[Symptoms]: Memory leaks at scan.
[Environment]: Both.
[Workaround]:
1. It seems to be fixed.
2. Use json instead of html.
>>146, pic2.
3. I'll support 4chan's native catalog to scan 8chan from 4chan.
[Details]:
Scanning all boards with HTML leaked memory. But very recently it seems to be fixed, so I can't find any memory leaks in scan now. Maybe you don't have to care this.
Last week I found memory leaks in parsing HTML by "DOMParser.parse" command. 8chan's javascript caught it in "MutationObserver" at line 2105 in main.js, and it caused memory leaks. The same matter was happened in
>>87, because 8chan's javascript acts as a king who thinks there are no script other than the native. In short, 8chan is not user friendly.
Anyway, all we can do now is to use json (in
>>146 pic2) when you feel there are memory leaks. But the json file has much more problems. You can't see thumbnails of webm and youtube, because the file doesn't have any information about those. And more, HW moved some thumbnails to other server, and it's dependent on size of original pic. In short, there are too much inconsistencies in json, so you can't see some thumbnails.
I'll support 4chan's native because of these matters, but it takes time. For now, it seems to be fixed, so the priority is low.