Two little ponies January 16, 2014 4:51 PM   Subscribe

I have a question about the feasability and desirability of two possible small ponies:

1) Is it possible and desirable to have the submit button execute a check for unclosed tags? Sometimes I'm dumb and forget to close bold or italics tags and then have to go back and edit. Is it possible for the MeFi software to warn about these things?

2) Every so often many of us will say something that is proven factually incorrect, but outside the edit window, and later comments saying "Oops I was wrong" may not come in time to prevent reactions to the Wrong Thing that we said. Is there a feasible option for a one-time-per-comment ability to add (not change the original comment) something saying "Sorry, I have been proven wrong about Thing I Have Said, so please disregard"? So people coming in later don't see the erroneous comment, respond to it, and contribute to a possible derail? Smarter minds than mine could see ways to make this functional and unabusable, I am sure. Is it possible? I'm thinking of something like the 'edit' button, but a one-time-only 'correct' button that only allows for addition to a comment, not changing the content of the original.

oh wait I have a third reallllllllllly tiny pony to ask about:

3) Could the flagging list have a separator (---) between the 'positive' flag reasons and the negative? Every time I flag something as fantastic I'm worried about a slight misclick and flagging incorrectly--the mod team has already received one message from me about an oopsie that way.

I am curious to see both community and Super Mod Team (and Mega Programmer God pb) thoughts on these. Feel free to tell me I'm a twit for poor mousing skills :)
posted by feckless fecal fear mongering to Feature Requests at 4:51 PM (41 comments total)

3) Could the flagging list have a separator (---) between the 'positive' flag reasons and the negative? Every time I flag something as fantastic I'm worried about a slight misclick and flagging incorrectly--the mod team has already received one message from me about an oopsie that way.

As an alternative; perhaps an 'unflag' option to rectify mistakes?

I totally accidentally just flagged this post while looking at the flagging options. Sorry mods.
posted by His thoughts were red thoughts at 4:54 PM on January 16, 2014 [7 favorites]


The server already does a check for unclosed tags and closes them up for you. It's kind of a brute-force method, but it keeps the entire page from becoming bold or blockquoted. If you happen to see that the format isn't what you expected you can edit your comment.

On number two, no, we already have editing. We're not going to amend comments further. I sympathize that it can be frustrating having a comment out there that isn't what you want. But people should read the comments following for any updates. We don't have threaded comments, and part of the idea behind that is that people can keep up with new information as the thread develops.

On three, yeah, I think that's something we can think about. But keep in mind that a mis-flag or even mistaken-flag here or there isn't a big deal.
posted by pb (staff) at 4:56 PM on January 16, 2014


1) Is it possible and desirable to have the submit button execute a check for unclosed tags?

The commenting apparatus does this already; it's why your un-closed tag only affects your comment instead of the entire thread. The 'notification' that you're after is that the entire rest of your comment is bold or italic or whatever, unless I'm correct in interpreting this request as specifically for a worded pop-up alert. Which, if I am, uh, no thanks. Already too many pop-up reminders in my life.

An alternate solution could be to use the highlight/button technique, where you highlight the text you want to change the display of, then use the appropriate formatting button from the selection under the comment box. Using those automatically inserts opening and closing tags, making the un-closed tag problem moot.
posted by carsonb at 4:58 PM on January 16, 2014


Shown up by pb, shoulda known he's a fast typer.
posted by carsonb at 4:58 PM on January 16, 2014 [1 favorite]


If you realize a comment was ill-thought-out before people have responded to it, you are welcome to drop us a line via the contact form and we can delete it for you. That's the simple way to handle that. If there's already discussion of it, there's no way to uncreate that discussion, so a follow-up comment is the way to go.
posted by restless_nomad (staff) at 4:59 PM on January 16, 2014 [1 favorite]


2. That's a technological solution to a social problem that isn't a big problem. Sure, it's not a clear and elegant as many would like, but ultimately the corrections get noted, if bumpily at times.

3. Individual flags don't matter much, so I'm guessing it probably isn't worth bp's time
posted by Brandon Blatcher at 5:00 PM on January 16, 2014


As an alternative; perhaps an 'unflag' option to rectify mistakes?

This would also be a great option, perhaps with the same 5 minute window.

I guess what I'm saying that I'd like to see is:

(flag)
fantastic comment
(and maybe someone else's suggestion of 'enlightening/I learned something,' because sometimes 'fantastic' just means hilarious or apropos)
----
HTML/display error
----
offensive/sexism/racism
(etc)

on preview:

pb: I get that, I was thinking more of the "Oh crap I didn't close the bold tag and now I have to edit my comment and basically I am asking the software to hold my hand a bit, but I understand if the people who actually write the software aren't interested in doing that."

On (2), I understand where you're coming from.

(3) I think is just a UX thing; some of us (me) are dumber than others and would like things a bit more delineated.

If you realize a comment was ill-thought-out before people have responded to it, you are welcome to drop us a line via the contact form and we can delete it for you.

That's not what I mean. I mean commenting "X is Y," and then 5:01 later someone drops by with proof that "X is actually Z," and being able to acknowledge such in the original comment to prevent derails.
posted by feckless fecal fear mongering at 5:02 PM on January 16, 2014


I favorite the hell out of anyone who makes a statement and later admits that they were wrong. In fact, I think we need more examples of people admitting that they were wrong than we need people who have commenting histories with a low number of errors.

A derail isn't an incorrect comment, A derail is when someone makes a comment that's really not on the subject of the post. Derails are to be avoided; people showing fallibility and the potential for learning and growth is to be cultivated.
posted by benito.strauss at 5:45 PM on January 16, 2014 [9 favorites]


I favorite the hell out of anyone who makes a statement and later admits that they were wrong. In fact, I think we need more examples of people admitting that they were wrong than we need people who have commenting histories with a low number of errors.

I am often wrong. I await your many favourites.
posted by His thoughts were red thoughts at 5:52 PM on January 16, 2014 [7 favorites]


Feckless, the problem with closing bold or other tags before the end of the comment is where to put the close rag? There is no reliable programmatic way to do that. Did you mean to bold one word or a whole sentence or the whole post?
posted by cCranium at 5:57 PM on January 16, 2014 [1 favorite]


On three, yeah, I think that's something we can think about. But keep in mind that a mis-flag or even mistaken-flag here or there isn't a big deal.

Personally, I think there are too many flag reasons if what I've been told about flags is true (and I believe I haven't been lied to!)

I'd like three: Display error, shitty comment, great comment.

Noise and derail are the same thing.

Display error covers double comment.

It breaks the guidelines, and offensive/sexist/racist could all be covered by shitty comment.

I guess I'd also be fine with other, but who really needs it?
posted by cjorgensen at 5:58 PM on January 16, 2014 [5 favorites]


I agree with your flagging scheme, but I think in practice most people would want further delineation. I think flagging something as a "derail" could be seen as a gentler measure than "shitty" or "offensive". So maybe someone would be less inclined to flag something because of the reduced nuance intended, even if the action to call the mods over is roughly the same. But I don't think derails need too much mod action anyway.
posted by planetesimal at 6:12 PM on January 16, 2014


Feckless, the problem with closing bold or other tags before the end of the comment is where to put the close rag? There is no reliable programmatic way to do that. Did you mean to bold one word or a whole sentence or the whole post?

What I am suggesting is like what happens when you don't include a link in the before [more inside] in an FPP: "You didn't close all your tags; did you mean to do this?" or something.

I'd like three: Display error, shitty comment, great comment

But Shitty comment because {racism|sexism|offtopic|noise} is useful for mods, as is great comment {on topic|funny|insightful|educational}. The former lets them know whether delete/keep an eye on/drop a discreet message is appropriate, the latter lets them know whether ok leave it alone/sidebar is appropriate. 'Great comment' is a wiiiiiiiiiiiiiide umbrella. 'Great comment because it taught me something important' is a much narrower one.
posted by feckless fecal fear mongering at 6:13 PM on January 16, 2014


there a feasible option for a one-time-per-comment ability to add (not change the original comment) something saying "Sorry, I have been proven wrong about Thing I Have Said, so please disregard"?

You know I think that would actually be more confusing, to be honest - and also would only apply to a vanishingly small minority of comments that are clearly wrong or right ("The capital of Sudan is Mogadishu... doh!"). Finer differences of opinion, far beyond the small kingdom of right and wrong, would not be well served by this. Additionally, and perhaps I'm being unfair to us mefites, but I could see it very easily being demanded of people when they have been smacked down by other commenters. That already happens to a degree and it's not a nice dynamic.

I mean, there's a mefite I regularly disagree with on climate change threads - but if his snark is too pissy he (and I, and anyone else engaging with him) gets deleted. And if it's not too pissy, it gets substantively and critically destroyed - which I think is a better outcome. He still doesn't believe in climate change, but that's okay; no one would start disbelieving on account of his nonsense.

I dunno, I think people should be allowed to be wrong, to hold wrong opinions about stuff you know? Even when it shits me up the wall and I disagree with it vehemently. Forcing someone to post a mea culpa on their original comment is not a very... affectionate way of engaging with each other, I think.

I'm wrong all the time, and I relish the opportunity to recognise I was wrong, as it demonstrates that I've learnt something. Sometimes, though, I'm right, and you fuckers are just being slow to come around. ;)
posted by smoke at 6:14 PM on January 16, 2014 [2 favorites]


feckless fecal fear mongering: "What I am suggesting is like what happens when you don't include a link in the before [more inside] in an FPP: "You didn't close all your tags; did you mean to do this?" or something."

This is a radical departure from the current fire and forget method of posting comments.
posted by Mitheral at 6:20 PM on January 16, 2014 [1 favorite]


Smoke, what I'm saying is this:

Initial comment: Well when Herbesoften said that Citizen Kane was the epitome of the germanwordforsomething he meant that it was silly.

Reply: Well no, if you look at this link Herbesoften actually said the opposite.

(me editing previous comment): Well when Herbesoften said that Citizen Kane was the epitome of the germanwordforsomething he meant that it was silly.

Sorry, I was wrong, please respond to the correct answer.


I think that could help prevent derails. If I'm in the minority here, cool. Just floating an idea that I think would help with painful and pointless derails, nothing more.

This is a radical departure from the current fire and forget method of posting comments.

Yep. Maybe "Are you sure you meant to fire there?" might be a good use for everyone. Or maybe I'm projecting my own failings and just need someone to write me a script.
posted by feckless fecal fear mongering at 6:26 PM on January 16, 2014


I just think that scenario happens much less frequently than you think it does on a number of levels:

1) It's fact-based thing where it's something that can objectively be seen as right and wrong.
2) People will interpret the same facts in the same way, leading to the same opinion
3) People will be totes happy to say that they are wrong and strike their comments voluntarily and in a highly visible way.

I just don't see that happening very often at all. Moreover - and this is unusual for me, as I typically favour aggressive Modly Intervention - I feel it really is a bridge to far when it comes to telling people what to say and how to say it.

I think it would cause more problems than it would solve, honestly.
posted by smoke at 6:34 PM on January 16, 2014


PS It would also generate derails where people insist someone modify their comment to admit they are wrong, and that person adamantly refuses.
posted by smoke at 6:34 PM on January 16, 2014 [1 favorite]


Yup. I see the problems you are positing. I guess what I'm saying is this would work if it were a thing nobody ever mentioned, and that'll never happen, so you have shown me explicitly why it wouldn't work.
posted by feckless fecal fear mongering at 6:36 PM on January 16, 2014 [1 favorite]


Didn't mean to harsh your buzz.
posted by smoke at 6:40 PM on January 16, 2014


Sometimes you have to bury the door to see the knob.

It's not a good metaphor, so what.
posted by planetesimal at 6:44 PM on January 16, 2014 [2 favorites]


You didn't harsh my buzz. You interjected some rationality to a pie in the sky ideal. Thank you :)
posted by feckless fecal fear mongering at 6:45 PM on January 16, 2014


You're not supposed to be making changes like that inside the editing window anyway, let alone after the fact.
posted by ODiV at 7:07 PM on January 16, 2014


I think you're missing what I'm saying, ODiV.

Comment: X is Y!

Another user comment, 5:00+n later: "Actually, X is Z, [link]"

Initial comment, edited: X is Y! Okay I was wrong, X is Z per [link]. please respond appropriately.

In my (probably fictional) headspace, the original comment would stand, with an addendum. No change to the original comment permitted.
posted by feckless fecal fear mongering at 7:16 PM on January 16, 2014


Yes, I understand your proposal.

Since you included the phrase "but outside the edit window" I assumed (incorrectly?) that you were unaware that substantive changes such as this are not what it's there for.
posted by ODiV at 7:23 PM on January 16, 2014


Initial comment, edited: X is Y! Okay I was wrong, X is Z per [link]. please respond appropriately.

But is this approach really more effective than just making a fresh comment, quoting yourself and admitting the error? For example:
X is Y!

Okay I was wrong, X is Z per [link]

This is (as I understand it) the status quo. You still get to admit you were wrong, but the thread structure remains linear and understandable. And you don't need to worry about non-typo edits to comments making the thread non-linear.
posted by His thoughts were red thoughts at 7:24 PM on January 16, 2014


One, two, ponies kneel before you
That's what I said now
Ponies, ponies who adore you
Just go ahead now



Well that's going to be stuck in my head until April, thanks.

posted by Horace Rumpole at 7:27 PM on January 16, 2014 [3 favorites]



One, two, ponies kneel before you
That's what I said now
Ponies, ponies who adore you
Just go ahead now


DOOBIE DOO, DOOBIEDOOBDOOBDOOBDOOBDOOBDOOBDOO...
posted by His thoughts were red thoughts at 7:33 PM on January 16, 2014


(me editing previous comment): Well when Herbesoften said that Citizen Kane was the epitome of the germanwordforsomething he meant that it was silly.

Sorry, I was wrong, please respond to the correct answer.


I don't think that this needs a software answer. If I'm reading the thread, I read your misstatement, I read your correction, I get down to the comment box. This is only a problem for someone who reads your misstatement and comments immediately, without reading the thread to see that you had already corrected yourself. What line of code is going to keep people from using the internet wrongly?
posted by ActingTheGoat at 7:34 PM on January 16, 2014 [1 favorite]


I admit I haven't thought through the actual timelines, but consider:

User 1: X is y!

U2: Well no

U3 I agree with U2 except

U4: YOU ARE SO WRONG BECAUSE LINKE

U5: WTF is wrong with you, U1?

U1: oops I was wrong

U6 - U6+n: yeah fuck what is wrong with you

vs

U1: X is Y!

U2: Actually no because [link]

(U1 edits to X is y! Oops, [link] proved me wrong, don't respond to my wrongness please)
posted by feckless fecal fear mongering at 7:41 PM on January 16, 2014


Yeah, we are really, really not going to consider anything that makes threads non-linear. The edit window took *years* to approve and is *solely* for typos, and that's on purpose.
posted by restless_nomad (staff) at 8:52 PM on January 16, 2014 [3 favorites]


I have absolutely no use for an 'I was wrong' button.
posted by mazola at 9:06 PM on January 16, 2014 [3 favorites]


I don't think any of these would be super-great ideas, except maybe for the last one. As 50 approaches rapidly, the old eyes are starting to go a bit wonky (I'm now one of those old farts who looks over his glasses when he's trying to see something up close), and the teeny-tiny, old-school tools for flagging and stuff are starting to get a little fidgety-small for me. I might even end up bumping my MeFi text size for the first time in like 14 years.

It amuses me mildly that the EXTRA LARGE size of everything I've been designing and redesigning lately in my Wonderchicken Industries Pile o' Websites is at least in part (beyond just being trendy) due to my old man eyes failing me.
posted by stavrosthewonderchicken at 9:09 PM on January 16, 2014


As 50 approaches rapidly

DRIVE THE OTHER WAY, STAVROS, DRIVE THE OTHER WAY!!! 50 IS A PAIN IN THE ASS! LITERALLY!!!
posted by It's Raining Florence Henderson at 12:14 AM on January 17, 2014 [1 favorite]


I am often wrong. I await your many favourites.

posted by His thoughts were red thoughts at 12:52 PM on January 17 [3 favorites +] [!]


My plan is working. It's WORKING! BWAHAHAHAHA!
posted by His thoughts were red thoughts at 1:45 AM on January 17, 2014


I'm with cjorgensen that there are too many flag options for dickish comments, but from a mod perspective does the number of options make a difference? For example, would a comment with 10 flags for noise show up quicker on the mod interface than one with 5 for noise and 5 for derail?

If not, I don't really see a problem with the scheme as is, and if some people want the option to be more specific then it doesn't hurt to give them that.
posted by Ned G at 3:08 AM on January 17, 2014


All flags show up together, by comment or post ID number. Only fantastic flags are separate.

So we have a queue of items that have been flagged at least once. We look at the post or comment that has been flagged, and we may delete the post or comment, or leave a note in the thread, or contact the member, or edit to fix an html mistake, or ignore, and then we delete that item from the queue. If it gets flagged again, it goes back to the queue, and we see the new flag, plus any older flags on it. If we don't delete it from the queue (we might be waiting to see what's going on with it) as the queue refreshes, we see any new flags that show up on that item.

From my point of view regarding individual flags, it's very useful to know right away if something is flagged for html error, because that's not necessarily obvious at all (a malfunctioning link, for example), also "double" for posts is very helpful (and for comments, as well, in case something has accidentally been posted twice, so I don't have to stare at it to figure out what's wrong). As for the rest, sometimes they are helpful, sometimes they aren't. In some cases, it literally doesn't matter how it's flagged, we just need to see it and know instantly what the problem is, and what to do.

Sometimes it's less clear. Occasionally I'll see something flagged and have no idea what the problem is. I'll spend more time trying to figure that out if it's flagged offensive/sexism/racism than if it's flagged "noise" (because some people are flagging because they think discussions on Mefi should be very, very strictly on-topic, similar to Ask Me, when it's actually a lot more flexible there). After checking it out a bit, if I still don't see what wrong, I'll usually ignore, since sometimes things get flagged by accident. If it gets more flags, then I'm really carefully checking out all the flag reasons, and who flagged, and going back deeply into the thread to see what happened there (this is usually going to be an instance of an earlier fight flaring back up, or an earlier mod warning ignored, and I need to find that extra context). In an instance like that, I'm using *all* the info I can get. In some other instances I don't need any more info other than a flag, any flag.

The other way that specific flags are helpful for me is when there are many items in the flag queue, I can sort of "triage" what to look at first. Something flagged "offensive" or "against the guidelines" might throw an entire discussion off course, while something flagged "html error" won't, usually, so I'll clear them in roughly this order, for comments: offensive, etc.; breaks the guidelines or derail; other or noise; html error or double.

Other moderators probably have different approaches. For me, personally, I wouldn't mind some paring down, but I wouldn't like it to be super minimal.
posted by taz (staff) at 4:25 AM on January 17, 2014 [3 favorites]


Didn't mean to harsh your buzz.

Sometimes you have to bury the door to see the knob.

You didn't harsh my buzz. You interjected some rationality to a pie in the sky ideal. Thank you :)


Things seem to be working out pretty well lately in Colorado (considers visit).
posted by Wordshore at 6:40 AM on January 17, 2014


I realize that this is already pretty much a settled issue, but here are my thoughts:

Regarding request 2: The temporal flow of MetaFilter is such that people assume that later content is always further down the page. The edit window messes with this a tiny bit, but under current usage it's hardly ever noticeable. Editing a comment to add an "Oops, I was wrong!" would mess up the way people expect MeFi threads to flow. The current solution is to just post a comment at the bottom of the thread saying "Whoops! Thanks for the correction!" and that meshes well with the way that threads work in general. Sure it doesn't cover inattentive readers who don't read far enough to see the correction (and apology) but I've personally never seen that sort of thing as a priority for MeFi. The expectation is that people will read threads and pay attention, and if you miss something because you're skimming then that's your own fault. I sort of like this this – I don't want the whole world to work like that, but I like that MeFi does.

Request 1: As far as checking for unclosed tags, I feel like we're already at a place where it's not much of an issue. There's Live Preview, Standard Preview, an edit window, and the contact form all there to help us prevent mistakes and correct them when they happen. I myself don't feel like there's really a need for additional layers of protection, especially since making an automated tag-checker that works 100% of the time is probably impossible. It might be cool to have a little "Hi! It looks like you have an unclosed tag, are you sure you want to post?" popup when an unclosed tag is detected, but there would inevitably be some annoying false positives (people leaving tags or tag-like things unclosed on purpose for whatever reason) and some annoying false negatives (unclosed tags that are missed by the checker). I feel like a partial solution there would be just as irritating overall as the current system, and therefore would prefer the simpler system that we currently have.

Request 3: While it would be nice if there was some protection against misflagging, it's always possible to contact the mods and tell them "Hey, that Offensive flag was meant to be a Fantastic flag, thanks very much". They'll get the idea. Where I might like this though is on mobile devices, because the flag dropdown (for me anyway, Firefox on Android) is small and fiddly. Touchscreens are a bit fiddly too at the best of times, and while I don't flag much on my phone I could see myself making a mistake. Having to then go and type out an email to the mods is a bit more laborious too on a phone, so any solution that helps to avoid that would be nice. I don't know about the dividers, but some kind of unflagging capability might be cool. It wouldn't have to have a five-minute window, a one-minute or thirty-second window would be fine. The mods would have to have something in place to prevent people from deliberately unflagging and reflagging something in order to be annoying, but I don't imagine that would be hard to implement (maybe only give people one chance to correct a flag, and if they blow it then it's off to the contact form for them). I'd definitely be in favor of something like that, but I don't think it's essential.
posted by Scientist at 10:52 AM on January 17, 2014


> As far as checking for unclosed tags, I feel like we're already at a place where it's not much of an issue. There's Live Preview, Standard Preview, an edit window, and the contact form all there to help us prevent mistakes and correct them when they happen.

QFT.
posted by desuetude at 10:14 PM on January 19, 2014



Yeah, we are really, really not going to consider anything that makes threads non-linear. The edit window took *years* to approve and is *solely* for typos, and that's on purpose.


*neighs and nibbles apples*
posted by infini at 12:17 AM on January 22, 2014


« Older Metafilter LinkedIn Group   |   When an Obama derail isn't actually an Obama... Newer »

You are not logged in, either login or create an account to post comments