« first day (1449 days earlier)      last day (2354 days later) » 
00:00 - 11:0011:00 - 00:00

12:00 AM
naa- by angussidney
 
particularly egregious but naa
 
Since it's the OP and they have rep, I decided naa
 
@SmokeDetector n
 
1:17 AM
Restart: API quota is 16762.
 
1:37 AM
merge conflict went away on its own
but I had to push another commit to get Travis to run again
clearly rails is to blame
@angussidney test failed on Travis
very interesting
...
and now that I'm running it again it also fails?
 
Yeah
 
OH
I'm so sorry
When I ran the tests I was on a different branch
I made from a tag
 
LOL
 
> Exit code 4: pytest command line usage error
lemme see which test this is
I have to count the dots, fml
Appears to be test_privileged
let's peep this out
> assert chatcommands.amiprivileged(original_msg=msg) == "\u2573 " + GlobalVars.not_privileged_warning
when I run it from the REPL
why does it exit with 4 tho
pretty sure the mock is just going to fake assigning the value
let me try
hmm
nope
huh
 
1:46 AM
...y'know, this could be an actual bug
 
waves
 
what up
 
right now... looks up ... the ceiling
 
ok it appears it is not a bug
at least not in the command
I just don't know what the test is doing wrong here
 
1:48 AM
sounds like a pytest command exploded
where're you having issues?
 
the test itself fails
for whatever reason pytest doesn't report it and exits with an exit code of 4 but I suspect that's not anything to worry about
something is definitely wrong with the test tho
 
link to the test?
or rather what you're seeing
 
a moment
what command line are you using to run the pytest tests if I may ask
nevermind i know
 
Before that assert, put print chatcommands.amiprivileged(original_msg=msg), "\u2573 " + GlobalVars.not_privileged_warning
Run the test again and you should get a line of out put
I haven't repaired by dev environment yet from my bad drive, so that's the best I can offer at the moment
 
1:56 AM
i am trying to run this in a dev env but it looks like i lost track of my unittests :/
 
@Andy I already did that
it outputs the regular "You are a privileged user"
I tried the command and it works fine though:
in Beep Boop Maggot, 9 mins ago, by Smokey-NG
@quartata ╳ You are not a privileged user. Please see the privileges wiki page for information on what privileges are and what is expected of privileged users.
 
You should have two things printed. The results of the amiprivileged command and the not_privileged_warning
 
I meant I already printed out the output of the function invocation
 
What's that unicode character? It's different than the other two in that block
 
The failure isn't caused by not_privileged_warning being empty if that's what you were concerned about
@Andy it's U+2573
I know, it looks weird
Oct 13 at 15:18, by SmokeDetector
@EricC ╳ You are not a privileged user. Please see the privileges wiki page for information on what privileges are and what is expected of privileged users.
 
1:58 AM
Sorry I broke the tests :/
 
no no it's not your fault
I mean
I should have checked that they work before I merged them
but more importantly if the test fails then something is wrong, that's not something to shy away from
there's about an 80% chance this is some odd Mock corner case
if so it's something we should know
 
Maybe some of the commands are trying to find a value from the msg object that we haven't mocked
 
that's actually possible
we can track that
at least I know we can obtain calls
mock.mock_calls
 
What branch are you using? What version of Python? I just cloned your master and ran pytest. Everything passed...two warnings:
python3 -W default::Warning -m pytest test/home/andy/.virtualenvs/se/lib/python3.5/site-packages/_pytest/assertion/rewrite.py:7: PendingDeprecationWarning: the imp module is deprecated in favour of importlib; see the module's documentation for alternative uses
  import imp
======================================================= test session starts ========================================================
platform linux -- Python 3.5.2, pytest-3.2.3, py-1.4.34, pluggy-0.4.0
rootdir: /home/andy/my_repos/q_detector, inifile:
 
@Andy it fails in master branch with 3.5 on my end, same assertion error
 
2:08 AM
@Andy sure you're on master?
because I don't see test/test_chatcommands in there
 
I don't know what to tell you. This environment is less than ten minutes old. I created a virtualenvironment, cloned quartata's master branch, pip installed the requirements, initialized the submodules and ran the command above
Oh, wait there is a switch to deploy in the readme that I copy/pasted in.
un momento
 
oh you know what
 
E ImportError: No module named 'yaml' - That needs to be fixed. Someone put that in requirements.txt
 
@quartata it's failing because it's saying you are privileged
 
right
@Andy should be in there
> pyyaml
at the end of requirements.txt
@ThomasWard the assert is after we change the ID to 2
 
2:10 AM
@quartata so then your privilege checker is broken
 
ugh. stupid deploy branch. Why is that in the readme?
 
@quartata I know. 1 moment
 
@ThomasWard it's not though
I just tried it
 
show me your code you used to test.
because ^ that suggests otherwise
 
No code. I ran !!/amiprivileged in a room that I wasn't privileged in
 
2:12 AM
i'm not asking you to test in a real world i'm asking you to test in the theoretical.
 
yeah, running the test it complains
 
because ^ that suggests that the test is going to fubar all the time
 
but keep in mind the test isn't passing in a real User object
 
well
then we have to consider the test broken
because your changes broke something.
just saying.
you could always revert just the tests but...
 
Well considering I just merged in those brand-new tests I'd rather make them work
:P
here let me run it with pdb
 
2:15 AM
test/test_chatcommands.py .......✓ You are a privileged user. ╳ You are not a privileged user. Please see [the privileges wiki page](https://charcoal-se.org/smokey/Privileges) for information on what privileges are and what is expected of privileged users.
That is the test output
 
that's what i'm seeing, yes.
 
wait, where's that print coming from?
did I commit something bad
 
I added the line I mentioned above, to the test and reran it
print(chatcommands.amiprivileged(original_msg=msg), "\u2573 " + GlobalVars.not_privileged_warning)
 
oh ok
pdb confirms that the only people that are privileged are 1 and 167070 so it's nothing wrong in the room config
> ('stackexchange.com', 11540): {1, 16070}
stepping through the test rn
hmm
ok
 
and what i'm seeing is the debug terminal output halted and suspended by PyCharm when it fails the test.
confirmed, is_privileged is failing with that specific mock item
 
2:21 AM
I'm almost positive the problem is because Mock maintains its own id field.
experimenting
 
@quartata make it be uid or something then?
id should be considered a protected name by the way
 
Blame Manishearth
OK I don't think that's being problematic:
<Mock name='mock.room' id='4321837968'> 11540 stackexchange.com
(from print(room, room.id, room._client.host))
Although I did change it so that it sets the fields manually rather than using Mock's constructor
since that makes more Mocks
ohhhh
I got it
It's because of the user.is_moderator clause
since we didn't set that to false it's returning a new Mock object
which is truthy
OK, that test now passes but pytest exits with exit code 4 anyways
 
Is that the same issue as before where it just exits halfway through the tests, or does it complete first then exit with 4?
 
exits halfway through
and I know why
 
\o/
 
2:29 AM
exit code 4 is what we use for crashes if I'm not mistaken
we're probably making a call into something that's expecthook'd and it's failing
it's not pytest itself that's exiting
yeah, ok
it's probably breaking on one of the report calls that's supposed to succeed
then it would try to post a message
and that would probably go afoul
commenting that one out for the time being
or at least just the parts where it gets past the initial checks
 
The issue with the current test format is that there isn't any way to test for actual chat output
 
woot, works
I just commented out the valid ones for the time being
the invalid ones passed
ok, one other failure
> assert chatcommands.allspam("http://stackexchange.com/users/me", original_msg=msg) == (
"The specified user has an abnormally high number of accounts. Please consider flagging for moderator "
"attention, otherwise use !!/report on the user's posts individually." # TODO: Update ID
)
fails
it instead returns "That doesn't look like a valid user URL"
 
Woops, I forgot to replace my userID in there
5869449
 
cheeky jon skeet test there btw
ok interesting
so something else failed
a globalvars test
for a pretty cool reason
when you ran allspam it actually blacklisted two users
but the globalvars test checks to see if the blacklisted user array is empty
>       assert len(GlobalVars.blacklisted_users) == 0
E       assert 2 == 0
 
Woops, should have deleted the blacklisted user pickle/array
 
2:41 AM
ok they pass now
although the blacklisted user one was skipped because I had a pickle
 
\o/
 
we'll see if it runs on travis, just pushed
@angussidney You can see the asserts I commented out for now: github.com/Charcoal-SE/SmokeDetector/pull/1135/commits/…
 
One of the allspam tests in there may break at some point, it's relying on a low-quality user with many posts but low rep. If they get upvoted at some point it may break, but I've provided instructions on how to fix the issue if it arises
 
travis taking its sweet time
 
\o/ it works
 
2:48 AM
waiting on the codecov report
went up about 7% if I'm not mistaken
mission_accomplished.jpeg.exe
 
7% from upstream or 7% from last commit?
 
went up 7 percentage points from last commit
 
I find codecov's interface so confusing
Anyway, 7% is a pretty good improvement
 
upstream coverage is 64% so we're still down 11 percentage points for passing
 
We should be able to increase that more if we add tests for the user whitelist as well, they're basically copy-paste from the blacklisted ones
 
2:53 AM
and I need to do chatcommunicate tests
 
Thanks for sorting those tests out
 
3:16 AM
@quartata Why are these commented out? Why not fix the underlying problem?
 
@Andy fixing them will be very difficult. they're trying to actually submit reports
 
Testing just the invalid cases is better than what we have currently
Now this isn't to say I won't try to make them work eventually, just not right now
 
Do they currently try to submit reports?
 
Hm?
The one's that aren't commented out error out before hitting the part where it's passed onto spamhandling (by design). It tests things like invalid URLs, more than 5 posts or in the case of allspam users with more than 100 rep
 
3:24 AM
Why do they pass currently, and with the code change they do not? What is the difference that makes them try to submit now that is currently being worked around?
 
These are all brand new tests
I'm not eliminating existing tests other than the old chatcommunicate tests which aren't compatible
 
ah
 
4:24 AM
[ SmokeDetector | MS ] Bad keyword with a link in answer: What is the user agent string for facebook app from android by Nguyễn Đình Giang on stackoverflow.com
 
sd k
 
@SmokeDetector @AlexanderO'Mara what's up with the TP on this one?
 
4:44 AM
@angussidney Sorry, typo
Did the following f not override it?
 
[ SmokeDetector | MS ] URL-only title, bad keyword in body, bad keyword in title, link at end of body, pattern-matching website in body, +1 more: nutritionofhealth.co.uk/claire-hydrafirm-cream/ by jirewase on askubuntu.com
 
sd k
 
@AlexanderO'Mara strange.... it should have
 
 
1 hour later…
6:05 AM
I just got my first Marshall badge. :)
 
Welcome to the club. :)
 
Nice work!
Next thing you know they'll come flooding in :P
 
Haha
 
[ SmokeDetector | MS ] Bad NS for domain in body, bad keyword in body, blacklisted website in body, link at end of body, pattern-matching website in body: : \\$1 This weight diminish persuading supplement by Fusent eas on drupal.SE
tpu- by angussidney
[ SmokeDetector | MS ] Bad NS for domain in body, bad keyword in body, bad keyword in title, blacklisted website in body, link at end of body, +1 more: It has no signs on slim plus garcinia body by Clave1991 on askubuntu.com
tpu- by angussidney
[ SmokeDetector | MS ] Few unique characters in answer, no whitespace in answer: A verb that gives a very close meaning to stubborn? by das on english.SE
tpu- by angussidney
 
6:53 AM
[ SmokeDetector | MS ] Bad NS for domain in body, bad keyword in body, blacklisted website in body, link at end of body, pattern-matching website in body, +1 more: It softens down your fat from within your body. by Whind19 on superuser.com
tpu- by angussidney
 
7:05 AM
[ SmokeDetector | MS ] Bad NS for domain in body, bad keyword in body, blacklisted website in body, link at end of body, pattern-matching website in body: HCA is the vivacious compound by Fohollea on workplace.SE
tpu- by angussidney
 
7:17 AM
Restart: API quota is 15112.
 
!!/queuestatus
sd del-force
 
[ SmokeDetector | MS ] Manually reported answer: Reducing Video Play Rate by Neeraj Bhatt on superuser.com
[ SmokeDetector | MS ] Bad NS for domain in body, blacklisted website in body, link at end of body, pattern-matching website in body, potentially bad keyword in body, +1 more: that disregarding the way by Hear ring on drupal.SE
@QPaysTaxes That pattern looks like it's already caught by Bad keyword in answer and Bad keyword in body; use !!/watch-force if you really want to do that.
 
@QPaysTaxes Yes. apowersoft 48 from 65 tp
 
[ SmokeDetector | MS ] Bad keyword in body, bad keyword in title, bad keyword in username, blacklisted website in body, link at end of body: Garcinia Fuel MumyBear by Body Slim Garcinia on askubuntu.com
 
7:30 AM
It was posted 2 hours ago. For some reason smokey didn't pick it up.
 
@DavidPostill Member since 2 months with positive reputation on multiple sites...
 
[ SmokeDetector | MS ] Bad NS for domain in body, bad keyword in body, blacklisted website in body, link at end of body, pattern-matching website in body, +1 more: It is a method to stock up healthy by laialbert35 on askubuntu.com
 
@SmokeDetector tpu-
 
7:36 AM
@NisseEngström Yeah, that's why I didn't nuke his account.
 
fp- by DavidPostill
 
@SmokeDetector Who reported this?
 
Rob
@DavidPostill I did
Rants are flaggable, as far as I've seen, as rude/abusive
 
@Rob Why? It's not spam. We don't report naa or rude/abusive manually to smokey. If smokey picks it up then by all means flag it.
 
Rob
If that's the case, I won't report the rude/abusive to smokey anymore. But I'm pretty sure there are regexes specifically to catch such things?
 
7:40 AM
@DavidPostill we do report rude/abusive to smokey
 
Rob
For example: metasmoke.erwaysoftware.com/post/92075 - marked as 'offensive answer'
 
Might be worth starting a GH conversation on that though, I know a few people have been a bit confused whether r/a is manually reportable
 
Fair enough for rude/abusive. But I think that one is naa. How is it rude? Everybody complains about Microsoft :)
 
[ SmokeDetector | MS ] Bad NS for domain in body, bad keyword in body, blacklisted website in body, link at end of body, pattern-matching website in body, +2 more: The fixings in testo rev by Misfor34 on askubuntu.com
tpu- by DavidPostill
 
@DavidPostill TBH I would also flag that one as n
 
@SmokeDetector n
Yeah, i changed my fp to naa
 
Rob
@DavidPostill I've been flagging them as r/a for a while now - haven't had any declined. As I see it, if it's a deliberately an attempt to derail a thread / rant - it's the same as typing gibberish.
 
Dec 6 '16 at 22:23, by ArtOfCode
@bwDraco we're not looking for NAAs
Dec 6 '16 at 22:23, by ArtOfCode
any NAA's we happen to catch are incidental, but we're not trying to catch 'em
 
Rob
Yep - and I disagree that it's NAA - it's abuse in my view :)
 
@Rob Fair enough.
 
7:51 AM
I can see why you may call it R/A
But the reason why I said NAA was due to Hanlon's Razor - it's probably someone who thinks SE is a forum and expressing your frustration on such a change is allowed
 
@SmokeDetector why
 
@DavidPostill Body - Position 1258-1269: millionaire
fp- by DavidPostill
tpu by Zanna on Garcinia Fuel MumyBear [MS]
tpu- by DavidPostill
[ SmokeDetector | MS ] Username similar to website in answer: How to run different website in a same domain with HTTP and HTTPS by Mahi Jat on stackoverflow.com
[ SmokeDetector | MS ] Blacklisted website in answer, blacklisted user: how to convert OST file PST file in java? by Shane will on stackoverflow.com
[ SmokeDetector | MS ] Bad keyword in body, bad keyword in title, blacklisted website in body, pattern-matching product name in body, pattern-matching website in body, +2 more: RLX Male Enhancement and omega (sunflower, walnut, Gra by user262542 on apple.SE
tpu- by angussidney
 
8:12 AM
related to the above discussion
0
Q: Should rant posts be considered against the "Be Nice" policy ("Rude/Abusive" flaggable)?

Thomas WardThis was brought up in another Meta thread about a disagreement over the use of Rude/Abusive flags on rants. In so much as the link for the Rude/Abusive flag points to the Be Nice policy. In my experience, rants are not respectful and ultimately should be considered violations of the Be Nice po...

 
[ SmokeDetector | MS ] Blacklisted website in answer, blacklisted user: Features Redemption use to work with OST files by Shane will on stackoverflow.com
tpu- by Makyen
 
@Zanna there's a good answer on that one
 
yeah I agree with Eliah Kagan
 
[ SmokeDetector | MS ] Bad NS for domain in body, blacklisted website in body, link at end of body, pattern-matching website in body, potentially bad keyword in body: In the long run I can work harder by Fountop on workplace.SE
tpu- by DavidPostill
 
8:44 AM
[ SmokeDetector | MS ] Bad NS for domain in body, bad keyword in body, blacklisted website in body, link at end of body, pattern-matching website in body, +3 more: This is a safe testo rev by Onner19 on superuser.com
[ SmokeDetector | MS ] Offensive answer detected: Counting occurrences of a word in a file by You dad on codereview.SE
tpu by Zanna
tpu by Zanna on This is a safe testo rev [MS]
 
9:16 AM
tpu- by rene
[ SmokeDetector | MS ] Bad keyword in body, bad keyword in title, bad keyword in username, blacklisted website in body, link at end of body: Garcinia Fuel MumyBear by Body Slim Garcinia on workplace.SE
tpu- by angussidney
[ SmokeDetector | MS ] Blacklisted website in answer: AdMob: Invoke Firebase method getInstance error by Shristy GC on stackoverflow.com
 
[ SmokeDetector | MS ] Manually reported question: REGARDER Thor: Ragnarok Streaming VF Film 2017 1080p by user374433 on meta.SE
[ SmokeDetector | MS ] Bad NS for domain in body, bad keyword in body, bad keyword in title, bad pattern in URL body, blacklisted website in body: Duoslim Forskolin - Directs digestion in the body by user170353 on webapps.SE
 
9:32 AM
!!/watch watchonmovies\.com
 
@rene You don't have code privileges, but I've created PR#1194 for you.
 
thanks dear
 
@SmokeDetector tpu-
 
Merged SmokeDetector #1194.
CI on 26fdb4c succeeded. Message contains 'autopull', pulling...
Restart: API quota is 14400.
 
Anonymous
9:36 AM
@SmokeDetector We all have those days
 
[ SmokeDetector | MS ] Username similar to website in answer: Gravitational time dilation on the surface of a planet by Knowledge Universe Online on physics.SE
tpu- by Mithrandir
 
!!/watch knowledgeuniverseonline\.com
 
@Glorfindel Added knowledgeuniverseonline\.com to watchlist
 
[ SmokeDetector | MS ] Bad NS for domain in body, bad NS for domain in title, bad keyword in body, bad keyword in title, blacklisted website in body, +6 more: www.7supplements.org/body-boost-garcinia/ by Insch1962 on askubuntu.com
tpu- by Glorfindel
CI on 9ff7f38 succeeded. Message contains 'autopull', pulling...
Restart: API quota is 14341.
 
9:47 AM
!!/watch farnian\.com
 
@Mithrandir Added farnian\.com to watchlist
 
(spammed Scifi.SE, wasn't caught as far as I can tell)
 
CI on 01e2b8e succeeded. Message contains 'autopull', pulling...
Restart: API quota is 14328.
 
10:04 AM
[ SmokeDetector | MS ] Bad NS for domain in body, bad keyword in body, blacklisted website in body, link at end of body, pattern-matching website in body, +2 more: on the web. Starting right now, by Dinto haddly on drupal.SE
tpu- by Mithrandir
 
10:15 AM
[ SmokeDetector | MS ] Username similar to website in answer: How to implement responsive web design and its best practices by Animesh Shrivastav on stackoverflow.com
[ SmokeDetector | MS ] Username similar to website in answer: Responsive Web Design by Animesh Shrivastav on stackoverflow.com
[ SmokeDetector | MS ] Username similar to website in answer, blacklisted user: How to run different website in a same domain with HTTP and HTTPS by Mahi Jat on stackoverflow.com
[ SmokeDetector | MS ] Answer similar to existing answer on post: Тесты по JavaScript by Alex Prytkov on ru.stackoverflow.com
fp- by Glorfindel
tpu- by Mithrandir
 
@SmokeDetector How is this similar to the other answer? I don't get it.
@SmokeDetector f
 
10:32 AM
sorry for the joke comment @bummi but I had to make it ...
 
2
Q: How to distinct ham from spam

bummiOn my way to familiarize myself with flagging I sometimes have to risk a decline to explore the limits. On that occasion, I stumbled over a border where I would like to ask for clarification. We use the spam flag, among other things, for excessive advertising (I know that strictly speaking, this ...

 
That is pretty much in context, thanks
 
[ SmokeDetector | MS ] Bad NS for domain in body, bad keyword in body, blacklisted website in body, link at end of body, pattern-matching website in body, +2 more: It is shown that this settling helps in the advance by Whant19 on superuser.com
tpu- by J F
[ SmokeDetector | MS ] Blacklisted website in body: Extracting mobile and working hours from website using regex by honeyboney on stackoverflow.com
fp- by Mithrandir
[ SmokeDetector | MS ] Blacklisted username: Google Assistant SDK to integrate it in an app by PUSSY on stackoverflow.com
naa- by Glorfindel
 
10:48 AM
@SmokeDetector @DavidPostill a moderator looked at that post and deemed the disclosure was sufficient ...
 
@Glorfindel Huh. Well it's NAA then ...
 
[ SmokeDetector | MS ] Few unique characters in body, no whitespace in body, repeating characters in body: How to create a new value our of the number of rows times another value by Stefan Gaillard on stackoverflow.com
tp- by J F
 
@SmokeDetector v
 
sd - v
 
@SmokeDetector n
 
10:52 AM
@rene perfectly fine :), like it
 
OK
 
@DavidPostill yeah, that's probably better.
 
00:00 - 11:0011:00 - 00:00

« first day (1449 days earlier)      last day (2354 days later) »