[Twisted-Python] Twisted Trac filters my messages as Spam when trying to update ticket #8102
Hello, I submitted some patches yesterday, got review from Adi and today I tried to respond. However I'm not able to respond because all my submissions are marked as spam and blocked by SpamBayes with super high probability. I wonder how many other users experience that problem? Can you please fix SpamBayes? I dont think there is anything in my messages that warrants calling it spam. Since I'm not able to update ticket in any way I'm going to post my message here. It's about ticket https://twistedmatrix.com/trac/ticket/8102 my response is following: Thanks for review adiroiban!
I assume that people will try to keep US-ASCII for their method names, so we might want to reject even valind latin1 methods.
For non US-ASCII or non latin1 encoding, I think that we should reject
I agree with that part. All standard HTTP method names are ascii, so I dont see much reason to support non-ascii HTTP methods. They were not supported before this patch (they caused failures), if we fix that and allow latin1 we will allow characters that are outside ascii in method names, e.g. users will be able to have non-standard latin1 method name: 'GET£'. In my opinion it's better to only allow ascii and nothing else. the request much earlier. That is when we first parse the request line, and not when we try to process the request Yes I agree, so patch should go somewhere to twisted.web.http.HTTPChannel.linereceived when we first get method (line 1709) right? Also is it ok we raise 501? With current patch we raise 501 Not Suppoted but maybe 400 Bad Request would be better here?
On Nov 16, 2015, at 12:43 PM, Paweł Miech
wrote: I submitted some patches yesterday, got review from Adi and today I tried to respond. However I'm not able to respond because all my submissions are marked as spam and blocked by SpamBayes with super high probability. I wonder how many other users experience that problem?
Can you please fix SpamBayes? I dont think there is anything in my messages that warrants calling it spam.
I have told the filter that your messages were not spam; apologies, but despite this occasional mishap, it's the best solution we've found so far. Now that it's had some data from you to look at it shouldn't be a problem for you in the future. -glyph
participants (2)
-
Glyph Lefkowitz
-
Paweł Miech