Ideas to clean & make faster

[Nigel_Garvey]

@I

I ran them both again earlier this evening and they both took longer ” no doubt because of the Net traffic at that time of day. The vanilla took 12:07 and reported two price drops; Chocoholic’s took 15:02 and didn’t report any drops. I don’t know which was right.

Ah. Both are probably right. Chocolic’s script doesn’t update any records modified less than twenty-four hours previously, so it’s not due to notice any changes for a couple of hours yet (as I write). I’ll try it again later.

I like the fact that sqlite3 can bulk-transfer values from one set of fields to another and can bulk-compare them to pick out any changes. That’s got to be better than the script checking records one-by-one.

An interesting read.

[Nigel_Garvey]

@Nigel Garvey

@I

I ran them both again earlier this evening and they both took longer ” no doubt because of the Net traffic at that time of day. The vanilla took 12:07 and reported two price drops; Chocoholic’s took 15:02 and didn’t report any drops. I don’t know which was right.

Ah. Both are probably right. Chocolic’s script doesn’t update any records modified less than twenty-four hours previously, so it’s not due to notice any changes for a couple of hours yet (as I write). I’ll try it again later.

No. It didn’t notice any price drops this time either ” unless the prices have gone up again!

I think the problem this time is that while “new” prices are only transferred to the Old_price fields when they’re over a day old, they’re overwritten anyway by the data from the Web pages. So if the script’s run less than a day after the previous time ” as when I ran it last night ” the old “new” prices are lost.