ran into a VSP bug that segues into what I consider a deficiency in our system for identifying stars within the AAVSO system.  I was trying to get a photometry chart for 32 Cas.  Yes, I know this star is now considered constant - I want to use it as a test target for PEP observers.  VSP does not recognize 32 Cas, even though it is in VSX under that name, and I had to specify, instead, HD 6972 to get a chart.  The chart, however, included HD 6972, itself, as a comparison.  This should clearly not happen.  The larger difficulty was that the chart identified the star as 000-BBC-486, and it took me a while to realize what was going on. To the best of my knowledge, there is no way to cross-reference AUID names of comparisons against any other catalog without pasting the RA/Dec from the chart into SIMBAD and performing a coordinate search.  In the past, I have needed such cross-references on dozens of stars to find comparsion information that was not in VSP.    It was a major headache.

Affiliation
American Association of Variable Star Observers (AAVSO)
vsp

It'd be really neat if VSP could include an option for upping (if that's the right word) the magnitude limit away from the variable, rather like the old-style d and e charts which showed faint comps within a close radius of the variable but only plotted the brighter field stars. Otherwise if you set the mag limit to 15 (say) then every single star above that mag is plotted on the chart, rather than just those you'd be interested in - i.e., those near the variable. I'm thinking of text boxes saying something like

"only show stars above magnitude..." (text box) outside a (text box) "arc minute radius from the variable"

Affiliation
American Association of Variable Star Observers (AAVSO)
Photometry Tables

I have discovered a disconnect between VSP and VSX.  VSP will sometimes include stars the VSX considers variable in lists of potential comparison stars.  It appears that there has not been a cross-check between the two databases.  Now, in the specific cases that came up, it was later determined that the stars were, in fact, constant, but whether this is always the case I think is unknown.  Even if all the VSP comps are truly constant, it would be good to feed that information back into VSX, and so improve its reliability.

Affiliation
American Association of Variable Star Observers (AAVSO)
Variable comparison stars

Most cases are the other way around: there are lots of comparison stars that eventually turned out to be variable and thus are now in VSX.
At the time VSX was launched and the years that followed, we did not have an option to check VSD for potential conflicts. Now we have, we identify if a new variable star is in VSD so we can delete it from the comparison star sequences.
Decades ago we did not have the number of resources to check for variability that we have now and there are lots of sequences that come from the past.

If you happen to find variable comparison stars please let us know.
We are aware of the problem and are always replacing those variable comparison stars for other constant stars of the same magnitude. I have a working list but if the observers report a specific problem of that ilk, that would be my first priority to correct.

In the cases you mention, when a star catalogued as variable was chosen as comp star and then found to be constant, one should check when the sequence was made because maybe the star was selected after its constancy was confirmed.
The Sequence Team and the VSX Team work closely on this matter, there are lots of NSV stars that are not variable so we may well include formerly suspected stars as comp stars. When we do that, we take the chance to classify the star as constant in VSX. Those ones are no different than the rest of the comparison stars.

On the other hand, we do have some small amplitude variable stars in some sequences but in those cases we add a note to the photometry table so they are only used for visual observations.
Sometimes, there are no better choices regarding position or color in a large area around a variable so selecting those stars as comparison stars is better than selecting another star too far away. Those variable comparison stars are usually variable by less than 0.05 mag. and do not show mean magnitude changes (strictily periodic variables, like ELL or small amplitude DSCT stars).

Cheers,
Sebastian

----------------------

Sebastian Otero

VSX Team

American Association of Variable Star Observers

Affiliation
Royal Astronomical Society of Canada (RASC)
Photometry Table Formats

One item I would like to see is the ability to have the photometry table outputted in multiple formats.  Saving HTML files locally for a chart does not work very well.  Saving as a PDF is ok but there's also the old method of using a simple text file. 

Vance