Notices by     n8   Doesn't follow you 🇺🇦 (n8@mastodon.social), page 3

I'm not as whiny as a lot of people are about the popular position "all search results are AI crap now" ... partly because I have always been selectively suspicious, but mostly because in 100% of the cases when I've needed to query what sort of adhesive is best for glueing a particular material, I have seen that the top 75 results are ALWAYS some humans on some web forum who say Super Glue and two-part epoxy, and both those answers are 100% incorrect.

I know it's a serious event, but every time the news mentions it, I can't help thinking that "Storm Bert" sounds like the title of a YouTube documentary about Scott Adams's downward public spiral (and one that takes a decidedly troubling noncommittal stance, at that).

#Font and #type nerds - where can I find something quasi-definitive about Adobe's "SING"/"Glyphlets" tech?

All that I turn up in normal search avenues is ancient and in non-standard material, like an isolated PDF that claims to be a presentation from a conference. Better than 50/50 that it's legit, but that's still far from an authoritative source of truth you would cite.

Yes, it seems like the format is no longer actively developed, but this is for reference purposes.

Your #DailyWikipediaTimewaste — the Bodélé Depression in central Africa, where North and South America get the majority of their dust.

https://en.wikipedia.org/wiki/Bod%C3%A9l%C3%A9_Depression

#SellersMarket #TheBalladsDontWriteThemselves

New HTML-form failure mode:

- Email is required, phone is optional.

- But when *updating* the email, phone is mandatory (well, assumed, but with the same result).

Guess the UK public service that has pioneered this technique and win 20, uh, shillings!

I wonder what it's like to report a issue and have a developer actually listen to what you say

@onpaperwings Shortcut to the actual converter tool is here: https://github.com/timhutton/twitter-archive-parser in case the full talk sounds overly detailed. The high points are that this utility produces clean output that's easily enhanced if you need to do that (and you likely will, since API rate limits restrict how much auto-repurposing any outside tool can do).

@onpaperwings I did a conference talk about this topic a month or so ago. The best utility I found for self-republishing is one that re-processes your official "account backup" to static files (like decoding internal IDs to account names), but even then I found that thoughtful post-processing was needed to make the result usable for the real world (e.g., to search or to be able to find the archive for a given old tweet link). I could point you to the talk page if you cared.

I hate it when I suddenly think of the PERFECT software-project name and also do not remotely have the time or working experience to actually start developing it.

@olasd @khinsen Slides are en-route. I uploaded them to the wiki page; they still need another hand to migrate them to the final location though.

Still haven't released my twarchive-to-Pelican converter script though. Lots of corner cases I haven't decided how to tackle (e.g. "reply to this via #Mastodon/#ActivityPub" sharelinks; multiple-authors for RTs; etc etc.) Still think there needs to be <link rev= > or microformat to pin URIs together. Fun stuff, not much prior art to study....

@afranke

* Should be "youse"

(or "yall" but that's a bit more of a stretch)

Okay why isn't this new movie called Gladiators

@TerryHancock (I had also just wanted to note from my own experience that the "search" process I assume they rely on is super buggy and inconclusive)

@TerryHancock Oh, believe me, I am solely limiting my comments there to "thinking out loud to try and interpret the wording they use", which says it was a renewal search. I'm not making any claim about their rationale being correct.

As you say, it does not really make a whole lot of sense. The only other possible angle I could come up with was that they think a loophole arose because of the multiple publications (magazine first) but I still don't get what that would be exactly.

Didn't notice it at first, but this museum is definitely using News Cycle. I recognize all the mistakes.

Also, nice sword design. #YouWouldNotPartAnOldManFromHisWalkingStick?

@TerryHancock But at least the search site has digitized the names of the registrants. So you can narrow down your search space by author names & publisher names.

@TerryHancock I noticed that as well. AFAICT, from the Gutenberg site, they're saying that the original copyright registration was not renewed (and it had to be, for all the works copyrighted between 1929 and 1964, after which it became automatic or whatever).

Searching for whether there was a renewal is generally a bear; only some of the records are OCRed & searchable. For the others, you have to step through images of handwritten renewal form scans. I did a ton of those for my PhD clearances.

@penguin42 Well, no spoilers from me, but if you look them up, I expect there'll be some "aha" moments.

It still feels like a student/beginner film in a number of ways (I think it was literally an expansion of a film-school project), but it absolutely has a bunch of moments and entire sequences that stand out for being original, interesting, well-executed and such.

Nice ideas unvarnished are far preferable to no ideas with a lot of glitz in my book....

@penguin42 Do you know all the "later credits by the creative team" stuff? Definitely you can see the influence later

Do I know anybody out there at Heroku?