joshmillard

Full disclosure: I used to run the site. This is, at this point, just my personal take; I handed it off to other folks in the community whose work I believe in entirely and I am glad they are working on this now.

So, MetaFilter already had and has an enormously higher than normal rate of voluntary contribution per capita; when the first call for community support went out in 2014 on the tail of the great web ad market fuckification, something like 20% of the userbase stepped up with contributions, which is an unheard of number.

It's obvious to say "just charge everyone" but it's unreflective of both the buy-in the community already has from folks willing to support the site, and of the degree to which slamming a door on a historically open community flies in the face of the ethos the site has tried to be increasingly inclusive and increasingly aware of the need to build a community out of not just folks with idle subscription cash sitting around.

A paywall is great if your goal is to use it as a stick against non-paying folks. If you want a community to include folks who want to be there vs. just folks inclined to pay to be there, it's a poor solution. And if you build a good place, people will pay even if you don't make them. That MeFi still exists at all post-2014 is proof of that.

I got to use the studio's supply, so I haven't had to really sort out glass buying myself. And yeah, I was lucky to benefit from access to a really great collection of glass. But there are some commercial firms (Spectrum does very flat, transparent glass; Delphi does stuff with more character) and you can go haunt estate sales, watch for hobbyists getting out from under their supply, check with art schools, etc.

Totally feasible! You could probably put a whole one-person workspace into about fifty square feet without getting cramped, less with some squeezing and multitasking, and I'd guess basic toolset would run you less than $500 retail.

The grinder I used is a pretty simple unit with a water reservoir tray under the grill that keeps the bit wet via a sponge feeder; probably about $100 new, used bargains if you go looking, but, yeah, water feed of some sort is key to the grinding working well on glass.

Yeah, this isometric rendering has nice hints of Necker Cube illusions to it.

I was surprised and please how much I took to it, honestly; I went in cold and was prepared to not really get a knack for it.

But easy-looking is also a product of writing it up after the fact, and learning under good instruction; it was still a solid 20+ hours of work with a fair amount of cursing and bits of messing up and redoing stuff.

In any case, whether or not I caught on quicker than average, the process itself feels very accessible to me. I can recommend giving it a shot a lot more readily than I'd have guessed before I tried it myself.

I'll agree that "a near impossibility" feels like a real big overstatement; curves are very doable and common in hand-cut glass. That's true even in my first piece, under discussion, which is chock full of curves. I talk a bit in the process diary about getting comfortable with doing curved breaks after being over-cautious with my first few bits of glass.

That said, I can see how a CNC approach could help with the speed of the process; shallow/simple curves are easy enough to do reliably, but deeper, more complex, and significantly concave curves require additional effort and planning, potentially a lot of manual work with grozing pilers to break out a concavity, and additional grinding to finish out the shape.

So being able to turn a machine process to some of those cuts would save a lot of manual effort and potentially get you right up to the final polish-grind stage in one go, which certainly sounds attractive for some kinds of design.

Also, if you like my mathy design instincts with this stained glass piece, you may enjoy looking through the painting work (mostly oils) I've done over the last couple years:

http://art.joshmillard.com/

It's a shockingly approachable medium! I was really surprised at how much it wasn't a series of frustrating technical or fiddly bits.

Well, hey, that's me! Howdy folks.

This is remarkable for the terse confidence with which it is misapprehends the actual structure and content of the site.

I'm still a member! For the record, we resist implementing threading or commenting by reference because it's makes for an unreadable discursive shitshow when trying to follow busy, active discussions.

Staying relatively on topic is an unrelated aspiration and one that we mostly let flex a lot depending on the specific thread and context.

Or to take it from another angle, Reddit couldn't be as large.

Which is sort of at the heart of what Matt was talking about with scaling community -- it's not that large is bad or that profit is bad, per se, but if your goal is to have a sense community ethos and some degree of coherence to the spirit of what being a site member means, you have to say no to some of the things that make big userbases and ever-upwards biz growth doable.

MetaFilter's major revenue growth up through 2012 was despite a lack of, not because of, any attempt to chase down big userbase growth. It was a ride on the AdSense wagon during a very big period of growth on that end, while the site maintained a reasonable slow increase of users. Likewise the AdSense crash (and the double whammy we got for the punitive misclassification) wasn't the result of shutting down or reversing user-growth measure. The site strategy of maintaining a community that doesn't grow or change too fast was consistent throughout that long stretch of years, and it's something that means we didn't explode by orders of magnitude but also that, even after a rough period, we also haven't collapsed under the weight of too much growth-chasing overleveraging our resources and debts and so on.

Reddit is a big, complicated topic about which I have lots of thoughts but no real interest in digging in on here or now. But there are some very complicated differences in the implications of what community, userbase, and moderation mean as words and ideas for a site like MetaFilter and a site like Reddit that make casual straight-across comparisons using those words essentially useless. The difference between volunteer and paid moderation is a very good, very big example of that.

It's a mix of things; Deck, AdSense (for all the trouble they're still a major and these days fairly steady source of revenue), mostly-passive Amazon affiliate, and direct support from the MetaFilter community.

We've looked at other ad rubrics and been disappointed with the specific performance we've seen in a couple test cases, but the volatility of the ad economy means that it's certainly something we'll be reexamining continuously as time goes by.

The warning was quoted from the source, so your beef is with The Guardian.

MetaFilter has no shortage of strong language and it's not against any kind of community guideline if it's not being used in an intentionally shitty way; as a moderation staff we've pretty clearly set those expectations straight on the rare occasions that someone has complained about sensibility-offending surfeit of fucks and shits and damns in this or that conversation. I mostly just expect it to be done with a degree of skill and thought.

The exception there is slurs and the like; there's strong language and then there's racist/misogynist/homophobic/etc epithets, and the latter subclass is a different beast entirely that in specific discussions of language can bear mention rather than use when the distinction is clear. But that sort of thing otherwise can, indeed, fuck right off.

> 2. The smart thing financially would have been to let the community die but, he didn't and paid the price

So, I'm a skosh biased, as the person currently running MetaFilter, with a moderation staff that's managed to grow back to close to the pre-crisis status quo, but I don't think that's really the best take on it.

With the benefit of hindsight, mostly what should have happened differently and for the better was to take the downslide more seriously more quickly in terms of changing our spending and saving and model for what the next year or three years looked like, and to have gotten the community involved in that discussion sooner. We'd likely be in more or less the same place we are today, which is smaller but pretty steady and the community doing well, but with fewer lurches and less pain and fumbling along the way.

I appreciate Matt's public frankness in this piece about some of the winging-it nature of the site's business history and the missteps along the way, because I know he's always tended to want to be more private or cautious about that sort of thing. And I think the nature of MetaFilter as a tiny business and a relatively close-knit community informs some of that dynamic; it's hard enough for a business to break bad news to customers, but to folks who are genuinely members of a community there's a whole other emotional commitment involved.

Heh. Hello there! This is a thing I made.

Your wager is good, on both counts; the current corpus is based on the first 14 months of strips, which comes out to about 10K total words for Calvin (who has more lines than the rest of the cast combined, in that chunk of strips). That's not enough to generate a lot of variety on anything other than very common word combinations. Strings of prepositions and articles are the most likely inflection points where you'll commonly see two distinct phrases glued together.

If I get the whole strip run into the corpus, that'll kick Calvin up to something more like 80-90K words, which will help with the variety a good bit, and the other characters will have more of a shot of current Calvin-like variety, but it's still a relatively small training set.

By comparison I've done some markov model experiments based on multi-million word corpora and that gets a lot farther into the territory of regularly producing satisfyingly weird disjunctures.

That's one of the fun little oddities Gary Larson talks about in "The Prehistory of the Far Side", where syndication captioning errors led to a couple of swaps of Far Side and Dennis the Menace cartoons, with weird results.

Heya, that's me! I also wrote up some details on the how and why of the whole thing, if you're interested:

http://www.joshmillard.com/2015/07/06/calvin-and-markov/

In the interest of truly returning to the days of Nintendo Power hijinks, I have added a high score table that is powered by people sending me screenshots of their high scores.

Early days yet; I'm looking forward to folks getting all kinds of stuff either installed or written from scratch.

Though I did spend about fifteen minutes working on a pure command-line turn-based version called Roguey Bird a few months ago before realizing I didn't actually feel like spending more than fifteen minutes on it at the time.

Decided we needed an arcade page for all the sweet games:

http://www.tilde.club/~cortex/games.html

No point in making an easy one, I reckon!

It's been a wonderful weird little thing so far, with folks having a mix of nostalgia for and first exposure to shell account stuff, playing around with oldschool webbery, hacking little things together across user accounts, and just generally being great people in a nice place.

I got inspired yesterday and wrote us up our own custom low-rent version of Flappy Bird:

http://www.tilde.club/~cortex/js/tildebird/

Man, I am sympathetic if you got an obnoxious answer like that, but for what it's worth as one of the folks who moderates it that as stated is a comment I'd delete from an Ask Metafilter thread in a heartbeat and would expect other users to have flagged a bunch as well. The system isn't perfect but "man up and grow a pair" is really precisely deletable as a crap non-answer in our moderation rubric.

It's definitely a matter of taste, and it's fine that different folks prefer different stuff. The flat model works well enough for Metafilter -- we're constantly seeing thoughtful, responsive conversations on all kinds of topics in the site's discussion threads -- but it's a virtue of the web that there's a variety of models out there so different folks can have their preferences met.

That is our existing paypal account, yes.

We aren't formally doing any fundraising at this point -- that's an account we've had set up for ever mostly for the once-a-month "hey, here's five bucks, buy yourself a beer" tip from a random happy/guilty user -- but if you toss money in that, it ends up with us, and I won't tell you no.

If you want to be extra sure about the paypal link, visit the one at the bottom of Metafilter's about page, here:

http://www.metafilter.com/about.mefi

Sure. For what it's worth, though, we've continued to see steady signups for years and years; the concern over the ossification of a conservative site design is a reasonable one and one we chew on plenty, trust me, but as far as that goes we haven't all else aside seen active new-user interest go away.

What teraflop said. cortex here, one of the mefi moderators.

I totally understand where arguments along the lines of "automate and eliminate human costs" come from and there is absolutely value in finding ways to simplify/streamline/remove work, and we've developed a lot of custom tools for the site over the years to do just that. It's something that we'll have to do more of now as we transition to the smaller staff, taking a lot from what we've learned over the last several years as guidance.

But the fact that Metafilter has attentive, thoughtful human eyes on the site and hands on the till is a defining part of the culture and nature of the site. It's not a question of whether it's a good business expense, it's a question of whether it's what makes for the kind of community we want to maintain.