Hello,
I would like to discuss about the limits of submissions to Variable Star Index. Currently, in guidelines it's written as:
The number of submissions per user/group is restricted to one per day.
That's the only one sentence about such limits and the interpretation is clear - you can submit 365/366 variables (and it doesn't really mean 365/366 acceptances by VSX moderators). But this is very confusing for many people and the reality is different.
The first thing, weekends. Based on the limit, we may submit 7 variables per week - a simple calculation. But stars are not reviewed in weekends and they are done once a day, so the queue gets longer, because checking 3 targets from one person in Monday isn't a reality. May we submit in Saturday and Sunday, how does this count? If I submit in Saturday, does it mean I have to wait for Tuesday? If the answer is "it's different, we just accept one of your stars everyday", so why do we get a warning when we submit 5 stars in a single day, if we meant to stop for the next week? That's the same situation. We know that we need to keep remaining stars as "Draft" (nothing about this in Guidelines!), but on every Saturday we think "May I submit today? Won't Sebastian be angry? What if I submit today, but I won't do that on Monday?". We are very aware of the rule, but it's a massive confusion.
One per day, what does it mean "per day". What's the determinant telling you when you can submit another variable. Universal time? Does that mean submitting at 23:55 UT and at 00:05 UT makes no problems? Variables are usually submitted between 15:00 UT and 04:00 UT, so it may look like we submitted two in the same day.
Sometimes people may submit two times in a single day by accident. I had this a few times, when "Draft" was not checked or I simply pressed "Enter" to send at half finished submission. We are aware that it's not a good idea to submit another one next day.
Another thing, revisions. Are these treated like normal submissions too? I believe yes, but without a good statement in Guidelines, we see this as a lost chance of submitting something previously unlisted.
And one more. There are groups of different people searching for their own, but sometimes they might find the same new variable. We are fair about this and one of us submit with two given discoverers (yes, we often show our findings, because it's a great way to learn about the variety of light curves). If person A submitted something by writting "Discoverer: A, B", does that disqualify person B to submit something else on the same day? That's a different situation than working as a group.
The best solution would be, if there was a timer showing how much time you need to wait before submitting again. For example, someone submits a variable at 21:22:00 UT, then he has a timer showing "Wait 02:38:00 before submitting again", and your submission will save automatically as a draft. If someone submits at 00:01:00 UT, then he needs to wait 23:59:00 for another one, because he lost his chance in previous day. This shows a situation of the determinant of "each day" was 00:00:00 UT time. Added content after edition: Or, which is very short and makes it all pretty clear, change to "The number of submissions per user/group is restricted to five in last seven days." (more clear than just "per week"). This rids out the problem with timing, submitting two in the same day by accident (or more on purpose), during weekends, also it doesn't disturb VSX moderators from reviewing once per day, giving a free weekend. Because we know if we submit 5 stars in Saturday, they will be submitted in Monday, Tuesday... to Friday. This also works if we do the same thing in mid of the week, let's say we did in Thursday: first one is in Friday, the second one in Monday etc. And then he is allowed to submit after 168 hours/7 days since 5th earliest submission, in this situation - in next Thursday. That even doesn't look difficult to implement as an automatic script to VSX (eg. an output error "Please wait 167:23:02 before submitting again" when trying to submit a 6th one).
This topic was meant to write an official statement from VSX moderators to get rid of the confusion that we have for last 15 months. We know that VSX moderators are sensitive about the timing of each submission, but after we make this clear, it will be good for us both. A revision of this point under guidelines will be a good idea, because the reality is different.
In the end, yes, I also have "paper" word in my mind. We know that is the best solution. There are many people that prefer submitting, because this is intuitive and easier for a person that doesn't have experience about papers. Also, sometimes there are unique targets where time matters - I would like to submit immediately, because we are monitoring the progress from other surveys, like ASAS-SN itself (eg. came back to the same ASAS-SN link to see "This star was already reported to the AAVSO", while two weeks ago there was nothing and there's still nothing on VSX).
Regards,
Gabriel Murawski
Hi Gabriel,
I think the main problem here is that you can't have (at least) 365 stars to submit and still think VSX is the publishing venue that you should have chosen.
Higher up in the list in our policy statement you can read:
"Observers undertaking systematic searchs of variable stars resulting in a large number of discoveries are encouraged to submit their results to journals like the JAAVSO."
Now I realize that we repeated this below in that same list:
"If you are performing a systematic search of variables resulting in lots of new objects, we encourage you to publish those results in a journal (e.g. JAAVSO or PZP) instead of submitting them individually to VSX."
This is not being taken into account. For instance the JAAVSO has a "Variable Star Data" type of paper that could include tables and light curves. A quick glance and a list import would be enough then if enough information is provided and e.g. 200 stars could be added to VSX at once instead of reviewing one by one (if we were up to date with our journal updates, something we are not because we are using too much time on moderating individual submissions)
In an e-mail sent to all VSX users last year, we wrote:
"Nowadays the number of submissions has grown exponentially and so has the number of objects published in the literature, especially as huge lists of variable stars. It is not possible to devote as much time as before to individual submissions when they do not have the required quaility to be accepted. It's beyond the VSX moderation process to explain data-mining technics or help users with their analysis. We do not have the time and resources to do that at this point since we have to focus our efforts in trying to keep VSX as up to date as possible in the new survey era when there are thousands (even hundreds of thousands) of already published stars waiting to be imported to VSX."
Reality is that even if they have the quaility to be accepted, there are many people submitting objects so moderation takes time. And the most common situation is that there is some feedback with the submitter for multiple reasons, even after the publication of our VSX Manual or the availability of help pages like the FAQ. E.g. difficult in classification, identification errors, misinterpretation of things, fields wrongly filled.
You go into detail about times when submissions are made and what may happen if more than one submission is sent in a single day.
We can simplify everything by saying that "5 stars per week submitted by the same observer will be reviewed."
That is the fact because -as you noted and is to be expected- we do not work on weekends.
Saving objects as drafts instead of submitting them helps us keeping track of the objects submitted by the same observer to avoid reviewing several stars by the same observer. The fact is that you can actually submit several stars in one day if you want and that's the only day of the week you can work on variable stars. But keep in mind that we will review one per day, that is five stars per week submitted by you.
The bottom line is always the same: if you have so many stars to publish, then you are undertaking a systematic search of variable stars. In that case, you are using VSX because it is easier for you than publishing in a journal but you are also causing our scarse resources to be devoted to individually moderating your stars which means less time devoted to literature updates, something that is very important nowadays.
So it is not about me being angry, it is about trying to spend limited time and resources the best possible way.
For every single person, their own stars will be the most important ones, but is that true in a general context? We need to be able to look beyond that. And even then we are delayed when it comes to upate VSX with the latest variable star lists published.
To answer other questions: yes, revisions are treated the same way as submissions. It is not about the type of submission, it is about the time it takes to review them.
About ASAS-SN variables not yet in VSX ot any other variable star that you may find in VizieR or in a publication, this is also contemplated in our policy:
"Check the literature (VizieR/SIMBAD) to confirm your variable star is a new object. We try to keep VSX up to date but there are lots of recent lists/papers that are still not included in VSX and you cannot be credited as the discoverer if the object has already been published (you are welcomed to submit it with the proper reference anyway)."
Some lists have lots of problems (e.g. false alarms, lack of information) so haven't been given a priority to be imported or some of their objects won't be added but that doesn't mean some bona-fide variables have already been found and thus the proper credit should be given.
I hope this answers your questions, we can replace the line "The number of submissions per user/group is restricted to one per day." to "5 stars per week submitted by the same observer will be reviewed." if that is more clear. But the main idea is the same: we don't encourage the use of VSX as a publishing venue when systematic searches of variable stars are being done resulting in hundreds of new objects to publish.
Cheers,
Sebastian