« first day (345 days earlier)      last day (2523 days later) » 

3:16 AM
[ SmokeDetector ] All-caps title: POWER PLUS QUESTIONS on math.stackexchange.com
 
 
2 hours later…
4:56 AM
[ SmokeDetector ] Bad keyword detected: weight loss new method on meta.stackexchange.com
 
 
3 hours later…
7:46 AM
[ SmokeDetector ] Bad keyword detected: Weight Loss Based on drupal.stackexchange.com
 
7:59 AM
@Undo There is a new SmokeDetector revision to pull.
 
8:57 AM
Reposting here on suggestion from Shadow wizard
Quick question. Do anyone know if anybody has ever built a constantly running (==hosted somewhere) chat bot? Or are all the bot run on demand by the developer?
 
[ SmokeDetector ] All-caps title: NEED HELP DESIGNING DATABASE FOR DONATING on dba.stackexchange.com
 
9:13 AM
@SPArchaeologist Yes, SmokeDetector runs constantly because it's hosted on Undo's server.
 
10:10 AM
[ SmokeDetector ] SmokeDetector aborted
 
eeek
!!/status
 
@ProgramFOX Running since 9:31:5 UTC
 
Aha, something to fix, it should show 9:31:05.
 
11:20 AM
[ SmokeDetector ] All-caps title: NEED HELP - DESIGNING DATABASE FOR DONATED SYSTEM on dba.stackexchange.com
 
12:08 PM
!!/status
 
@ProgramFOX Running since 11:31:35 UTC
[ SmokeDetector ] Phone number detected: +91-9950211818 Bla ck Magic Mantra For Money on english.stackexchange.com
 
@SmokeDetector true
 
@ProgramFOX Registered as true positive: added title to Bayesian doctype 'bad'.
 
 
2 hours later…
2:58 PM
Wild off-the-wall idea incoming, inspired partly by meta.stackoverflow.com/questions/275354/…:
Could we write a bot that every x minutes takes in new SO answers then runs them through some rules about how it "should" be written?
 
Sounds cool.
 
I'm thinking we could write some simple rules to at least help with grammar, etc. Then we collect data on it, make adjustments, etc.
 
Isn't that a bit what Flare is supposed to do?
 
Not actually apply the edits (maybe we could suggest them?)
@ProgramFOX nah, that was supposed to be purely data-gathering to see if something interesting shows up
 
@Undo ah
 
3:01 PM
I'm thinking we stay away from code, naturally.
 
@Undo We could suggest them, but only after running it some time to see what it does, so it does not get suggest-banned.
 
Then have some kind of a thing to capitalize stuff that really should be capitalized (language names, etc)
 
And the word "I".
 
@ProgramFOX yeah, we'd want to be sure it works perfectly
@ProgramFOX Right. And we'd probably want to know what usually comes before/after the word "I" in code and in normal speech so we can tell if maybe it's unformatted code.
> SIOSocket did not move smooth and have much delay
There must be a way to automatically grammarify this
 
If we would find a way to get the subject, that should be doable.
 
3:04 PM
 
Nice!
 
I don't know if it tells us the plurality of the subject, though.
Language is a terrible thing :P
 
3:11 PM
!!/status
 
@Undo Running since 15:10:55 UTC
 
Huh, nice, Zephyr also detects Blaze's messages:
in Low Quality Posts HQ on Meta Stack Exchange Chat, 2 mins ago, by Zephyr
Flag Request for http://stackoverflow.com/a/26611626 by Blaze the Fiery from Charcoal HQ
 
3:31 PM
Yup
 
 
1 hour later…
4:54 PM
@Undo New SmokeDetector revision to pull.
It now also performs checks on usernames.
 
Oooh neat
 
And you can enable/disable specific regexes for titles and usernames.
@Undo Thanks! :)
 
[ SmokeDetector ] SmokeDetector started at rev 73cf36b (ProgramFOX: Perform checks on username) (owned by Undo)
[ SmokeDetector ] SmokeDetector aborted
 
... welp
 
uh oh
 
4:57 PM
That's not supposed to happen.
 
My connection is crappy, might be a few seconds before I get you an error
!!/alive?
 
[ SmokeDetector ] SmokeDetector started at rev 73cf36b (ProgramFOX: Perform checks on username) (owned by Undo)
[ SmokeDetector ] SmokeDetector aborted
 
<type 'exceptions.KeyError'>
That's it :P
 
okay
 
5:02 PM
[ SmokeDetector ] SmokeDetector aborted
 
launches debugger
Ah, I tried to fetch the wrong value from the web sockets.
I tried owner while it had to be ownerUrl.
 
ah
 
But I actually don't want a URL, so I'll try whether ownerName exists.
No, ownerName does not exist.
 
@ProgramFOX ownerDisplayName
 
5:17 PM
@Undo ah, thanks!
 
[ SmokeDetector ] SmokeDetector aborted
 
^ That was just a run to print all keys on the console.
Posting it here for future reference:
 
I just opened Chrome's network inspector on the realtime page and looked at what comes through
 
[u'siteBaseHostAddress', u'tags', u'url', u'lastActivityDate', u'ownerDisplayName', u'apiSiteParameter', u'bodySummary', u'ownerUrl', u'titleEncodedFancy', u'id']
@Undo That's what I also tried, but as I use Firefox and not Chrome, I failed when trying it in Chrome :P
 
ah ;)
 
5:19 PM
And Firefox doesn't support it.
 
Chrome is nice
 
Firefox is nicer.
 
:/
I tried it once and text looked all weird and blurry
 
oh?
For me, Firefox's rendering is much better than Chrome's.
 
firefox on right
Just seems like text is blurrier
 
5:23 PM
Firefox has better rendering then Chrome for me too.
 
Errors fixed @Undo, you can try pulling again.
 
Yay!
 
The message looks quite weird actually, you can read it wrongly and think that the errors fixed @Undo...
:P
 
[ SmokeDetector ] SmokeDetector started at rev bc153b2 (ProgramFOX: Removed debugging line) (owned by Undo)
 
no crash!
yay!
 
5:25 PM
yay!
 
Today I learn about in-app purchase receipt validation. Oh joy.
This is gonna be fun
 
Do you need it to validate it when people purchase one of your apps?
 
No, this is for in-app purchases. Apple takes care of actual app purchases.
I could run without it, but then the *&(*&@ dishonest jailbreakers would have too much fun
 
Ah, now I looked up "in-app purchase", I know what it means.
 
hello @Undo @ProgramFOX
 
5:28 PM
Hello @AwalGarg
 
oy
 
what's buzzing?
 
Bees, servers, normal things that buzz
 
ohh
and what is "buzzing"?
 
bitcoin miners‌​, if you're lucky.
 
5:31 PM
hehe
 
6:03 PM
[ SmokeDetector ] Bad keyword in username: What was the size of surface of a cereal crop needed per man per year during the Dark Ages in Western Europe? by bob - Master of black magic on history.stackexchange.com
 
^ It works! :)
 
Cool!
It might need some tweaking, but still cool
 
6:24 PM
@Undo There is a new SmokeDetector revision to pull.
 
Just a minute, have to go eat.
 
Infinite Recursion requested to give all commands the same format (ending in question mark or not), so I made the question mark optional.
cya
 
okay back
 
!!/alive
 
6:35 PM
@ProgramFOX Of course
 
6:54 PM
[ SmokeDetector ] Offensive title detected: Como fazer scrapping no Google Scholar? by Karla Sessin Dilascio on pt.stackoverflow.com
 
7:32 PM
[ SmokeDetector ] Offensive title detected: How to add try exception in scrapy spider? by Jake on stackoverflow.com
[ SmokeDetector ] Offensive title detected: Como fazer scrapping no Google Scholar? by Karla Sessin Dilascio on pt.stackoverflow.com
 
8:41 PM
@ProgramFOX around?
 
@hichris123 points at activity chart
 
@Undo Too much work. ;)
 
;)
 
@Seth Go file/bug me about a bug. ;)
 
9:05 PM
[ SmokeDetector ] URL in title: Can forwarding example.com requests to example.com be a security risk? by salt on webmasters.stackexchange.com
 
@hichris123 huh?
 
@Seth I can look into it/bug someone to look into it. If you can find/report a bug on Chrome's bug tracker.
 
meh, I don't feel like putting any effort into it.
I stopped using Chromium for other reasons.
 
like...?
 
9:22 PM
My profile kept getting corrupted, for starters.
 
9:32 PM
[ SmokeDetector ] Offensive title detected: Utilization of Steam APIs and web-scraping by Vishwa Iyer on codereview.stackexchange.com
 
@SmokeDetector false
 
@Undo Registered as false positive and added title to Bayesian doctype 'good'.
 
9:48 PM
@SmokeDetector false
 
@Undo Registered as false positive and added title to Bayesian doctype 'good'.
 
10:29 PM
[ SmokeDetector ] Offensive title detected: How to add try exception in scrapy spider? by Jake on stackoverflow.com
 
11:08 PM
[ SmokeDetector ] Bad keyword in title: Copying Linux NFS files to a standard consumer external hard drive by Andrew on unix.stackexchange.com
 
11:25 PM
[ SmokeDetector ] Offensive title detected: Utilization of Steam APIs and web-scraping by Vishwa Iyer on codereview.stackexchange.com
 

« first day (345 days earlier)      last day (2523 days later) »