[MeFi Site Update] July 19th

July 19, 2021 12:15 PM

Hello Metafilter! Please find another update on the state of the site. Reminder: I will be the only mod monitoring this thread so please be patient as I reply to your feedback and questions. If you have any questions or feedback not related to this particular update, please Contact Us instead. If you want to discuss a particular subject not covered here with the community, you’re welcome to open a separate MetaTalk thread for it.

Technical Changes

- Several small behind-the-scene dev tweaks to better support changes to the moderation schedule and help with flag-related alerts for the team
- Rectified an old anonymous-question-related block affecting a community member

General Projects we are working on

- Privacy Policy has been completed and you can read more about that here.
- Changes to moderation coverage seem to be going well so far.
- We’ll start working on the Terms of Use once we have completed the upcoming Fundraising.

Flag feature updates
cortex, on the possibility of accepting user assistance in implementing changes to the flag feature:

the principle is good and the offer is appreciated; putting it into practice is more complicated. The main holdup in getting a change implemented here isn’t conceptual at this point, it’s the legacy mess of code, and that’s not trivial to bring volunteer developers in on in a way that would save time. frimble has identified two paths toward a change there: fixing the code in many dozens of places distributed throughout the code-base, or rework how the site handles bylines everywhere to support a much easier-to-maintain common template system in the future.

I’m up for talking about what a volunteer effort to get acquainted with the codebase would look like and what labor it would involve, so interested folks are welcome to contact me to talk about it. As a long-term plan to get more hands in the community doing useful maintenance, it’s absolutely worth exploring. But I need to set expectations on what big projects it would be to (a) get folks usefully familiar with the codebase, and (b) coordinate volunteer coders in a way that won’t impact frimble’s ability to get stuff done or the mod team’s limited resources. It’s not trivial to do and it’s not a quick dip-in, dip-out sort of project.

- The new CRM System has been put on the back burner to prioritize the completion of the Privacy Policy and to work on the Fundraising.

Site finances & Fundraising
- We’ll be talking more about fundraising soon in a dedicated MetaTalk post.

- From cortex, about the discussion of Autostraddle’s fundraising processes in the previous site update:

I think there's some good stuff in Autostraddle's documentation of their fundraising rationale and processes; there's a lot of ways that they're a different organization and site than us that make a straight-across comparison unhelpful, but the core notion of disciplined, focused fundraising is a good example and something we haven't yet accomplished well enough with MetaFilter. A lot of what they talk about in terms of things they *don't* do despite recurring recommendations from readers and community members also resonate with me; there's a lot of "well, why don't you just do x?" stuff that they explain well that translate pretty well to similar conversations we've had in MetaTalk over the years.

The core issue with MeFi and fundraising -- that none of the team are experienced fundraisers -- is a real one, and something we haven't done enough so far to mitigate, despite making bits of progress over the last couple years. I’m trying to reach out more about that; to the extent that folks in the community have the capacity and inclination to put their own experience to work there in support of the site, I am absolutely open to having a more detailed one-on-one conversation about it. Please reach out to the contact form or to me directly at [e-mail] if you'd like to discuss that or to just more concretely and personally share ideas or brainstorm, etc.

Feel free to start adding questions and feedback to this thread and I will do my best to address them as promptly as possible.

posted by loup (staff) to MetaFilter-Related at 12:15 PM (11 comments total)

Re: the flag feature updates: Is it accurate to say that you've decided how you want the new tags to operate (i.e. that you know what you want the new flag feature to look like to the users, how you can make its function more intuitive, and what [if any] expanded options you want to present to the users), but you're waiting to see how much frimble work-hours can be devoted to it?

OR are those questions still open and dependent on a frimble work-hour estimate in order to be answered?
posted by The Pluto Gangsta at 8:43 PM on July 19, 2021

Accepting coding volunteers is going to be counterproductive even in a shiny clean codebase.
posted by michaelh at 10:31 AM on July 20, 2021 [3 favorites]

Being able to ask for help is hard, but a necessary part of any healthy relationship.
posted by aniola at 11:12 AM on July 20, 2021

It's accurate to say we have a reasonably settled take on what changes to the flag function we should try to make—better visibility overall, easier tap targeting on mobile in particular, basically hewing toward better accessibility of the feature in line with the past discussions about it—and that the holdup is that implementation of any site-wide change to flags is a big job for messy legacy codebase reasons.

To expand a little on what I passed on to loup for the post above: the bylines for every post and comment on the site are handled in (sorry, 2000s-era Matt and pb, but) a really ugly tangle of a way. Even though they're all basically identical in look and function, the way mathowie implemented them back in the day led to them each being an independent chunk of code for different subsites, contexts, etc. Definitely frimble would be better able to lay out the specifics, but the short version is making a change to the flag function (or most other things in the bylines) currently involves a bunch of sorting through a bunch of different chunks of site code to make nearly-but-not identical changes simultaneously, every time.

So our two options for making a change to the flag function are:

1. frimble settles on a specific change of format/function, and goes through and makes that change many times in slightly different ways, and if anything needs change, does that all again. For an emergency (e.g. we discovered some sort of security issue with the byline code or whatever), that's the way to go. Tedious, doable as a one-off, doesn't scale to ongoing tweaking.

2. frimble reworks the site's byline code so everything is operating off of a common template, and then changes can be made to that template rather than a bunch of different places. This is what they've been putting some time into; with that done, things like changing the format or function of the flagging feature will be far simpler in the long run and much more reasonable to make multiple iterative adjustments to. Unfortunately, getting that implemented and tested is a time-consuming job in its own right, since it involves touching and testing all those points in the code from option 1 above more thoroughly as an upfront effort. I don't think frimble has a good estimate on the time there, but I'll check in with them about it.
posted by cortex (staff) at 9:34 AM on July 21, 2021 [4 favorites]

Uh why not just JavaScript hack it on the front end and then hack it on the backend until the ROI justifies rewrite?
posted by geoff. at 7:02 PM on July 21, 2021

Open source the code and let people contribute.
posted by blue_beetle at 7:50 PM on July 21, 2021

Alright I'll bite and I won't thread sit. If this is too long or off-topic mods feel free to delete. Quick and dirty background: I've worked for large management consulting firms and specifically large web presences. Yeah I worked for big firms that are evil banks and the "big 3" white shoe firms in some capacity and yes they're good at righting ships. I'm not here to argue that's right or wrong and obviously no one wants to gut Metafilter and squeeze blood out of a stone, but I unfortunately have learned things in my career. First and foremost successful organizations: fundraising or not, usually hire outside executives that have a track record of turning things around. This is NOT a slight on Metafilter, cortext or anyone else. This also doesn't always work, but I can only say what I've seen. What I've seen fail is bottom up, let's all pitch in like a farmer market. It tends to yield to the most vocal members winning over, lack of a mission statement, etc.

Open source the code and let people contribute.

Have you been involved in a non-trivial open source product? Metafilter is having trouble fundraising, running a successful open source project is very, very difficult. it also relies on talented, well paid individuals from in the industry. It involves a board and usually has a well defined technical goal with well heeled backers (Kubernetes, the variety of Linux projects, etc.). Pull requests are complex requiring showing actual improvement to a specific problem. Moving from from websockets to browser events. Anyone involved in an open source project knows even the most well run ones result to politics and in-fighting. Open sourcing might help frimble here or there but the overhead is enormous.

Re: Rewriting code

In my experience and what I've learned from those smarter than I is rewriting old code bases without clear metrics (load times keeping users down? cost an issue? keep patching it), nearly always is useless. The MO usually is identifying if there's significant real liability and determine a roadmap for new features having an interop with old code. That's a huge simplification but how large software projects generally work without gobs of investor money. Silicon Valley has graveyards with well meaning programmers rewriting something from scratch. Even Facebook chose to spend time optimizing the PHP engine than just writing it all in CPP or equivalent at first.

Re: Metrics and goals

I cannot speak for the mods but usually goals and metrics are established for something like this. More comments? More users? Longer page view time for users that are here? Gamification of existing features such as Reddit stars? There's a ton of things that are non-intrusive and don't really harm the site that let users pay for things beyond an NPR subscription model while measuring success. Similarly diversity initiatives are great, but would love to see an ombudsman that's not connected with the site actually auditing the diversity initiatives.

Well that's pretty much all I had and I said my peace. I'd love to see more activity on the site and a wide range of opinions and a happy mod staff. They're good at community outreach but running a business is hard and not really rewarding. This is why you see high paid executives at non-profits. I really like how Metafilter wants to come together and spend the weekend painting this like it is a local school but to get Metafilter to succeed we need to define what success is, realize this is going to not make everyone happy and fundamentally shift away from a really off-putting interface. Just because IRC or Usenet works and people use it doesn't mean it is really engaging users.

And since this is long anyway, the above rant was more towards what I see a lot of people suggest and what I've see not work, if I had my way:

- Complete redesign by a leading digital/product agency and UX ($50-100k?)
- Deep audit of existing technical debt and what code can be used as an interop vs estimate for complete rehaul
- Content audit, including aggressively removing non-high engagement portions of the site
- Removal of $5 fee and institution of Reddit style stars
- Removal, generally, of the vestiges of scalability issues like max posting on sites at least until it becomes a de facto issue and defining when that threshold should be. Other sites don't no longer do it and survive.
- Editing in some form which I believe was again an issue with 1999 ColdFusion technology and fear people would wholesale edit comments. We now can keep versions easily and cheaply to prevent this.
- Allow users to delete comments, again negative reactions to comments are pushed down a memory hole and sometimes can be confusing with a cryptic mod message. Reddit makes it hard to see but doesn't remove it. This contributes in my opinion to a confusing in-group mentality to new users.
- A move away from non-linear home page to filtering popular posts if not algorithmically but through editors.
- Greater multimedia in the front page which is more engaging and again, something not possibly in 1999.

These are huge and overwhelming asks. I can't imagine that Metafilter wouldn't have some value where an outsider investor could help ameliorate these costs. I used Reddit extensively as an example as I'm being a bit lazy but nothing i suggested is technically difficult. Volunteers can certainly buttress this but maybe embracing monetizing our user base, some less intrusive advertising and gamification can go along way. Me paying $5 so a gold star can appear next to a name is pretty simple ask.

Anyway that's all I'll say I hope I didn't hijack the thread. I wouldn't spend this much time outlining this if I didn't love the site and wanted to it to succeed. This is not at all a criticism, and I feel as a lot of suggestions are taken personal. Take this lightly, like if I said, "If you had a pool you'd have cooler backyard parties!" along with some general advice on how I've seen projects done right. Please if I came across as crass or rude I did not mean to, and to keep mod sanity I won't be responding but if anyone wants to PM me feel free too.

(Also wouldn't it be cool if we had a discord or something more modern than a text chat? I feel as if conversations like this would be great if there was a more fluid realtime chat and not locked into a comment section)
posted by geoff. at 10:34 PM on July 21, 2021 [12 favorites]

Uh why not just JavaScript hack it on the front end and then hack it on the backend until the ROI justifies rewrite?

I'm not sure if this was intended as a joke, but in case it wasn't:

The reason not to do this is that every hack you add on top of an existing system will make both the potential rewrite and developing future features more expensive, more time-consuming, and more likely to introduce bugs. If you go down this road the result is the ROI will never justify the rewrite, because the rewrite will become more and more expensive. Instead, the application will just accumulate more and more ugly hacks, and become more and more incomprehensible, until new feature development in a timely manner and without breaking existing functionality becomes virtually impossible.
posted by firechicago at 4:32 AM on July 22, 2021 [8 favorites]

The california disclosures are interesting. You don't explicitly refute that you are gathering thermal or olfactory information, just listing audiovisual stuff, which I find mildly amusing for some reason.

On a more serious note, is the voluntary use of pronouns on profiles not considered gathering information on gender, as it's defined under California law?
posted by Chrysopoeia at 3:58 PM on July 22, 2021

which I find mildly amusing for some reason.

Mild amusement is know to the state of California to cause etc etc

is the voluntary use of pronouns on profiles not considered gathering information on gender

The legal folks working through this with us made a clear distinction between information required by or involuntarily collected by the site (e.g. the minimal stuff we collect at signup via PayPal) and stuff folks share voluntarily. So folks can include pronouns or other gender stuff on their profile page; they can also remove it afterward if they choose; we don't require anyone to do those things; and we don't do anything with that information except allow a user to optionally display it. Likewise folks can share literally anything else on their profile page, or in comments or posts on the site.

That's ultimately another "this is why we hired lawyers to do this" item. Trying to navigate the space between literal reading and prudent legal interpretation is not really in my wheelhouse and I'm glad to let a professional do the steering there.
posted by cortex (staff) at 4:16 PM on July 22, 2021 [16 favorites]

carmicha:

Eschew technology upgrades and, instead, use available resources (or fundraise) to identify and recruit the people who want what MF offers and are charmed by the site's anachronisms, like a completely text-based experience

First up, not remotely a comment intended as aimed directly at you carmicha and apologies in advance if it comes across that way. I'm just off on a rant.

Bolded the bit I'm referencing as a way of trying not to take the idea completely out of context, but IMHO a "completely text-based experience" on the 2021 internet is still not all that anachronistic. www.reddit.com is a semi-popular site which from a user's perspective leads with a picture of a cat or an embedded link to a video of a cat or the terrible site video player failing to show a cat. The content however is almost entirely text comments (with the very occasional inline gif, which excites remark as inline images can here) and the participants engage extendedly both with those comments and each other (threaded obvs, but I'm not here to argue that)

I don't think Metafilter is Reddit, TikTok, Vine, Facebook, Twitter nor tbh Gopher or Usenet if it comes to that, but it is definitely still within spitting distance of the online zeitgeist. Granted I haven't been around here much for a good while but people pushed back against flipping titles on the Blue and yet a frontpage thumbnail/inline image isn't gonna kill anyone and might make the Reddit-proven, internet-still-equals-WORDSWORDSWORDS, text-heavy conversations more initially accessible.

(full disclosure: I use the Modern Dark theme. Others use others)
posted by I'm always feeling, Blue at 9:33 PM on July 24, 2021 [3 favorites]