Mikhail Korobov

> The problem is there is a race condition Sorry @whalebot-helmsman, could you please elaborate? Are you saying that there is a bug in telnet extension, and that's the reason...

The warnings in test_engine are addressed in https://github.com/scrapy/scrapy/pull/5561, using pytest.warns.

for redirection loops there is http://doc.scrapy.org/en/latest/topics/settings.html#redirect-max-times

//cc @sibiryakov - this is related to 'canonical URLs' feature you're working on for frontera; how are you solving it there?

Encountered this myself. To clarify: ``` In [1]: from scrapy.utils.request import request_fingerprint In [2]: import scrapy In [3]: req1 = scrapy.Request('http://scrapy.org/foo') In [4]: req2 = scrapy.Request('http://scrapy.org/foo/') In [5]: request_fingerprint(req2) ==...

@lborowczak this is not necessarily an issue with canonicalize_url, it looks correct. The problem is that just using canonicalize_url makes Scrapy drop requests which shouldn't be dropped in case of...

@kingname can you workaround it by using dont_filter=True?

@kingname Yeah, that's a bug, and it should be solved. There are 300+ open issues in the bug tracker, and, as any other large open source project, we should prioritize...

Hey! Unlike request dict conversion, which is used by Scrapy e.g. for the disk queues, here it seems the newly introduced method and function are not used by Scrapy itself....

An idea: we can try solving memory issues by setting higher priority for inline requests. This way Scrapy will prefer to finish current requests in `inline_requests` method first, before starting...