Project Midnight is interesting, but I do not see anything that tells me of the vision.
I have started building my own social networking system, but it is very different. It will be:
Anonymous
High trust
All user data will be held by the user, and is the basis of that trust.As such, although I will be using WordPress as a platform, the network will operate largely outside its user and content management system.
My system will be designed to operate outside of the internet, largely, as tracking, user data collection, and commercialism will not be allowed. Security will be entirely distributed, and based on levels of trust collected via a participating network entry point. Here, users will effectively judge themselves and others, controlling how trusted each other is, but helped to rise up those levels of trust, based on a gamified interaction zone.
There will never be a data breach, as my network will have no personal data, ever.
Encryption will also be at the heart of this new solution. I want to create a place where people can come to trust each other, and start working together again. They need to know that nobody is watching them, collecting their data, excluding them based on financial or other exclusionary tactics, other than, then staying as a part of the network and having sufficient positive interactions in the journey to have collected sufficient trust keys for a level of trust to be given to that user, whoever they are.
My psychology website, self-transcendence.org will be one of those entry points. I am currently moving it into a 25 trust level content management system. Within that architecture, I will have specific training courses for people who want to use (safely) moderate or manage social networks. Completing those courses will ultimately be part of that trust model.
Hi George,
Welcome back.Thank you for opening the discussion and thank you for sharing about your network.
What I understand from your reply is:- You are building a federated system like Mastodon where users may run their own server(or may be you are storing it in browser if they don’t need server) and uses heavy encryption.
Our Project Midnight Sun(codename), is not a federated social network system. It is more like how the popular social networks like Facebook/Twitter(X)/Linkedn etc. deal with user and data.
You can think of it as modern software for building community with WordPress. Our main focus has been developer productivity and site admin’s flexibility. We have architected the system in a way to scale. I have been personally working on it for more than a year now, have built custom ORM, form builder, background system and a lot more as the infrastructure layer. Anything that you can do with current available systems, you can do it in 1/10th time if you are a developer.
The project is in heavy development state(I have the infrastructure layer ready, routing, navigations, core API’s working.) I am working on modules and UI and my expectation is to go live in next 20 days. The current system is not something that we can show you as UI is a work in progress. For now, You may need to wait until 20th for the first demo/availability of the software.
PS:- I understand that different us cases have different requirement for data processing. We are more like traditional community in the sense. We are not anonomyzing data at user level but it can be made so at display level as anonymous/incognito is a built in feature for all kind of data.
Regards
BrajeshI’ve started to discuss the design philosophy – https://venutius.com/project-anonnet/
Hi George,
Congratulations! The ideation looks fantastic(we are on different path). In our case, we believe site admins(who run the network) should have the control over what they want to do with user data and how much freedom/features they want to give to the users. It is not our job to decide that, our job is to empower site admins/community managers with the tools and let them decide.I wish you all the best and I am looking forward to a better future for community building with WordPress!
Well, my user trust levels allow for 30 different levels of user access. New Admin, for example, start as trainees, at admin level 1, and they progress, based on the agreement of the super-admin/site owner, through five levels of trust, until they can manage everything, should the site admin wish. In addition, the site admin gets to choose which features become manageable, at each level of trust. They have complete control over every nuance of responsibility that they hand over.
In addition, users have 25 levels of trust, and the site admin gets to control everything, from the WordPress side of things, all data stored on their WordPress site, all links, cats, posts, and any other feature they choose to apply a trust setting to, will control exactly what level of knowledge, paid access, whatever, they just get a finer level of granularity to that access control.
AnonNet, however, is a social network, not a user network, not a commercially driven, biases and controlling network. And there, the owner of the data, is the person that owns it, the user. AnonNet is the back-end network to relax and spend time with friends. It also avoids me having to collect proof of age for every site that I run, simply because some content could be classified as too complex for a younger person to understand.
But also, AnonNet, by its architecture, moves like the wind – servers relocating themselves the minute they think they are under attack.
I was wondering what your architecture is going to look like? I mean, as far as I can see, there is no information on it at all. AnonNet may be able to use it. You see, as I said, AnonNet is a back-end, social, gaming and learning network. It is not public, as such, all users have to earn their trust, it can’t be paid for, and that is why it cannot be commercial, in any way. But, you can have your own AnonNet, and you can use it exactly how you like, anyone can. But if it becomes a commercial, then the linkage with the core anonNet would be reduced, directory listings, too, are based on trust. Anyone that walks in the room looking for sales, is not authentic, and is not going to play nice in our network.
The great thing about trust based systems, is that you get to trust your users, and can do things like automatically reduced levels of trust for a user who gets blocked by too many users with a set trust level.
I’ve completed the user trust level manager, It’s actually working on a live site, but only in the background. Part of its functionality is assigning trust levels to bots, and part of that, is that bots also get assigned trust levels, and the less we trust them, the less of our site they get to see.
You see, what I’m interested in, is the idea of having deep conversations online again. Facebook has killed that.
Imagine you had a WordPress site, with a shop or other commercial content. You might be selling training, or anything. You may well have buddypress as part of that, but that social network is a sales grift network, it is not about relaxing and letting your hair down. Where do you go if you want to chat to friends? Do you want all that personal chat recorded and tracked? Shipped off to a third party site for “processing”? Why do you want to know what your users are talking about? I don’t, personally, and I find their personal data to be an issue if I am the one holding it. These privacy laws are designed to kick all of the small sites out of the net, by forcing them to collect identity info.
I’m designing a system that avoids all of that. An open, free system that works side by side with WordPress and any public social network that they may have, and it’s systems and services, could benefit from the resilience and security features that I’m developing. A server with minimal data, can move like the wind, it may run in just a handful of megabytes, but yes, not your bloats WordPress installation, that is far too large to be agile, and that is the point – I’m decoupling the data from the access and control mechanism. The login server is the agile, no user data AnonNet polling server, whose location changes every day, and only you have the history, encoded in shared keys, to prove that you are that person, with that earned trust, on our, or your network.
Is your plan to use the existing user and security architecture of WordPress? I personally think that lack of granularity of access and trust levels meant BP was never scalable.
But also, WordPress is going to die soon. It’s moving commercial, this is why my ultimate goal is to have my mini servers hosted on peoples own PCs, and not need hosting, not from a hosting provider. They will be able to make backup agreements with admins that they trust. They would be friends that they met in the network, and had gained a very high level of personal trust with them. That is a trust that working with AnonNet, with its systems, policies and training, is always going to be far higher.
That is a bit of the vision. Better than Crypto network levels of security and resilience, that to a trust model that does not take cash as any sign of trustworthiness. Most of the action is outside of WP, it’s a solution that will ultimately integrate onto any platform, and sit there, side by side, ensuring a coherent application of a much more secure and resilience, yet highly flexible network data sharing architecture.
- This reply was modified 4 days, 20 hours ago by
Venutius.
- This reply was modified 4 days, 20 hours ago by
Hi George,
Tahnk you for the reply.I don’t think we have a competition here and we need to compare our goals with the project.
I understand your goal is to create a new experience and trust based network(what is trust? different people have different perspective on it and it is not the right forum to discuss it philosophically).We(BuddyDev) are not creating a social network, we are creating a software that allows site admins to create social network. What is that network and how it works is not our concern. We are not here to decide for site admins, our goal is to provide the tools and let the site admins decide what they want to do with it.
As of data ownership, unless your users are storing their data on their own systems, there is no real data ownership. The client server approach(where one or more servers store data) and not all users run their own server is not going to solve data ownership(Mastodon does allow users to run their own server, but users may be part of a server run by someone else too).
There are always going to be legal rules around world(Like EU/Australia’s current rules), and they are dynamic, they will change over time. Our goal is not to incorporate them into our software. Our goal is to enable site admins with tools to make the compliance simpler/easier. For that reason, our identity is not around these compliance, compliance is just one of the many things our system will provide easier tools/integrations.
As of your question:-
What is the architecture of this software
Client/server with partial inspiration from DDD.
Is your plan to use the existing user and security architecture of WordPress? I personally think that lack of granularity of access and trust levels meant BP was never scalable.
Yes, we are and we do not see any issue. WordPress provides granular role/capability based system which can be extended for all use cases. BuddyPress’s problem is not use of WordPress users, It is how they store/fetch data.
But also, WordPress is going to die soon. It’s moving commercial, this is why my ultimate goal is to have my mini servers hosted on peoples own PCs, and not need hosting, not from a hosting provider.
You are entitled to have your opinion. I have full faith in the future of WordPress and we are sticking with it. Also, you have all the rights to decide how you envision your software being used, for us, WordPress suits our requirement pretty well and we have full faith in the future potential of it.
To sum it up, we are creating a modern social network creation tool that will allow users(WordPress admins) to create social communities with various use case. Since our system is modular, It allows future extension without being limited by the current constraints. We are not re-inventing social networking(We believe the existing social networks like FB, X, LinkedIn, Reddit, Instagram etc.) is a good indicator of how a social network should work , we are rebuilding for a better experience and rapid development with WordPress.
Regards
BrajeshMay I join this discussion.
I was a little bit surprised when I read your anouncement about project midnight sun. I mean everyone can see right now that the development of Buddypress has stalled. But why don’t you help the buddypress team to “make it great again”? I mean if not you, who else can help pumbing up buddypress.Do you wanna make buddypress die? And you know that there is already someone on the market called buddyboss.
When I decided to go with buddypress 18 months ago I looked around more than two weeks. Some of the community plattforms are to expensive, some of them lack on documentation and some of them are not easy to extend.
Buddypress doesn’t look fancy and its a hard way to style it and adapt it to your needs. But in the end something nice came out and there also must be a reason why such a system is so long on the market.Nevertheless I am curious about your new projekt MS. But did you considered customers like me who are looking for improvement but already have a lot of plugins in use to extend bp. How can I transfer them to your system? And whats with plugins which are essential to me f.e. the one of Andrea Tatiani. Do they run on your system as well?
And finally, what will happen to your existing bp customers like me who bought your bp plugins. Do you not maintain your plugins anymore because just working for your own project now?
Honestly, I would have preferred if you had forked and further developed buddypress as a premium plugin instead of doing your own thing now which will be probably incompatible to a lot of things.
BuddyPress is far too complicated, and its complexity is, for the most part, undocumented and secret sauce.
It was designed for developers, not for individuals, it’s not been managed well, the few volunteers that have raised their hands have typically been judged and sidelined. Now those core members are going. The ones with that secret sauce, it’s not likely that anyone would want to take on that mountain.
Wordpress is going the same way. The recent changes to the editor and the underlying architecture typically make it much more difficult to develop on the platform. Both WordPress and BuddyPress have huge privacy and security issues. Simple changes needed to clamp down on breaches have never been carried out, and instead it is down to each and every plugin to be updated. Thousands of hours of work, instead of a few lines of code in the core to make sure that everything is escaped, for example.
Another example – how many WP sites are infected the minute they are created, because WordPress comes, out of the box with minimal security and well known holes?
Hi Alex,
Thank you for the reply.You have raised some real fair questions and concerns and I will address each of them.
1. Future of BuddyPress:-
Since 12.0, BuddyPress development has been going downhills. Fragmented eco-system, lack of developers(active), limited capability, old code base, these are the negatives. Unless the team decides to break away from their legacy/ecosystem, It is going to be really difficult to create positive experience with BuddyPress. BuddyPress was the first software for WordPress that allowed creating social network/community and that’s why it is still being used(the active installs have been falling sharply for last 5 years). Their architecture was good(modular)for that time but it has too many bottlenecks for the current era.BuddyBoss is affected by similar issues + Bulky code base. The good thing abut BuddyBoss is their User experience(UI/UX).
We wanted to avoid the past baggage. That provides us with huge advantage of building something from a new perspective and design the architecture as we see fit for today/future. It also helps us provide better user/developer experience.
We have been focused on low entry barrier/resistance and we hope to achieve that.
2. Why not fork BuddyPress:-
Forking will not help much as we will have the same issues that current BuddyPress/BuddyBoss suffers. Also, we will be stuck with the ecos system and create negative experience. Our goal is to create net positive experience for Site admins/Developers/Users and creating a new system from scratch is the best way to achieve it.3. What happen’s to our BuddyPress/BuddyBoss plugins:-
We will not develop any new plugins but we will keep updating and supporting all our plugins for atleast 3 years from now. After that, we will either put all code on github or let nanother developer/company acquire them.4. Will there be a way to migrate:-
Yes. We will be providing migration from BuddyPress, BuddyBoss a couple other plugins for migrating user/data.Many of the current features form our plugins/BuddyPress eco system are going to be in core. You may not need plugins. Search is part of core(still to be implemented). Most of the features from the current system will have better implementation available as art of core/extension.
It is not going to be a premium only plugin. After initial release and user feedback/updates, we plan to put the core on WordPeress.org and develop via github. It is a freemium model, not a paid only solution(Initial release is for paid users only).Regards
BrajeshPS:- Please open a new topic in future to keep the discussion focused on your actual query/concerns.
You must be logged in to reply to this topic.