"Why don't browsers allocate bigger stack?" Because shitty online games which rely on stack overflows for obfuscation are breaking then, that's why. For reals. https://bugzilla.mozilla.org/show_bug.cgi?id=1537609
@tomayac oh god.. can we not have AMP in emails? There probably will be a whitelist or something and it already requires showing images by default. It's like DRM in your emails, nice.
Gitlab also makes you solve Google's captcha if you edited snippet couple of times. Ugh. One of the first things to purge Google from the internet is to replace reCaptcha.
But because Pleroma encodes their IDs in Base62 (62! Have you heard anyone using it before? With 64 it would be soo much easier) we cannot just to that. I was thinking about implementing bad converted but I think I have universal idea: because we sort things as strings we might as well matipulate strings. So maybe we can just increment string representation directly. Might be as efficient as before. #tusky#tuskydev#pleroma
Okay, I was pretty sure that we don't convert IDs to numbers in Tusky and Pleroma devs even tested it. Turns out it breaks with notifications and with caching (which is in the Nightly now). The thing is, for caching we try to come up with ID for placeholder and we request statuses ordered by these IDs (as strings). The way I did it was parsing it as integer, incrementing and converting back to string. This way it's always bigger/smaller. #tusky#tuskydev#pleroma
Opinion: Mastodon should show image descriptions not only on hover but below the image as well (as done in Tusky e.g.). This would: 1) Raise general awareness that they exists (it shouldn't be an obscure feature) 2) Will remind people that they exist 3) Will increase usability in mouseless scenarios