Show in Reference Highlights Referenced Entity more clearly

[B_Sp]

My team extensively uses References/Highlights, it’s a real differentiating feature of Fibery.

However, we find the biggest drawback is the inability for the References to capture context when you are writing more elaborate text, such as bullets, multiple paragraphs, etc. But these types of writings are commonplace and really can bring out the benefit of References.

We are struggling with this right now. The main problem is that the only way you can ensure that in a Reference you get the entire text “around” an entity that was mentioned is to make sure you do not execute a “hard” return in that text. You could also potentially describe this as saying “make sure the referenced Entity is in the same ‘block’ of text that you want to show up in the Reference.”

***Footnote there are discussions around here about whether “blocks” are indeed a part of Fibery as they are in Notion, Roam, etc.

So this means right now, we have to write long paragraphs that have artificial spaces between them executed with a “soft” return of “shift” + “enter.” We also can only write bullet points the same way. This leads to awkward long segments of texts, but at least we can see the whole text within the Reference.

Another issue here is that you wind up having to reference several Entities at once in this scenario, because you have to make sure to not do a “hard” return anywhere, and if the text in question is long, and has many references, but you want to have it all showing in a Reference of one entity, you wind up with hard-to-follow References due to a proliferation of Entities in them, and the accompanying colors, states, assignees, etc.

This would be solved with this request:

In Highlights, "Roam-like" inclusion of Hierarchy of Text Blocks

Hey, thanks once more for that! You raise a great point - is Fibery based around Blocks in the Rich Text? I actually thought it was, in particular because Highlights draw from what appears to be a “block” as I understand them - and I’m basing this on the two tools out there that seem to have the deepest concept of a “block” - Notion and Roam. For example, in Fibery you can have a reference pick up text that is within this “block” but in another row below if you create your new row by hitting …

Another thing that would help would be limiting what you see in References via configuration of that, discussed here:

[DONE] Customize Fields in Inline Entities in Rich Text Boxes

Hi again, I would like to revive a subject from the April 20 Changelog re: customization of the great inline mentions we have in Fibery (one of the best features, please guys don’t change this it’s good as is!) My request is only about what you can see about the Inline Entity, following on @mdubakov assertion here from that post: My team makes great use of the Inline mentions, and at times we are seeing that we could definitely use some customizations. For example, we reference Users a lot,…

However, short of these improvements being made, I have another suggestion: It would be very useful if you could see highlighted, perhaps in yellow like you get when you click on the “arrow” in the top right corner of References, the actual Entity itself in the reference.

Here is an illustration:

“Winnie Null” is the Entity with the Reference. If “Winnie Null” was highlighted, say in yellow, where I indicated, the reader could at once see where the actual reference is. This is an example with a small amount of text, but we have many that are paragraphs long with many, many Entities references.

This would solve a big issue for us now, which is if a Reference contains many other entities, it can be a burden on the reader to find the actual Entity that is triggering that particular reference. This is mainly because the actual Text Description of Entities in References blends in with the rest of the text, a feature I like in fact as it aids readability. But when you have many entities in an Reference, it makes it hard to find the actual referenced Entity.

Thanks and hope that’s clear!

By the way @rothnic @Matt_Blais @Oshyan and others would particularly like to hear if you’re big users of References and having the same pain we are? Cheers!

[Oshyan]

Fortunately for me my use of references is much more modest. I’m having trouble picturing A: why these extremely lengthy reference quotes are necessary (or at least why there isn’t another way to handle them) and B: what would be a better way to handle them design-wise given limited space in the entity view. I believe I’ve proposed before a little Discourse-like click-to-expand-quote arrow (conveniently exemplified in your post above ). Would that solve your need?

The thing is I can understand why reference text quoting is limited, at least by default. It makes sense to me from a design/UI perspective. I for one certainly wouldn’t want the References section to be just massively long, at least by default. But I also understand this has been a big frustration for you (although I note that not many other people I can recall seem to have this particular need, am I forgetting?), and want to address the pain point in a reasonable way if possible. Do you see the concern Fibery team or other users might have with much longer reference quotes? Do you have a specific suggestion for how to balance this feature request vs. the need to use space effectively, balance with other elements on the Entity view, and also for the system to determine exactly what you want quoted? Maybe I’m overly-broadening your request in my mind, but as I said I have trouble picturing it.

[B_Sp]

Thanks!

So this would be a great addition, but I don’t know if it’s on the Fibery team’s radar?

@Oshyan

I believe I’ve proposed before a little Discourse-like click-to-expand-quote arrow (conveniently exemplified in your post above ). Would that solve your need?

I have often been posting in the community and longed for more Discourse-like functionality in Fibery References. The ability to highlight text and quote it, then have it immediately pick up the reference to the originating post is so smooth! To do the same in Fibery, I have to copy text, paste it, then wrap in quotes. And if there are references, they will then duplicate here…I need to do this often when responding in comments to another user’s post…which is what we do all the time here in Discourse, right?

Anyway, References are a show-stopping feature of Fibery even with the limitations we still deal with, one of which I described in this post. We use them liberally and they really help my team track the myriad of stuff we are bringing into Fibery, and then all the activity around that - meetings, decisions, tasks, projects, you name it. We write Wiki-style in Fibery - if there is something somebody wants to mention, and it exists as an entity, we reference it. There is real power here, and the concept I’m talking about is for sure the reason for the success of the Networked Thought tools out there, starting with Roam. They are all about liberally referencing with a simple “[[” or other easy keyboard shortcut.

I will add to you question here:

@Oshyan

Do you see the concern Fibery team or other users might have with much longer reference quotes? Do you have a specific suggestion for how to balance this feature request vs. the need to use space effectively, balance with other elements on the Entity view, and also for the system to determine exactly what you want quoted?

that yes, there are many ways this could be handled:

• Toggle References, for example at the level of the referring entity, so you don’t have to see the entire referred text (this is what you are talking about, right?);

• Group references by Type that references. We often reference other Types in our Meeting Type. It would be useful to see the references grouped by Type

• Show a snippet of the reference with a “more…” to expand. This is a bit more context than the first point above.

You are speaking of a UI approach here, but my point is that the UX experience really needs work right now. The references wouldn’t need to be this long if blocks were handled like in Roam, which I’m getting at with that request above. Right now if I want to write something like this:

"In today’s meeting, when discussing [Project X], we decided that:

• we are changing the due date
• we are cancelling this [Task Y]
• we are bringing in [Resource Z]
• We are adding this [Phase A]"

I have to use “shift” + “return” to make sure each of the entities in the bullets pick up the statement about [Project X]. That means the reference will have all of those Entities mentioned. This is where it would be useful to have Roam-like handling of the indent, so each of the sub-bullets would relate to the initial statement in its reference without the need to insure, via soft return, that you don’t add a new “block” and thus break the reference.

Not sure if that’s clear, but I hope you follow. I would much rather Fibery start to handle References as blocks and sub-blocks like Roam, this request would be moot if that was the case! And our references would be much less lengthy! But short of that, it would be helpful now to be able to pick out the referenced entity within the whole text at a glance, if it were highlighted.

Thanks again!

[Oshyan]

@B_Sp

So this would be a great addition, but I don’t know if it’s on the Fibery team’s radar?

Yeah, I’m not sure if it is, but I did call it to at least Polina’s attention here:

In Highlights, "Roam-like" inclusion of Hierarchy of Text Blocks

Thanks for the encouraging response! I want to just say one other thing, since my feelings about highlights are a bit different than @B_Sp. I personally do sometimes run into “not enough context”, but also run into the opposite: more context than I need, the backlink quote takes up more space than necessary. I think regardless of what you do you’ll be “guessing” how much context to include. So when you get back to looking closely at this and how to improve it, just keep in mind what I mentioned…

@B_Sp

• Show a snippet of the reference with a “more…” to expand. This is a bit more context than the first point above.

This is, I think, what I mean when I talk about Discourse-like expandable references. Not sure how this differs in your thinking from your first point above:

@B_Sp

• Toggle References, for example at the level of the referring entity, so you don’t have to see the entire referred text (this is what you are talking about, right?);

[rothnic]

@B_Sp

"In today’s meeting, when discussing [Project X], we decided that:

• we are changing the due date
• we are cancelling this [Task Y]
• we are bringing in [Resource Z]
• We are adding this [Phase A]"

I have to use “shift” + “return” to make sure each of the entities in the bullets pick up the statement about [Project X]. That means the reference will have all of those Entities mentioned. This is where it would be useful to have Roam-like handling of the indent, so each of the sub-bullets would relate to the initial statement in its reference without the need to insure, via soft return, that you don’t add a new “block” and thus break the reference.

I did not realize at first what you were doing here about shift+return because of the bullets, but now I think I’m following. Yes, to me this would be a great addition to the references. We are more in the getting started phase, so haven’t had too many others using Fibery at the moment, but I have used this type of feature in Roam and really like it. It is so much easier to quickly loosely associate two things using hierarchy.

[B_Sp]

Thanks Nick yes you have grasped exactly what I meant!

[antoniokov]

Let me mysteriously say that we have something in the early design phase that addresses this very valid problem…

We’ll get less mysterious once we build a prototype

[B_Sp]

This is truly music to my ears!

Also @antoniokov, I had been meaning to post another example of this problem and now that you’ve jumped in here, let me do so:

You guys posted a nice video about how to use backlinks/references for meeting notes. What’s interesting is in the video you can see a great example of the problem. If you watch in the link below, I’ve started the clip where @mdubakov is making a note about a project:

Meeting Notes

The screen looks like this:

He then shows us the example of the backlink. However, if you look at the backlink as I’ve highlighted it right here…

you’ll notice that the key part of the context, the bullets points that are actually the update, are missing!

Likewise when Michael creates an entity out of one of the bullets here:

You have lost the context that the particular action item represented in that entity is from the Project that is linked at the top of the outline/bullet list.

@Oshyan and @rothnic would love to hear if you guys are following my example here, and if this is also inline with the pain points you have with Highlights/Backlinks as they currently work!

Regardless, back at @antoniokov, it sounds like you guys get this issue, so that’s great to hear!

I really think if you guys brought in a solution similar to Roam, you would make a major leap with Fibery as a product in the Networked Thought space - I should say something you already have a good foothold in. My team’s #1 reason for staying in Fibery is that we can document our entire operation in a much superior way than in Confluence, Roam, Notion, not to mention Evernote, Obsidian, etc. And I’m not even talking about the Work Management advantages of Fibery!

[Oshyan]

@B_Sp

@Oshyan and @rothnic would love to hear if you guys are following my example here, and if this is also inline with the pain points you have with Highlights/Backlinks as they currently work!

Yeah, I think you’ve pointed out a couple of great examples here! Often I’m unsure exactly how this would best be solved in a programmatic way. Even if it’s easy as a human to say “well it should include those things in the context, obviously”, it is not necessarily as easy to tell a computer program how to do it in a majority of cases. That’s why I continue to favor some combination of “good defaults, but expandable context view”. I don’t want it to default to “too much” context just to address issues like this, and end up with a really long and cluttered reference view, but I also agree there are plenty of times when more context is needed. As I’ve mentioned before, the quote expansion functions here in Discourse are a nice example of how this can work.

I’m not sure I quite agree that a Roam-like approach makes sense for this particular use case, or at least in the way that backlinks are currently setup. The advantage Roam has there is that it’s all outline format and otherwise “free form”. It can show all child blocks because it’s all just in the same “document context”. The references in Fibery are more specialized and I think a different approach is needed to jive with that approach. Not to say the fundamental approach could not be changed and a more Roam-like one adopted… But I like my Discourse quote-alike approach concept.

[B_Sp]

Ok cool thanks for weighing in!

@Oshyan

As I’ve mentioned before, the quote expansion functions here in Discourse are a nice example of how this can work.

I would love to see this in Fibery, makes it so easy to write in context in Discourse.

I agree there is a lot of potential in toggled content in Fibery. I would love to in fact be able to group references by Type, then have a toggle to open and close, instead of the current system where if there are say tens of references to an Entity (we are getting to that point as our Fibery instance gets bigger), instead of just having the first few from any type showing up, then a number you have to click to see the rest, you could group by Type, see all the Types that are referencing, then drill in with a toggle to see the individual entities in each Type group. This is again useful for meetings - you could see how many meetings for example referenced an Entity.

Thanks!

[Oshyan]

Mm, this is an interesting idea of how to approach it. Grouping by type could well be a good addition. What I had in mind was each reference being as we already have it, with the shorter existing quote, but expandable if that didn’t show enough context (each reference expands individually). Combining these two could be interesting, trying to think how it would work in the UI, but conceptually it’s a good idea.

[mdubakov]

Thank you for the interesting insights. Let me shed SOME light on the Fibery 2.0 that we are thinking and digging into right now. Essentially, we want to introduce blocks, but not Notion-style blocks, more high-level. For example, a quite long part of a conversation could be a block (with several paragraphs and bullet-points). Then you can include/transclude this block into any entity and it will be fully visible by default. Moreover, we are going to introduce a concept of IMPORTANT fields and these fields will always present on a referenced entity. For example, here is a Meeting entity and you see important Task details here.

In a similar way, on a Task you will see the same block, but with more info about a Meeting.

In short, Fibery 2.0 will include Blocks, new Navigation and deeper/more advanced relations (we are thinking about polymorphic relations as well BTW). However, this is in design phase right now and we’d like to start implementation in 2-3 weeks.

[B_Sp]

Michael, I know there is a lot of discussion about blocks around here now, I am excited about them as everybody else.

I was wondering if one thing you guys are going to try to solve is the ability for bullets under a reference “pick up” the “top level” reference when they are indented? Basically the subject of this thread and the examples I pointed out from your video? My team still has to use a bunch of “soft” returns to make sure we don’t lose the reference to an entity when including the entity in text, because if we do a “hard” return the reference is lost. You could loosely compare what I’m talking about to how Roam Research handles references that are within a hierarchy.

Hope that explanation makes sense, and glad for any insight you can add, as this is something that is very important to my team’s daily activity in Fibery.

And @Oshyan would be curious to get your input, in particular did I explain the issue sufficiently here?

[Oshyan]

I think it’s pretty well understood at this point, yeah. It’ll be interesting to see how the forthcoming Blocks functionality interacts with References. I know you’ll be able to sort of link to or transclude a block elsewhere, from what we’ve heard/seen so far, but how that will relate to References I am not sure.