Pipeline release! nf-core/mag v3.0.1 - mag
Turns out you can add Mastodon-style content warnings to comments on both Hubzilla and (streams).
There's a
#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)
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:
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
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
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
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)
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 #AII've got over 500 connections, most of which are on Mastodon. Yet, it feels like I've got much less reach in the Fediverse now than I used to have when I only had some 200 or 300 connections. This can only mean that especially lots of Mastodon users and maybe even entire Mastodon instances have muted, blocked or shadow-banned me. I wouldn't even be surprised if some public Mastodon instances have secretly muted or blocked hub.netzgemeinde.eu as a whole just because of my actions.
I can't really tell from the delivery reports, for one difference between Mastodon and Hubzilla is that Hubzilla rejects blocked or filtered messages at the server level whereas Mastodon accepts them and then discards them before they can be delivered to the muting, blocking or filtering recipient. Otherwise I think I could weed out a whole lot of connections.
And here's a list of reasons for which I've probably been muted, blocked or shadow-banned:
Essentially, this boils down to two points:
All because the Fediquette is entirely defined by Mastodon users, most of whom at that point didn't and often still don't know that the Fediverse is more than just Mastodon.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Mastodon #MastodonCulture #Fediquette #Mute #Muted #Block #Blocked #NotMastodon
Certainly not all of them. How about:
Also, if "Mastodon" means that the ActivityPub side is only built and tested against Mastodon and only supports Mastodon's features, chances are it might not work with dozens of other Fediverse projects, including but not limited to Pleroma and Akkoma as well as Misskey and its many forks.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Openvibe #Mastodon #Pleroma #Akkoma #Misskey #Forkey #Forkeys #Hubzilla #Streams #(streams) #Diaspora* #NotOnlyMastodon #FediverseIsNotMastodon #MastodonIsNotTheFediverse
CC:
#CW has been hijacked by amateur radio where it stands for continuous wave. The few exceptions are about TV and refer to the CW Network.
#CWS has recently been hijacked by a commercial advertising account.
#ContentWarning has recently been hijacked for posts about the eponymous video game.
And both #ContentWarnings and #ContentWarningMeta are too long to type on a screen keyboard.
Not to mention that Mastodon is largely falling back into Twitter mannerisms, i.e. neglecting hashtags altogether, and most of the rest of the Fediverse has never had a hashtag culture to begin with.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Hashtag #Hashtags #HashtagMeta #CWHashtagMeta
Da gibt's keinen Konsens. Das kann ich aus Erfahrung sagen, auch wenn ich fast kein Feedback bekomme.
Ich schreibe die mit Abstand lngsten Bildbeschreibungen im ganzen Fediverse und halte das auch gerechtfertigt. Einige feiern sie tatschlich, auch wenn es bis zu einer Stunde dauert, eine davon zu lesen. Inzwischen wundert mich das regelrecht.
Andere, kurioserweise durchweg ohne nennenswerte Behinderung, sind strikt dagegen und sagen auch mal, absolut jedes Bild lt sich adquat in 200 Zeichen oder weniger beschreiben. Und wer etwas nicht versteht, soll eben fragen oder googlen.
Ich versuche tatschlich, alles zu beschreiben, was sich innerhalb des Bildes befindet und nicht komplett verdeckt ist. Allerdings zeigen meine Bilder ein so obskures Nischenthema, da meines Erachtens so detaillierte Beschreibungen gerechtfertigt sind. Nur dauert das eben seine Zeit.
Kenne ich auch. Ich htte eigentlich viel zu zeigen. Aber die Bildbeschreibungen, wie ich sie schreibe, fressen derart Zeit und Energie, da zwischen meinen letzten zwei Bildposts mehr als drei Monate lagen.
Das letzte Mal habe ich mir schon berlegt, was ich zeigen knnte, was interessant ist, aber nicht so aufwendig zu beschreiben. Ich lag mal wieder komplett falsch und schrieb letztlich ber zwei Tage verteilt meine mit Abstand lngste Bildbeschreibung.
#Long #LongPost #CWLong #CWLongPost #LangerPost #CWLangerPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #BiBesch #Bildbeschreibung #Bildbeschreibungen #BildbeschreibungenMeta
The central resource is the actor that started the tree.
Traditionally, all replies go to this actor and this actor alone, and this actor distributes them to all other participants.
The whole tree with all its nodes is ideally owned by the actor who started it.
Obviously, the actor who started the tree must not be muted or blocked, otherwise the entire tree is inaccessible.
As for comments, I'm not sure what happens when a blocked actor replies, and someone else replies to this reply. Most likely, the reply from the blocked actor will not appear because it still came from the blocked actor the actor who started the tree only forwarded it automatically. But I can't say how replies to a blocked reply are handled.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Fediverse #Threadiverse
Yes, because nodeBB, Flarum, Lemmy, /kbin, Mbin, PieFed, Sublinks, Friendica, Hubzilla, (streams) and everything else that knows what a context is will interact with each other. They often already do.
It doesn't make sense for each one of them to try and patch incompatibilites with the others one by one, all by themselves, without even talking to each other. That might only cause more incompatibilites, not to mention loads of extra code just to address the quirks of individual projects.
It's better to have one agreed-upon standard.
Safety can and certainly will come with consistency.
Look at Mike Macgirvin 's , the result of 14 years of work that started in 2010 with Mistpark, now known as Friendica. (streams) knows what a context is. (streams) can hold a discussion together. (streams) supports groups/forums. And so did all its predecessors, two of which (Friendica, Hubzilla) are still alive and maintained.
At the same time, (streams) has to be the safest place in the whole Fediverse. Not necessarily by culture or policy, but by technology. (streams) has ramped up its permission controls beyond what Hubzilla has to offer, and Hubzilla's permission controls are already extensive. In the meantime, Mastodon and everything else in the Fediverse that's based on ActivityPub doesn't even know what permissions are.
On Mastodon, all you have is mute and block for users or instances. Hubzilla goes much, much further, (streams) goes even further. Imagine being able to delete posts from a thread that you've started (Hubzilla, (streams)) or even banning users from participating in your thread ((streams)). Imagine being able to define who exactly is allowed to send you posts, who exactly is allowed to see your profile (or which one of your multiple profiles even), who exactly is allowed to comment on your posts, who exactly is allowed to send you DMs etc. etc.
Now comes the consistency part: Mike is right now working on porting some of (streams)' key technologies to ActivityPub, namely conversation containers, nomadic identity (currently exclusive to Hubzilla and (streams)) and permissions on (streams)' level.
All this is planned to become possible using only ActivityPub with a number of already existing FEPs. This means that any project that mainly or only uses ActivityPub can adopt all these features in the same way as (streams). Even Mastodon could, and the only way that'd stand in the way of adoption would be Gargron's unwillingness to adopt anything developed by someone else.
But if Mastodon did adopt all this, you would be able to moderate your own threads. You would have more power over your own connections and threads then than Mastodon moderators and admins have now.
CC:
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Fediverse #Threadiverse #Hubzilla #Streams #(streams) #PermissionsTIL that (streams) has a defined character limit of 24 million characters, but until now, the database limits the size of the object to a bit over 65,000 characters. The latter limit is being raised for new installations, and existing servers can raise that limit, too.
That's over 130 vs 48,000 Mastodon toots or more than 20 vs 8,000 Misskey notes.
Mind you, (streams) counts characters differently from Mastodon. Links aren't always 23 characters, they're as long as the URL plus the link code. The embedding code for in-line images, image URLs and alt-text included, is part of the post text. In general, all BBcode/HTML/Markdown counts into the character count. I'm not sure whether titles and summaries add to the overall character number, but at 24 million characters, that shouldn't make a difference.
Unfortunately, I no longer have the delivery report of the one time I exceeded the 65,000 character limit so I don't know how (streams) handles longer posts from outside.
The long post triggered something we haven't seen before... we allow the post body to be up to 24M, but the obj is capped at 65K. This was acceptable a year ago, but with conversation containers we put the entire post body into the object when sending a collection activity. So you're getting capped at 65K. You can change the item.obj column in the database and make it 'mediumtext'. I'll put this in the install database schema for new installs.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #QuotePost #QuoteTweet #QuoteToot #QuoteBoost #CharacterLimit #CharacterLimits This, however, is exactly how Mike Macgirvin 's creations have been working since Mistpark was launched 14 years ago.
This is how all of Mistpark's offspring works. This is how Friendica works, this is how Hubzilla works, this is how what's in the streams repository works. The thread starter owns the thread. All comments go to the thread starter, and the thread starter distributes them to all participants in the thread, mentioned or not.
Especially in (streams)' case, this goes all the way to moderation. Since the thread starter owns the whole thread including all comments, the thread starter even has moderational power over the thread. The thread starter can both delete comments regardless of who posted them. and ban actors from participating in a thread. On Hubzilla and probably also on Friendica, the thread starter can still delete any comments.
For someone who only knows Twitter and Mastodon, this sounds awfully complicated. In reality, it's an absolute no-brainer because this is exactly how Facebook works.
Also, unlike Mastodon, Mike's creations give you a counter and list of unread activities in your stream in addition to the counter and list of notifications. You know when someone has replied to you even if that someone hasn't mentioned you. You even know when someone has participated in a thread that you've seen before even if you have neither been mentioned nor replied to.
All this makes discussions with more than two participants a great deal more convenient than the Mastodon way. In fact, both Friendica, Hubzilla and (streams) have always had groups/forums as a feature which are nothing but accounts (Friendica)/channels (Hubzilla, (streams)) with particular settings.
Again, this is neither new nor experimental. It predates Mastodon by almost six years, and it has been in daily stable use for longer than Mastodon.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta
One, there's one Mastodon bot that scrapes cat pictures from various sources, slaps AI-generated alt-text on them with no human interaction, churns them out hourly and adds the #AltText hashtag because they've got alt-text.
Two, #AltText should be used for discussing alt-text rather than image posts with alt-text which is why hardly any human user puts that hashtag on image posts.
Three, hardly anyone seems to discuss alt-text anymore. That, or more and more Mastodon users have fallen back into Twitter mode and stopped using hashtags altogether. And outside of Mastodon, almost nobody knows about or cares for alt-text anyway.
Thus, the bot-generated cat pictures dominate that hashtag.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Hashtag #Hashtags #HashtagMeta #CWHashtagMeta #AltTextMeta #CWAltTextMeta
The only thing you can do is reply with an alt-text, add one or multiple of these hashtags: #Alt4U, #Alt4You, #AltText4U, #AltText4You, and hope the user gets the cue and adds the missing alt-text.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #AltText #AltTextMeta #CWAltTextMeta Does it absolutely have to be as close to Mastodon as possible
If not, allow me to introduce you to . Created and developed down under by Mike Macgirvin and the result of 14 years of evolution, longer than Mastodon has been around.
Main advantage: no media caching whatsoever. (streams) simply doesn't do that, neither do its predecessors, Friendica and Hubzilla.
Other advantages include nomadic identity, security and permissions on a level unimaginable for your average Mastodon user, many extra features and a smaller server footprint than Mastodon.
Okay, fair warnings, disadvantage #1: It has to re-download media over and over again.
Disadvantage #2: It comes with its own built-in WebDAV cloud storage, and people might be tempted to use it as such.
Disadvantage #3: It handles nothing like Mastodon. I'd say switching from Twitter to Mastodon is easier than switching from Mastodon to (streams), also due to (streams) not trying to mimic Twitter.
Disadvantage #4: no Mastodon API support which means absolutely no mobile apps available whatsoever. You can use it as a PWA instead.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Streams #(streams)
Da Friendica unter der grafischen Oberflche vllig anders funktioniert als Mastodon und Hubzilla sogar radikal anders, ist vllig unvorstellbar, wenn man bisher nur Mastodon benutzt hat.
Auerdem sind es typischerweise Friendica- und Hubzilla-Nutzer, die immer wieder auf Mastodon erwhnen, da das Fediverse nicht nur Mastodon ist. Und dabei treffen sie dann auf Mastodon-Nutzer, die nichts davon hren wollen.
#Long #LongPost #CWLong #CWLongPost #LangerPost #CWLangerPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Fediverse #NichtNurMastodon
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
Add your location to a Google MapBut 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 #AII've got over 500 connections, most of which are on Mastodon. Yet, it feels like I've got much less reach in the Fediverse now than I used to have when I only had some 200 or 300 connections. This can only mean that especially lots of Mastodon users and maybe even entire Mastodon instances have muted, blocked or shadow-banned me. I wouldn't even be surprised if some public Mastodon instances have secretly muted or blocked hub.netzgemeinde.eu as a whole just because of my actions.
I can't really tell from the delivery reports, for one difference between Mastodon and Hubzilla is that Hubzilla rejects blocked or filtered messages at the server level whereas Mastodon accepts them and then discards them before they can be delivered to the muting, blocking or filtering recipient. Otherwise I think I could weed out a whole lot of connections.
And here's a list of reasons for which I've probably been muted, blocked or shadow-banned:
- Posting over 500 characters. Regardless of whether or not I put them behind a Mastodon-style content warning.
- Posting over 500 characters without a Mastodon-style content warning. That's because I simply can't put Mastodon-style content warnings on comments. Still, I'm pretty sure that just about everyone who has ever come across an over-500-character comment from me while using the official Mastodon app has taken countermeasures.
- A summary in the content warning. Mastodon's content warning field is the summary field, but 99% of all Mastodon users don't know.
- More than four hashtags. On Mastodon, more than four hashtags are spam, regardless of whether they're spread across a 150-character toot or a 10,000-character essay or at the very end of the post. Also, Mastodon largely doesn't know the concept of filtering hashtags, it knows filter-generated content warnings even less, and it knows the concept of adding hashtags in order to trigger filters another great deal less.
- A few may still be upset about text formatting, even though I guess many mobile apps still don't show it.
- Talking about virtual worlds, but the Mastodon users in question don't know how to filter.
- Talking about the Fediverse, but actually talking about the Fediverse outside of Mastodon and/or making Mastodon look bad in comparison to non-Mastodon Fediverse projects.
- Questioning Mastodon practices that are set in stone and never discussed on Mastodon such as certain aspects of describing images.
Essentially, this boils down to two points:
- not adopting the entirety of Mastodon's culture as a Hubzilla user
- not abandoning Hubzilla's culture as a Hubzilla user
All because the Fediquette is entirely defined by Mastodon users, most of whom at that point didn't and often still don't know that the Fediverse is more than just Mastodon.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Mastodon #MastodonCulture #Fediquette #Mute #Muted #Block #Blocked #NotMastodon
all decentralized social networks
Certainly not all of them. How about:
- Hubzilla (part of the Fediverse because it can optionally use ActivityPub, but no Mastodon API support that's where I'm posting from)
- (streams) (part of the Fediverse because it can optionally use ActivityPub and does so by default, but no Mastodon API support)
- Diaspora* (part of the Fediverse by proxy, namely by being federated with Friendica, Hubzilla and GoToSocial which, in turn, can use or are based on ActivityPub)
Also, if "Mastodon" means that the ActivityPub side is only built and tested against Mastodon and only supports Mastodon's features, chances are it might not work with dozens of other Fediverse projects, including but not limited to Pleroma and Akkoma as well as Misskey and its many forks.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Openvibe #Mastodon #Pleroma #Akkoma #Misskey #Forkey #Forkeys #Hubzilla #Streams #(streams) #Diaspora* #NotOnlyMastodon #FediverseIsNotMastodon #MastodonIsNotTheFediverse
CC:
Franck Muller Long Island Crazy Color Dreams WhiteRG
Frank Muller Exclusively Crazy Hour Edition Design For Mens Now Available & Ready to ship same day # Frank Muller # For Men # Master Quality Island Crazy Color Dreams Non Comparable Original Japanese Power Resurve Automatic Jumping Machinery following Working Jumping Hour Stainless Steel rg Case -Black Genuine Leather Strap White Dial With Luminus Hands Top High End Back Original
It has become next to impossible to discuss content warnings in the Fediverse using hashtags.#CW has been hijacked by amateur radio where it stands for continuous wave. The few exceptions are about TV and refer to the CW Network.
#CWS has recently been hijacked by a commercial advertising account.
#ContentWarning has recently been hijacked for posts about the eponymous video game.
And both #ContentWarnings and #ContentWarningMeta are too long to type on a screen keyboard.
Not to mention that Mastodon is largely falling back into Twitter mannerisms, i.e. neglecting hashtags altogether, and most of the rest of the Fediverse has never had a hashtag culture to begin with.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Hashtag #Hashtags #HashtagMeta #CWHashtagMeta
Wie detailliert sollten sie sein
Da gibt's keinen Konsens. Das kann ich aus Erfahrung sagen, auch wenn ich fast kein Feedback bekomme.
Ich schreibe die mit Abstand lngsten Bildbeschreibungen im ganzen Fediverse und halte das auch gerechtfertigt. Einige feiern sie tatschlich, auch wenn es bis zu einer Stunde dauert, eine davon zu lesen. Inzwischen wundert mich das regelrecht.
Andere, kurioserweise durchweg ohne nennenswerte Behinderung, sind strikt dagegen und sagen auch mal, absolut jedes Bild lt sich adquat in 200 Zeichen oder weniger beschreiben. Und wer etwas nicht versteht, soll eben fragen oder googlen.
Eigentlich ist auf den Bildern meist noch mehr zu sehen, als ich schreibe. Und ich schreibe tendenziell viel.
Ich versuche tatschlich, alles zu beschreiben, was sich innerhalb des Bildes befindet und nicht komplett verdeckt ist. Allerdings zeigen meine Bilder ein so obskures Nischenthema, da meines Erachtens so detaillierte Beschreibungen gerechtfertigt sind. Nur dauert das eben seine Zeit.
Wenn die BildBeschreibungen nicht wren, wrde ich mehr Fotos zeigen (insbesondere auf pixelfed).
Kenne ich auch. Ich htte eigentlich viel zu zeigen. Aber die Bildbeschreibungen, wie ich sie schreibe, fressen derart Zeit und Energie, da zwischen meinen letzten zwei Bildposts mehr als drei Monate lagen.
Das letzte Mal habe ich mir schon berlegt, was ich zeigen knnte, was interessant ist, aber nicht so aufwendig zu beschreiben. Ich lag mal wieder komplett falsch und schrieb letztlich ber zwei Tage verteilt meine mit Abstand lngste Bildbeschreibung.
#Long #LongPost #CWLong #CWLongPost #LangerPost #CWLangerPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #BiBesch #Bildbeschreibung #Bildbeschreibungen #BildbeschreibungenMeta
Heat Check: Discussing Heats free agents and Jimmy Butler
does a consistent tree require a central resource to define the tree
The central resource is the actor that started the tree.
Traditionally, all replies go to this actor and this actor alone, and this actor distributes them to all other participants.
The whole tree with all its nodes is ideally owned by the actor who started it.
Does it require everyone have access to all nodes in the tree
Obviously, the actor who started the tree must not be muted or blocked, otherwise the entire tree is inaccessible.
As for comments, I'm not sure what happens when a blocked actor replies, and someone else replies to this reply. Most likely, the reply from the blocked actor will not appear because it still came from the blocked actor the actor who started the tree only forwarded it automatically. But I can't say how replies to a blocked reply are handled.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Fediverse #Threadiverse
even for forums, does the Fediverse really need a consistent tree
Yes, because nodeBB, Flarum, Lemmy, /kbin, Mbin, PieFed, Sublinks, Friendica, Hubzilla, (streams) and everything else that knows what a context is will interact with each other. They often already do.
It doesn't make sense for each one of them to try and patch incompatibilites with the others one by one, all by themselves, without even talking to each other. That might only cause more incompatibilites, not to mention loads of extra code just to address the quirks of individual projects.
It's better to have one agreed-upon standard.
I understand wanting a consistent tree, so everyone has access to all the posts, but I think the Fediverse focus on safety should override an attempt at consistency
Safety can and certainly will come with consistency.
Look at Mike Macgirvin 's , the result of 14 years of work that started in 2010 with Mistpark, now known as Friendica. (streams) knows what a context is. (streams) can hold a discussion together. (streams) supports groups/forums. And so did all its predecessors, two of which (Friendica, Hubzilla) are still alive and maintained.
At the same time, (streams) has to be the safest place in the whole Fediverse. Not necessarily by culture or policy, but by technology. (streams) has ramped up its permission controls beyond what Hubzilla has to offer, and Hubzilla's permission controls are already extensive. In the meantime, Mastodon and everything else in the Fediverse that's based on ActivityPub doesn't even know what permissions are.
On Mastodon, all you have is mute and block for users or instances. Hubzilla goes much, much further, (streams) goes even further. Imagine being able to delete posts from a thread that you've started (Hubzilla, (streams)) or even banning users from participating in your thread ((streams)). Imagine being able to define who exactly is allowed to send you posts, who exactly is allowed to see your profile (or which one of your multiple profiles even), who exactly is allowed to comment on your posts, who exactly is allowed to send you DMs etc. etc.
Now comes the consistency part: Mike is right now working on porting some of (streams)' key technologies to ActivityPub, namely conversation containers, nomadic identity (currently exclusive to Hubzilla and (streams)) and permissions on (streams)' level.
All this is planned to become possible using only ActivityPub with a number of already existing FEPs. This means that any project that mainly or only uses ActivityPub can adopt all these features in the same way as (streams). Even Mastodon could, and the only way that'd stand in the way of adoption would be Gargron's unwillingness to adopt anything developed by someone else.
But if Mastodon did adopt all this, you would be able to moderate your own threads. You would have more power over your own connections and threads then than Mastodon moderators and admins have now.
CC:
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Fediverse #Threadiverse #Hubzilla #Streams #(streams) #PermissionsTIL that (streams) has a defined character limit of 24 million characters, but until now, the database limits the size of the object to a bit over 65,000 characters. The latter limit is being raised for new installations, and existing servers can raise that limit, too.
That's over 130 vs 48,000 Mastodon toots or more than 20 vs 8,000 Misskey notes.
Mind you, (streams) counts characters differently from Mastodon. Links aren't always 23 characters, they're as long as the URL plus the link code. The embedding code for in-line images, image URLs and alt-text included, is part of the post text. In general, all BBcode/HTML/Markdown counts into the character count. I'm not sure whether titles and summaries add to the overall character number, but at 24 million characters, that shouldn't make a difference.
Unfortunately, I no longer have the delivery report of the one time I exceeded the 65,000 character limit so I don't know how (streams) handles longer posts from outside.
The long post triggered something we haven't seen before... we allow the post body to be up to 24M, but the obj is capped at 65K. This was acceptable a year ago, but with conversation containers we put the entire post body into the object when sending a collection activity. So you're getting capped at 65K. You can change the item.obj column in the database and make it 'mediumtext'. I'll put this in the install database schema for new installs.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #QuotePost #QuoteTweet #QuoteToot #QuoteBoost #CharacterLimit #CharacterLimits This, however, is exactly how Mike Macgirvin 's creations have been working since Mistpark was launched 14 years ago.
This is how all of Mistpark's offspring works. This is how Friendica works, this is how Hubzilla works, this is how what's in the streams repository works. The thread starter owns the thread. All comments go to the thread starter, and the thread starter distributes them to all participants in the thread, mentioned or not.
Especially in (streams)' case, this goes all the way to moderation. Since the thread starter owns the whole thread including all comments, the thread starter even has moderational power over the thread. The thread starter can both delete comments regardless of who posted them. and ban actors from participating in a thread. On Hubzilla and probably also on Friendica, the thread starter can still delete any comments.
For someone who only knows Twitter and Mastodon, this sounds awfully complicated. In reality, it's an absolute no-brainer because this is exactly how Facebook works.
Also, unlike Mastodon, Mike's creations give you a counter and list of unread activities in your stream in addition to the counter and list of notifications. You know when someone has replied to you even if that someone hasn't mentioned you. You even know when someone has participated in a thread that you've seen before even if you have neither been mentioned nor replied to.
All this makes discussions with more than two participants a great deal more convenient than the Mastodon way. In fact, both Friendica, Hubzilla and (streams) have always had groups/forums as a feature which are nothing but accounts (Friendica)/channels (Hubzilla, (streams)) with particular settings.
Again, this is neither new nor experimental. It predates Mastodon by almost six years, and it has been in daily stable use for longer than Mastodon.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta
GPT-4 supera gli analisti finanziari nella previsione degli utili aziendali
ChicagoBooth
Three reasons.One, there's one Mastodon bot that scrapes cat pictures from various sources, slaps AI-generated alt-text on them with no human interaction, churns them out hourly and adds the #AltText hashtag because they've got alt-text.
Two, #AltText should be used for discussing alt-text rather than image posts with alt-text which is why hardly any human user puts that hashtag on image posts.
Three, hardly anyone seems to discuss alt-text anymore. That, or more and more Mastodon users have fallen back into Twitter mode and stopped using hashtags altogether. And outside of Mastodon, almost nobody knows about or cares for alt-text anyway.
Thus, the bot-generated cat pictures dominate that hashtag.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Hashtag #Hashtags #HashtagMeta #CWHashtagMeta #AltTextMeta #CWAltTextMeta
Ambience at Hanoi's 1000-year Anniversary
It's technologically impossible to add alt-text to someone else's media in someone else's posts. This is extremely unlikely to change, and when that someone else is on certain non-Mastodon instances, this will definitely never change.The only thing you can do is reply with an alt-text, add one or multiple of these hashtags: #Alt4U, #Alt4You, #AltText4U, #AltText4You, and hope the user gets the cue and adds the missing alt-text.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #AltText #AltTextMeta #CWAltTextMeta Does it absolutely have to be as close to Mastodon as possible
If not, allow me to introduce you to . Created and developed down under by Mike Macgirvin and the result of 14 years of evolution, longer than Mastodon has been around.
Main advantage: no media caching whatsoever. (streams) simply doesn't do that, neither do its predecessors, Friendica and Hubzilla.
Other advantages include nomadic identity, security and permissions on a level unimaginable for your average Mastodon user, many extra features and a smaller server footprint than Mastodon.
Okay, fair warnings, disadvantage #1: It has to re-download media over and over again.
Disadvantage #2: It comes with its own built-in WebDAV cloud storage, and people might be tempted to use it as such.
Disadvantage #3: It handles nothing like Mastodon. I'd say switching from Twitter to Mastodon is easier than switching from Mastodon to (streams), also due to (streams) not trying to mimic Twitter.
Disadvantage #4: no Mastodon API support which means absolutely no mobile apps available whatsoever. You can use it as a PWA instead.
#Long #LongPost #CWLong #CWLongPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Streams #(streams)
MIDSUMMER SCREAM, the Worlds Largest Halloween and Horror Convention, Launches 2024 App and Schedule
Halloween may be five months away, but the Halloween season begins July 26 to July 28 with MIDSUMMER SCREAM. The worlds largest Halloween and horror convention, has unveiled its official app, which allows attendees to plan and schedule their spook-filled days at this years event and with all there is to do, a schedule is a must....
scream beach july 2024
Lets Get Our First REAL Mage Weapon Terraria #
hair
Wanderer Fr viele, die nur Mastodon kennen, sind Mastodon, Akkoma, Misskey, Friendica, Hubzilla usw. nichts weiter als verschiedene Webinterfaces fr dasselbe System darunter und haben denselben Stellenwert wie Mona, IceCubes, Tusky, Fedilab & Co. fr Smartphones.Da Friendica unter der grafischen Oberflche vllig anders funktioniert als Mastodon und Hubzilla sogar radikal anders, ist vllig unvorstellbar, wenn man bisher nur Mastodon benutzt hat.
Auerdem sind es typischerweise Friendica- und Hubzilla-Nutzer, die immer wieder auf Mastodon erwhnen, da das Fediverse nicht nur Mastodon ist. Und dabei treffen sie dann auf Mastodon-Nutzer, die nichts davon hren wollen.
#Long #LongPost #CWLong #CWLongPost #LangerPost #CWLangerPost #FediMeta #FediverseMeta #CWFediMeta #CWFediverseMeta #Fediverse #NichtNurMastodon
Who needs when you have a ...
Last updated: 2024-06-13
All text and images © eLoc8.is-fabulous.com