I don't know what's going on inside the scraper. I make a request directly from the recent posts page, and between each page it pauses for 3 seconds. The entire cycle also includes Merit scraping and lasts about 3 minutes. I'm obviously doing something wrong somewhere.
Time for some troubleshooting? Log all scrape-requests with it's date, see what you're scraping.
[2026-10-20T19:39:33.534Z] OK 923ms https://bitcointalk.org/index.php?action=merit;stats=recent
[2026-10-20T19:39:34.174Z] OK 155ms https://bitcointalk.org/index.php?action=recent;start=0
[2026-10-20T19:40:07.845Z] OK 303ms https://bitcointalk.org/index.php?action=recent;start=10
[2026-10-20T19:40:26.249Z] OK 204ms https://bitcointalk.org/index.php?action=recent;start=20
[2026-10-20T19:40:29.418Z] OK 147ms https://bitcointalk.org/index.php?action=recent;start=30
[2026-10-20T19:40:32.599Z] OK 156ms https://bitcointalk.org/index.php?action=recent;start=40
[2026-10-20T19:40:35.793Z] OK 177ms https://bitcointalk.org/index.php?action=recent;start=50
[2026-10-20T19:40:38.975Z] OK 159ms https://bitcointalk.org/index.php?action=recent;start=60
[2026-10-20T19:40:42.149Z] OK 150ms https://bitcointalk.org/index.php?action=recent;start=70
[2026-10-20T19:40:45.334Z] OK 166ms https://bitcointalk.org/index.php?action=recent;start=80
[2026-10-20T19:40:48.521Z] OK 164ms https://bitcointalk.org/index.php?action=recent;start=90
This is what happens during a normal scraping cycle for merit+post.
On average, it takes about 3 minutes, sometimes even 4 or 5, but in the meantime I’ve encountered the problem we’ve already discussed, even though I’m well below 1 request per second. I make about 11–12 requests over the entire cycle, all spaced about 3 seconds apart, except when the ‘merit’ scraping finishes and the ‘posts’ scraping starts – there’s a cooldown of about 30 seconds.
I saw something strange today:
On a
thread, I saw this very strange score for a well-known user:

But of course, when I went into the profile, the result was different, and probably correct.

Going back to the post, I saw the wrong score again.
It was most likely a temporary glitch; I can see the correct score in that thread. Something went wrong within the database at that moment, as the BRDB score doesn’t start at 0.0 but ranges from 1.0 to 10.0; so the fact that 0 appears suggests that the userscript didn’t receive the correct data, though I don’t know why – I’ll have to look into it.