My personal addition to this would be (if CORS isn't possible for configuration or security reasons), a simple requestable JSONP wrapper. So, adding "&jsonp=itemData" would wrap the entire thing in:
itemData(
content...
)
This way, we can load the data directly into our pages client-side, allowing for much more convenient data retrieval.
Additionally, it would be brilliant if we could get a single batch file containing the G.E.'s entire price data, either in a .csv or a .json file.
In an old fansite staff forum thread, I checked the validity of such a feature. In order to prevent a huge number of database calls, the file could be generated once per G.E. update. With the current ~4k items in the G.E., if 6 values were given:
ID, name, price, 1-day-change, 30-day-change, 90-day-change, 180-day-change
...then the entire file would be smaller than ~250kB. With JSON, this might go up to ~500kB at most. With gzip transfer compression, this could then go down to 50-100kB. That's smaller than most images on the website, and in addition, it would only require a single request, rather than 4,000 requests - massively reducing overhead. Everyone ends up benefiting.
Either of these two solutions would be brilliant for us, and most likely all other fansites.
03-Jun-2014 13:12:43