The above light curve of the object V2492Cyg looks like any of the ones we have shown before. That is true! But the way the light curve was made is very different to all the ones we have show before.
Some context: Just over two years ago the raid system on our server failed and we lost all the data, including the entire database.We did recover a complete backup of it, but basically restarted data collection from fresh. For this reason, the light curve search on the website only returns data that have been added since the 29th of March, 2023.
The old database contained about 313 million photometry measurements. Since the restart we have gathered another 126 million measurements. Even now, if you use the online light curve search, the process of extracting the light curve takes up to 30s. This process was much longer on the original, much larger dataset. This was caused by a less than optimal organisation of the data and database tables. One could argue this has been due to bad planning, but in reality we did not anticipate the project being such a huge success.
In any case, we decided to not just add back the original data to the main database, but to create a new version with much improved structure. In particular, all the photometry will get cross matched against the Gaia catalogue of stars. This has now finally been done for the original, pre-2023 data. The search to extract the light curve now takes of the order of a few 10 milli-seconds. Thus, it is about 1000 times faster.
The next steps now are to re-organise the last two years worth of data. After that, the backbone code of the web-interface needs to be adapted so it works with the new database structure, including the Gaia cross-match, without any changes to the interface the users see. After that, the full dataset will be available online for searches again, which will then be much faster. 🙂