I've had a little process running for the past year or so, totting up Mastodon "active monthly users". Occasional bumps up (one outstanding one last summer when Elon did something particularly idiotic) followed by long slow declines.
Was somewhat disappointed to see recent growth of about 250k turning out to be a phantom instance (telling lies via the API).
Elk ist ein Fediverse-Client als alternative #Weboberfläche und nutzt die #Mastodon-#Client-#API, um sich mit Konten auf kompatiblen Instanzen zu verbinden.
Nach eigenen Angaben befindet sich Elk im #Alpha-Status. Bugs und fehlenden Funktionen sind zu erwarten.
Just had to design an API for work, and thanks to the painful lessons of ActivityPub, we made sure that specializations of a base type occupied different fields on a wrapper object so it will be easy to deserialize.
Instead of requesting new YouTube videos from a channel by searching the channel for videos sorted by date, I retrieve the "uploads" playlist id and THEN get the newest video in that playlist. The first approach ate 101 Google API quota, the new approach only 3, with the same result xD I increased the amount of API requests I can do by 3000% by choosing a slightly different approach c: (It's for my discord bot) #coding#developer#python#google#api#programming#development#cloud#youtube
I’ve had a few people reach out to me to ask how secure Argos Translate is.
I try to take security seriously; but of course all software has bugs. To my knowledge, to date, no one has found a serious security vulnerability in Argos Translate or LibreTranslate. It’s probably inevitable though that ArgosTranslate/LibreTranslate or one of it’s dependencies will have a security vulnerability at some point.
A question for #API historians (and software archivists in general):
Who coined the term, "API-First"?
I assume there is probably no one single author, but I'm just as interested in the chain of who popularized the term. Researching this sort of thing can be hard, as search algorithms heavily favor recency.
Trying to trace things through slide decks, it seems like there was a big bang of mentions around 2013, but then things get muddled.
{ Reddit said in a filing to the Securities and Exchange Commission that its users’ posts are “a valuable source of conversation data and knowledge” that has been and will continue to be an important mechanism for training AI and large language models. The filing also states that the company believes “we are in the early stages of monetizing our user base,” and proceeds to say that it will continue to sell users’ content to companies that want to train LLMs and that it will also begin “increased use of artificial intelligence in our advertising solutions.” }
Another option for getting out of Reddit and messing with the #LLM training - in an insignificant, yet satisfying way - is the Redact app. It can edit/replace all your posts and comments with gibberish before mass deleting them. https://redact.dev/
I was a daily user of #Reddit for ten years but quit in July 2023 when they killed third party apps. (After I exported my data, Redact wiped out everything in just a few minutes.) And honestly, there is nothing else like Reddit in the fediverse or anywhere else. No substitute for the network effects of such an enormous userbase.
It makes me sad, and I do miss some of the communities and interesting content over there. I don't blame the company for trying to make money thru ads or charging a reasonable rate for #API access - but they are unarguably, to my mind, taking the shittiest route possible. Alas.
mal ne frage
ist es sinnvoll #CRUD operationen über eine #API (ich denke #graphql) durchzuführen?
basically brauche ich eigentlich einen GUI für die #datenbank
ach ja, ich brauche primär keine tipps zur umsetzung, sondern es geht mir primär darum ob bzw. weshalb das konzept sinnvoll ist oder nicht
#MicrosoftGraph has recently deployed support for #http2 . 🚀🚀🚀
In this blog post I share a retrospective on how we made this happen for such a large #API
Heads up in case any of you are playing with Kitten¹, I’m going to implement a major API change today (it’s still pre-release/experimental so expect these things… though there should be fewer with time):
Instead of your routes receiving positional arguments like this:
(2/2) The main reasons are consistency and ease of authoring:
Currently, components get parameter objects. This will make them seem less alien when developing Kitten apps.
The onConnect() handlers in the new Streaming HTML workflow currently get positional arguments (request, page) – ordered for consistency with page routes – where you’re mostly interested in the page argument. Moving to a parameter object will hopefully mean fewer errors.
What are some good resources for building #API clients/SDKs in #PHP? E.g. what should I consider, how much customization is required, what expectations are there when you pull in an API client?
Das es bis heute einfach keine gute Card Collection App gibt für #Pokemon#TradingCardGame, in dem die deutschen Karten vorhanden sind, ist irgendwie traurig.
Oder gar ne Webseite, das macht die Inventarisierung echt umständlich. 😔
Una apparente #maledizione si è appena abbattuta su di me oggi: non riesco a far funzionare un #programma C!!! 😭️
Il problema è il seguente: il programmino di ieri per pubblicare su WordPress dal #3DS funziona benissimo, ma mi sono ricordata che dal browser del 3DS non si può caricare come #foto la schermata del gioco in pausa, ma solo cose dall’album… e non tutti i giochi permettono di salvare lì. Per salvare #screenshot da qualunque gioco c’è il menù #Rosalina, ma salva foto non-JPEG e in una cartella diversa. 😪️
Ho pensato per un attimo di forkare #LumaCFW (ho controllato e nessuno lo ha fatto, almeno non per versioni recenti), e modificare la funzione di screenshot per salvare in #JPEG, ma dopo qualche prova mi sa che evito, perché è un casino: l’ambiente di esecuzione è talmente ristretto che non vanno nemmeno tutte le funzioni della #stdlib#C, quindi non è immediato far funzionare tutto. 😮💨️
La seconda opzione, più sfavorevole ma comunque decente, sarebbe un #homebrew che converte le foto da #BMP a #JPG e le mette nella cartella giusta… e ovviamente non esiste, esistono cose simili ma non questo, quindi devo farlo io. Ho trovato questa repo da cui posso semplicemente rubare il #codice di #conversione, che sul 3DS compila e quindi dovrebbe tranquillamente fungere. E però mi sto incartando su tutto il resto, cioè la navigazione tra le cartelle e la gestione dei #file (automatica, non pianifico di mettere menù o cose, deve semplicemente aprirsi, convertire, e chiudersi). #Mannaggia! 🤢️
Purtroppo C per fare cose di #utilità pratica (ossia non videogiochini, o roba tipo demo audiovisive, e così via) è #tedioso… soprattutto perché la #API della stdlib fa proprio schifo, con tutti i nomi di funzioni e tipi scritti tutti in minuscolo e spesso senza vocali perché a qualcuno pesava il culo di scrivere parole intere. E quindi finisci per creare millemila macro e funzioni intermedie di astrazioni per evitare di ripetere di continuo le stesse criptiche righe ma così vai a creare astrazioni ancora più astruse per passare i dati in giro e così via. Mi chiedo come sia possibile usarlo per fare addirittura interi sistemi operativi, #pezzi. Vabbè, oggi si soffre. 💀️