Proposing development of a Fibery <-> Obsidian sync plugin (independent, open source)

First an apology for going silent in this thread for a while. I’ve had a lot going on in my personal life, including some family health issues (everyone is OK, just things to work through!). But I remain very interested in this idea and hopeful that an actual sync/integration plugin could be developed in time.

For me this would be one of the principle reasons to integrate with Obsidian. I think it’s clear at this point that Fibery will probably never have the vibrancy of development community that Obsidian does, and that’s OK. It would be even more “OK” though if we could have at least some access to all the capability in those plugins, while also taking advantage of what Fibery has to offer (which I agree is generally better than Obsidian in many respects).

  • Offline and local access to my core data (text, markdown, hopefully structured data too, i.e. databases, although this seems much harder to sync properly)
  • Use of Obsidian’s advanced text editing environment, customizable hotkeys, etc.
  • Alternate and powerful views and UI affordances for working with text, e.g. Tabs, flexible Panes, Themes, and perhaps one of my biggest motivators: Canvas view
  • Since this one is important enough, I’ll repeat it as its own point: Canvas View. Because of its importance to me, and what I think it represents, I’ll elaborate further on it separately below.
  • Ability to leverage local tools to operate on my content/data (making an Obsidian sync inherently creates a local archive of markdown files that can be manipulated by any local or even cloud tool, provided it connects to Obsidian, and which then syncs legibly back into Obsidian)
  • Ability to utilize any number of mobile markdown/Obsidian apps for a far superior mobile experience, especially for the kind of quick note-taking, list management (e.g. shopping list), etc. that I mostly want for mobile (and which is fairly horrible with Fibery at the moment)
  • Access to the development community and all Obsidian plugins (since almost all Obsidian plugins fundamentally operate on markdown content, their manipulations/functions are mostly codified in a way that could be synced into Fibery legibly)

So about Canvas View: to me this is an (unfortunately) far superior “Whiteboard” function that sprang from nowhere in Obsidian and already does the most important knowledge organization things better than Fibery (for the most part), including embedding full contents of notes/docs, better and smoother visual connections, frames, embedding active websites, and more. Honestly this one feature was one of the most motivating things for me to want a sync and I started using Obsidian exclusively for tracking my personal therapy work due to the ability to create visual maps of my “Parts” (IFS therapy model), but it is equally if not more powerful and applicable to many business cases.

I’ve tried to use the Fibery whiteboard for similar things in the past and it’s just too buggy, inconsistent, and inadequate. And unfortunately it does not appear to be getting much development focus for the moment, understandably given the many competing priorities, but this just gets right back to the main point: Fibery team can’t and won’t “do it all”, but what if they don’t have to? What if we can reliably sync and in general operate compatibly on the same content with another tool that massively expands Fibery’s accessible functionality? Then the Fibery devs don’t have to do everything.

Imagine if the Obsidian ↔ Fibery sync plugin had existed already a year ago. Then all of the cool AI tools that immediately sprung up (some within days or a mere week of ChatGPT API availability) could have been usable on Fibery content from the moment that plugin became available. Likewise as soon as Canvas became available I could start my IFS mapping, but those docs/pages that I created in Obsidian would also be available in Fibery right away, and I could link to and search on them for other purposes for which Fibery is better, while still maintaining the advantages of Obsidian’s Canvas.

I also hope that such a direct and well-developed sync could actually make Fibery the de facto/default “other” tool that Obsidian users turn to and become comfortable with when Obsidian alone doesn’t do all they need. Right now this is Notion mostly, and a few others. What if it were Fibery? The massive and popular Obsidian community could actually help popularize Fibery for individual PKM use cases, which could in turn make Fibery more popular overall (just as personal Notion use seems to have translated heavily into organizational use).

In short, this whole idea is - to me - fairly significant in its potential impact, and I think to a degree it should be considered with a little distance from the fact that it would be syncing with one specific tool (Obsidian). If you think of it more like a local backup and interop system which allows you to reliably operate on Fibery content with any number of other Markdown-compatible tools, both offline and online, then it gets closer to the hoped-for reality and potential impact. It just happens to be enabled by working through Obsidian, which makes sense because it’s a mature, robust, and popular tool, with a good API, and strong development community. So while a true stand-alone local markdown sync tool for Fibery could be made, I think it makes more sense to leverage the Obsidian advantages (given it is free anyway), speeding up development and popularity/reach with minimal downsides. Even if the only reason to make it sync with Obsidian was to be able to access that community, it’s reason enough: a Fibery-only local sync would only be of real interest to existing Fibery customers, whereas an Obsidian plugin that syncs directly and reliably with a powerful, collaborative cloud dev/proj management and database tool would potentially be of great interest to many Obsidians, i.e. expanding Fibery’s market. We just need a developer or two to take on the creation of the plugin…

Fantastic, thank you! Because of your interest and that of a few others (including myself), I am strongly considering setting up an OpenCollective for this project. That would allow us to safely contribute money to it in a way that is quantifiable and can be clearly and reliably demonstrated to any potentially interested developer(s) (as opposed to the more nebulous - though greatly appreciated - pledges of support here). Is that something you would consider supporting, even with just an initial contribution of say $100 or something?

You’re probably right, but why wait? :wink: In a distant future I’m hopeful that Fibery will encompass everything I need/want Obsidian for right now, and implement it all in a superior way. But realistically that’s probably not going to happen, and even if it does, in the meantime I have to make do with a lot of limitations vs. my needs. An Obsidian sync plugin could, I think, solve a lot of challenges in the short-medium term. Maybe we can even get Fibery team itself to contribute a little toward the effort? :grin:

1 Like

Sorry for the double-reply, but I forgot this one…

That’s quite cool, and goes a long way toward what I am hoping for from this plugin! Although relying on a 3rd party tool/platform with non-trivial costs/less accessible self-host requirements (cost/complexity vs. a simple Obsidian plugin). Nonetheless it’s very cool and something of a proof of concept.

Is there a major (technical) reason it’s not two-way? Or you just didn’t need/want to develop that?

i setup the n8n with the (indirect) help of @Chr1sG in a few hours. i wanted to proof that i can get my obsidian files into fibery with all data extracted into fields and the markdown as a rich text. that worked and is for now where i need to be.

it can be for sure expanded to a two way sync. although then there needs to be some kind of check so that it does not create endless loops.

true, having to rely on yet another external tool is something i also dont really like about this setup. although n8n can be hosted on your own machine for free and the setup in a dockers is of managable compexity. also something that took me less than an hour and i really have no idea what i am doing. probably a security risk though. but compared to pipedream, zapier and make (i tried them all), n8n is by far more superior (a bit harder to setup, but with if/else splits better than pipedream in that regard. and with being free way cheaper than make and zapier)

1 Like

I see what you’re saying and I get it. I have about 5,000 notes on a topic and I have no need to put them in fibery. No desire to do anything with them other than search, and a homegrown LLM+vectors is superior to whatever Fibery can do in that regard.

I see your point.

Regarding this: * task management. its so easy to create inline tasks and have these accumulate at other places.

I’ve crafted a pretty efficient way to do this in Fibery (at least for me).

Oshyan,

I see what you’re saying. And one point in particular make some want to cry :sob: :sob::

  • Alternate and powerful views and UI affordances for working with text, e.g. Tabs, flexible Panes, Themes, and perhaps one of my biggest motivators: Canvas view
  • Since this one is important enough, I’ll repeat it as its own point: Canvas View. Because of its importance to me, and what I think it represents, I’ll elaborate further on it separately below.

I could not agree more. In fact, I was so excited to see it, and only superficially tested it in Fibery, and ultimately went in this direction. Then, I have run into significant useability problems. I wrote a post about it but somehow deleted it.

I read a blog page form the CEO where he shows a Fibery whiteboard and it’s look really good, and I though why doesn’t mine work that way??? Perhaps because I’m on PC it is mostly useable, so I’m just using Miro and considering having my team do more Pipedream (like N8N, which we also use) work to integrated Fibery and Clickup JUST for a WORKABLE whiteboard.

What happens for me??? Well, when I zoom in or out and goes too far, either by 25% or 50% and the appearance (again on the PC) when it gets smaller, is unreadable because they do something weird (that looks like they’re rendering is a blurry image).

So - the beauty of canvases/whiteboards is the ability to zoom in and work on a particular node, but moreso (at least for me), the ability to zoom out and recognized and appreciate the relationships and the flow of things. I cannot do that in Fibery and I’m bummed about it. So I’m with you 1,000 % on that.

And I think the one thing fibery does better (potentially) than anyone is provide a visual learning/building experience - seeing the relations the way they’re shown in Fibery, incorporating a whiteboard that does entities (at least in theory) is very intuitive for me and would be AMAZING if it worked, though it’s still good for me even though it seems like I have to do a lot of hacks.

The local tools, I totally agree with. For instance, I have a very powerful dual 4090 setup specifically for LLM training and all you can do with that. Now that you mention it, the wheels are spinning, and I’m seeing just how powerful that could be if one could use both.

Yes I’ll contribute $100. I’ve already got some team members building things for me (not the Obsidian/Fibery integration), and I’m going to share those as well. For instance, I made this post - Automatic icon for every entity in a table - Get Help - Fibery Community - assuming that was pretty easy to do, apparently it’s not.

So I have a team member putting that together in Pipedream. We use Pipedream, Albato, Zapier, and N8N. If you guys have a preference for N8N, let me know and I’ll them emulate stuff there.

You know that N8N has a white labeling component. Maybe we could make a Fibery-N8N suite of plugins!? Is that what you have in mind? You’re obviously way ahead of me and much more technically-capable, so I’m still catching up.

One last thing: A while ago I was pretty heavily into Obsidian and making that work for me. (In the end I think it nearly drove me insane, lol.) During that time we developed a private plugin that 2-way synced with Google Docs and Google Sheets. So I already have some experience there.

That said, I think we should put our heads together to either (1) persuade the Fibery team to fix their whiteboard and/or (2) figure out a solution to improve that, which might be the Fibery/Obsidian thing you’re talking about.

1 Like

I used that. In fact, I think I was his earliest tester!

Do you prefer N8N over Pipedream? I’ve noticed that N8N has the ability to white label an app and incorporate all its integrations. I mentioned (to the wrong person) that I had done Obsidian to Google Docs 2-way sync and Obsidian to Google Sheets 2-way sync previously. I have rebelled against Obsidian so don’t use that (it was just a private plugin) but we also use a lot iSaaS and that N8N white labeling thing looks interesting.

i tried pipedream and it is great. but if you want to realise more complex things (if else splits for example) you can not do that in pipedream.

also the fact that i can host it on my machine and its free is a killer argument. especially in usecases where i dont need uptime.

i tried make, zapier, pipedream and they all had their ups and downs in integrations. n8n is the only one where i have a functioning google contacts update with all fields - something i tried to get with the pipedream AI but was not successful.

also the super easy integration of http requests with cURL makes it a blast to work with fibery. also their community is really active and you get answers on their forum nearly as fast as here :slight_smile:

so n8n is the winner for me

1 Like

I wasn’t aware of that white labeling option, that’s interesting. I’m not opposed to going that route, as long as it’s long-term maintainable. Complexity would be my concern there. My hope would be for as turnkey of an end result application/plugin as possible. I was originally thinking it would just be an Obsidian plugin.

Why not both? :grin: I think Fibery team definitely needs to improve the whiteboard, while I see the Fibery ↔ Obsidian sync as separately worthwhile. I want to make clear something clear though: The Canvas Obsidian feature is what I consider to be a very good implementation of a PKM-focused “Whiteboard” (as opposed to a more design or sketch or “brainstorming” focused one like Miro or even the Excalidraw plugin for Obsidian). But it is at least as important if not more so as an example of the benefits of syncing with a tool that has a robust feature set and plugin ecosystem.

In other words yes, Canvas is an advantage right now for Obsidian, but I don’t want anyone to think that the solution - instead of syncing with Obsidian - is just to make Fibery whiteboard better. Yes, that would solve that one problem, but Obsidian meanwhile solves dozens if not hundreds more that Fibery does not now solve, and perhaps never will. Some things are outside the scope of what they even want to focus on or have time for, but there are people like me (and perhaps you) who still want and would benefit from those features that Fibery team will likely never implement. An Obsidian sync has advantages for Fibery’s core market, but also for many, many more uses besides.

So I continue to think this is a good idea. I don’t quite have time to lead this project though, to reach out to potential devs, etc. I was hoping that some discussion here might connect with a dev in the Fibery community who might want to tackle at least a proof of concept (and to some degree that already exists with e.g. N8N stuff, etc.). The other prong of this approach here in the discussion is to perhaps generate enough pledged funding to attract a dev who is needing some payment to dedicate the time to it. And ultimately I’d hope for some ongoing funding for support/update for it anyway. Still just trying to build support/momentum here for the idea. We’re getting there…