Pony Request - Dismiss Header Notifications Across Account June 29, 2020 11:19 AM   Subscribe

Occasionally there are site-wide messages that pop up in the header of the site (e.g. currently there's one about a BIPOC advisory board). These messages can be really tiny on mobile, and currently the hide button isn't clickable on my mobile device. Two requests below the fold.

1. Since I've already dismissed this message on desktop, would it be possible to dismiss it on an account-basis vs a device-basis?
2. Is is possible to improve the layout of the message on mobile? This is my view, which means the "hide" button is obscured by the "skip to menu" button.

Thanks!
posted by too bad you're not me to Feature Requests at 11:19 AM (79 comments total) 1 user marked this as a favorite

[I'm curious to hear thoughts from the community about this. Implementation details will have to wait on frimble or cortex, though. ]
posted by loup (staff) at 11:20 AM on June 29


I don't have that many devices but I still feel like I end up dismissing any given notification a million times. Which seems like more than is helpful. I'm notified already!
posted by aubilenon at 11:23 AM on June 29 [1 favorite]


Definitely agree that this should be a one-and-done dismissal per account.
posted by obfuscation at 11:33 AM on June 29 [4 favorites]


It's probably a time-based cookie that gets set to whenever you click "hide" and any banner message newer than the time in your cookie is shown you. This requires no per-user database space, and no code running on the server, and is probably just a couple lines of javascript in the HTML. It's understandable why it's implemented this way, certainly, but yes, it is annoying to have to click "hide" on each unique browser you use to access Metafilter.
posted by seanmpuckett at 11:42 AM on June 29 [1 favorite]


Omg yes please. Dismissing a notification also shouldn't need to reload the page.
posted by oulipian at 11:46 AM on June 29 [3 favorites]


Yes please to this as I am having the same issue right now.
posted by Kitchen Witch at 11:54 AM on June 29


Agree. It's like playing whack-a-mole every time I open MeFi on a new device.
posted by bondcliff at 12:41 PM on June 29


It’s not a big deal, but I’d prefer account-based dismissal tracking to browser-based, given the choice.
posted by michaelh at 12:56 PM on June 29


In a perfect world, sure, but please put it after all the more important stuff on frimble's list - I don't have that many devices. On mobile, I usually zoom in to click on the little x if I miss it the first time.
posted by zamboni at 1:28 PM on June 29 [2 favorites]


yes, please fix this, but do this first.
posted by scruss at 1:44 PM on June 29 [6 favorites]


This has come up in MeTa before, the answer is basically that this is a nontrivial change that hasn't seemed worth the work. Gotta say, having some Feelings about this coming up again now that the header message is about doing anti-racism work.
posted by stoneweaver at 2:18 PM on June 29 [41 favorites]


I get the account- rather than device-based setting pony. In the meantime, could you use the "spread" technique to enlarge the screen on your phone to access that (hide), then "pinch" back?
posted by Iris Gambol at 2:18 PM on June 29


This is a trivial change.
posted by lucidium at 4:06 PM on June 29


stoneweaver - I wasn't aware of any previous conversation around this topic, and hope that you can give the benefit of the doubt around my intentions on this request. Was truly just a top-of-mind site improvement thing vs wanting to silence important conversations that need to be had on this site about race.
posted by too bad you're not me at 4:30 PM on June 29 [6 favorites]


I care less about moving it per account and more about fixing the CSS that hides Skip to Menu on Chrome/iPhone in mobile mode. I think the extra lines on the new banner breaks the CSS. Thanks!
posted by geoff. at 6:35 PM on June 29


I'd also like it if clicking the button used JavaScript or AJAX to hide the header notification instead of doing a server-side change. It would be better user experience and probably help a little with server resources. (Same thing applies to other site configuration options.)
posted by kirkaracha at 7:07 PM on June 29


It would be nice if it was per account rather than per device, but I agree with everyone saying that this is way, way, way less important than the other issues.
posted by Dip Flash at 7:13 PM on June 29 [1 favorite]


yeah, it's annoying on mobile and barely visable.
posted by clavdivs at 7:14 PM on June 29


While we are talking banner, has anyone else had issues clicking on the links contained therein? I have, I'm using Chrome on Android.

(I like leaving my banner open, and agree that it's painful framing in the current time, despite being "oh hey banner, (which doesn't get seen very often) I wonder if it makes sense to close across accounts")
posted by freethefeet at 9:11 PM on June 29


I use my phone primarily for metafiler, but I use the standard desktop layout because I cannot stand the mobile layout. I haven't had a problem with the banner. Agreeing with others here that other work takes precedence over this change. Personally I like that it's there. It reminds me to check on threads that are happening in metatalk that I may not have read in a few days, and see what is going on and being discussed.
posted by AlexiaSky at 11:48 PM on June 29 [1 favorite]


> hope that you can give the benefit of the doubt around my intentions on this request

Intentions. This language keeps coming up. The anti-racism thread linked in the aforementioned banner is marred from top to bottom with the words intent and intentions. Intentions are not relevant. The outcome is relevant. The harm done is relevant.

It is incredibly frustrating to see intentions being catered to. Intentions being invoked as a reason. Intentions being brought into what needs to be a productive discussion. The metafilter I want to be a part of rejects this feeble and centering language that serves only to deflect. It is employed as a defense against a perceived attack. Stop taking things personally and start taking time to really think about it.

Why was this particular banner the one that resulted in this request? It is worth considering that this post does serve to minimize, in more than one way, discussions that are critical to the future of the site. More care and thought are required than simply saying, oh it wasn't about that. We do not need to prioritize the feelings of one party over the feelings others have taken the time to actively show resulted from that party's actions. Sit with it. Do some self-searching and come to the discussion with those reflections. That is the anti-racist move, here.

This is a pattern of defensiveness I have seen in other threads. We members of metafilter can do better.
posted by kaelynski at 3:00 AM on June 30 [22 favorites]


Considering that there was already a conversation about how unfortunate it is when key race conversations get pushed down the page by light-hearted posts (plant share, music swap! seriously!), I'm curious as to why the mod crew thought it was a great idea to let this one out of the gate at all.

At least the plant share and the music swap are less "on the nose" than a post about how to get rid of this specific header, especially when the post doesn't acknowledge that context (which the OP may not know, but the mods do), and then it's followed by a bunch of folks cheerily agreeing how irritating that banner is and why won't it just go away, in comments which are still there the next day.

This is not a good look.

PS. No I don't need explaining to me that fun swap posts are a traditional feature of metatalk, nor that deleting comments in metatalks has historically been avoided. Actual people are more important than tradition and history.
posted by quacks like a duck at 4:00 AM on June 30 [12 favorites]


Gotta say, having some Feelings about this coming up again now that the header message is about doing anti-racism work.

I mean, I'm bothered by the banner and also stand firmly with the racism efforts that are underway on the site.

I've been bothered by it before on numerous occasions but this is the first time (perhaps because I've just updated phones in the last few weeks) where I share the issue mentioned by OP where the 'hide' button was completely obscured by the 'skip to menu' option such that I couldn't even hide the banner without going into my phone's settings, allowing auto rotate (because I keep that preference disabled for phone use in general), go back to the browser in question, turn my phone, and (thank goodness) it was more accessible that way (though still not perfect if I recall correctly) such that I could hide the banner finally. Then I went back in and reset my view options in my phone settings.

Sure, it took just a few moments but it was a definite pain in the ass that I'd love to have addressed such that if I click once on my PC browser on this newsworthy type of information it'd be percolated through all my devices. I swear that I've even been asked multiple times on the same device but that could be red car syndrome on my end.

I mean, I get folks saying they'd rather prioritize limited development/testing/design resources on the other request for filtering out racial slurs but this isn't at all an unreasonable thing to want addressed from a user interaction perspective.

I want to say that I understand having "Feelings" at this point, it's rough all over and I hope the above isn't dismissive of those feelings.

Considering that there was already a conversation about how unfortunate it is when key race conversations get pushed down the page by light-hearted posts (plant share, music swap! seriously!)

Well shit, I'm the Poarch Creek individual who started the metatalk post about the plant swap for the distinct reason (as stated in that post) of celebrating the recent mod momentum we were seeing on the race front, not to mention that, ya'know, growing your own things/food/etc are key features that I was raised with in that culture as well. I feel really shitty that my effort has turned into a weird dogwhistle or assumed subtext that those actual inclusive and much needed things shouldn't be handled with the proper gusto and expediency.

Pretty disheartening to be honest.
posted by RolandOfEld at 6:20 AM on June 30 [17 favorites]


Oh, do you really?

Yea, I do. I mean, the fact that I posted something specifically in the header for the swap that I was pumped by recent efforts and those were the inspiration for the post:

"I had the idea, inspired by the username of none other than one of our new mods, who seems to have their hands full spearheading new anti-racism initiatives on the site, for a seed/plant swap."

I mean, you can question my efficacy or usefulness or methodology or outcomes or what the fuck ever but questioning my intent is well and truly fucked up.
posted by RolandOfEld at 6:30 AM on June 30 [10 favorites]


Given the moderator controlled queue for MetaTalk, it's hard to say for certain when the original post was created.
posted by kendrak at 6:32 AM on June 30 [11 favorites]


Also in hopes of not making this about me in particular I'm stepping away from this thread, and the plant swap, by the way. Potentially the site as a whole for a while because damn....
posted by RolandOfEld at 6:32 AM on June 30 [2 favorites]


Given the moderator controlled queue for MetaTalk, it's hard to say for certain when the original post was created.

Last note to give support to kendrak, this comment is valid rather than the one attacking me. I found records of my draft from when I submitted my metatalk post the the queue that confirms I submitted well before the post mayurasana accuses me of trying to steal thunder from or derail appeared on the site. So I'm expected to be clairvoyant or risk being labeled as someone who isn't supportive of anti-racism efforts I guess. This is surreal.
posted by RolandOfEld at 6:45 AM on June 30 [24 favorites]


I think this discussion really highlights the need for a MetaTalk sidebar or other pinning feature for important posts. There really have been a lot of MetaTalk posts lately besides the two in the banner, which has pushed them way down or off the main page. And I don't think it's only about white people being at best unthoughtful -- several of the lighter posts are by POC, at least two posts are about feature requests relevant to the racism discussions, and apparently the queue means that timing isn't predictable -- but ultimately it's a real problem, and it's a point of serious friction and a really unnecessary one. Without a way to keep important, site-defining conversations highlighted and easily accessible, this kind of friction is going to keep happening. And I don't know if a sidebar is the perfect solution, but it seems like a relatively easy starting place since the code already exists for the blue.

Since increased transparency has been one of the recurring points of discussion for a long time now, I think it might help for frimble or cortex to maintain a publicly-visible list somewhere of code-related tasks that are in the works, along with order of precedence and at least a rough time frame. The features that have been been discussed recently that would be part of the site's approach to dealing with racism are (iirc)

- preventing people from posting the n-word
- preventing people from posting other slurs
- keeping important MetaTalk posts from being buried
- improving the flagging interface so that it's more accessible and can enable more users to point out problematic posts

Are there others? What is current plan regarding these in terms of precedence and time frame? For features that are not first in terms of precedence, what intermediate steps will be taken to address the problem? (e.g. delaying some posts from the queue, or displaying more posts on the MetaTalk front page -- I take the banner as an intermediate step, but it still only has room to highlight one or two posts and so isn't enough on its own in this situation.)

I think having a publicly viewable list like that would be a step toward increasing confidence that (a) progress is happening, and (b) racism-related work is prioritized over other features and tasks.
posted by trig at 7:06 AM on June 30 [7 favorites]


I also ran into the "hide and skip to menu buttons are overlapped" issue with the current banner on mobile and would appreciate a fix for that. This feedback isn't intended as a comment on site conversations around racism, which I support; I'm just a compulsive notification-dismisser and want the interface to behave as expected when I'm being presented with the option to hide the notification.
posted by terretu at 7:06 AM on June 30 [1 favorite]


RolandOfEld - I’m sorry to see you go. Your presence here has always made me feel less lonely on the site.
posted by stoneweaver at 7:09 AM on June 30 [11 favorites]


Speaking of efforts for greater transparency: this exact kind of outcome is tied to the metatalk queue system itself is not transparent. Its existence as site processes and procedures is obfuscated--there's a note leading to explanations if you hit new post, but otherwise? Knowing it's there involves closer reading and familiarity with this section of the site than most members have, and it's very understandable that many (probably most!) members don't even realize it's a thing.

So the assumption is made that posts go up as people post them leading to, again, exactly this.

With the queue remaining a thing, I'd suggest verbiage correcting the false timestamp of "posted by $user at $time" (false because that user did not post at that date and time--they posted any kind of stretch of time before that). Something like, "submitted by $user at $time. Reviewed in metatalk queue and posted by $moderator at $time."

With explanatory link of what that metatalk queue is all about. This will sometimes inspire some uncomfortable heat about especially odd gaps between submission time and posting time--but it will be heat with less friendly fire.
posted by Drastic at 7:17 AM on June 30 [6 favorites]


RolandOfEld, I wholeheartedly apologize for my comment. My "Oh, do you really?" was shitty in tone and also completely unnecessary. I have noticed other MeTas about racism being pushed down and I made the very harmful decision to lump you and your intentions in with devaluing the racism threads. I snapped at you, which wasn't fair to you. I am sorry. I am on edge with what is happening in the race threads, and with the MeTa queue and I shouldn't have taken out my frustration on you. You deserve better.

I appreciate you pointing out kendrak's comment about the timing of the MeTa post releases from the queue. My frustration actually has nothing to do with you, RolandOfEld, but with the timing of when mods release posts from the queue. My problem is with the structural issues of Metafilter, not you.

You post about the plant swaps is a joyful one, and I appreciate the spirit in which you proposed it. I am so sorry I snapped at you, was unkind to you, and have potentially driven you away.
posted by mayurasana at 7:35 AM on June 30 [7 favorites]


(oh, and if cortex/frimble do take the suggestion to post such a list, please don't spend time making it fancy - just a plain text file at a permanent link would do for now.)

RolandOfEld, you're one of those users I noticed a long time ago first for your username and then for your comments, and I really hope you come back. I probably wouldn't have been able to participate in the seed swap due to geography, but I'll be thinking of you as I'm planting mine.
posted by trig at 7:45 AM on June 30 [1 favorite]


RolandOfEld, you did nothing wrong here and I'm sorry about how this last chunk of conversation has played out. I hope you can have a breather and then come back.

I've heard folks' interest the last few days in a way to pin important posts up top or in the MetaTalk sidebar. I think that's a good idea, and we'll do it. Current priority is the slur filter project, but then frimble can work on this.
posted by cortex (staff) at 8:17 AM on June 30 [3 favorites]


Legit been thinking about this all day yesterday and this morning.

To the extent that my request was insensitive to the conversation this community's been having around race, I apologize and take ownership. I read the most recent thread (after checking out the link in the header and then dismissing the notification, which led to me noticing the bug on mobile and wanting an account-based option), but haven't read all of them and I mis-judged.

That said, I'm genuinely wondering, given the climate of the site right now: what's the best way to request bug fixes/site improvements? Or do we just not?

Maybe some of the other suggestions (i.e. pinning posts, more transparency from mods on the request queue) will make this less of an issue going forward.
posted by too bad you're not me at 8:23 AM on June 30 [1 favorite]


Might be better to MeMail Roland with your comments. If they are anything like me, they just close the tab and never look back again so they probably aren't seeing any of these comments here.
posted by hippybear at 8:29 AM on June 30 [1 favorite]


This place is a mess.
posted by banshee at 8:35 AM on June 30 [12 favorites]


Speaking of efforts for greater transparency: this exact kind of outcome is tied to the metatalk queue system itself is not transparent.

The queue is a new thing. It was setup for mods to have a higher level of control, and veto power, in this section of the site. In the bigger picture the queue is another expression of the trend towards policing rather than facilitating mindset. It is convenient for the mods, but it is anti-transparent.
posted by Meatbomb at 9:20 AM on June 30 [3 favorites]




The queue was started in 2014. It is not a new thing.
posted by hippybear at 9:42 AM on June 30 [6 favorites]


Can I Nth the point that frustration with the ordering of things on the metatalk page, and the effect on that of the overall metadiscussion, really can't be on individuals. Folks don't have control over their post ordering, or of what other posts may land around the same time, and even if they did, they can hardly be expected to take responsibility for the effect of that overall ordering on the tone of discussion.

A sequence of topics including both "racism" and later "plant swap" sounds like a really beneficial way to switch things up between heavy material and a chance to take a break and talk about plants (as RolandOfEld said, and framed the plant post neatly to make that topic transition less jarring).
But when you continue and it goes, racism, plant swap, Instagram, music swap, oh how can we get rid of the racism banner? Then it's starting to tell quite a different story that can really NOT be attributed to any of the individuals who enjoy plants or music or who are struggling with the UI of the heading. None of the topics have to be problematic of themselves, to end up telling an unfortunate story. That's a product of the site itself and the way it works. The issue doesn't have to be anyone's fault for it be detrimental to Metafilter.

Sorry RolandOfEld that my earlier comment came out snarky and in a pileon, your plant post is lovely.
posted by quacks like a duck at 9:53 AM on June 30 [5 favorites]


I'll be honest, I notice the banners so little that I don't even read them half the time. Especially now, there seems to always be an active banner. Early on, I realized that they all point to MetaTalk posts, and so I just got in the habit of regularly checking in here to catch whatever I might miss by ignoring banners. The banners are just part of the design of the site to me now. I don't think any more about them then I do about any other color or layout item.
posted by kevinbelt at 10:19 AM on June 30 [1 favorite]


I think it's important that we don't turn this discussion onto a referendum on the banners in the first place; the request was for a fix to a fairly annoying edge-case usability problem some people were running into, specifically on mobile (and probably relatively small-screened mobile).

You can, and I do, think that the banners serve a really good purpose, but also want to be able to acknowledge/hide them once read.
posted by Kadin2048 at 10:44 AM on June 30 [5 favorites]


The queue siderail is something that might need its own MeTa. I find it frustrating that the community used to be able to discuss the things that it wanted to discuss about the community but now the community only gets to discuss the things the mods decide are allowed to be discussed about the community.
posted by hippybear at 10:54 AM on June 30 [4 favorites]


RolandOfEld — I hope you come back.

For the people complaining about the “lighter MeTas,” remember that you are taking aim at at least some POCs in doing that. If you’re going to shoot at the Man, make sure it’s the Man you’re shooting at (I’ve made this mistake myself; I am more careful with zingers these days and slower to comment).

As far as the queue goes, at least we haven’t had a “I want to relitigate my comment deletion” MeTa in quite a while; that is worth a queue. I do wish the mods had delayed this particular pony request for a bit though.

I think the anti-racist work being done on the site is the most important thing being done on the site right now and is critical to the site’s future. We also have about five threads going on right now but deal with that work, and it is stressful trying to keep track of everything in every thread, even with recent activity to help. I don’t suppose there’s a way we could set up a temporary “anti-racism subsite” to help keep it together? Sidebars don’t appear on mobile, and the header is an impractical in place to keep all the links. I also wouldn’t want to pull the threads off of meta-, because some people would not find their way to the new sub site.

Lastly, I just finished listening to the audiobook of How to Be an Antiracist by Ibram X. Kendi; a lot of what he talks about in that book relates directly to what we are trying to do here.
posted by GenjiandProust at 11:13 AM on June 30 [2 favorites]


As far as the queue goes, at least we haven’t had a “I want to relitigate my comment deletion” MeTa in quite a while; that is worth a queue.

A lot of the relitigation of posting discussion are how much of MetaFilter formed its community standards, though. And at this point it's not a community discussion in a lot of ways, it's the mods and nothing but the mods.

I'm not saying the mods are doing a bad job, but there used to be a feedback mechanism between the mods and the community which is now gone because of this queue. I don't think it's necessarily good, but I get why it's going on. I just sort of wish it were different because I feel like things were maybe better when it was more open. It's possible a lot of the anti-racism things could also be addressed through the same now-missing mechanism.
posted by hippybear at 12:09 PM on June 30 [8 favorites]


I wonder if there's a way to use a locked thread as a stop-gap to a sidebar. Use a single thread to collect links to the various anti-racism posts, but keep it locked so the discussion stays in the posts. Whenever there's a new item to go in the banner, include both the new item and a link to the standing post that's collecting the recent conversations. That would allow mobile users to access it more readily than a sidebar and not require waiting until a sidebar could be coded for MeTa.
Or something similar by rolling up a (largely) static page like the FAQ pages, and use that link in the banners.
I'm not proposing not posting a banner update whenever there's a new anti-racism thread, to be clear, just a possible way to push back against the older ones scrolling off the front page of MeTa and getting less attention as a result.
posted by jacquilynne at 12:12 PM on June 30 [3 favorites]


I am in favor of greater transparency on MetaFilter in general, but if we get rid of the queue we should be prepared for fragile white people (and other privileged folk along various spectrums) to show up whining about their how their brilliant and necessary microaggression was deleted. Yes, those threads can be shut down once they've been posted, but we should ask ourselves if there's a net benefit to that kind of noise being actively filtered out. Or would it be preferable to know who is throwing this flavor of tantrum and to see how the mods and the community are responding? I genuinely don't know but figured it was worth asking. Sorry if this is a derail.
posted by zeusianfog at 12:14 PM on June 30 [10 favorites]


Metafilter: brilliant and necessary micro-aggression
posted by seanmpuckett at 12:56 PM on June 30 [1 favorite]


Sidebars don’t appear on mobile

That's a good point. (They do appear in the modern theme, but only at the bottom of the page which is not good in this case.)

These comments about the queue are reminding me that a lot of the rationale for MetaTalk UI decisions has historically been informed by a desire to reduce engagement, whether to try to discourage pile-ons and flameouts or just to keep the number of simultaneous potential headaches low. I think that makes sense for some kinds of MetaTalk threads, especially the kind that is or was often sort of dismissively described in terms of 'everybody gets sore at something sometimes', 'people just need to let off some steam', etc. The kind of threads that were mostly born out of fragility. For things like that maybe it's okay to have UI design that encourages conversations and interactions to die out on their own.

But for some conversations about fundamental issues, I think the very last thing you want is for them to peter out before things get much closer to resolved. Because then things fester unaddressed, eventually explode, peter out unresolved again, fester some more, and years later the cycle is still going, with a lot of damage done. For issues like that the historical thinking about how MetaTalk should work seems like a disservice to the site.

There's also historically been, afaict, a kind of thinking that's common in software development where users are seen as inherently annoying and impossible to satisfy, so the value of any given feature needs to be weighed against the headache of users nitpicking and criticizing it. Often that kind of thinking leads to conclusions like "doing X will just increase chatter, so we won't do it". Which again makes sense for some things, but is the opposite of helpful when dealing with issues that really do need to be criticized and discussed. And it's not like that "chatter" doesn't then boil up in other places when it's actually important.

All of which is to back up what has been said a lot in these recent threads, namely that the approach the site is taking to racism and to community relations just doesn't feel truly proactive, and it should. So I hope there's a lot of proactive thinking going on behind the scenes about how to keep crucial conversations alive with an aim to using them to both discover and address issues; how to increase mod responsiveness and the perception of mod responsiveness; how to let the community be able to check at any point and see which specific issues the site has decided to work on, where progress on those issues stands and what timeline exists for the future, without requiring everybody to keep on top of mod announcements buried in various threads; etc. Even if doing some of those things might feel like opening the site up to too much user oversight and input. And I hope the results of that proactive backstage thinking get placed in front of the scenes quickly, proactively. Because I think we're missing a sense that MetaFilter is not just struggling to keep up with the community's needs but actively working to be ahead of the curve and do more than the minimum at any point.

(For those of us who've been here for a while, remember how often it happened that someone would post a feature request and pb would pop in really quickly to say "sounds good, I've implemented it" or "sounds good, I should have that done by tomorrow"? This isn't meant as a comparison to frimble - pb was full-time, was working under very different conditions, most of the codebase was his design, etc. I'm also not saying it to highlight technical ponies specifically. But the sense of responsiveness that gave, that sense of being impressively on top of things, contributed a lot to the site. That's been missing for a while.)

MetaTalk needs to be conceived of as a place that helps the mods stay ahead of the curve, proactively keeping the community informed, proactively eliciting and listening to community needs, and proactively prioritizing those needs. Getting there is going to be an iterative process, but there are things that can be done quickly, and they should. And we should be consistently aware of the changes that are coming and of how they're coming along.
posted by trig at 12:59 PM on June 30 [6 favorites]


I am in favor of greater transparency on MetaFilter in general

Yeah, I hear that. The queue showed up when I left and money was tight and moderation went down to a slightly-lower level. And I think there's usefulness to the queue, but I also think that in the absence of better information coming out of Mod Town, the general message seems to be implying "Hey we're pretty busy" and "We'll get to things as we can" which is good as far as it goes, but it leads to a lot of community arguing feeling like if there is truly super-limited mods-and-dev time, they want their issue to take priority. Which makes sense. I think if people heard things like "This will take five days because of schedules" they'd (usually) wait, and not feel like they had to constantly lobby and clamor the whole time to make sure their thing took priority. It's bad for the community and stressful for everyone. And the queue can sometimes lead to "Post approved and here's a wall-of-text answer at nearly the same time" which can be good if the wall of text is "We're doing this, here's the plan" and less good if it's a "Well here's what we're thinking...." which can feel a lot like a softball way to say "no" or "not yet"

I know in some ways mods are trying to avoid the mathowie promise-and-not-deliver pattern (TravelFilter RIP ) from a long time ago, which was bad in its own way, but I think there's been a creeping lag and stagnation, and feature requests that either sort of happen immediately or kind of drag on forever. I know there are reasons, and I know those are even reasons I sympathize with, but part of being active in anti-racist work is just that, being active. It's not being non-active just because there's a good reason. And there's literally no way to determine the difference between well-meaning-thinking-and-inaction and plain old uncaring inaction. And I feel bad for everyone, but as with all power dynamic situations, the burden should be on the people in the positions of power to find a way to handle what needs handling.

Plant swaps and MetaTalkTails and anti-racist work MeTa posts should be able to coexist, and it's too bad that the community doesn't have the information that they need in order to be able to feel confident that the anti-racist work is being done (or worrying if it rolls off the main page). I know it can be stressful and sucky to have a public "to do" list, but as a confidence-instiller, it might be a good idea.
posted by jessamyn (retired) at 1:04 PM on June 30 [23 favorites]


I hate to add a level of additional work, but this is exactly why I and others use Confluence or something similar to track tickets and prioritize. When you get into the cadence of doing two week sprints and transparency to stakeholders, bundling up feature requests/bugs in one release suddenly doesn’t look like your giving precedence to one thing over the other. I know even if I had a mission critical feature, if it was an 80 hour task I might take a 1 hour feature as a mental break “I just want to get something done to feel like I did something.”

So everything gets bundled into say a July 1st release (arbitrary date), but if it makes sense from a development perspective to get this feature done first or just is a really quick fix, the perception isn’t one is getting priority as all the features promised get released into one go.

Open source-ish projects backed by commercial companies have a similar release cycle but they tend to just show if something is assigned and what the status is with a general roadmap. That may be better as I don’t think anyone here really needs to know every item and when it will exactly be done, just that it is getting done. We don’t need everyone on metafilter to be a PM.
posted by geoff. at 2:50 PM on June 30


Addendum: It might help to look at the transparency on how Github and Red Hat are moving away from sensitive language like master/slave or blacklist/whitelist and how they’re fitting it into release schedules, along with the transparency given. I have not seen any bug reports being called insensitive simply because diversity were concurrently taken place, so they might have a different way of community engagement but I have no insight into that unfortunately.
posted by geoff. at 2:55 PM on June 30


I’d much prefer frimble’s limited time and energy be protected from member scrutiny. Well intentioned communities just love to burn out part time help.
posted by michaelh at 3:18 PM on June 30 [9 favorites]


Yeah I guess what I was trying to suggest was transparency to avoid member scrutiny and also give members an idea of what’s going on. Right now these threads are dominated with a lot of calls on what to work on first, how to implement, etc. so maybe general roadmaps “fall of 2020” with a confluence board would help. That’s what Red Hat and Github appear to be doing while acknowledging they can’t just outright break vendor APIs and completely eliminate everything.
posted by geoff. at 4:00 PM on June 30 [1 favorite]


Rereading folks' suggestions and requests about various things, want to share a few specific thoughts:

We have talked some previously about the idea of a public "what's in progress" todo list and changelog as far as dev work goes, and I think it'd be good to come back around to it and make that happen. I don't think that's likely to take the form of an elaborate public-facing task management system, because we have one dev and I want them to spend their time doing dev work rather than doing paperwork, but something simple and publicly maintained as far as "this, and then this, and then this" and timeframe estimates where possible seems pretty doable without a lot of overhead labor, and it's something the mod team can pitch in on updating. Basically publicly surfacing the kind of what/when/how stuff we tend to already be discussing internally on the team about dev priorities in one predictable place.

Sidebars not being visible on the Classic mobile theme is worth taking into account as far as MetaTalk visibility stuff goes; that also gets back to the extra dev effort and difficulty of supporting multiple themes and in particular how poorly Classic mobile was designed for long-term use vs. the Modern theme's reactive mobile-friendly design.

I think the point about just having more entries on the MetaTalk front page is a good one, and will check with frimble about increasing the size of the index. I suspect that's a genuinely simple change to make, but I'll see. Bumping it up from the 10 or so it has been for years and years to 20 or 25 would help with overall longevity of posts remaining on the front page to see at a skim.

Disabling the MetaTalk queue is a complicated discussion but one I think we should go ahead and have soon. My basic feeling on it is (a) we can do it and I like the idea of doing so to remove that point of concern about transparency, but (b) I want to make sure we have set good expectations about what happens when, sans queue, stuff goes wrong that the queue would have prevented. That's enough of a conversation that I don't want to get into it in here, so probably just its own dedicated thread some time soon. But I want to at least explicitly acknowledge that it's something we've been exploring and hear the interest in.
posted by cortex (staff) at 9:42 PM on June 30 [1 favorite]


Part of the problem is that MetaTalk is doing too many jobs. Even if we weren't prioritising the racism post, the light-hearted community posts don't sit well with either the technical posts (e.g., pony requests) or the posts about moderation and community attitudes.

I suggest that MeTa be used solely for the latter, while everything else (gift swaps, post disaster check-ins, podcasts, reminiscences, etc.) get their own new category. MetaChat, maybe. Without those posts, MetaTalk would probably be quiet enough that it would take literal months for posts like the current racism one to leave the front page.
posted by Joe in Australia at 12:30 AM on July 1 [8 favorites]


Or maybe shunted into IRL, as it has a clear "fun" and "community building" theme? Agree with Joe that there is too much disparate stuff in here.
posted by Meatbomb at 1:13 AM on July 1 [3 favorites]


That's not really what IRL is for. Maybe MetaSocial?
posted by GenjiandProust at 8:18 AM on July 1


I like the idea of having all the disparate stuff in one place. Besides my general preference for disparate-ness, I think that if we roped off a special "this is where we talk about racism" section, that's an easy way to ensure that a lot of people would never interact with it. Here, if I come in looking for a fun post about plants, I'll be exposed to the racism discussions.
posted by kevinbelt at 8:30 AM on July 1 [10 favorites]


Maybe TellMe?
posted by GenjiandProust at 9:15 AM on July 1


I like that MeTa has so many kinds of topics about the site and community in one feed. The likely expansion to more than 10 topics per page (a good change) should address concern about any particular topic’s visibility.
posted by michaelh at 11:05 AM on July 1 [3 favorites]


I agree with Kevinbelt that we should not tuck the racism talk away in a corner.
posted by meaty shoe puppet at 12:13 PM on July 1 [4 favorites]


If it is decided to spin off feature requests into their own thing, how you could do that and not call it MetaPony is beyond me.
posted by pdb at 2:21 PM on July 1 [7 favorites]


stuff goes wrong that the queue would have prevented

Does this mean stuff that never made it out of the queue and onto the MeTa front page, as opposed to stuff that made it onto the MeTa front page, but not until the mods were good and ready? How much stuff does enter the queue but gets deleted instead of posted? And how much of that is just flameouts that I don't really mind missing?
posted by aubilenon at 10:19 PM on July 1


My biggest concern is the stuff that was a bad idea in the first place, yeah. That includes both the now quite rare "this is a terrible idea, period" meltdown/flameout stuff and the more likely these days examples of a well-motivated or understandable post that's framed badly enough that it'd get in its own way or cause the poster frustration or harm from the inevitable blowback. So there's a small percentage of MetaTalk threads that we end up talking through and tweaking with the poster, or work with them to redraft after an initial hot reaction has worn off so they can aim more the actual problem they want to solve or thing they want to address.

Most stuff that enters the queue gets posted. What doesn't is usually a mix of some bad idea posts and email-level requests/questions that didn't need to be a MetaTalk, and the occasional more odd and complicated situation. I've done a roundup a couple times before of what's been rejected and why over a given previous six month period, and I could try and knock another out tomorrow.

The stuff that we hold sometimes just for timing wouldn't be a catastrophe to not hold, and I'm not as worried about it. That'd mostly need folks to keep expectations in check about mod availability to give a definitive answer on things.

There's a lot of nitty gritty if I dig into it, and I don't want to do that right here and now, so again probably best something for its own MetaTalk discussion in the near future. I think looking at moving away from the queue is worth talking about; I think revisiting some assumptions about how MetaTalk works is too. I prefer this as an all-inclusive topical space and would rather focus on how to make it work better for everybody than to try and radically restructure or partition up what is here, is my general feeling.
posted by cortex (staff) at 10:59 PM on July 1


Looking at the change in MeTa posts and comments over time, it doesn't look like the queue introduction in 2014 had a massive impact, quantitatively at least.
posted by Klipspringer at 11:20 PM on July 1


The stuff that we hold sometimes just for timing wouldn't be a catastrophe to not hold, and I'm not as worried about it. That'd mostly need folks to keep expectations in check about mod availability to give a definitive answer on things.

Honestly for some stuff I think I'd prefer to not have a multi-paragraph comment from a mod right up front, to let the post breathe a bit first. Maybe that wouldn't end up being best for some threads, I can't say, but sometimes it can feel a bit stifling.
posted by ODiV at 8:44 AM on July 2 [3 favorites]


I hear that and I've tried to get away from doing that as often over the last few years. Obviously I still have some work to do there, and I appreciate that it's been a source of frustration for some folks and I apologize for that.
posted by cortex (staff) at 9:09 AM on July 2


looking at moving away from the queue is worth talking about
again probably best something for its own MetaTalk discussion in the near future

Just a friendly reminder to be mindful of the message this kind of language can send, especially in well-trodden areas like the queue. It does not send the message that stuff will happen in a timely way.

I believe you that you'll eventually do this, cortex, because I like thinking you're a stand-up guy. But language like this—sentences full of filler someday-we-should-probably-maybe "looking at it is worth talking about"—really doesn't help inspire confidence.

What happens after we talk about looking at something? Then does the "real" discussion culminating in a hard decision about whether it will result in action happen? Can we just skip ahead to that one instead?
posted by heyho at 9:19 AM on July 2 [2 favorites]


I gotta defend cortex here, because I'm also a probably/should/hopefully communicator. It drives my project managers (and my wife!) crazy, but I don't want to commit to something I don't know if I can deliver. It doesn't inspire confidence to say "we'll look at it", but it also doesn't inspire confidence to say "yes, we will have it done by _____", and then miss the deadline. If that happens more than once or twice, a firm commitment means nothing more than the non-committal response, except that the non-committal response is honest. Even in a well-staffed, corporate organization, planning is tricky and prone to error. Personally, what cortex is saying is exactly how I'd want the planning process to proceed: add to agenda, discuss high-level feasibility and impact, then work on specifics.
posted by kevinbelt at 9:54 AM on July 2 [1 favorite]


Yeah, I hear you there. That kind of reflexive qualified phrasing is a long habit for me and one I've been trying to edit out over the years. Still a work in progress, and one I don't manage consistently when I'm tired, which is when I should be double-checking it more.

To redraft: I specifically don't yet have a timeline for a new post about MetaTalk queue stuff. I'd like to do so within the next two weeks and I think that's a responsible estimate. The mod team has discussed removing the queue a bunch previously and we've been revisiting those discussions this week to refresh ourselves and outline the major considerations involved and figure out to what extent we can fold in other MetaTalk-centric issues folks have been discussing without adding more delay to the process.

There will be some amount of implementation involved in making that change, and frimble's priority right now is a couple other things we've been talking about the last few days, and that affects the lack of a more specific timeline as well.
posted by cortex (staff) at 9:55 AM on July 2 [2 favorites]


kevinbelt, Yep, I've worked with people like this all my life; hence, my suggestion to stop using that kind of rambly language that sounds empty after a while (even when it's not).

Thank you for a clearer response, cortex. Much appreciated. (And really, that was a friendly reminder. I like you, and I really like this site and want to see it thrive in the 21st century; I have a feeling we're all going to need it.)
posted by heyho at 9:59 AM on July 2 [1 favorite]


add to agenda, discuss high-level feasibility and impact, then work on specifics.

Further banging-on about transparency, and stated goal to improve it: the practical approach to corporate organization (which metafilter absolutely is: it's a privately owned business, ad- and donation-driven revenue with employee payroll etc expenses; the community aspect makes that easy to forget; said reminder applies to rando member me as well!) decision-making are all areas where greater transparency can be applied.

The basic structurelessness of metatalk is often/can be a structural and systemic problem, and it's one that can be addressed. Rather than maybe sometime possibly likely probably we'll discuss sometime in 1d6 weeks: set a schedule. Bi-weekly, monthly, whatever--state of the site MeTas. Agenda items transparently presented; use a section of the wiki perhaps, or (when technical solutions in place) a 'stickied' Upcoming SoS Agenda locked post. 'Meeting notes' of last official thread. Presentation precis of where ownership and moderator thoughts are on last official thread's points. Etc.

The idea isn't locked-in deadlines or specific-date-and-time promises or to roberts-rules-of-ordering up the joint, but to add more, or rather healthier, structure than free-for-all wheelspinning without direction, where it often seems like progress is only being made by traction gained by fed-up bodies throwing themselves under the spinning tires.
posted by Drastic at 10:46 AM on July 2 [1 favorite]


cortex, can you please comment on some of the other suggestions made in this thread? there's the original bug fix + feature request plus lots of solid ideas for improving this area of the site (eg pinning threads, more transparency with dev efforts, etc).
posted by too bad you're not me at 11:22 AM on July 2


cortex, can you please comment on some of the other suggestions made in this thread? there's the original bug fix + feature request plus lots of solid ideas for improving this area of the site (eg pinning threads, more transparency with dev efforts, etc).

Sure, here's an attempted run-down in brief (or as brief as I can make a 7-point list). Obvious note here that a comment like this way down a thread is a less elegant solution than a notional roadmap and changelog sort of deal (see point 3 below), but something is better than nothing and I can do this something right now, so:

1. Banner UI issues

The post had two requests: (a) add an account-wide rather than device-specific banner dismissal button, and (b) fix the bug that makes it hard to hide long banners on mobile.

frimble is looking at (b) this week. We retrofitted the banner code at one point to accommodate longer, mutli-line messages so we could include things like multiple links practically; this bug is a side-effect of that. Needs fixing, hope to find a solution soon. In the mean time we may aim for one-line banners to avoid the problem. No timeline for the bug fix, because it's a bug and we have to successfully identify the cause.

I don't know how practical a change (a) is, will need to discuss with frimble. The current function is cookie-based and light-weight to manage. Lower-priority issue; I appreciate the annoyance, but it's not breaking the UI and we don't change banners very often, so for now expect to continue to click once on each device. When frimble is able to look at how that would work, we can look at an actual timeline for implementation then.

2. A pinning solution for long-term visibility of especially important MetaTalk threads

I agree this would be good. We've got at least two major routes right now we could take: pinning a thread or threads to the top of the MetaTalk front page, or creating a MetaTalk-specific or site-wide sidebar widget for important discussions. The sidebar approach has the advantage of fitting a model we already have: the US Politics sidebar on the front page has most or all of the code functionality we'd need and could be cloned, tweaked, and made visible site-wide. Pinning threads to the top of a subsite page we don't have a process for right now, but we can explore.

I don't have a timeline for either of these; I'm hopeful that cloning up a sidebar widget would be relatively easy, but incorporating it on multiple subsites would be additional work vs. just putting it in the MetaTalk sidebar. Both of those also have mobile view considerations to take into account as well, which means extra design time, but if we can get something that works well on desktop and well-enough on mobile sooner rather than later we'll aim for that as a start and then polish later.

3. Dev roadmap (i.e. "what's in the works, and when?") page and site changelog stuff

I think these are good ideas. I agree with folks arguing to just do something simple with this: we can take the internal discussions we have about this stuff during team meetings and through the week and put them somewhere visible (a page linked in the MetaTalk sidebar, probably). Just a thing updated weekly-ish would be fairly simple and a big improvement in visibility. I'm envisioning a static page with (a) a list of upcoming/in progress dev and mod projects with approximate timelines where possible, and (b) a list of recent completed changes, bug fixes, features, etc.

I am guessing we could knock out a basic version of that this week, frimble's availability pending. Ideally it is something that the rest of the mod team can do the text maintenance on so it's not an additional frimble responsibility beyond getting the page installed to begin with. We have a team meeting every Sunday so that'd be a natural time to collate updates for this on a regular basis.

4. Updating the flagging toolset

This is still in progress as something we plan to do something with fairly soon. I don't have a more specific timeline, but I'd like it to be the next couple weeks. I would prefer to make a smaller change sooner rather than get bogged down in trying to perfect the changes, so aiming to change [!] to [flag] and update the documentation could be the whole scope of the actual visible change to start. This is slightly lower priority than some of the other stuff we're working on.

5. Timestamping MetaTalk posts with submission vs. approval time

I get the logic behind this and I think it'd make sense to consider if the queue as it exists now stays in place longer term. I currently expect us to discuss moving away from the MetaTalk queue in the next couple weeks and am going to reserve frimble's working time for stuff less likely to become moot in the near future, but we can revisit if something changes there.

6. Setting up a temporary anti-racism subsite; setting up not-serious-stuff alternate subsite for chatty MeTa stuff

I don't think this is a practical scope of solution for the current problem of several concurrent important MetaTalk threads. I appreciate the motivation, but spinning up a subsite is a ton of work and I'd rather improve MetaTalk's function and visiblity/sidebar stuff to better support how it works now. I feel our problem is with visibility of key threads, not with the all-in-one nature of MetaTalk as a community hub.

7. Larger number of posts on the front page of MetaTalk.

This is very doable, frimble has since confirmed it should be an easy change to implement. I expect we can do so in the next couple days if other dev emergencies don't come up.
posted by cortex (staff) at 4:23 PM on July 2 [3 favorites]


Quick update: frimble was able to make that change in (7) last night; there are now 20 metatalk posts on the front page, and we can see how that feels for a while.
posted by cortex (staff) at 7:58 AM on July 3


Sweet. It feels like the ship's engine has started up again. This is good.
posted by heyho at 8:42 AM on July 3 [1 favorite]


20 feels right on mobile.
posted by michaelh at 9:23 AM on July 3


« Older Music Swap 2020 | now in beta!   |   Singapore Elections post? Newer »

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