Hookay, kicking conflict resolution... this'll be fun
I, [2018-10-05T01:21:59.358757 #28959] INFO -- : [ActiveJob] [ConflictingFeedbackJob] [f71c0536-5fb7-4c68-926b-aa75ec8cdbd3] Total 2375 existing conflicts.
I, [2018-10-05T01:22:03.530209 #28959] INFO -- : [ActiveJob] [ConflictingFeedbackJob] [f71c0536-5fb7-4c68-926b-aa75ec8cdbd3] Found 0 immediately resolvable conflicts.
I, [2018-10-05T01:22:03.668098 #28959] INFO -- : [ActiveJob] [ConflictingFeedbackJob] [f71c0536-5fb7-4c68-926b-aa75ec8cdbd3] Found 1396 further conflicts resolvable by user precedence.
@tripleee That pattern looks like it's already caught by Pattern-matching website in answer and Pattern-matching website in body; append -force if you really want to do that.
[ SmokeDetector | MS ] Bad keyword in link text in answer, bad pattern in url answer, blacklisted website in answer, pattern-matching website in answer, potentially bad ns for domain in answer (317): CPU temp - Dell vs Intel by Apple id support on serverfault.com
@Shree That pattern looks like it's already caught by Pattern-matching website in body and Pattern-matching website in answer; append -force if you really want to do that.
@iBug That pattern looks like it's already caught by Mostly non-latin body, Bad keyword in body, Bad keyword in answer, and Mostly non-latin answer; append -force if you really want to do that.
[ SmokeDetector | MS ] Link at end of body, potentially bad ns for domain in body, repeated url at end of long post (200): Surge RX Review by Fofopaan on askubuntu.com
@Shree That pattern looks like it's already caught by Pattern-matching website in body and Pattern-matching website in answer; append -force if you really want to do that.
[ SmokeDetector | MS ] Bad keyword in body, bad keyword in title, bad pattern in url body, blacklisted website in body, pattern-matching website in body, +1 more (488): primalis-rx-reviews/ by ElizabeBurt on apple.SE
@iBug That might work. It's not clear to me if that change will cause problems with some more complex regexes, but I'm not coming up with any off the top of my head, as most of those are written more to get around the bookending \b. While it's significantly better than using WORD for all of our regexes, it's possible that it will result in some confusion in how our regexes match under various conditions.
@Makyen The primary advantage we get is boundaries for languages like Chinese that don't separate words with spaces. Other things are more removal than addition, e.g. the "boundary" in l'italiano and 123.456
The latter case is already covered by "non-word" boundaries, so this "word"boundary focuses on other languages.
@iBug I'm aware, generally, what it does. I don't know enough details as to exactly what it does to be able to say what it's impact is on existing regexes or any we use in the future. While I understand that it makes adding regexes for specific non-Latin based words significantly easier, I'm unsure of the exact impact on other regexes. In addition, it's unclear to me if the spam we see really complies with what's needed to make the non-Latin words match, just with that addition.
Regex are often about the details. It concerns me to be using something that isn't something that the majority of people writing the regexes are familiar with.
@Makyen AFAICT, this addition only adds support for eastern typographies. Other differences between ASCII boundary and Unicode boundary focus on linguistical "word"s and miscellaneous stuff like emojis.
@iBug It's my understanding that it's intended to cover both Western and Eastern typographies. To a large extent, it is adding support for Eastern typographies. While I'm slightly concerned about that and it's impact on some regexes, my main concern is what other things it is doing other than just adding support for Eastern typographies. That it is doing other things has been clear from both your own testing and the example used in the SE question you posted.
[ SmokeDetector | MS ] Bad keyword in body, bad keyword in title, bad pattern in url body, blacklisted website in body, link at end of body, +3 more (686): What Is One Two Slim? by jenifarlopessh on meta.SE
@ArtOfCode IMO, this automatic conflict resolution was too aggressive. IMO, automatic resolution should only resolve the clear and obvious cases. Going through some of those resolved, I really don't agree with a significant percentage of those I checked. For instance, it does not take into account that there may be a comment on the post providing information which wasn't known to the majority, or just the one, other person(s) giving the majority feedback.
Even though it's more work for multiple people, I'd much rather see all of these posts put into a review queue to accumulate additional feedback until the verdict is more clear.
In addition, using this resolution system makes it useless to provide feedback on posts where you think that the current feedback is wrong. The only effective things to do are either "request admin attention" through the MS interface, or post in chat asking other people to look at it. However, other people are also disincentivized from providing such feedback because it will only result in the feedback being invalidated, without anyone reviewing it, unless at least one other person joins them.
@Makyen And that's all. Since it's written as (\b|(?w:\b), it means the union of two sets. Therefore its impact on existing western typography can be safely ignored. Note that Unicode boundary adds nothing to western typos, and ANSI boundary adds nothing to eastern typos.
> Chinese character in title, mostly non-latin body, mostly non-latin title ---------- Title - Position 0-20: 办理毕业证学历认证 哈哈哈Keto哈哈哈 Title - Text contains 15 non-Latin characters out of 19 Body - Text contains 15 non-Latin characters out of 19
> Bad keyword in body, bad keyword in title, bad keyword in username, pattern-matching product name in body, pattern-matching product name in title, potentially bad keyword in body, potentially bad keyword in title, potentially bad keyword in username ---------- Title - Position 0-10: rapid tone, Position 11-20: keto diet, Position 32-42: bella rose Title - Position 11-15: keto Title - Pattern-matching product name: Position 0-31: rapid tone keto diet shark tank Body - Position 0-10: rapid tone, Position 11-20: keto diet, Position 32-42: bella rose
[ SmokeDetector | MS ] Blacklisted website in body, body starts with title and ends in url, potentially bad ns for domain in body, repeated url at end of long post (297): The denial of this area of human by user305004 on apple.SE
@TetsuyaYamamoto 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.
> No whitespace in title, potentially bad keyword in body, potentially bad keyword in title, potentially bad keyword in username ---------- Title - Position 0-21: williamsloanfundshelp Title - No whitespace or formatting in title Body - Position 0-21: williamsloanfundshelp Username - Position 0-21: williamsloanfundshelp
[ SmokeDetector | MS ] Blacklisted website in body, link at end of body, potentially bad ns for domain in body, repeated url at end of long post (299): Alpha Flex XT Review by Jatehoney on askubuntu.com
> Bad keyword in body, bad keyword in title, bad keyword in username ---------- Title - Position 4-13: forskolin Body - Position 4-13: forskolin Username - Position 4-13: forskolin
> Bad keyword in body, bad keyword in title, bad keyword in username ---------- Title - Position 0-7: cbd oil Body - Position 0-7: cbd oil Username - Position 0-7: cbd oil
> Bad keyword in body, bad keyword in title, bad keyword in username ---------- Title - Position 2-11: antiaging Body - Position 2-11: antiaging Username - Position 2-11: antiaging
> Bad keyword in body, bad keyword in title, bad keyword in username ---------- Title - Position 4-10: cbdoil Body - Position 4-10: cbdoil Username - Position 4-10: cbdoil
room topic changed to Charcoal HQ: Where diamonds are made, smoke is detected, and we break things by developing on production. 109k true positives and counting. For more information, see charcoal-se.org/about. Handy links: metasmoke.erwaysoftware.com, github.com/Charcoal-SE [dev-on-prod] [panic-driven-development] [plastic-knives] [programmatic-crowbars]
> Pattern-matching website in body, pattern-matching website in title, potentially bad ns for domain in body, potentially bad ns for domain in title ---------- Title - Position 3-21: supplementlist.com Title - topsupplementlist.com NS suspicious ns44.domaincontrol.com.,ns43.domaincontrol.com. Body - Position 3-21: supplementlist.com Body - topsupplementlist.com NS suspicious ns44.domaincontrol.com.,ns43.domaincontrol.com.
[ SmokeDetector | MS ] Bad keyword in body, bad keyword in title, bad pattern in url body, pattern-matching product name in body (389): What is Thermo Burn by perttypery on askubuntu.com
[ SmokeDetector | MS ] Bad keyword in body, bad keyword in title, blacklisted website in body, blacklisted website in title, pattern-matching product name in body, +5 more (694): garciniamarket.com by RegalKetoDietReviews on webapps.SE
@Shree 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.
[ SmokeDetector | MS ] Url in title, bad keyword in body, bad keyword in title, pattern-matching product name in body, pattern-matching website in body, +3 more (590): click2fitness.com/keto-x-factor/ by pipshirow on graphicdesign.SE
[ SmokeDetector | MS ] Bad keyword in body, bad keyword in title, blacklisted website in body, link at end of body, link following arrow in body, +2 more (590): Surge RX Surge RX by platitunerr on askubuntu.com
