@Guero Looks broken to me.

@Shog9 @Shokhet hey guys sorry too much too drink to do anything with code tonight will deal with all tomorrow good night and ty for all your help

@JasonC @hichris123 Can you make me an RO in Chimney as well or at least give write permission there?

Are moderators supposed to be impartial

Or they can exhibit bias towards users

I am honestly curious

Can they say "well, rules A and B only apply to so-and-so, but for you, I won't apply it"

sure, they are human, not robots. They can make judgement calls.

Is singling out users really a judgmeent call

I'm having trouble understand -- I thought moderators were supposed to impartially apply the law to all

Not say "well here's our undesired question list ... don't worry, it doesn't apply to you ^.^"

Without context I can't say and might be different on some sites so this is better brought up on the site meta or maybe in chat with said moderator

I've already sent a message to the email

community email or w/e

I'm just confused -- can site moderators simply ignore the rules when talking about a specific group of users

"theses site rules do not apply to you"

I thought rules were meant to be enforced down to a T

If you have send the email I'm sure the CM team will look into it,

I sure would. I'm not biased towards dogs

Or cats

Would you treat dogs/cats/raccoons differently?

Would you treat dogs/cats differently

afk

Who says you can't leash a cat?

No one's stopping me from leashing my own kitties :)

Just because they're not dogs doesn't mean they don't have to be a cat

Not every raccoon is the embodiment of chaos

Not every dog/cat is the embodiment of good

And you sir, are just an animal.

Take my word, really :)

I'm not biased -- I don't classify people just to be cute

So why did you classify people into 4 categories?

Surely there's a reason

You sir, are a troll then

That's fine with me, I'm just classifying you based on mythical creatures

:)

@JasonC great work on chat topbar 1.11! However, the changelog dialog is vulnerable to mouse-wheel overscrolling :p

anyway, that's very uncritical, ignore it...

Seems UK Election is a good thing to watch now...

...hung parliament?

@Elephant Did they try to reboot it?

I'm not surprised by people calling @Telkitty a troll. Sir however ...

TIL.

→ 16 messages moved to Chimney

brb, I'm going back to nature.

> It sounds like you’re going through a hard time. I’d really like to help you, but unfortunately, we’re not well equipped to do so here. Your best option is probably to call a suicide hotline. People are on call there to talk to people struggling with the same kind of issues you are, regardless of location. If calling is not good, you can chat with them live online. It might not help, but what’s the harm?

that's scary. just like having suicidal thought if not enough SE

So ... go exercise

Shaaaaaaaaaaaaun!!!!

@JasonC There's no rush, and no need to apologize! Take your time and do it right.

...and if the script keeps working the way it is, it's still much better than using chat without it. If I have to learn to scroll a little bit, then so be it :)

Does anybody have a second to help me test something in chat for the topbar? It requires a second person.

@JasonC sure

Sweet meet me in chat.meta.stackexchange.com/rooms/1034/shadows-sandbox

"sweet meets"

Not unlike hot meat posts.

Or exchanging meat stacks.

Ew. Nevermind.

meat.SE

Mmmm

@peterh Welcome to Tavern on the Meat.

EWWWW

@JasonC Welcome! :-)

@Guero I never understood why the SE is working so actively to avoid its growth

@Elephant To be fair I would totally be in to a butcher's se site.

Although it's probably too narrow.

there's already Seasoned Advice

Oh right, duh.

Which I'm already totally in to.

@Guero If the creators of the Data Science SE could have got the possibility to freely select from the anyways closed questions of other sites, and import them, all of the SE sites would be much bigger

@Guero Because the owner didn't wash the bowl enough well, and the old meat remains are rotting on it and stink. The cat senses the odor, and understand very well that it is a stinky bowl, but with fresh meat.

@Guero And the cats hate the rotten/stinky things. They like most the mouses and similar, small things what they've just killed and they are still warm and bloody.

@Guero Most dogs aren't so choosy, except if they had some bad experience, for example they've ate rat poison already or tried to eat the food from a mouse trap what clased their noose. But cats are by default choosy.

@Elephant @Shokhet Everything's fixed and published, will be there when you update. Changes.

@Guero Alternatively, it is also possible that the owner washed the bowl well, and the odor of the washing materials deter the cat. Cats hate also the non-biological odors, their genetical program says it is poison.

@peterh Holy metaphors, batman

@JasonC Cool! Thanks so much! :D

Chat reply fix: confirmed. 👍

fist pump

@Guero I've never used them before but I just turned them on and there doesn't seem to be a p in the g list

@Shog9 I'm looking at this firefox patch now. Why is changing the setting stuff into window events needed for firefox? Just curious. The GM_ stuff is supposed to work in FF I thought.

@JasonC it's not. It's yak-shaving.

Ah. Why do you prefer doing it that way?

The bulk of the script has no need to run in a privileged context (access to GM_* API). When injected into the page itself, it has direct access to other scripts (jQuery), events, expandos, etc. - so no need to load jQuery twice, no need for unsafewindow.

Also, it's a lot easier to debug.

Ahhh

This is great

The only privileged APIs the script needs while running are those to save / delete settings - since you were already wrapping the GM_* API for this anyway we can just proxy those through a custom event on window, and use a little stub in the privileged userscript to handle 'em.

The alternate solution is to load jQuery into the privileged context or not use jQuery, both of which are more messy than this solution and also don't have the debugging bonus.

I see

And fundamentally, tampermonkey vs. greasemonkey sandboxing differences were the root of the firefox issues?

Now, there's a wee little downside: settings changed in one page don't automatically update in other pages open at the same time. I don't consider this particularly important, but if we wanted to work around it there's a simple solution to that.

@JasonC yeah, pretty much; I was actually kinda surprised it worked in Tampermonkey.

...sorta makes me uneasy about the sandboxing Tampermonkey does

but, that's a problem for another day

From what I was reading gm changed their sandboxing behavior a while ago to make it more "proper", but also broke a bunch of scripts and irritated people in the process (probably for a good cause though). And afaict tampermonkey retained the old gm behavior.

But I'm fuzzy on the details

@JasonC well... That happened a long time ago

@Shog9 That's OK it never really behaved quite that way anyways. I was never monitoring for external changes to settings.

Atm in the current version if you change settings in one window, it'll "update" the checkboxes in the settings dialog in another but it'll be inconsistent with what that page is actually doing. A little odd but I didn't really care about it.

When Greasemonkey was brand-spanking new, it pretty much just injected the script into the page with only enough of a wrapper to prevent namespace collisions. Then it added some APIs, some of which were potentially dangerous and some of which were obviously dangerous.

This is good, I have a couple other scripts I probably need to make similar changes to.

At the same time, there were pages that were actively trying to detect userscripts and mess with them (because programmers are dumb and were doing dumb things client-side without thinking about it, and so naturally took a dumb approach to fixing that when folks decided to take advantage of it)

Heh

and don't run userscripts on pages you don't trust

Ah, so they separated window and such to keep it isolated?

(again, this was all a very long time ago; you can find similar discussions happening right now WRT extensions though, since modern browser extensions are spiritually similar to userscripts)

or by authors you don't trust, for that matter.

@Shog9 aren't most browsers moving towards super locked down APIs now though?

I know of several firefox extensions, for e.g., that stopped dev because Firefox's new API made their extensions impossible.

Yeah, the problem with extensions in chrome, especially with regards to Stack Overflow, is that installing one bears the risk that someone can access your fkey.

@JasonC not just window. All untrusted objects in Greasemonkey are "wrapped" in a special proxy object that blocks access to potentially dangerous things (read: things that can be modified by unprivileged scripts) by default.

@ɥʇǝS yeah, that's what I'm obliquely referring to with "spiritually similar to userscripts" - the new API is a lot closer in terms of privilege restrictions to Greasemonkey than to the old XPCom stuff.

old FF extensions had essentially unfettered access to the internals of the browser - Firefox was more or less an "extension" that just went whole-hog and put an entirely new UI on the old Mozilla Seamonkey browser.

You can hijack windows from an FF extension

Not sure why anyone really uses that browser. The annual Pwn2Own has shown some really scary vulnerabilities over the years.

It's a nice browser. Still a lot lighter than Chrome (though Chrome has gotten a bit better recently)

Org kinda lost its way philosophically a while back though; too many priorities == no priorities.

Maaaaybe coming back from that, donno

also no Google looking over your shoulder.

@TravisJ if memory serves, IE/Edge on Windows still had some of the scariest attacks there last time around.

Not '00s-era scary, but still plenty bad

The rendering engine is done well, but the basis of the browser still shares the problems that the original set of browsers had. It doesn't have a strong enough sandbox, and as a result is risky to navigate with in my opinion. The security issues far outweigh anything else, and I would rather have Google inspect (you can look at wireshark/router traffic to see what is sent or collected) a small amount of traffic than have my registry queue up some rando code and have at the system.

@Shog9 IE and FF had the same worst case scenario, open the calculator and make a calculation. Which, while benign, represents a serious security flaw.

yep

The problem is that the underlying runtime for both IE and FF is the system runtime. So once you are out of the browser, you execute directly at OS level. Chrome, and I think Opera, are properly sandboxed to avoid that.

oh, sandboxing, that's where we were before I got sidetracked here...

:D #fullcircle

It is friday afterall, we are allowed a little sidetrack

someone queue up mix master mike

So anyway, @JasonC, in GM everything is wrapped in that proxy to maintain the integrity of the GM sandbox. window is the obvious one (because it's so often a PitA), but so are all of the objects hanging off of window, the DOM, events, etc.

Assuming you don't punch through this intentionally, it means you can trust a property somewhere not to do unexpected things when you access it (like, run unprivileged code)

Makes sense. Also happy to get rid of unsafeWindow in that script. It felt weird when I used it too but I didn't know any other solutions to what I was trying to do.

This is also why I pass key / value names along with the event instead of just reading them off of an object on window that's directly modified by the injected script.

If you're curious, here's the MDN page on that proxy: developer.mozilla.org/en-US/docs/Mozilla/Tech/Xray_vision

I wish that JavaScript gave out access to the EnvironmentRecord in code.

Oh word, thanks

That would solve a lot of problems.

> As well as these two levels of privilege, chrome code is able to create sandboxes. The security principal defined for the sandbox determines its privilege level. If an Expanded Principal is used, the sandbox is granted certain privileges over content code and is protected from direct access by content code. For example, the Add-on SDK runs content scripts inside sandboxes.

gotta kinda read between the lines there, but Greasemonkey is one of those sandboxes

Ahh

Lots of til on this project

It's funny looking at my code too because I can tell what stage of noobness I was in at any given point. There's subtle style changes depending on the day I committed the changes, lol.

"Oh god, that looks like a Wednesday line. How times have changed."

"Back in the day when I was writing height: `${h}px` instead of height: h".

Can someone check this rude answer, I'm not 100% sure but I think that is not OK, despite the F-word seems to be in context. Also username is ...

→ 9 messages moved to Chimney

@rene - tbh, the translation may be somewhat accurate as far as what it means, although using "idiots" at the end (which seems to not be part of the translation) is certainly rude. Moreover, that username =/ ... All that said, why is that post even allowed?! Are people allowed to just post whatever chinese sentence they want, be it racist, offensive, or excessively crude, and then allow the community to post the english equivalent?

Translating / explaining an idiom might make sense in cases where a translation would not

@Shog9 Btw, take your allman braces and go home.

Leave your K&R in the '70s where it belongs

Never

I have a 30" monitor - I can afford a few newlines

@TravisJ fair enough.

@Adam why did you delete your answer with log out screenshot?

because the dupe is fine

@AdamLear but not complete (yet), lacking SO screenshot.

should edit the SO screenshot into it, though, now that I think about it

be the change you want to see :P

Well, @rene got animation there so adding ordinary screenshot would look weird. :/

Unless adding new answer, but better have it all in one place.

not a big deal, IMO

you're overthinking it

I am over thinking, period. ;)

it is Friday :)

Right!

there, edit made. if someone feels like making a gif, go nuts. but I'm gonna get a beer instead ;)

@AdamLear animated beer?

I wish

beerlabelsinmotion.com - really neat stuff

I found a $40 bottle of Guinness at BevMo and I can't stop thinking about it

Is it worth it? Why is it so much?! Confounded

@TravisJ this one?

yup

that is the one haunting me

> range of limited edition luxury beers made from rare and highly sought after ingredients. It is brewed with both traditional beer malt and the same peated whisky malt used in the world’s finest Scotch and Irish whiskies

Reason ^

Worth it? If you want something special, then yes.

Every weekend is special :)

Same as with art, people but paintings for millions of dollars

Yeah but art (at least the million dollar ones) is timeless.

This premium concept now rise in Israel, almost every brand come with some fancy "premium" product, which cost double or thrice the usual price.

@TravisJ so is the bottle. ;)

People also buy wine bottles for thousands of dollars, never intending to actually open or drink it.

:D

And, you can boast with your friends with this. :P

Bill Gates alludes to the fact that even as a billionaire the only burgers available are the same ones that anyone else can get. Perhaps a premium $500 burger would have a market in certain places.

@ShadowWizard - Boast? More like toast! :D

@TravisJ hmm? Gates? Not sure I follow... ?

In reference to your comment about fancy "premium" products

@TravisJ toast after boast then! :D

> "I can understand about having millions of dollars. There's meaningful freedom that comes with that, but once you get much beyond that I have to tell you, it's the same hamburger. Dick's has not raised their prices enough," Gates said.

I do have a small collection of scotch bottles, perhaps I could buy the guinness just to add it to the collection. They are all empty though, naturally. No sense in staring at perfectly good alcohol all the time and not doing anything about it.

@ShadowWizard OK, I get it done ...

@TravisJ hehe, many got such collection, think @Shog also mentioned something like that once. ;)

@rene you don't have to, @Adam added the screenshot to your answer after all.... :)

@TravisJ oh, missed that.

I did read though that he don't give any money to his children.

Which is good.

@ShadowWizard sorry @AdamLear I replaced your screenshot with an animated GIF. I was pressured for doing so, this was not my free will.

BLASPHEMY

yw

Not sure if that is the appropriate response but it is the best I had ...

@ShadowWizard I don't have anything too crazy, nothing over $1000. Have to drink it after all :D Although, I did once see a bottle of Johnny Walker from 1972 at costco for $50,000.

Feel free to buy a full bottle. I'm always happy to help empty it .... if that is what bothers you ...

Bring one for the happy hour ...

Emptying them is where all the fun is ;)

I thought so :D

hey

building my meta cred today...

but everybody went home...

@rene lol... rebelling flower! :D

@TravisJ whoa... probably behind 100 locks? ;)

@AaronHall or to parties...

@ShadowWizard It was in a locked glass case similar to the jewelry.

@TravisJ makes sense, worth more than most jewellery there, probably...

Yeah, I would say. I doubt they have anything above $20k as far as jewelry goes.

@Shog9 Firefox is awesome. I love how, instead of reporting syntax errors in a script on the console, it just silently doesn't load the script, so I have to keep pasting it into Chrome to find out what's going on. Super handy. They really nailed it.

@JasonC that's Greasemonkey. You can attach a separate instance of firefox as a remote debugger, but that's a pain. Script injection is better.

It's freakin' annoying is what it is.

Hah, welcome to cross browser support @JasonC

Ugh

Hey, FF is nice to support really. You should try IE8

being open source you can always send an PR to get FF fixed ...

Well, look at that.

@Guero Math and Physics both have ads.

@Guero Yeah, on mobile it's unlinked except in a couple of places. If you hit it with a query, then it's all good. An empty search, though...

@Shog9 Woot, Firefox update tested and released. Thanks again, man. Really appreciate it.

Speaking of search... have you seen this? meta.stackexchange.com/questions/280668/…

@TravisJ It's even worse on Arqade

IIRC right now we only run that search when the title is edited... possibly only once? I'd have to check. Using tags is a good idea, but most of the time they're the last thing filled in, so we don't actually have them handy until someone already spent effort typing out the question anyway.

@Guero - So the math "cut the knot" thing was a community ad?

Would be interesting to experiment with, though.

On that note, I'm calling it a day here. Y'all have a great weekend :)

It wouldn't take much, just trigger the search update on tag edits as well as title edits.

there is so much guidance about not placing tags in titles that often users will fill the tags in with the title

The technical implementation isn't that hard. I'm just saying it possibly wouldn't make much of a practical impact without redesigning the Ask Question page overall.

Perhaps, however duplicate suggestions being accurate would make a massive difference. It would save a lot of time up front, and not even require the use of the actual ask question feature.

The implementation would also be very lightweight.

@TravisJ That's the goal, yes. It doesn't do much if the question is already typed up.

I'm on your side here - I agree the current results don't make much sense, you don't need to sell me on it.

Even still, many low quality questions don't have much in them as far as what was written. These are people who write a very simple example of "has array, but no work" and can't tell that it is a common question.

@TravisJ Totally. I think you overestimate how many of those people would even notice that dupe suggestions have changed.

There's more to this than just wiring up additional searches in 30 seconds if what I'm saying. Some UX concerns, impact measuring, etc. Nothing unsolvable.

(Caveat: I'm not 100% sure how searching within tags works off the top of my head. Most of the time we're looking up tag stuff, the tag engine is involved. Title search, on the other hand, is purely in elastic. I assume we have a way to do that since [javascript] how to sort an array in the search bar would do the right thing, IIRC, but I don't recall off-hand how that works.)

@AdamLear I think that it would be significant. I understand you would have to A/B test it which is a pain.

I think the impact would be not as significant as we'd like without a page redesign. But there's only one way to know for sure :)

If it saved you the cost of a page redesign, it would be worth it :P

Eh?

I'm saying you need both.

I don't think the interface is that unfriendly. The main issue is that users post questions which already have answers. A wizard, or similar, will not be able to prevent lower quality submission.

Anyway, this is... not really my department at the moment. I'll try to carve out some time to at least see what the UX would look like with multiple search refreshes from tags and/or nag folks internally, but the actual change isn't in the "knock it out in a friday afternoon and yolo" ballpark IMO

Fair enough :)

To note, I am from Yolo County.

I doubt that many people would notice or care if the dupe suggestions refreshed after they already typed their question. That's the core issue IMO. I could be wrong... that's where an a/b test would come in handy. :)

or some sort of a test, anyway. not sure a/b is the way to go. ANYWAY. later all.

Have a good weekend

You too

(a/b test is a good idea) ;)

@Guero Given that math is the second most active site on the exchange, it seems odd not to have ads.

lol

I am making a second colored multi column bar chart in d3.js on this view.

One is a stuff planned, and one is a stuff done.

(working on stuff done)

In 14 minutes though I will vanish until Monday.

At least the coffee has survived

Somebody just pinged me from somewhere and I can't figure out where.

#firstworldproblems

It doesn't make any sense.

heh, there is no way to stop that from happening. even if you were to fashion some sort of cardboard table to protect the keyboard it would be in the way of the view. just accept the gifts from the tree :P

You should have got a msft surface ;)

Correct, but the structure to protect the keyboard would be in the way of the screen.

Most devs can literally type with their eyes closed :D So long as they are on a keyboard they are used to.

Some keyboards are really annoying. At least to me. I don't like the flat keyed ones. Must be able to get solid tangible feedback when typing.

Happy.

Hour.

@Shog9 It turns out I ended up having to add the unsafeWindow grant back anyways. Apparently on Chrome when the script is injected that way, sometimes window.load events get missed. I'm guessing its something to do with tampermonkey load timing.

So on Chrome I made it load the old way, directly, and use unsafeWindow for adding the js object and triggering resize, and on Firefox, it does it the inject + load event way.

@JasonC oh. You can probably drop the window-load thing. Forgot I left that in.

Isn't that how the script loads though?

You can just call the fn directly

Oh; via that little stub script?

Oh that makes more sense.

I had that in before I dropped unsafewindow since there was a timing issue on Firefox, but ended up working around it anyway

Can someone advice what the best response is here, if any? Is it just trolling or is there a major misunderstanding / miscommunication?

Oh yeah that's way better; loads faster on ff too.

@rene Best response: Back away slowly and close the tab.

Also that's pretty broad.

OK

I back away to my bed then. The room is yours @JasonC

Lol peace

@soup @bookend @words @redredwine @zaq @normalhuman @Guero = one user