Rebrand ZOT6+ ?

2024-11-03 11:34:15

𝓒𝓱𝓻𝓲𝓼

chris@im.allmendenetz.de

There may be reasons to rebrand "ZOT6+" to "Nomad"?

What are the thoughts of the community about that?

Mike Macgirvin 🖥️ wrote the following post Sat, 02 Nov 2024 22:24:48 +0100

(..)

'Zot6' is sometimes mentioned in discussions with Hubzilla folks to distinguish that project's frozen protocol implementation (nearly always displayed as 'Zot6/Nomad'); but if you're speaking in general terms about the underlying nomadic protocol in documentation, be advised that some folks in the fediverse get pretty riled up over dead-naming.

2024-11-03 12:29:46

Mario Vavti

mario@hub.somaton.com

A lot of work for what reason?

2024-11-03 13:43:41

Der Pepe (Hubzilla) ⁂ ⚝

pepecyb@hub.hubzilla.hu

I understand the approach behind the question. There is also confusion when people talk about Zot and Nomad at the same time.

However, I don't see any real necessity. I'm still not sure why Mike renamed Zot.

Zot is an established term. Nomad, on the other hand, is not. When I first read about Nomad as a protocol, web searches led me to pages about ‘something to do with blockchain and digital wallets’. For me, that's actually the moment when I stop. Now the hits in the search had nothing to do with the Nomad protocol from e.g. (streams)... but it can contribute to confusion. A search for Zot, on the other hand, leads to information about Hubzilla's communication protocol... so it fits.

And there is even a ‘Zot protocol community group’ at W3C. 😉

My personal opinion: stick with the term ‘Zot’ and avoid ‘Nomad’ as far as possible in connection with Hubzilla.

2024-11-03 14:09:44

Scott M. Stolz

scott@loves.tech

Up until this point, the development of the protocol has been somewhat linear. It was called Zot when Hubzilla implemented it, and the version we use is Zot 6. Mike kept on developing new versions of Zot, but Hubzilla stayed with Zot 6. And then at some point Zot was rebranded as Nomad. Because of this name change, one could argue that we are running Nomad 6. And that seems to be Mike's argument.

Following this linear timeline, Zot and Nomad are the same thing, and it is just a matter of what we prefer to call it.

~~But the upcoming version of Hubzilla 10 is going to update Zot 6 to support Mike's conversation containers rather than upgrade to the latest version of Nomad.~~

~~What that means is that the progression is no longer linear. Zot 6.1 (or whatever you want to call it) is effectively a fork. A small stubby one, but a fork none-the-less.~~

So, the question becomes, is Hubzilla going to eventually adopt Nomad 12 (or whatever version it is now), or are we going to ~~continue to~~ modify Zot 6 so that it deviates into being a separate protocol from Nomad/Zot6?

Edit: inaccurate information struck out.

2024-11-03 14:21:23

*_jayrope

jrp@hub.kliklak.net

@Scott M. Stolz I guess, that (streams) is already - and now not anymore - a Hubzilla fork, so it seems just logical, that the language was developed separately as well. I prefer, when Hubzilla goes it's own way there for the simple reason, that readapting a former fork might cause more trouble, that a thourough analysis of the good things, that have been found there - and then coding the good stuff again into the new version of Zot6.

Sent to the #Fediverse from a #Hubzilla instance.

2024-11-03 14:23:49

Scott M. Stolz

scott@loves.tech

@_jayrope It's interesting. Streams is a fork of Hubzilla, ~~and by adding conversations containers to Zot6, we are basically forking Nomad/Zot by doing so.~~

Edit: Inaccurate. Post will probably be deleted.

2024-11-03 14:28:04

Der Pepe (Hubzilla) ⁂ ⚝

pepecyb@hub.hubzilla.hu

@Scott M. Stolz

we are basically forking Nomad/Zot by doing so

I would rather call it a "backport".

2024-11-03 14:31:51

Mario Vavti

mario@hub.somaton.com

@Scott M. Stolz

But the upcoming version of Hubzilla 10 is going to update Zot 6 to support Mike's conversation containers rather than upgrade to the latest version of Nomad.

That is not a change at the protocol level

2024-11-03 16:26:13

𝓒𝓱𝓻𝓲𝓼

chris@im.allmendenetz.de

@Scott M. Stolz

So, the question becomes, is Hubzilla going to eventually adopt Nomad 12 (or whatever version it is now), or are we going to continue to modify Zot 6 so that it deviates into being a separate protocol from Nomad/Zot6?

(..)

If we upgrade, then it becomes Nomad.

When i read this i wonder what is meant by that... are you actually aware of the differences of ZOT and Nomad?
Would an upgrade make sense or is it even possible in the first place to upgrade HZ to "Nomad" without loosing part of what HZ is?

After all there were significant reasons why Mike started a new project and found a new name for the protocol he rewrote in parts. - Just a guess.

I`m not able to tell what the differences are, maybe someone can explain - would be kind. Thanks.

2024-11-03 14:16:57

Scott M. Stolz

scott@loves.tech

It also should be noted that what you call it internally in the software does not have to be the same was what you call it publicly and in documentation.

But before we decide what to call it, we need to decide whether we will eventually upgrade or fork. If we upgrade, then it becomes Nomad. If we fork, it should not be called Nomad to avoid confusion.

Unless we want to do some sort of "Nomad family of protocols" sort of thing.

2024-11-03 14:37:07

Scott M. Stolz

scott@loves.tech

Another relevant question is the future of Nomad. Mike said he retired from working on Streams, and he is now working on a project called Forte that does not even have Nomad or Zot6 in it.

If Nomad is no longer being developed, and we are the only actively developed project still using it, then how does that affect things? If that is the case, does that mean we become the de facto maintainers of the protocol?

I hope that Mike continues to develop Nomad and Streams, but his latest announcements seem to indicate otherwise. Hopefully he is just taking a break and will come back to Nomad and Streams.

2024-11-03 16:52:17

Der Pepe (Hubzilla) ⁂ ⚝

pepecyb@hub.hubzilla.hu

I also ‘quickly’ tried to grasp the differences. You can forget it. Converting Hubzilla to Nomad would probably be a very, very complex project. Is it worth it? I have no idea.

In my opinion, it should be limited to keeping Zot as Zot (Zot7, Zot8, Zot9...) and only transferring the appropriate and useful parts from Nomad to Zot. This is easier to manage and would prevent a break in Hubzilla's compatibility and functionality.

A look at the specifications of Zot6 and Nomad does not reveal the actual differences.

2024-11-03 17:32:37

Der Pepe (Hubzilla) ⁂ ⚝

pepecyb@hub.hubzilla.hu

Zot is not a ‘brand name’, but a name for a protocol. What else should we call it? ‘A protocol’ perhaps?

Mistpark, Osada, Zap etc. are dead. Since the name Zot finally came to an end with version 12 (and has been called Nomad since then and had developed incompatibly anyway), it would only make sense to name the further development of Zot6 either by skipping the version numbers up to 12 and making it Zot13 (nothing against the 13, that's a completely normal number 😉), or to continue with Zot61, Zot62 etc. in order to keep the ‘6’ as the root of the new Zot branch.

2024-11-03 22:09:36

𝓒𝓱𝓻𝓲𝓼

chris@im.allmendenetz.de

so there are different names and follow up projects with different features but still no one here could name me the actually differences which the protocol make and what part they have by the noticeable changes

2024-11-03 22:14:39

Giac El Vecio

giacelvecio@unio.diversispiritus.net.br

Well, a notable difference is that you in streams can post using markdown, BBCode, and also html, using zot just the first 2.

2024-11-03 22:35:33

Der Pepe (Hubzilla) ⁂ ⚝

pepecyb@hub.hubzilla.hu

@Giac El Vecio

using zot just the first 2

Are you absolutely sure about
that?

So if I use HTML to write a post, then that works too. It is converted to bbCode anyway.

Are you absolutely sure about
that?

So if I [b]use[/b] [size=large]HTML[/size] to write a [i]post[/i], then that works too. It is converted to [url=https://hzhelp.pepecyb.hu/en/usermanual/bbcode.html]bbCode[/url] anyway.

2024-11-03 22:59:33

Giac El Vecio

giacelvecio@unio.diversispiritus.net.br

Is not the same, you can use pure html.

2024-11-03 23:10:51

Giac El Vecio

giacelvecio@unio.diversispiritus.net.br

Now it works yes ,also in hz, I tested months ago and didn't work, when I tried to explain this for friend.

2024-11-03 23:38:15

Giac El Vecio

giacelvecio@unio.diversispiritus.net.br

In the early days of nomad,
i asked Mike what the difference was between zot and nomad and he answered in a comment, if you want to look for that conversation...i don't even remember if it was a conversation started in a forum or not, but i remember a part where he says he didn't want to "put all the cards on the table" where everyone could see "our secrets" or something like that 😆

2024-11-04 09:28:07

𝓒𝓱𝓻𝓲𝓼

chris@im.allmendenetz.de

It is all about communication :-)

The term which makes things more clear and gets understood by the relevant folks the best may be the term to use.

Time after time the significant parts of Nomad will get part of AP and Nomad will dissolve and vanish as an own protocol - this is the project Mike is working on.

How will HZ follow here or will take its own path and place?

If we want to emphasize that HZ has and will have it's own protocol than the term ZOT is just right.

If we want to make clear that HZ uses since a long time parts of the technology which in the future will become a feature by other apps as well ( in that what is called the Fediverse ) than it may be not a bad idea to use Nomad as term because that is the name other devs may understand the best, without studding all the history.

But for what does the protocol stand at the end anyway?

For instants: streams has few more permissions to apply - how does this option depend on Nomad or could the changes be part of HZ as well without touching the protocol?

or

The HZ and streams IDs can´t be ported and cloned between each other any more - is this because of the protocol or because of other reasons?

I do not know and can't tell, but we should know.

It is all about communication.

2024-11-04 09:45:34

𝓒𝓱𝓻𝓲𝓼

chris@im.allmendenetz.de

Mike explained to me a while ago:

Nomad sends messages encoded in ActivityStreams (the same serialisation format used by ActivityPub). As does Zot6. So technically the only difference between an ActivityPub message and a Nomad message and a Zot6 message is that they use different delivery services. In the case of Zot6 and Nomad, these delivery systems support nomadic identity.

So according to this statement "just" the nomadic identity feature makes all the difference.

2024-11-04 09:56:25

𝓒𝓱𝓻𝓲𝓼

chris@im.allmendenetz.de

Mike Macgirvin 🖥️ wrote the following post Sun, 03 Nov 2024 21:24:30 +0100

(...)

(*) ActivityStreams is a data serialisation protocol. ActivityPub and Nomad are message transfer protocols. Both of these are consumers of ActivityStreams formatted data. The most notable difference is that ActivityPub requires all identities to be DNS-based URLs. Nomad identities are cryptographically derived and are not permanently tied to any single DNS address.

2024-11-04 10:18:48

Der Pepe (Hubzilla) ⁂ ⚝

pepecyb@hub.hubzilla.hu

My current idea: I would leave it as it is for now, even if there might be a little confusion here and there about whether and how Nomad and Zot are connected. I think that's negligible.

I also believe that we are making the mistake here of equating the actual protocol and the implementation. A communication protocol only describes the rules for communication between servers. How this is ultimately implemented in code is independent of this. In my opinion, however, the published specifications are quite sparse as far as the concrete syntax is concerned. This is not an accusation against Mike, but quite normal that the person who develops such a protocol does not present many things abstractly enough for others ‘because he himself knows exactly how it is meant’.

As I said, I think the equation is unfortunate.... but nevertheless it inevitably results from the naming of the libraries, the programme routines that ultimately implement the protocol. It is completely logical and natural that if the protocol is called ‘Zot’, routines and libraries are somehow also called something with ‘Zot’. This makes it clear what they do... namely to implement communication based on the Zot protocol as software.

As far as I know, the fact is:

After Mike left and handed over the Hubzilla project, the protocol (in the above sense, i.e. not the software implementation) of Hubzilla was and is Zot6. And it still is today. I don't know whether the specification for the actual protocol was also changed for changes, improvements and new functions (of the software, and thus also of the libraries/routines). I suspect that this did not happen. New features were only a programmatic matter, but they utilise and adhere to the original protocol.

It is probably also a fact that Nomad is possibly a completely different protocol (with regard to the technical specifications, which unfortunately cannot be found anywhere). This seems to be the case because there are clear incompatibilities.

The question is whether the actual protocol was changed in the technical specification during Mike's further developments with and after Osada, Zap (and whatever the projects were called). Were there new protocol specifications, or was it also the case that the consecutive numbering was more due to the implementation of new functions based on the protocol in the form of new routines?

In any case, as I see it, Hubzilla actually has nothing to do with Nomad, and I don't see ‘Nomad’ as a general umbrella term for Zot and the actual Nomad protocol, because that doesn't correspond to the facts. Zot is Zot and Nomad is Nomad.

This could also lead to real problems. If we say Hubzilla also uses a ‘Nomad protocol’ (a, not THE), and someone (whether from the original repo or a fork) sets up a ‘(streams)’ instance. Presumably he (and users of the Hub) will be disappointed if it doesn't interact with Hubzilla as hoped (cloning, for example).

Oh... and we really shouldn't orientate ourselves on all the projects (Zap, Osada... etc.) that only had a short lifespan. The protocol versions (or are they just software versions that always implement the same technical protocol?) are now meaningless because the projects have effectively been cancelled and are hardly running anywhere.

2024-11-04 13:29:08

Giac El Vecio

giacelvecio@unio.diversispiritus.net.br

What I think (if anyone is interested):
having 2 different protocols was nice and useful, having a single protocol used by incompatible platforms (incompatible at 25% or 40% or more) is believing in a utopia of a perfect fediverse in terms of communication and results, so it is finding yourself in a sea of trouble.
Having a single protocol means being exposed to an ocean of shit.
You can invent something powerful to keep this shit out, ok, but this translated means: communicating with just 5 or at most 10% of existing projects.
So it's a bit like going back to square one.

2024-11-04 20:45:56

Scott M. Stolz

scott@loves.tech

One question floating in my head is this: If they are the same protocol, why are they incompatible?

Is it simply lack of if then statements to detect whether the other version supports a certain feature, or has the protocol itself changed.

For example, a lot of platforms are running very different versions of ActivityPub, yet they can talk to each other regardless, albeit with varying levels of support for certain functionality.

Is it the implementations that are incompatible, or is it a protocol change that made it incompatible?

2024-11-05 09:22:53

Der Pepe (Hubzilla) ⁂ ⚝

pepecyb@hub.hubzilla.hu

@Scott M. Stolz

With a fresh head in the morning, in the context of what I wrote yesterday (writing something down often helps to sort your thoughts) and the comment Mike wrote (I had seen it, but hadn't realised the actual statement behind it), I now have a fairly concrete idea, which I will put into words here again (not now, for now the dogs and horses demand my love and care 😉).

2024-11-05 09:48:41

𝓒𝓱𝓻𝓲𝓼

chris@im.allmendenetz.de

@Mario Vavti could you point us to snipe of code where ZOT is active and mark the lines for us? That would be kind. Thanks

Than we could have a look into the steams repo and find out what is different

2024-11-05 10:26:52

Der Pepe (Hubzilla) ⁂ ⚝

pepecyb@hub.hubzilla.hu

I have now interpreted it like this...

Mike has communicated very clearly: The protocol that used to be called Zot is now definitely called Nomad.

He also distinguishes between the protocol and the implementation in a software and mentions the implementation of the protocol with the name ‘Nomad’ at Hubzilla in the software modules and libraries that bear/contain the name ‘Zot6’ (‘...that project's frozen protocol implementation...’). He acknowledges, but does not contradict, the fact that ‘Zot6’ or (as is now more common) ‘Zot6/Nomad’ is spoken/written in connection with the code.

So we have a communication protocol called ‘Nomad’ and we have the protocol implemented in Hubzilla, which is done by libraries and functions with the name ‘Zot6’.

In my opinion, this means that it is irrelevant what the programmed implementation of the protocol is ultimately called. The user is not interested in this either (I don't care what name any function in the source code of an application that I use has). And it follows that we would have to replace the word Zot with Nomad in all public documents (project page, help pages, info pages, etc.) if we are referring to the protocol. And that is what we are referring to almost exclusively. Only in the developer documentation should the term ‘Zot’ or ‘Zot6’ or more precisely ‘Zot6/Nomad’ appear, because the code parts simply have this name.

I am of the opinion that renaming in the code is out of the question because it would be a very risky (and possibly also lead to backwards compatibility problems) and extremely costly thing to do (is disproportionate to the benefit and would tie up developer capacities that would be much more important for other things).

I also see a clear advantage in the fact that the software implementation of the Nomad protocol does not also have "Nomad" in its name: the clearer separation of the protocol and the software implementation of the protocol. This has led to confusion in the past (yes, even today). When Mario added new functions to Hubzilla that represented a permitted implementation of the protocol, many people had the feeling that Zot6 had been expanded and a new version had arrived. However, the protocol has not changed at all, only the library, which provides new functions based on the same protocol.

Are these thoughts understandable? Am I wrong somewhere?

2024-11-05 16:31:20

Der Pepe (Hubzilla) ⁂ ⚝

pepecyb@hub.hubzilla.hu

@Scott M. Stolz

For Streams the corresponding file is called NomadHandler.php and for Hubzilla Zot6Handler.php:

However, I think we could (and should for the reasons mentioned) stick with Zot6. Mike probably named the handler library after ‘Nomad’ because he renamed the protocol the same way. But in the end, it doesn't really matter what the programme files are called. For Hubzilla, the Zot6 routines are the ones that implement the Nomad protocol in Hubzilla. The name can be left as it is for ‘historical reasons’.

2024-11-26 12:55:06

𝓒𝓱𝓻𝓲𝓼

chris@im.allmendenetz.de

Mike Kutetz shows Nomad and ZOT as two different protocols. :-(
__

#hubzilla wrote the following post Tue, 26 Nov 2024 12:35:17 +0100

https://social.tchncs.de/@kuketzblog/113549044465749902

Wir haben das Fediverse-Schaubild erneut aktualisiert! Ein herzliches Dankeschön an @imke für die Überarbeitung und die wertvolle Unterstützung. Hier ist die neueste Version des Fediverse-Schaubilds – schaut es euch an! 👇

Gerne teilen!

#fediverse #schaubild #mastodon #pleroma #misskey #hubzilla #diaspora #lemmy #kbin #friendica #bookworm #funkwhale #peertube #pixelfed #activitypub

2024-11-26 13:55:30

Der Pepe (Hubzilla) ⁂ ⚝

pepecyb@hub.hubzilla.hu

Ich weise ihn mal darauf hin, dass Zot auch Nomad heißt (unter Verweis auf die Assage von Mike, wo er das ja explizit so geäußert hat).

2024-11-26 13:58:34

𝓒𝓱𝓻𝓲𝓼

chris@im.allmendenetz.de

hab es gerade schon in dem thread gepostet

2024-11-26 14:14:43

Der Pepe (Hubzilla) ⁂ ⚝

pepecyb@hub.hubzilla.hu

Bin mal gespannt...

2024-11-26 14:23:51

Der Pepe (Hubzilla) ⁂ ⚝

pepecyb@hub.hubzilla.hu

Im Endeffekt müsste er nicht mit unterschiedlichen Farben etc. arbeiten, sondern einfach nur Nomad nennen. Die interne Bezeichnung der Bibliothek, welche das Protokoll umsetzt, interessiert niemanden und will auch keiner wissen.

Also z.B. so:

Image/photo

2024-11-26 17:09:40

Der Pepe (Hubzilla) ⁂ ⚝

pepecyb@hub.hubzilla.hu

@𝓒𝓱𝓻𝓲𝓼 @Scott M. Stolz

That seems ok to me so far (so write Nomad/Zot6+). Ultimately, however, the name Zot is just the name of the library that implements the Nomad protocol.

2024-11-26 22:46:30

Der Pepe (Hubzilla) ⁂ ⚝

pepecyb@hub.hubzilla.hu

@Scott M. Stolz
I think your idea regarding the procedure is a smart one. That's how we should do it.

I'll think about the best way to incorporate and implement this in the help (current status of the new version). I already have some ideas.

2024-12-09 23:47:29

𝓒𝓱𝓻𝓲𝓼

chris@im.allmendenetz.de

Mario Vavti wrote the following post Mon, 09 Dec 2024 23:14:43 +0100

Just to clarify, as I have seen claims recently that Hubzilla uses the nomad protocol: it does not.

Hubzilla uses the zot protocol. While zot and nomad are somewhat related, they are not the same and are not fully compatible.

Thanks for your attention!

2024-12-10 00:48:04

Der Pepe (Hubzilla) ⁂ ⚝

pepecyb@hub.hubzilla.hu

@Mario Vavti @𝓒𝓱𝓻𝓲𝓼 @Scott M. Stolz

As I understand Mike's comments, the protocol has been called Nomad for some time (since Zot was renamed Nomad). Hubzilla uses the subset of Nomad that is implemented by the Zot library (which has not been renamed and is available in version 6).

Streams, on the other hand, uses the Nomad protocol completely and the programme library there is also called Nomad.

This would not be an inconsistency either. Fediverse services that use AP natively do not always use all AP protocol specifications, but subsets of them. But they still use AP.

Is that wrong again? Which statement can I rely on? Do I have to rewrite the documentation in which I have now made the distinction between the protocol (now Nomad) and the programme routines (Zot) that implement the protocol? Can Scott take his website opennomad.net off the net again?

2024-12-10 04:50:54

Scott M. Stolz

scott@loves.tech

And there is another consideration. Mario is maintaining Zot 6, and Mike has declared that he retired. If that is the case, then we should defer to the active developer.