Because of how poor the API has performed, with about 10% gateway errors and noticeable random long requests (and the fact that we are already abusing the "one API request per 15 minute"), I put in enforcement of the number of days downloaded when catching up an old SCID also, so that if you haven't used the program for six months or if your SCID ends with 2011 data from previous fail versions, it doesn't go crazy and hammer the server, unless you specifically tell it to get more history at the command line.

If you have already created an SCID with a gap, the program can't go back and fill it in, as data is expected chronologically.

Besides the method two posts above to remove data after a gap (my version of Sierrachart doesn't have that exact option), from the edit menu you can also "export and edit to txt", cut lines off the end, and then "import and load".

The file format is 54 bytes of header + 40 bytes per record, so you can also use a utility like GNU split to experimentally cut off extra new data from the tail of the SCID file if you have received bad data at the end, from the original version failing, or from not seeing my change.