@user12986714 Yes, something like that. However, I'd prefer to use [\W_]*+ and s?+, because A) "_" isn't a character in \W, but is a character that's commonly used as a word separator. Without it, nass_cleaners will not be detected. The use of the underscore as a fairly common inter-word separator is why we tend to use [\W_]*+ instead of \W*+.

B) I wrote the look-behind to be prior to the optional final s?, as having it prior to the s? results in the look-behind being checked only once, but, given that the look-behind is far into the regex and there is little other possible backtracking, that's a minor optimization, in this case. A better alternative is to use s?+ prior to the look-behind, so the s is possessive, which is a better choice due to improved readability.

!!/watch nass?[\W_]*+cleaners?+(?!\.com(?<=nasscleaners\.com))

@user12986714 You don't have blacklist manager privileges, but I've created PR#4125 for you.

PR#4125 ("user12986714: Watch nass?[\W_]*+cleaners?+(?!\.com(?<=nasscleaners\.com))") opened by SmokeDetector

[ SmokeDetector ] Blacklists reloaded at rev 4dc70ff (SmokeDetector: Merge pull request #4123 from Charcoal-SE/auto-blacklist-1593451097) (running on teward/Osiris)

[ SmokeDetector ] Blacklists reloaded at rev b3e60c4 (SmokeDetector: Merge pull request #4119 from Charcoal-SE/auto-blacklist-1593450469) (running on teward/Osiris)

[ SmokeDetector | MS ] Body starts with title and ends in url (98): Can I run direct burial line in a conduit from the ground to the electrical box? by OFFLINEX on diy.SE

fp- feedback received

[ SmokeDetector | MS ] Bad keyword in link text in body (69): Reading review text files ✏️ by Hemanth kumar on stackoverflow.com

fp- feedback received

[ SmokeDetector | MS ] Bad keyword in answer, potentially bad keyword in answer (87): Correct way to start Meditation and Fitness Exercise by rajat semwal on fitness.SE

[ SmokeDetector | MS ] Link at beginning of answer (35): How to insert an emoji into a text in Ubuntu 18.04 and later? by kissu on askubuntu.com

[ SmokeDetector | MS ] Potentially bad keyword in body (1): API to get Coinbase exchange buy / sell prices ✏️ by Mining Amazing on stackoverflow.com

fp- feedback received

fp- feedback received on How to insert an emoji into a text in Ubuntu 18.04 and later? [MS]

fp feedback received on How to insert an emoji into a text in Ubuntu 18.04 and later? [MS]

fp feedback received on How to insert an emoji into a text in Ubuntu 18.04 and later? [MS]

[ SmokeDetector | MS ] Bad pattern in url answer (64): .Net Core 2.1 with a .Net Framework EF6 Repository - The type initializer for 'MethodCallTranslator' threw an exception by Sai Ram on stackoverflow.com

fp- feedback received

[ SmokeDetector ] SmokeDetector started in standby mode at rev ca0760e (SmokeDetector: Merge pull request #3980 from Charcoal-SE/auto-blacklist-1592031427) (running on iBug/Coral)

@Daniil considering that it's a watch & not a blacklist, it might be worthwhile

[ SmokeDetector ] Blacklists reloaded at rev 4b29466 (SmokeDetector: Merge pull request #4120 from Charcoal-SE/auto-blacklist-1593450754) (running on teward/Osiris)

@Daniil 0 in the last year

70% accuracy ain't bad

[ SmokeDetector ] Blacklists reloaded at rev 65cb738 (SmokeDetector: Auto watch of visit our website by Daniil) (running on teward/Osiris)

Also

!!/watch visit my website

@user12986714 You don't have blacklist manager privileges, but I've created PR#4126 for you.

PR#4126 ("user12986714: Watch visit my website") opened by SmokeDetector

@user12986714 that one's got a lot less TPs and a lot more FPs

@user12986714 That one may be getting over the line of what we want to watch wrt. the number of occurrences on SE.

@NobodyNada Yeah, so that is the point to separate them

@user12986714 we also have some special cases for it already: github.com/Charcoal-SE/SmokeDetector/blob/…

[ SmokeDetector ] Blacklists reloaded at rev cf571aa (SmokeDetector: Merge pull request #4115 from Charcoal-SE/auto-blacklist-1593450095) (running on teward/Osiris)

[ SmokeDetector ] Blacklists reloaded at rev 983ebb7 (SmokeDetector: Merge pull request #4114 from Charcoal-SE/auto-blacklist-1593450030) (running on teward/Osiris)

actually it might be worth taking "visit our website" off the watchlist and just adding "our" to the regex in bad keyword with link

[ SmokeDetector | MS ] Offensive answer detected, potentially bad keyword in answer, toxic answer detected (159): SessionNotCreatedException: Message: session not created: This version of ChromeDriver only supports Chrome version 81 by iProgrammer on stackoverflow.com

@NobodyNada I'm certain that regex is problematic; it uses (blah) rather than (?:blah)

@user12986714 that's because it captures the keyword for the why data

naa- feedback received

although it might have more capture groups than it actually needs

@NobodyNada But are there need to use nested capturing groups?

[ SmokeDetector ] Blacklists reloaded at rev a800f0d (SmokeDetector: Merge pull request #4113 from Charcoal-SE/auto-blacklist-1593449850) (running on teward/Osiris)

@SmokeDetector This is definitely NAA, but I wouldn't have a problem with it being blocked. IMO, it fits "inappropriate for respectful discourse."

[ SmokeDetector ] Blacklists reloaded at rev e7898ef (SmokeDetector: Merge pull request #4122 from Charcoal-SE/auto-blacklist-1593450878) (running on teward/Osiris)

[ SmokeDetector ] Blacklists reloaded at rev 3fb4a15 (SmokeDetector: Merge pull request #4125 from Charcoal-SE/auto-blacklist-1593454205) (running on teward/Osiris)

[ SmokeDetector | MS ] Mostly punctuation marks in answer (35): Raspberry Pi 3 model B "dtoverlay=gpio-shutdown" ✏️ by user121583 on raspberrypi.SE

tp- feedback received

[ SmokeDetector | MS ] Mostly punctuation marks in title, repeating characters in title (92): Why-----------------C+++---------------ccc by Mathew MacDonals on stackoverflow.com

fp- feedback received

PR#4127 ("findspam.py: Partial performance optimization for regex") opened by user12986714

[ SmokeDetector | MS ] Link at beginning of body (40): Anatomically Correct Sliver by Someone404 on worldbuilding.SE

fp- feedback received

One of the things that we should be looking at with adding watches of things that have existing uses is how well the regex finds things which we are not currently detecting. The goal isn't to have a regex covering every possible thing that might be in a spam message, but is to detect all spam.

If there's something we should add because it will detect posts we're not otherwise detecting, that's usually a good thing, but adding watches which have a significant number of FP uses, when that watch doesn't detect things we are missing (i.e. that had to be manually reported) doesn't really help us. Adding such can make our task a bit harder, because it increases the number of FP we see.

We already don't have sufficient reviewers to consistently get 2 feedbacks on every post, so the number of additional FP we introduce is of concern. So, we should consider carefully if watching some bit of text actually helps us detect things we're missing or if it's much more likely to to be detecting things that are just going to be FP.

For example, the "visit our website" text might be better if we limit it to detecting that text only when it's within X characters of a link. Now, we may not be able to determine that without watching just the text for a while to begin with, in order to get detection data, but, after some time, someone, commonly the person who added it initially, needs to go back through and determine if the regex can be refined to more accurately discriminate for only TP.

New metasmoke user 'πάντα ῥεῖ' created

@ArtOfCode Please activate my account at Metasmoke.

@πάνταῥεῖ Already done.

naa feedback received on Needle Hotend for nonplanar printing [MS]

fp feedback received on Resetting TCP connection programmingly in Scrapy [MS]

fp feedback received on Front Page design feedback please [MS]

fp feedback received on Is using "Selenium & Test NG" the best approach to Automate Regression Test Cases for Web based Application? [MS]

fp feedback received on Currency Converter button on woocommerce product [MS]

tp feedback received on Add/edit string in compiled C program? [MS]

tpu- feedback received on Correct way to start Meditation and Fitness Exercise [MS]

[ SmokeDetector | MS ] Potentially bad ip for hostname in body, potentially bad keyword in body (2): How can I upgrade from 18.10 to current 20.04 version of ubuntu? by Principle on askubuntu.com

fp- feedback received

[ SmokeDetector | MS ] Link at beginning of body (40): Visualizing Logistic Regression Predictions by simonel garrad on stackoverflow.com

fp- feedback received

@SmokeDetector fp

@πάνταῥεῖ 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.

@SmokeDetector fp

@πάνταῥεῖ 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.

@πάνταῥεῖ Elsewhere, a suggestion was made that the feedback you give be in code format (i.e. enclose it in ``), which should prevent SD from repeatedly posting the "not a privileged user" message.

@art @Makyen @undo @tripleee @Andy @angussidney One more question about improving my feedbacks. Do these count only in this room, or could these be done in other rooms (e..g. Tavern on the Meta) also?

@πάνταῥεῖ I expect the suggestion ^^ is probably not something you saw, so, please, don't take that comment as meaning that you should have already known that. It's only intent was to give credit for the suggestion to someone besides myself.

@Makyen OK, sorry. Will do so in the future.

No reason to ping the entire universe; Makyen is competent to answer questions on this

[ SmokeDetector | MS ] Few unique characters in answer (87): How to hide window border of the game? by user254034 on gaming.SE

tpu- feedback received

@πάνταῥεῖ You can do them elsewhere, but we're much less likely to see them or go looking for them. Having them in here also allows there to be more people available to provide feedback on your feedback. We all exchange comments on the feedback we each give, which makes for a situation where all of us learn and/or find out things about a post or a larger situation which we missed.

@SmokeDetector k

@Makyen OK, ty. I will stay here.

np

[ SmokeDetector | MS ] Potentially bad ip for hostname in answer, potentially bad ns for domain in answer (2): Delayed messages in Slack by Josh on superuser.com

[ SmokeDetector | MS ] Potentially bad keyword in body (1): Test the security of a smb port exposed on the internet by a deleted user on security.SE

@SmokeDetector fp

[ SmokeDetector | MS ] Potentially bad ns for domain in body (1): Is it possible to implement old AVR library onto arduino (specifically, NodeMCU esp8266) ✏️ by Nobel Edgar on arduino.SE

@SmokeDetector fp

fp- feedback received on Test the security of a smb port exposed on the internet [MS]

[ SmokeDetector | MS ] Potentially bad keyword in answer (1): How do I create a messaging tab in my app I want it so users can interact with each other by Esc Official on stackoverflow.com

fp- feedback received on Ubuntu 20.04 LTS does not mount NFS when booting PXE [MS]

fp- feedback received on Is it possible to implement old AVR library onto arduino (specifically, NodeMCU esp8266) [MS]

!!/watch- nascleaners\.com\.au

@Makyen That pattern looks like it's already caught by Potentially bad keyword in answer and Potentially bad keyword in body; append -force if you really want to do that.

fp feedback received on How was this site made? It is using some advanced JS and I would love to know more [MS]

How?

"I broke MS ...again" - Implies that you did.

@Daniil I'd have to go check the transcript to be sure, but you're one of the people I considered a high probability for the person that originated it. To be honest, I primarily didn't want to appropriate the idea to make it appear to be mine, but felt that I was in a bit of a rush to get the message posted, so didn't feel I had the time to go check who actually originated it.

I'd mentally planned on looking in the transcript for the correct person and giving them explicit credit in a reply to my own message. Unfortunately, I got sidetracked (on the several things I'm context switching between today) and didn't actually do that part. I'm sorry that I let that drop.

iBug/Sandy: Executing automatic scheduled reboot.

[ SmokeDetector ] SmokeDetector started in standby mode at rev ca0760e (SmokeDetector: Merge pull request #3980 from Charcoal-SE/auto-blacklist-1592031427) (running on iBug/Sandy)

@Daniil Thanks for the reminder. Currently that process is swapped out, but in the run queue. :)

fp- feedback received on Delayed messages in Slack [MS]

fp- feedback received

@Daniil Thanks for the bug report. That's fixable.

@Daniil Cool. Thanks for finding that.

@Makyen The suggestion for using code format came from @Daniil here.

Should I rather use tpu- explicitely?

@πάνταῥεῖ - doesn't suppress error messages, so I don't think it would be silent

@NobodyNada In code format.

@πάνταῥεῖ I don't see why it would matter; after all, k and tpu- are synonymous

@πάνταῥεῖ I'm not sure what you're asking. If you're asking if you should use tpu- instead of k, I'd say use what you would normally.

@NobodyNada I know.

@ThomasWard @Undo @ArtOfCode It looks like MS is down/unresponsive. It probably needs a kick.

Metasmoke status is now set to down.

[ SmokeDetector | MS ] Potentially bad ip for hostname in answer (1): Passing data in application/x-www-form-urlencoded in HttpClient by Arnab Roy on stackoverflow.com

I really liked having an MS reboot button. :(

sd f

@Makyen Not working for now?

Notice. teward/Osiris: Last 2 connection(s) to metasmoke succeeded. Setting metasmoke status to up.

Metasmoke status is now set to up.

@user12986714 It has never been re-created since MS was moved to ThomasWard's infrastructure.

Annnd we're back.

!!/scan-force stackoverflow.com/a/62646383 "Backfill"

@user12986714 Post 1: Looks like spam but not reported: Post has already been reported

[ SmokeDetector | MS ] Potentially bad ip for hostname in answer (1): Passing data in application/x-www-form-urlencoded in HttpClient by Arnab Roy on stackoverflow.com

@SmokeDetector f

@Daniil The expectation was that it could, but that the work has not been done.

@Daniil There was an independent access list controlled by Undo.

@Daniil metasmoke is up. Current failure count: 0 (teward/Osiris)

!!/scan-force stackoverflow.com/a/62646383 "Backfill"

[ SmokeDetector | MS ] Potentially bad ip for hostname in answer (1): Passing data in application/x-www-form-urlencoded in HttpClient by Arnab Roy on stackoverflow.com

fp- feedback received

@user12986714 It actually already existed.

Metasmoke post 250438 (Passing data in application/x-www-form-urlencoded in HttpClient) destroyed by Makyen

[ SmokeDetector | MS ] Potentially bad keyword in answer (1): iMac FaceTime HD camera shows no image/black image by Christopher Moore on apple.SE

@Makyen Why can't we simply have a user reboot@metasmoke.erwaysoftware.com and allow authenticated users to ssh to it, whose login shell is a reboot script?

@user12986714 that's probably a lot more of a pain to deal with than the "reboot" button

@Makyen Or, maybe it didn't exist when that command was executed? Either way, there were two copies.

@SmokeDetector naa

@NobodyNada Not really; ssh reboot@metasmoke.erwaysoftware.com is all one need to do

@user12986714 That won't help if the system is non-responsive, which is much of the time we want to reboot.

@Makyen Then the lambda won't help either...

@Makyen does rebootms not run on the same box as Metasmoke itself?

@user12986714 yes, but then you have to muck about with public keys to manage access, and you can't easily reboot MS from your phone

@user12986714 Yes, it will, if it tells the machine that controls the VM to reboot the VM, which is what the lambda used to do, effectively, by rebooting MS' AWS instance.

@NobodyNada It was an AWS lambda, so, no, it didn't run on the same box. It used to send a command to AWS to reboot the MS instance, which rebooted MS regardless of the current functional state of the instance.

Hmm... But it appears that Art can somehow ssh to MS and reboot it when it is down.

Well... OK... There are times on AWS when you actually need a "stop" and then a "start". I'm not sure if it did that or just a "reboot".

@user12986714 Yes, that works sometimes, but not all the time. It's also not known to me if the ssh access is to MS, the controlling machine, or potentially both (depending on what they log into).

fp- feedback received

Makyen/EC2-linux: Executing automatic scheduled reboot.

[ SmokeDetector ] SmokeDetector started in standby mode at rev 0ae51c8 (SmokeDetector: Auto watch of theforexbroker\.co\.uk by Machavity) (running on Makyen/EC2-linux)

@Makyen brief network blip on my side it looks like

[ SmokeDetector | MS ] Few unique characters in body, no whitespace in body, potentially bad keyword in body (193): Is there a way of "merging" NFS shares? ✏️ by Gabriel Milan on unix.SE

But i can kick MS again once i am home

@Makyen SSH access is to the VM that houses Metasmoke. NOT the host node because thats my server farm lol

sd v

Edit was immediately reverted; looks like a cat-on-keyboard or something

@ThomasWard OK. np. It looks like it's back and working, so no priority either way. Thanks.

@ThomasWard That's what I really expected, but wasn't certain. I would have been surprised if you had given that access.

[ SmokeDetector | MS ] Potentially bad ns for domain in body (1): How to use displaylink adapter with ubuntu 18.4.4 by Beans On Toast on askubuntu.com

@SmokeDetector f

fp- feedback received

[ SmokeDetector | MS ] Post is mostly images (0): renaming files according to barcode using python by Ijaz Ahmad on meta.SE

@SmokeDetector k

[ SmokeDetector | MS ] Repeating characters in title (65): having problem with if statementsjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjjj by Krystian25 on stackoverflow.com

fp- feedback received

[ SmokeDetector | MS ] Pattern-matching website in body, pattern-matching website in title, potentially bad asn for hostname in body, potentially bad asn for hostname in title, +2 more (201): Unable to find element for website "www.blsindia-canada.com/trackapp.php" Submission Type drop down by Sudeep Panshikar on stackoverflow.com

fp- feedback received

@SmokeDetector It looks like the title was repaired in a grace period edit.

Metasmoke status is now set to down.

@SmokeDetector Why?

@Makyen That's not a report.

@Makyen Can you still access MS?

Smokey and I don't think so

@user12986714 Slow, but yes. It looks like the WebSocket...was working when I started typing this.

@user12986714 I'm able to load MS pages.

Notice. teward/Osiris: Last 2 connection(s) to metasmoke succeeded. Setting metasmoke status to up.

Metasmoke status is now set to up.

@SmokeDetector This had 6 detection reasons. I'm not sure why it wasn't autoflagged. I'm glad it wasn't, but...

Where are flag reports posted?

PR#4128 ("findspam.py: Anchor pattern_websites") opened by user12986714

@dalearn metasmoke.erwaysoftware.com/flagging/logs

[ SmokeDetector | MS ] Bad keyword with email in answer, email in answer (162): How could I trust on bitcoin investment websites such as Amfeix by felix on money.SE

tpu- feedback received

!!/watch bitcointechnologist@gmail\.com

@user12986714 You don't have blacklist manager privileges, but I've created PR#4129 for you.

PR#4129 ("user12986714: Watch bitcointechnologist@gmail\.com") opened by SmokeDetector

@SmokeDetector k

[ SmokeDetector | MS ] Few unique characters in body (94): GFYS GFYS GFYS GFYS GFYS GFYS ✏️ by Vistance on workplace.SE

sd v

@user12986714 While this isn't wrong, for a gmail.com email address which has a sufficiently unique username (i.e. the username is unlikely to match other things on its own, we usually make the "@gmail\.com" part optional, so bitcointechnologist(?:@gmail\.com)? in this case.

That is, however, a judgement call as to if you feel that the username, "bitcointechnologist", will get FP if matching is permitted without the "@gmail.com".

!!/watch bitcointechnologist(?:@gmail\.com)?

@user12986714 You don't have blacklist manager privileges, but I've created PR#4130 for you.

PR#4130 ("user12986714: Watch bitcointechnologist(?:@gmail\.com)?") opened by SmokeDetector

!!/approve 4130

[ SmokeDetector ] Blacklists reloaded at rev d675d00 (SmokeDetector: Merge pull request #4130 from Charcoal-SE/auto-blacklist-1593470156) (running on teward/Osiris)

PR#4131 ("findspam.py: Add reason user unregistered or non-existent") opened by user12986714

@SmokeDetector fp-

@Makyen Flagged now

@SmokeDetector fp (To let SD recognize that it is a fp)

@user12986714 Registered question as true positive. If you want to blacklist the poster, use trueu or tpu.

@user12986714 Registered question as false positive.

[ SmokeDetector | MS ] Potentially bad keyword in answer (1): How can I respectfully ask my landlady to use english when in my home? by name on interpersonal.SE

Autoflagged FP: flagged by @SmokeDetector, @ArtOfCode, @FrenzyLi (Manual notification)

@SmokeDetector k

@user12986714 Thanks. That's a ridiculous delay.

Looks like we still need a ms reboot

@Makyen MS is claiming that it flagged 45s after the post was created. I am fairly sure I refreshed the /flag_logs page well after that and there were no flags recorded. There is content now. I'm also fairly sure I opened the page on SE after that point and did not see the -3 from the flags. I think there might be something wrong with the time recorded for when the flags were raised.

I flagged it as well if that makes a difference

@dalearn You raised a spam flag on this question?

no, another one. sorry about the confusion

np

How do I link my account so that it auto flags? I'm a little new here.

@dalearn If you haven't already read it, there's an Introduction to Participating in Charcoal., which should explain that.

@SmokeDetector @ArtOfCode @Undo There's definitely something wrong/inaccurate with the date-time recorded for when the spam flags were raised on this post. The post was created at 21:44:41 UTC. MS claims that all three flags were raised at 21:45:26 UTC. My browser history shows I ctrl-F5 refreshed the post's page on MS at 21:51 UTC and no flags were shown at that time.

Looking on SO, at the record of the flag raised for SD, it shows the flag was actually raised at 22:18:07 UTC, or ~33 minutes after MS claims.

Also, that's ~33 minutes of lag time between when a post was reported and when it was autoflagged. That's really not good. That delay used to be a couple/few seconds (faster than most human feedback/flagging).

BTW: @ArtOfCode, You were one of the autoflaggers. This is the question on SO. If it's not too late, you probably will want to retract your flag.

Already done :)

Good. :) I guess a couple of people downvoted in a similar timeframe, because I was watching for changes in the post's score, but didn't see the +1 from removing the flag.

@FrenzyLi An autoflag of yours was used on this question [MS]. If you get this notification in time, you will probably want to retract your flag.

!!/report superuser.com/a/1564464/337631

[ SmokeDetector | MS ] Manually reported answer (94): How to get youtube like timed-bookmark effect on local mp4 files? by Minha khan on superuser.com

tpu- feedback received

!!/watch (?-i:HF2TaZtFOL4)(?#youtu.be)

@user12986714 You don't have blacklist manager privileges, but I've created PR#4132 for you.

PR#4132 ("user12986714: Watch (?-i:HF2TaZtFOL4)(?#youtu.be)") opened by SmokeDetector

!!/approve 4132

[ SmokeDetector ] Blacklists reloaded at rev 5f49a5d (SmokeDetector: Merge pull request #4132 from Charcoal-SE/auto-blacklist-1593473124) (running on teward/Osiris)

[ SmokeDetector | MS ] Potentially bad ns for domain in body (1): Can't receive email from certain office 365 organizations - untrusted root by LorFerza on serverfault.com

fp- feedback received

!!/watch visit my channel

@user12986714 You don't have blacklist manager privileges, but I've created PR#4133 for you.

PR#4133 ("user12986714: Watch visit my channel") opened by SmokeDetector