For what it's worth, banter on social media with someone you're not familiar with is almost always playing with fire. It's really easy for something to come across wrong or just be kind of exhausting, and this effect is magnified the more of a spotlight that person has. You're just one of thousands of interactions they've had that day/week/month, and so unless you know they enjoy that kind of playfulness, I find it's worth assuming they don't. This is, ironically, especially true with people who publicly post in that tone, because they get it coming back at them all the more frequently.
It really doesn't have to. I thought I was being clever when in a thread I likened something Michael Godwin said to being Nazi, because I thought it was a funny self-reference, and he just gave me the Twitter equivalent of an eye roll and moved on.
It is, though the other side of the coin was always uploading an image for your niche hobby sub and being inundated by sneering imgur users who wanted to know why your keyboard looked "stupid." The site developed a reputation (among reddit users) for being full of thoroughly incurious people with very confident opinions on your niche interest.
They function differently, but in terms of what gets communities to actually form and stick around, I think they're very similar:
- Making a new space for your community is trivial for anyone; it can be done in seconds with a few clicks and all you have to do is choose a name.
- Many people already have an account, so you don't need to convince everyone to sign up for a new platform. (Which scales with the platform's size, like all network effects.)
- Communities have their own space they can adjust to their liking, rather than being a vague cluster of nodes with a similar interest like in other social networks.
- Owners of those spaces have a lot of leeway to run things as they see fit.
Personally, I don't like the growing trend of every community being a Discord server that is going to collect dust in the corner of my chat window unless I commit to keeping up with it every day, but I understand why it's happening. Discord is an adequate social hub for any project or hobby group with a very low barrier to entry, which is more important than the actual functionality being the best IMO.
> So if tests are bad for developers, they won’t write tests, duh. Paradox solved. There need to be tests written, no exceptions, for some time to gain the benefits and make tests good for developers. Make tests work for them and they will write more.
This doesn't actually fix the incentive problem, though. If the reason developers don't like writing tests is because finding bugs puts them under additional pressure and isn't rewarded, requiring this so doesn't actually help them. Unless, of course, management comes to understand and accept that more time will be spent fixing bugs/improving code quality as a result.
If there was a placebo effect, I would expect there to be a significant difference in the participants' self-reported sleep quality, but there was not:
> Sleep diary data revealed no differences in the number of hours slept while wearing the eye mask (7.15 ± 16.66) or the control mask (7.18 ± 16.82; t(23) = −0.11, p = .914, N = 24). Likewise, there was no significant difference in self-rating of sleep quality (eye mask: 3.13 ± 0.19 vs control: 2.84 ± 0.16; Z = −1.53, p = .131, N = 31).
Those sleep quality measurements barely overlap at 1 SD. I don't care if anyone says it's not statistically significant; the trend is enough that I would not be convinced without a lot of additional data which pushes the mean values much closer together.
I do like some elements of the algorithmic feed, like surfacing things that several of my friends like or other interesting tweets from outside my immediate network. But the downside I've noticed is that this view is also far more likely to put contentious or abrasive tweets in front of me, because that's the kind of content that gets "engagement." For me, the rise in blood pressure isn't worth sometimes seeing an interesting tweet.
The bar example is a still a bit vexing to me, because it's the sort of problem/solution I often see brought up as an example, but I still fail to see the utility. A digital signature (e.g. PGP) already can act as a certificate of authenticity from a trusted party and all you need is their public key. Could you explain what the blockchain improves about that, or what it solves in that application that digital signatures don't already?
That its difficult to do today, you need to manage your own private key infrastructure. So it becomes reserved for a few important things, like a digital ID.
What if it were easy and dirt cheap to make a provable claim about anything, in seconds, with a fraction of the infrastructure cost?
Like "I am signed into this application". Or "I have received a covid vaccine", or "i authorized this payment". Theres probably a lot more, my background is in identity so those are the ones I know.
Yes you can do those today but they all carry significant cost, and as a result there are significant moats around established players.
Just like Letterman asking "why do I need such information, have you heard of magazines?". Its true, magazines can do that. Just like PKI can do all those things.
reply