How to scrape MAME best?

vbs

Hi Gentlemen,

what is the best way to scrape MAME ROM sets for you? I tried scraping them with UXS which worked out ok, but the problem is that you do not get proper names for clones. All clones (and the parent) will get the same name.
Then I tried SSelph's Scraper. What is the source to use here? OpenVGDB or thegamesdb? I tried both but I only had 1 match in my 30 games MAME test set.
Then I tried the built-in scraper with thegamesdb and there I had ~10 matches from my 30 games set (still not very good).

So could you give me a hint how to match as much games as possible and also get the clone titles correct? Thanks!

I am currently using ROM set 0.139.

Lyle_JP

@vbs My best results have been by using SSelph's scraper first (with thegamesdb as the source), and then using the ES scraper to fill in holes. But I will say that SSelph's scraper works MUCH better on console games (where it uses the checksum of the binary file) than arcade games (where I believe it relies on filename).

Also (and I may be getting it confused with KODI's scraper here) I think it helps if the file-name has the year of the game's release in parentheses.

BuZz

I believe SSelph's scraper uses mamedb.com which is now down hence the problem. Also thegamesdb.net has been unusably slow recently.

https://github.com/sselph/scraper/issues/115

vbs

@Lyle_JP
But the MAME files have a well defined and unique filename, as far as I know. So hashing would not be needed. You could say the filename itself is the hash.

@BuZz
Oh, thanks for the info!

Seems like this is just a bad time for scraping MAME. I am hoping that it is getting better soon! Summoning Sselph for the rescue! ;)