Nebular Night Sky by unciepaul

Nebular Night Sky by unciepaul

Ganz und gar nicht. ActivityPub ist ja nicht nur Mastodon und Mastodon-Alternativen, die nicht so gewaltig viel anders sind als Mastodon.
(streams) ist fast schon das diametrale Gegenteil von Mastodon. Es ist die fortschrittlichste, sicherste, wahrscheinlich revolutionrste Serveranwendung im Fediverse und eine der mchtigsten.
Und Mike Macgirvin ist aktuell dabei, alle Features von (streams), die ja momentan noch Nomad als Protokoll bentigen, mit ActivityPub nachzubilden. Von (streams)' Berechtigungen, die noch ber die von Hubzilla hinausgehen, bis hin zur nomadischen Identitt, die mit ActivityPub die Grenzen von Serveranwendungen berschreiten soll.
Ob andere Projekte diese Features bernehmen werden, sei dahingestellt. Aber wenn er fertig ist, dann knnen sie es. Mike wird es mglich gemacht haben, und das streams-Repository ist die Referenzimplementation, die zeigt, wie es in der Praxis geht.
#Long #LongPost #CWLong #CWLongPost #LangerPost #CWLangerPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Fediverse #ActivityPub #Streams #(streams) Das, was ich geschildert habe, drfte sich letztlich fast komplett auf Mastodon austragen. Je weniger etwas im Fediverse wie Mastodon ist, desto weniger spielt Barrierefreiheit eine Rolle.
Auf Pleroma, Misskey und Forks davon drften fast nur Mastodon-Konvertiten Bilder beschreiben. Friendica, Hubzilla und meines Wissens auch (streams) haben kein dediziertes Alt-Text-Feld. Da mu man Bilder erst aus dem eingebauten Filespace in-line in den Post einfgen, statt sie wie bei Mastodon als Dateien anzuhngen, und dann hndisch den Einbettungscode um- und den Alt-Text einbauen. Soweit ich wei, hat nur Friendica berhaupt Dokumentation dafr. Entsprechend, und weil das als "Mastodon-Marotte" abgetan wird, beschreibt da auer mir fast niemand irgendwelche Bilder. Und Lemmy hat sich gerade erst vorgenommen, berhaupt Alt-Text offiziell zu implementieren.
Worin natrlich alles, was nicht Mastodon ist, besser ist als Mastodon, ist die Zeichenbegrenzung fr Posts. Die geht da los mit hartgecodeten 3000 Zeichen auf Misskey und endet bei 65000 Zeichen auf lteren, noch nicht umgestellten (streams)-Instanzen wegen Datenbankeinstellungen und praktisch berhaupt keiner Begrenzung auf Friendica, Hubzilla und (streams)-Instanzen mit angepater Datenbank.
Das heit in der Praxis, man mu nicht alles in den Alt-Text pressen. Erklrungen fr Bilder z. B. gehren berhaupt nicht in den Alt-Text, weil es eine Menge Grnde geben kann, warum jemand den Alt-Text nicht einsehen kann. Die gehren in den Post.
Auf Mastodon werden sie im allgemeinen mit in den Alt-Text gequetscht, ursprnglich, weil ein Trt nicht genug Platz hat, und inzwischen, weil "man das auf Mastodon/im Fediverse eben so macht", ohne es zu hinterfragen. berall sonst gibt es dafr aber keine technische Begrndung dafr.
Zum Thema leichte Sprache habe ich noch nichts gelesen. Ehrlich gesagt wrde mir das noch fehlen, da das auch noch "erzwungen" wird. Ich beschreibe meine Bilder auf Englisch und extrem detailliert und habe so schon gengend Aufwand damit.
Und in meinem Umfeld bin ich so ziemlich der einzige, der berhaupt Bilder beschreibt.
#Long #LongPost #CWLong #CWLongPost #LangerPost #CWLangerPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #AltText #AltTextMeta #CWAltTextMeta #Bildbeschreibung #Bildbeschreibungen #BildbeschreibungenMeta #CWBildbeschreibungenMeta Zumindest gibt es diese Grabenkmpfe momentan nur zwischen Alt-Text-Befrwortern und -Verweigerern. Alt-Text ist immer noch so selten, da fast alle froh sind, wenn es berhaupt mal brauchbaren Alt-Text gibt.
Wenn es mit Alt-Text aber viel mehr wird, was ja an sich wnschenswert ist, fr den Fall rechne ich mit einer Ausweitung der Grabenkmpfe und neue Fraktionen. Dann geht's nmlich um die Definition von "brauchbar".
Zunchst mal wird man die Qualitt und den Informationsgehalt der Bildbeschreibungen steigern. Vor allem auf Mastodon gelten da heute schon teilweise hhere Standards als auf Websites, in Blogs oder in kommerziellen Silos wie . Das ist ja an und fr sich auch nicht schlecht.
Man knnte sagen, einem Blinden zuzumuten, einen Poster nach Detailinformationen ber ein Bild zu fragen, ist genauso anmaend und ableistisch, wie einem Blinden zuzumuten, einen Poster berhaupt erst nach einer Beschreibung des Bildes zu fragen.
Das drfte aber zu zweierlei fhren. Zum einen eine Spirale, in der die Bildbeschreibungen generell immer lnger und detaillierter werden, auch weil die, die in der glcklichen Lage sind, eine Hardwaretastatur zu haben, sich nicht sagen lassen wollen werden, sie seien keine guten Allies, weil ihre Bildbeschreibungen "nachlssig" sind.
Zum anderen Proteste von Leuten, denen die Alt-Texte schlicht und ergreifend zu lang werden. Die berufen sich auf Alt-Text-Howtos, die sich eigentlich auf Websites, Blogs oder vielleicht noch kommerzielle Silos beziehen und erklren, Alt-Text sollte nicht lnger als 200 Zeichen sein.
Dann stehen sich zwei Seiten gegenber. Die eine Seite sagt, kurze Alt-Texte sind nicht inklusiv, weil sie nicht informativ genug sind und zuviel zwangsweise weglassen. Und die 1500-Zeichen-Limits auf Mastodon, Misskey und deren Forks sind viel zuwenig. Wenn man denen sagt, Friendica, Hubzilla und (streams) haben berhaupt kein Zeichenlimit fr Alt-Text*, dann kriegen die feuchte Augen.
Die andere Seite sagt, lange Bildbeschreibungen, egal, ob im Alt-Text oder nicht, sind nicht inklusiv, weil viel zu lang zum Lesen bzw. Vorlesenlassen. Das ganze Fediverse sollte eine harte Beschrnkung auf nur 200 Zeichen im Alt-Text einfhren. Und wenn sie wissen, da es nicht nur Mastodon mit seinem 500-Zeichen-Limit gibt und es durchaus auch mglich wre, Bilder im Post zu beschreiben, sollte auch das sanktioniert werden.
Da drften dann letztlich viele sagen, wie man's macht, man macht es falsch, und wieder ganz mit Bildbeschreibungen aufhren, wenn sie dafr sowieso einen auf den Deckel bekommen.
*Theoretisch jedenfalls. Theoretisch knnte ich hier auf Hubzilla Alt-Text mit hunderttausenden oder vielleicht sogar Millionen Zeichen schreiben. In der Praxis aber mte ich meinen Monitor hochkant drehen, um auch nur 4000 davon lesen zu knnen, weil Alt-Text nicht scrollbar ist. Ganz zu schweigen davon, da Screenreader Alt-Text nicht wie anderen Text navigieren, sondern nur in einem Stck von Anfang an vortragen knnen.
#Long #LongPost #CWLong #CWLongPost #LangerPost #CWLangerPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #AltText #AltTextMeta #CWAltTextMeta #ImageDescription #ImageDescriptions #ImageDescriptionMeta #CWImageDescriptionMetaWhat really gave me to think when looking at was the varying discrepancy between active users, Hypergridders included, and land mass on the various grids.
The biggest grids in land mass are both special cases. Being the second-largest grid with 29,248 standard regions, the is one because it's largely varsims, and at least the mainland is varsims from 4x4 upward. Whereas other grids are dominated by standard-region-sized sims, a typical Wolf Territories sim is 16 of these. Also, himself is probably still the owner of the most land in the metaverse. At the same time, the Wolf Territories Grid tops the list of most active grids with 6,232 MAUs.
, largest with 32,547 standard regions, easily leaving Second Life in the dust, and second-busiest with 5,186 MAUs, is a special case because it doesn't offer land rentals. Everything that isn't an official sim is attached to the grid and hosted by its users. Size-wise, that goes from single personal sims hosted at home to whole archipelagos of varsims like Tropicana Estates (which used to be a whole lot bigger even) or Nautilus Estates. Technically speaking, OSgrid is the one grid with the most region servers. Also, for this reason, OSgrid's land mass always takes a little dive whenever they do a clean-up and remove dead sims from their map because many don't shut their sims down properly when removing them. At the same time, of course, OSgrid is a popular destination for Hypergridders even beyond parking your avatar at Lbsa Plaza.
Now let's take a look at some other grids.
is the third-largest grid with 18,077 standard regions, but the 20th-busiest grid with only 627 MAU. But it's a popular builders' grid due to its stability. Kitely's trick to cope with that much land is to shut all sims down when they aren't populated and start them up again when someone wants to enter them, a trick that was later re-implemented by DreamGrid.
is the fourth-largest grid with 12,910 standard regions, but only the 21st-busiest grid with only 596 MAU. I doubt that this includes the 3rd Wind community. Still, one reason ZetaWorlds is so large and underpopulated is its huge ocean, consisting of dozens upon dozens of varsims I think it was 3x3s. It's so big that it takes well over five minutes to cross in a motor boat at full speed. Even Stark, a nudism-encouraging archipelago of 14 4x4 vars, that's still 224 standard regions in the hands of three owners, doesn't make up a large percentage of ZetaWorlds' land area. The lack of MAU can be explained by ZetaWorlds not letting avatars from lots of grids in for reliability reasons.
is actually balanced. It's the fifth-busiest grid with 2,025 MAU and the fifth-largest grid with 10,178 standard regions. This may not seem balanced, but AMV residents love to build big, too. Grid owners Cataplexia Numbers and Clifford Hanger seem to almost always build vars because they always need lots of room. Welcome and Annex are exceptions because they're practically entirely indoors. Chris McCracken builds vars so big that even he gets lost without a proper map. And AMV was where Jimmy Olsen inflated Norway-themed Alfheim from an already respectable 4x4 with lots of details to a 10x10 giant before it sadly vanished. At the same time, AMV has loads of events which also attract Hypergridders.
, sixth-largest at 7,837 standard regions and 23rd-busiest with only 514 MAUs, is a community effort in building a whole lot of land. It's actually building a sim-crossing railway network which by now may be larger than the one in the Wolf Territories, and that one already takes quite some time to travel.
One reason for the lack of activity may be because only few sims are advertised on OpenSimWorld, and so only few people know about them in the first place. That's partly because of 's very justified love-hate relationship towards OSW to say the least. In fact, she "loves" OSW so much that she is working on an alternative to replace it.
is another interesting case: The tenth-largest grid with 1,072 standard regions isn't even in the top 25 of busiest grids. The reason becomes apparent if you look at its grid map, and the grid name is a dead giveaway: It's a sailing grid with only 49 adjacent sims, six of which are 8x8 vars, the other 43 being 4x4 vars. Over half a dozen sims don't even seem to have any actual land on them.
The , formerly GreekLife Breath Grid, is the opposite case. It's the third-busiest grid with 2,239 MAU, but only the tenth-largest grid with only 956 standard regions. This is astounding for two reasons. One, GBG is the result of two grid mergers. First, the Tranquility Grid was merged into the younger Little Breath Grid. And then Little Breath merged with GreekLife. Two, GreekLife used to make renting at least one sim mandatory for every resident. GBG has switched to a different model: Your inventory is limited to 5,000 items unless you rent land or donate. Still, I wonder where the discrepancy comes from, seeing as GBG doesn't have any super-popular events or freebie sims.
Even more extreme: and are number seven and eight in MAU, both between 1,600 and 1,700. But neither is even in the top 40 of largest grids. However, both are famous for one very popular freebie sim each, Darkheart's Boutiques and iPleasure respectively, the latter being the home of the R. Lion "brand". I think both grids have only got about half a dozen sims each or so. So either the stats are vastly dominated by Hypergridders, or people create alts on these grid to make grubbing freebies and passing them on to their mains easier and circumvent grid blocks. For example, AFAIK, ZetaWorlds has blocked both, and in the case of Darkheart's Playground, the block is mutual. I'd really like to see the faces of the users when they discover that most content offered on both grids is no-transfer.
Last but not least, makes me wonder, too. It reported 1,101 MAUs, ranking ninth, on only 343 standard regions, ranking 22th. And that's considering this grid is mostly vars, too. Most of it is joined together in one big mainland connected by a network of streets. A lot of it is owned by grid owner Govega Sachertorte and split into parcels for everyone to rent for free, even Hypergridders. Still, most of that land is vacant, maybe also because you can't set your home out in the Hypergrid, so it isn't quite useful as a dwelling-place unless you're a Neverworlder. What probably causes the most traffic, however, are Nexus Storm's several large freebie sims.
#Long #LongPost #CWLong #CWLongPost #OpenSim #OpenSimulator #Metaverse #VirtualWorlds #Grid #WolfTerritories #WolfTerritoriesGrid #WolfGrid #OSgrid #Kitely #ZetaWorlds #AlternateMetaverse #AMV #GroovyVerse #ShoalwaterBay #GBG #DarkheartsPlayground #WaterSplash #Neverworld #OpenSimStatisticsJupiter Rowland It's interesting that you claim the Fediverse to be defined by ActivityPub in a reply to . The man who invented the Fediverse in 2008 by launching the Identi.ca site, the Laconi.ca project (later StatusNet, now merged into GNU social) and the OpenMicroBlog protocol (later superseded by his own OStatus).
The term "Fediverse" was first used in 2012, some five years before the first two implementations of ActivityPub and some six years before its spec was established as a standard. Back then, it was StatusNet, Friendica and the fledgling, experimental, in-development Friendica fork Red that were connected through OpenMicroBlog. In addition, Friendica and Red were federated with Diaspora* through its own protocol.
Several other projects joined in before the creation of ActivityPub. Red/the Red Matrix and Hubzilla inherited their OpenMicroBlog connector from Friendica even though it had to be rewritten. And just like Mastodon, Pleroma started as an alternative frontend for StatusNet. The only one in the bunch whose goal it actually was to pull everything together in one big network was Friendica.
Interestingly, it took "if it exists, we federate with it" Friendica until 2019 to adopt ActivityPub. Hubzilla was the first to do so in July, 2017, followed by Mastodon in September, both before official standardisation. So for a while, it was only these two which have never really gotten along with each other too well until today.

I am trying to think of a platform that incorporated ActivityPub from the start.

I think the whole Threadiverse did. Lemmy certainly did. /kbin definitely did because it was still in an early experimental stage when the Redditors came rushing in. And I guess so did Sublinks which is too new to have started out without ActivityPub.
Other stuff like Plume, Funkwhale, Mitra and micro.blog might have been built on top of ActivityPub right away, too.
If you count forks, then this applies to all Forkeys because forking the hell out of Misskey hadn't been a thing before 2018, also all three incarnations of Osada (all now defunct), Mistpark 2020 (now defunct), Redmatrix 2020 (now defunct), Roadhouse (now defunct) and the streams repository.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Fediverse #OStatus #ActivityPub

Which is to say that I think that in *theory* Hubzilla ought to be able to offer this for Hubzilla-Hubzilla federations.

That'd still mean two things.
One, even if you as a Hubzilla user have only got Hubzilla connections, your Hubzilla connections may have connections to elsewhere from Mastodon to Diaspora*. And they may repeat (= boost =RT) your stuff. Good luck going after that.
Two, Hubzilla didn't have repeats until version 9.0 which is only a few months old. Some hubs still haven't upgraded. Until then, shares (= quote-posts = quote-tweets) were the way to go. And I guess not everyone has gotten used to being able to repeat now. So even within itself, Hubzilla would have to try and track down shares of posts whose settings are being changed.
Unlike Mastodon, Hubzilla promises water-tight privacy, and it's renowned for water-tight privacy. This is actually one of its killer features.
Sacrificing absolute water-tightness to becoming able to change privacy settings after the fact would put a huge dent into this water-tight privacy to the point of the promise of water-tight privacy becoming an out-right lie. That is, unless Hubzilla finds a way to turn back time.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Hubzilla #Privacy #Permissions Hubzilla takes privacy and permission controls very seriously. This is the exact reason why privacy settings can't be changed on Hubzilla after the fact.
If you posted something in public, and you wanted to change the permissions to private, Hubzilla would have to take away your post from all those whom you don't want to see it.
It would have to track the post down all across the Fediverse. Follow boosts, boosts of boosts, boosts of boosts of boosts etc. It would even have to follow quote-posts which are possible just about everywhere that isn't Mastodon. It would have to invade people's Mastodon apps and Web browser caches.
And it would have to completely purge that post, even including in quote-posted form, from the timelines of everyone who isn't supposed to have received it after the change of permissions. And it would have to purge it from the databases and caches of all instances where nobody is supposed to have received it.
It would essentially have to undo itself completely without leaving any traces whatsoever of its prior existence. And even then people would remember it. And even then Mastodon users might have tooted screenshots of that post because they can't quote-toot.
What's out in public stays in public. Even Hubzilla can't change that. Any claim or implication otherwise would be a lie into the faces of those who use Hubzilla for its privacy features.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Hubzilla #Privacy #Permissions Keep one thing in mind, though:
Like everything else, hashtags can be filtered. In fact, hashtags are good for filtering. Depending on where in the Fediverse you are, especially far away from Mastodon, filters can remove posts, filters can out-right block posts (not on Mastodon, but elsewhere), and filters can automatically generate content warnings, since last year even on Mastodon.
Many of us aren't on Mastodon. Some of us are in places where the use of filters, especially to generate content warnings, has been part of the culture since the very beginning. And that beginning was often before Mastodon's beginning.
That's part of why I use so many hashtags. If I put the four hashtags #FediMeta, #FediverseMeta, #CWFediMeta and #CWFediverseMeta on a post or a comment, it's about the Fediverse, and the hashtags serve to trigger the filters of those who don't want to read the stuff I write about the Fediverse.
I use four hashtags instead of only one because there are these four possible hashtags that can be used or filtered for this topic, and I can't possibly know who uses which hashtag or hashtags in their filters. So I use all four to catch them all.
If I use the four hashtags #Long, #LongPost, #CWLong and #CWLongPost, it's because a post or a comment exceeds 500 characters which disturbs many Mastodon users deeply. So they can at least filter out the latter three to get rid of my over-500-characters content without having to mute or block me entirely. This is also necessary because I don't have the technical means to put Mastodon-style content warnings on posts.
If I use #Fedisplaining, I'm not talking about fedisplaining. I'm rather doing something that may be considered fedisplaining. Whoever doesn't want to see that can filter that hashtag.
In the case of this comment, only #Hashtag and #Hashtags would be hashtags for the purpose of discoverability. Long Hashtags breiten sich nicht durch Fderation aus, sondern durch Kontakte, genau wie alles andere an Content.
Wenn du ein neues Hashtag erzeugst, wird das nicht geteilt mit allen Instanzen, mit denen das troet.cafe fderiert ist, sondern mit allen Instanzen, auf denen du Followers hast. Und wenn einer davon deinen Trt mit dem neuen Hashtag boostet, landet das Hashtag auch auf den Instanzen, auf denen derjenige Followers hat. Und so weiter.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Fediverse #Hashtag #Hashtags

Dolphins in Depth: Dolphins struggling to get a deal done with Tua as QB salaries rise

After 70 years, DNA test reunites 3 long-lost siblings across the pond lost siblings Lost Siblings Reunite

How is an AI supposed to have extensive, in-depth, up-to-date detail knowledge of any imaginable super-obscure niche topic that even Google can't find so that it can describe images about that topic at the same level as a human expert
How is an AI supposed to see the same details in an 800x533-pixel image that I can see when looking at the real deal with a nigh-infinite resolution and without a fixed point of view Or look around objects in front of signs in order to transcribe what's written on them
And that's why I will always describe my images myself by hand, no matter how big an effort it is, no matter how long it takes me.
I've actually let an AI describe two of my images which I've described myself before. I've even asked the AI for a detailed description the second time. The results were always pathetic, not only in comparison, and the AI kept getting things utterly wrong. Also, the images required extensive explanations to understand, but the AI didn't explain anything because it didn't really understand the images itself.
Nobody can tell me that AI can describe any and every image out there with the same accuracy, faithfulness, informativity and level of detail as a human expert on what the image shows.
#Long #LongPost #CWLong #CWLongPost #AltText #AltTextMeta #CWAltTextMeta #ImageDescription #ImageDescriptions #ImageDescriptionMeta #CWImageDescriptionMeta #AI

If the idea is to work around limitations like not enough space for text, well, Mastodon is young enough that we should change the design specs rather than use it in an unintended way with unintended consequences.

If you regularly run out of characters due to the 500-character limit, the solution that works best right now is not to wait for Mastodon to remove that limitation on all instances. Of course, it isn't to put excess characters into the alt-text either.
It's rather to move to e.g. Akkoma or Iceshrimp or Sharkey or Friendica or so, anything in the Fediverse that offers way more characters on all instances so that you don't have to ask around which instance has more than 500 characters.
CC:
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #CharacterLimit #CharacterLimits #500Characters #NotOnlyMastodon #FediverseIsNotMastodon #MastodonIsNotTheFediverse

Badge: Source:

I too had no idea how terrible hospitals are treating the staff. And this was 10 and 15 years ago.

The Fediverse does have invite systems. Hubzilla definitely does. Mastodon doesn't.
I think there are two major obstacles. One, the Fediverse being as decentralised as it is makes it too difficult for the casual, tech-illiterate masses. It'd still be tiny without the relentless railroading of Twitter refugees onto mastodon.social without telling them what the Fediverse really is, implying that a) the Fediverse is only Mastodon, and b) mastodon.social is the Mastodon website.
Two, nothing in the Fediverse looks and feels nearly like any corporate American (or Chinese) social media giant. This is what makes Bluesky so popular: While Mastodon tries to mimic Twitter's functionality with all its shortcomings, Bluesky has even largely cloned Twitter's UI and UX and its entire look and feel.
People don't want to learn something new. And they don't want to have to choose, especially not when they have to educate themselves about what's the best pick for them first.
It's the same with Linux: Many say it won't take off unless it's unified into one distro simply named "Linux" with one graphics toolkit and one desktop environment that's an absolutely 100% identical clone of the currently most popular Windows version in look and feel.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Fediverse #Mastodon #Linux #Onboarding Depends on the content.
You post real-life photographs which show a certain action going on. Nothing much has to be described or explained.
I post renderings from very obscure 3-D virtual worlds. I can't expect anything in my images to be known or familiar to anyone. And since my images don't focus on one particular element or action, neither can the descriptions.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #AltText #AltTextMeta #CWAltTextMeta #ImageDescription #ImageDescriptions #ImageDescriptionMeta #CWImageDescriptionMeta

Heat Check: Exploring all angles of Jimmy Butler situation

There are two hashtags for that: #Alt4me and #AltText4me.
But they're more miss than hit. Even simple images usually don't get any community description. And the more obscure the knowledge required to describe the image is, the less likely it is for someone to provide a description.
I myself would rather spend another three months not posting a single image because describing it appropriately is so tedious and time-consuming than rely on someone else trying to describe an image they don't even get because it's too niche.
CC:
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #AltText #AltTextMeta #CWAltTextMeta #ImageDescription #ImageDescriptions #ImageDescriptionMeta #CWImageDescriptionMeta

* provide extra context and information perhaps not covered in the main body of your toot, making your toot more engaging

Never offer information exclusively in alt-text!
There are people who can't access alt-text. They may have a tremor that keeps them from steadily hovering a mouse cursor over an image. They may be unable to use their hands, poking their computer keyboard with a headpointer or a stick in their mouth, so they can't use pointing devices like mice, trackballs or touchpads at all.
Or they may be bernerds who use Linux with a minimal tiling window manager that doesn't even support pointing devices.
Any information that is only available in the alt-text and neither in the post itself nor in the image is lost to them because they can't access it.
Explanations always go into the post text body where everyone can read them.
If you regularly have to explain images, but you can't because Mastodon only lets you toot 500 characters, the solution is not alt-text. The solution is not really a thread either because description and explanation should always be in the same place als the image.
The solution is to move somewhere with a higher character limit.
If you don't know any Mastodon instances with a higher limit, play it safe and move to Iceshrimp or Sharkey. Thousands of characters guaranteed on all instances. Or if you don't depend on mobile apps, try Friendica. Practically no character limit at all.
But explaining stuff in alt-text and nowhere else is a huge no-no.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #AltText #AltTextMeta #CWAltTextMeta #ImageDescription #ImageDescriptions #ImageDescriptionMeta #CWImageDescriptionMetaTurns out you can add Mastodon-style content warnings to comments on both Hubzilla and (streams).
There's a summary/summary pair of tags which, I think, is undocumented on Hubzilla. The summary and/or content warning goes between the tags, and everything that follows the tags is hidden behind it. Still, you might want to put it at the beginning because I think Mastodon would be irritated otherwise.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #CW #CWs #ContentWarning #ContentWarnings #CWMeta #ContentWarningMeta #FediTips #Hubzilla #Streams #(streams) I do something similar, but only a tiny part of it in the alt-text and the vast majority of it in the post CW: long (almost 2,000 characters), Fediverse meta, non-Mastodon Fediverse meta, alt-text/image description meta View article View summary
I do something similar.
I give a fairly "short", purely visual description in the alt-text. Also in the alt-text, I mention that a full, detailed description with explanations and text transcripts can be found in the post itself. I usually add that for Mastodon users, it's hidden behind a content warning which they have to open first, and for users of e.g. Friendica, Hubzilla and (streams), the full description follows immediately after the image which is embedded in the post. I didn't do that the last time because the alt-text was already precisely 1,500 characters long without it.
You've read that right. I don't put the full, detailed description into the alt-text. It's way, way, way too long for the alt-text. I put it into the post text body. I'm not bound to Mastodon's 500-character limit. I'm on Hubzilla, and I don't have any character limit that's worth worrying about.
But when a Mastodon user opens the content warning and exposes the post with the full image description in it, it turns out that the long post warning and especially the character count that's part of the long post warning is actually accurate. My full, detailed image descriptions can be anything between 25,000 and over 60,000 characters long for only one image. And I'm certain that this level of detail is actually necessary.
Now, the problem is that it is this hyper-massive wall of text that contains all explanations. You wouldn't be able to get the image just from the alt-text. You'd have to read this monster. It also contains all text transcripts, so if someone wants to know what's written in the picture, they have to endure going through this behemoth.
But I couldn't possibly fit either into the alt-text, let alone both.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #AltText #AltTextMeta #CWAltTextMeta #ImageDescription #ImageDescriptions #ImageDescriptionMeta #CWImageDescriptionMeta I guess posting images about obscure niche topics in the Fediverse has to be ableist all by itself. That's because it's impossible to do in a fully inclusive and fully accessible way.
On the one hand, a short image description is not nearly informative enough to understand the image, and therefore it's ableist. Having people ask you for information about your image is always bad. It doesn't matter if they have to ask you for basic information or for detail information. Everything necessary to understand the image must be delivered with the image.
On the other hand, a sufficiently detailed image description is way too long for people with a short attention span to read and for blind or visually-impaired people to sit through and have it read to them, and therefore it's ableist, too.
Anything in-between is both still too long and not informative enough, it's watering down two concepts of accessibility, and therefore it's ableist for two reasons.
Essentially, you can't be inclusive if you post content that can't be understood by everyone out there without over 200 characters of image description.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #AltText #AltTextMeta #CWAltTextMeta #ImageDescription #ImageDescriptions #ImageDescriptionMeta #CWImageDescriptionMeta #Accessibility #A11y #AccessibilityMatters #Inclusion #Inclusivity #Ableist #Ableism I'd love to join that thread, but somehow, I can't.
So good advice here: If you really depend on having consistent conversations, don't hope for Mastodon duct-taping support for them on somehow. Instead, look into the Fediverse beyond Mastodon and at those corners of the Fediverse that do not try to mimic Twitter with all its shortcomings.
and its descendants are just as much part of the Fediverse as Mastodon is, and they've partly been for longer than Mastodon itself. But they work fundamentally differently because Friendica is a replacement for Facebook with a side of full macroblogging capability.
This also means that their threads are structured fundamentally differently from those on Mastodon. On Mastodon, a thread consists of single posts which are pretty much only tied together by referencing the parent post and the child post.
On Friendica, and url=https://codeberg.org/streams/streams(streams)/code, a thread consists of exactly one post, the start post, and otherwise only comments. It's like on Facebook or Tumblr or in a blog. Comments may reply to other comments, but fundamentally, all comments reference the start post, and the start post knows all comments in the whole thread.
A side-effect is that you can say bye-bye to character limits.
Fair warnings, though: When I say these three work differently from Mastodon, I mean they also handle differently from Mastodon. The difference between and Mastodon is smaller than the difference between Mastodon and either of the three.
Also, all three are mainly for desktop browsers. Friendica was conceived in 2010 when using social media on phones was a gimmick, and Hubzilla and (streams) are its still existing direct offspring made by its own creator.
So don't get your hopes high for mobile use. Friendica has only partial support for the Mastodon API plus, I think, one flaky Android app. Hubzilla and (streams) have no Mastodon API support, and they aren't even based on ActivityPub they connect to it through an official add-on. Hubzilla has one Android app that was last updated in 2018, and connecting it to your channel has been reported to be hit-and-miss. (streams) is not supported by any mobile app you have to use it as a PWA.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Friendica #Hubzilla #Streams #(streams)

I work with visual digital media and have to write a *lot* of (and multiple versions of the same alt text), so I try to be as inoffensive as possible. It takes a crapton of energy some days.

That's actually difficult because what one Mastodon user demands to be mandatory all over the Fediverse, another demands to be a bannable offence all over the Fediverse.
Sometimes general demands or demands from the same person contradict each other. On the one hand, image descriptions in the Fediverse must be informative and explanatory enough that nobody will ever have to ask the author or look something up themselves if they don't understand an image. No matter how obscure whatever the image shows is.
On the other hand, no image description should be longer than a few hundred characters. Even though images in the Fediverse aren't only social media screenshots or cat photos.
Never mind that explanations must never, under any circumstances, be put into the alt-text. Some people can't access alt-text, and any information exclusively available in alt-text is inaccessible and therefore lost to them.
But Mastodon has basically made it a Fediverse-wide rule that explanations must be crammed into the alt-text as well because Mastodon has that puny 500-character limit. Everything else has higher limits from 3,000 characters (Misskey) to 24,000,000 characters ((streams) if installed this month or with a manually upgraded database), but what can't be done on Mastodon mustn't be done anywhere else.
So even if you have virtually no character limit, you still have to cram explanations into the alt-text.
#Long #LongPost #CWLong #CWLongPost #AltText #AltTextMeta #CWAltTextMeta #ImageDescription #ImageDescriptions #ImageDescriptionMeta #CWImageDescriptionMeta The more obscure the content of an image is, the more difficult is it to describe it with enough information that nobody will ever have to look anything up or ask me anything in order to be able to understand the image, but at the same time at a length typical for alt-text outside the Fediverse.
I know that from personal experience because the images I post show stuff that's extremely obscure. So obscure, in fact, that extensive explanations are necessary.
You wrote:

Who are Ipswich Town Why are we assuming that a blind person will understand their club colours If Portman Road is relevant, it needs to be described. If it isn't, it shouldn't be there.

Well, I've read somewhere that it's good style to always mention where an image was made, where the place is which the image shows, unless this information would hurt someone's privacy, unless it's classified, or unless it really doesn't matter because the image was taken at someone's home.
The thing is, I don't post real-life photographs. My images are rendering from very obscure 3-D virtual worlds. Very obscure as in maybe one in over 200,000 Fediverse users even knows the underlying system of these worlds.
So, to take a theoretical example that I've never actually posted and described, I could mention that an image shows the Sendalonde Community Library.
Of course, nobody knows what and where that is. Or do you
It's in Sendalonde.
Of course, nobody knows what and where that is. Or do you
It's a sim in DigiWorldz.
Of course, again, nobody knows what and where that is. Or do you
DigiWorldz is a 3-D virtual world, a so-called grid, based on OpenSimulator. And a sim is a simulation, basically a large piece of virtual land.
Of course, nobody has ever heard of OpenSimulator, nobody knows what it is, nobody knows why a grid is called a grid etc. etc. Or do you
So I have to explain all that as well.
I regularly need well over 1,000 characters alone to explain where I've made an image so that nobody who's actually interested in the image will have to ask me anything to be able to understand the image.
But I never explain my images in the alt-text. Explanations don't belong in alt-text. There are people who can't access alt-text for various reasons, including certain disabilities that make it impossible for them to hover a mouse cursor steadily over an image or even use pointing devices altogether. Any information exclusively only available in alt-text is inaccessible and therefore lost to them. Thus, explaining stuff in the alt-text is ableist. Just as ableist as expecting disabled users to ask me things which I could just as well have explained in my post right away.
Instead, I always put a full, detailed image description, including all necessary explanations to understand the image and the described elements, in the post text body itself. I'm not on Mastodon, I don't have a 500-character limit. In fact, I don't really have any character limit. Or if I do, it's probably in the tens of millions.
I have to describe my images at high details because I can't assume any blind or visually-impaired user to be familiar with what anything in the image looks like. The topic is just too obscure, too niche.
But I always have to expect someone to be curious about the topic and wanting to know exactly what the image shows. And if I merely mentioned what is there, I'd expect blind or visually-impaired users to inquire immediately, "Yeah, but what does it look like!" Which is fully justified because, again, how are they supposed to know
So I give detailed visual descriptions, as detailed as I possibly can. These go so far that I even prefer not to show buildings in my images because they're way too tedious to describe.
Also, if there's text somewhere within the borders of the image, I transcribe it. If it's too small to be legible, I transcribe it. If it's so tiny that it's actually invisible, I still transcribe it. If it's partially or even mostly obscured by something standing in front of it, I still transcribe it verbatim. Telling blind users who don't have a concept of image resolution that a piece of text is too small to be legible and using that as a justification to not transcribe it feels like weaseling out.
Granted, the last time, I've limited myself. I did not transcribe about a dozen tiny bits of text in an image in my image, nor did I transcribe hundreds of microscopic bits of text on dozens of images in that image in my image. I was writing my longest image description ever already, and it was the second day of writing it. Had I not limited myself, I would have spent the entire rest of the week describing and explaining dozens of more images, each one of them only a infinitely tiny fraction of a pixel in my image, and transcribing hundreds of bits of text. And my image description would have grown so long that Mastodon would have rejected the whole post, rendering the image description itself largely moot.
Thus, I had to stop myself at that point and declare the image description complete at the end of the second day.
By the way, in order to satisfy those who demand there always be an image description in the alt-text, I always describe my images twice: once at full detail in the post, once only visually and not quite as detailed in the alt-text. In addition, the alt-text always mentions that a full and detailed image description with explanations and text transcripts can be found in the post text. It mentions that users of Mastodon and the like can find the full, detailed description by opening the long post content warning, and that users of Friendica, Hubzilla and (streams) wil find the full, detailed description right after the image itself.
It's tedious and time-consuming, and even then I'm never fully satisfied with my own image descriptions. But at that point, it's just about the best I can do.
It also leads to unimaginably hyper-massive image descriptions in a place where Mastodon users would never expect them, image descriptions so long that hardly anyone is willing to read them or have them read by a screen reader. But I don't really think my descriptions are critically lacking detail.
That is, if I'm wrong about the latter, I appreciate feedback so I can further improve my describing style.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #AltText #AltTextMeta #CWAltTextMeta #ImageDescription #ImageDescriptions #ImageDescriptionMeta #CWImageDescriptionMeta

Pipeline release! nf-core/mag v3.0.1 - mag 3.0.1!

Please see the changelog:

-read-sequencing -sequencing

But they won't suspend them because they're "mainstream" and "do it the normal social media way".

No, they won't suspend them because they don't even know they exist.
All this looks like the actions of Mastodonians. And what do they largely know about the Fediverse outside of Mastodon
If these people really were as Fediverse-savvy as some of them claim to be, they'd know how Friendica and especially Hubzilla and (streams) work. While a few would have switched to (streams) which really offers the privacy and security they thought Mastodon had, many many more would block any and all instances of all three they could find or even try and have them Fediblocked.
These three do such "weird stuff", they bombard "the Fediverse" with "long posts" over 500 characters, and they don't follow the "Fediverse rules" which are largely only Mastodon's culture that doesn't even always technically translate to anything else.
Misskey and all the Forkeys would be the next. While they aren't as alien as Mike's creations, they're still weird, and their users unapologetically pump out overly long posts, too.
I'm still convinced that many Mastodon users would welcome a switch that blocks everything that isn't (vanilla) Mastodon, both at instance level and at account level.
CC:
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #FediblockMeta #Fediverse #Mastodon #NotMastodon

Amazing Japanese Kite Festival! (in Japan!) After visiting Hamamatsu in Shizuoka last year and learning about their annual kite festival

I beg to differ. AI is far from able to create image descriptions for all images, especially images about super-obscure niche topics, that are perfectly on the same level of quality, accuracy and detail as what an expert in that field can write by hand.
I myself often write about such a super-obscure niche topic. I always describe my images by hand which always takes me hours. But the image descriptions that come out of this effort are sufficiently detailed, which means extremely detailed, and as highly accurate as I can possibly write them. And they should be informative enough that even someone who knows nothing about what the image shows doesn't need to look anything up to understand the image. They often include information that you couldn't possibly find anywhere else on the Web.
And I've actually let an AI describe an image after describing it myself twice already.
The actual analysis of the comparison is in a comment.
Again, the analysis is below, and this time, the image with my own description is way up in the start post.
Both AI descriptions are bordering on utter train wrecks in comparison with my self-researched, hand-written descriptions. Not only are they utterly incomplete to degrees that could partly be considered ableist, but they're even wrong in several points.
Granted, I always have another unfair advantage. The AI describes what's in the image by examining a picture of 800x533 pixels. I describe it by examining the real deal.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #AltText #AltTextMeta #CWAltTextMeta #ImageDescription #ImageDescriptions #ImageDescriptionMeta #CWImageDescriptionMeta #AI Ich war 2022 mal auf einer Party in einer Art Ballsaal im Dereos-Grid. Da haben sie mit dem Trick einen Spiegel realisiert, also quasi den ganzen Raum zweimal gebaut, einmal davon hinter einer Glasscheibe. Nur da der eine Vampir in unserer Runde nicht der einzige war, der sich nicht gespiegelt hat.
Bei den Mbeln hinterm Glas hatte man sich brigens nicht die Mhe gemacht, die Sitzskripte rauszunehmen. Eine von uns hat dann einfach mal hinter die Glasscheibe gecammt und sich auf einen von den Sthlen gesetzt...
#Long #LongPost #CWLong #CWLongPost #LangerPost #CWLangerPost #OpenSim #OpenSimulator #Metaverse #VirtuelleWelten

Gilgo Beach murders: Blueprint of crimes leads prosecutors to lay 2 more charges News beach murders island serial killer Heuermann killer

Fr mich liest sich das, als wenn es die Mastodon API an WordPress dranschraubt.
Das hat also nichts mit Fderation zu tun, sondern damit, da du eine Mastodon-Smartphone-App ans Blog anhngen und dann mit der Mastodon-App bloggen kannst. Das ist natrlich nur dann sinnvoll, wenn die App nicht hart nur gegen Mastodon gebaut ist und nicht nur Mastodons Features untersttzt, sondern z. B. auch unbegrenzt viele Zeichen und Markdown mit allen Schikanen bis hin zu eingebetteten Inline-Bildern.
#Long #LongPost #CWLong #CWLongPost #LangerPost #CWLangerPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #WordPress #Mastodon #MastodonAPI #MastodonApps Are you familiar with Mike Macgirvin 's creations Friendica Hubzilla The streams repository even
He has been working on some of these points since longer than Mastodon has been around, even since befor the term "Fediverse" was coined.
Safe spaces: I dare say that (streams) can be the safest space in the whole Fediverse. That's because it has the most advanced, most extensive permission settings in the Fediverse. And it's one of the very few Fediverse server applications that actually know permissions.
Portable identity: Has been reality since Mike invented the Zot protocol and nomadic identity in 2011 and first implemented it by forking Friendica into Red and porting it to Zot in 2012. Except for the first Osada, everything Mike made afterwards supported or still supports nomadic identity amongst its own server instances, including Hubzilla and (streams).
Right now, Mike is working on porting both nomadic identity and (streams)' permission model to 100% ActivityPub so that other already existing projects can adopt them. His goal is to go even further and stretch the same identity across server instances of different projects.
Channels/feeds: Stuff like this should be possible with Hubzilla and (streams) right now already. In their case, "channel" refers to a Fediverse identity, of which one can have multiple separate ones on the same account. But a Hubzilla or (streams) channel can do much, much more than Mastodon-style microblogging, even more than what a Friendica account can do. It can serve as a discussion group/forum, complete with moderation, even with multiple moderators. It can also serve as a news aggregator, and a Hubzilla channel can even act as a fully automatic reposter for one or multiple sources of various kinds, including RSS or Atom feeds.
In general, if you think something should be developed in the Fediverse, ask whether Hubzilla or (streams) already has it implemented.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Hubzilla #Streams #(streams)

Nicht , aber hier neu

"Three years after their initial bouts with , patients whod once been hospitalized with the virus remained at 'significantly elevated' risk of death or worsening health from COVID complications, according to a paper published May 30 in Nature Medicine."

To add to your table:
supports nodeinfo 2.0 and 2.1.
Examples from a stable release, just to show that they can differ, depending on how a hub is configured:


Example from a development release:

However, it looks like nodeinfo can be turned off entirely by the hubmin. At least hubzilla.org has a blank nodeinfo page.
always has a blank nodeinfo page. Most nodeinfo code has intentionally been removed. It understands nodeinfo, but it seems to send something else instead which it only understands itself, which is limited in what information it offers, and which isn't parsed by Fediverse stats/instance-listing websites.
By the way: Both Hubzilla and (streams) support both OAuth and OAuth2 both as a client and as a server. However, Hubzilla's documentation is painfully outdated in this regard parts of it still refer to Red, so they were last touched before the name change to Red Matrix. And (streams) doesn't have any documentation.
Also, Hubzilla and (streams) are the only Fediverse server applications with full, i.e. both server-side and client-side support for OpenWebAuth single sign-on.
CC:
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Hubzilla #Streams #(streams) #OAuth #OAuth2 #Nodeinfo When it comes to describing images, Mozilla is still thinking in dimensions for either static websites or commercial/scientific blogs or corporate American social media, namely no more than 150-200 characters.
They've had their own Mastodon instance for a while now, but it seems nobody at Mozilla actually follows any other Mastodon user. Otherwise they'd know a thing or two about Mastodon's high standards for image descriptions.
Do my image posts ever make it over to Misskey, or does it actually reject posts over 10,000 characters or so Because there are Mastodon users who don't even find these overwhelming, and I probably write the longest image descriptions in the whole Fediverse.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #AltText #AltTextMeta #CWAltTextMeta #ImageDescription #ImageDescriptions #ImageDescriptionMeta #CWImageDescriptionMeta One solution that hasn't been mentioned yet: Get yourself a Bluetooth hardware keyboard that's large enough for touch-typing but compact enough to take it with you if you have a bag or something.
I myself wouldn't trust any AI to properly describe my own images, not after actually pitting one against my own hand-written image descriptions twice and seeing it fail miserably. But then again, describing my images in a way that I deem sufficiently is practically impossible for AI unless it's absolutely omniscient.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #AltText #AltTextMeta #CWAltTextMeta #ImageDescription #ImageDescriptions #ImageDescriptionMeta #CWImageDescriptionMeta #AI

Last updated: 2024-06-19

Art for my home