How to do block referencing in Obsidian

[ryanjamurphy]

@FriSti I couldn’t say for sure, either way, but my instinct is that the kinds of generated IDs you’re describing wouldn’t be simpler than the text they’re standing in for. The only way to get a simpler ID for a given line in a given file would be to generate it, and then you’re either appending that generated ID to the plain text somehow or maintaining a separate index data file in your Obsidian vault somehow. The latter is likely the direction a successful block referencing plugin will take, once externally-developed plugins are possible.

@Hitchhiker90 Sorry, good point. I hadn’t thought of two-way editing in my “solution.” I wouldn’t expect the filename to change when you changed the [[reference link]]. It’s possible, though, that an embed would work? These are of the form ![[reference link]] (note the preceding exclamation point). I haven’t tested this myself.

[Hitchhiker90]

@ryanjamurphy Thanks for the clarification. It’s impossible to change the file name this way, even with embedding.

Embedding does work but the embedded bit really stands out, so it could be useful in some cases, I guess, but it does have a different feel to it.

Anyway, maybe it’s best to wait and hope for a plugin.

[ryanjamurphy]

Ah, there are some CSS themes that clean up the look of embedded blocks.

But yes, there are definitely use cases that would benefit from a specialized plugin.

[Dor]

@FriSti

I like the idea of obsidian working just with plain text and that adding a feature, which ads ID’s to every text line just doesn’t seem to work with this sort of file.

Actually every text line has a number (even if hidden), so surely,if it were parsed, this could be used. Line instead of block should be okay.

[Sellaro]

@ryanjamurphy

This is actually the same thing Roam does, under the hood. The only difference is that block linking here requires an extra step when creating the block reference (initiating the link) and when going back to find the original note (this requires two steps instead of one).

My thoughts on that in another forum topic: Block reference

[amirography]

huh! Interesting approach! I think for those of use using a strict folder structure, there would be a hassle of finding a place to put these newly made files, but it sounds interesting. Though For longer strings of text, as you mentioned, there is the limitation imposed by the OS (for windows it’s 255), and the use of most special characters is out of the question. But I think it will do. At least the price of using a Database is probably too high of a price for having something more convenient.

[macedotavares]

The problem with referencing a line number is that as soon as you add a line anywhere above it, the whole thing falls apart.

I’ve been entertaining the idea of plain text block referencing for a while now, and I believe that some ID has to be added to the referenced line. Maybe just a few characters.

[Dor]

I think in theory it could be done by tracking, in the same way that Obsidian tracks and adjusts for name changes.

[ryanjamurphy]

@macedotavares @Dor Indeed, I think the best plugins for this will do both instantiate a block reference with a few characters and then cache and track that instantiation.

If I were to make such a plugin, I would use something like ((double parentheses)) to declare a statement reference. The statement itself can then be the identifier. (This assumes such statements would be unique, which I think is safe.)

[typo180]

Adding a couple thoughts. I think we could get most of the way there with Aliases if there were a way to embed a Page or Header without actually displaying the page title or header (e.g. embed contents only) and to display the contents in-line instead of inside a box.

Example:

• Make a file called BlockTest.md
• Type this out: [[blockRef123abc|Content I want to embed.]]
• This would create a file called ‘blockRef123abc.md’ linked to a line of text “Content I want to embed.”
• Now I can manually add the text “Content I want to embed.” to the file ‘blockRef123abc.md’.
• Finally, I can embed this in another file with ![[blockRef123abc]]

The remaining problems are that the whole thing is in a box and that there’s a big “blockRef123abc” title above it.That should just be a display thing. Remove the title and display the content in-line with the rest of the note. We might even be able to do this with custom css? I’m not sure.

Now let’s modify this a bit:

• Keep your BlockTest.md file, but also make BlockTest_BlockRef.md
• Open BlockTest_BlockRef.md and type the following:

# blockRef123abc
Content I want to embed.

• Now, in another file, type: [[BlockTest_BlockRef.md#blockRef123abc|Content I want to embed.]]

This works almost the same way as the first example with the same problems, except that, instead of creating a new markdown file for every single block you want to reference, you’re creating one additional file for each file that contains blocks you want to reference. The number of block reference files are, at most, going to equal the number of non-reference files in your vault instead of the total number of “blocks” that you want to reference.

Having gone through all that, I think a specialized syntax for creating a block is probably simpler (e.g. [[%abc123|Content I want to embed.]]. Though you’d probably need a way to generate a unique ID - at least unique per page, since you would probably do something like ![[BlockTest%abc123]] to embed it

The other problem I see is that any of these solutions would make the text pretty messy and difficult to read in edit mode.

[RikD]

@typo180

We might even be able to do this with custom css?

We are able to do such thing by CSS. Many examples in the current CSS library.

[RikD]

@typo180

problem

I don’t like ‘problem(s)’

[argentum]

I’ll be quoting a few posts from Block reference to summarize best practices found in that thread (which is a bit long).

@deftdeg

It appears the debate on block referencing has quite a bit of focus on the mechanics of how it could be achieved, with perhaps more than some comparisons to Roam and some probing as to why you would, or wouldn’t want to do it.

Apologies for Stating the Obvious
If we consider the essence of the use-case where we want to achieve a graph of linked thoughts, each thought in Roam is, at it’s most atomic, a single bullet. This is conveniently lazy (not meant to be pejorative) in that we could ‘brain dump’ in our daily note page with all kinds of links and the subsequent nodes of the thought graph would all reflect the thoughts we had - because bullet node links to bullet node.

In Obsidian, each thought (node) is either an .md file or a header within that file. Thus the same brain dump in our daily notes file would yield connections that all backlink to the daily file - which has a certain utility, like when you want to recall when you had a particular brain-wave, but isn’t a real robust knowledge graph.

In Roam I’ve noticed most all of the thought ‘pages’ I’ve seen in examples - confession: I’m not a Roam user - are empty save for all the backlinks, which i where the magic happens.

In Obsidian this cannot be - well it could , but it wouldn’t be terribly useful. Pages link to pages or headers in pages.

@Silver and @Licat are showing us we need to be more deliberate in expressing our thoughts - which is hard, at least for me, which is why I keep reading everyone’s posts for inspiration .

I’m Making This up as I Go
Since pages are file names, we are constrained by OS rules as to legal characters and length. This is where header referencing and multi-page panes are our friends.
If we want to brain-dump in our daily files, we could briefly pause to consider the topic or context of our thought, quickly link to that [[topic or context]], immediately open it in a new pane and dump our thought in there, which may contain further links - in which case, rinse-and-repeat.

We can then if we want to further contextualize our thought, place it under a header. We can add the header to our original link [[topic or context#further context]]. At that point we can be as wordy as we want since the header isn’t constrained by OS file-naming rules.

We preserve the initial link to when we had the thought, but we are also creating a knowledge graph, from the thought itself, by populating the thought page with the substance of the thought and links to further thoughts.

A Feature to be Requested
I’ve noticed though that if we preempt the header before it’s created - [[topic or context#further context not-yet-created]] and click to the page to create it, the header is not created.

Further, as @Lachlan points out, while file name changes are reflected in all links, header changes are not. And while what I have outlined is deliberate/intentional thought mapping - I’m going to have to try it myself today - I would think having to manually change header links, after we have carefully revised our thought-map, would just get in the way and cause a break in our focus - counterproductive.

[argentum]

Using H6 headers as blocks (albeit that can be a bit resource-intensive from what I recall).

@benperkins

To piggyback on @ja_rule’s suggestion, one way of referencing blocks that currently exists is to enclose text in an h6 (######) ‘block’:

###### If you dislike folding
If you dislike folding for whatever reason, it can be turned off in Settings.
######

###### A really great quote
> Obsidian rocks!
> \- Anonymous
######

The second ###### acts as a terminator to delineate the block.

Then use obsidian.css to hide h6 (whoever used an h6 with a straight face though, really?):

h6 {
    display: none;
}

Using the first few words of the sentence/block in the h6 heading works nicely when autocomplete kicks in during link creation (as opposed to using something like an ID).

I’ve never had the need to ‘title’ a block reference, so it seems fine for my use that the h6 is used as a reference, but never rendered.

@goodsignal

@Théry in the Obsidian CSS discord offered a little more refinement. I’m just copying it here because this is where people will find it. (I haven’t tested it yet)

For basic hiding of H6 in Preview:

.markdown-preview-view h6
{
  display: none;
}

Refining margins with the embedded text:

h6 + p { margin-top: 0; }
h6 + ul { margin-top: 0; }

[argentum]

Block-level references for mathematical proofs:

@jwelch

When I am taking notes from a mathematical text, there are many definitions that one has to keep track of when attempting to write proofs. For example the following definitions are in my page corresponding to notes from a math text:

Some way down through the notes, there is another entry which refers to a theorem and linked note page containing the proof.

On the page which contains the actual proof, I use block references which pull in relevant definitions or theorems that are required for actually proving the theorem.

Using header-level embeds:

@WhiteNoise

# Definition
your definition

and in the other file you embed as

![[firstfile#definition]]

[argentum]

Using CSS to make embeds inline:

@mrjackphil

For those who like inline embed…

.internal-embed {
	display: inline-flex;
}

@mrjackphil

Multi-line support

.internal-embed, 
.internal-embed > .markdown-embed,
.internal-embed > .markdown-embed > .markdown-embed-content,
.markdown-embed-content > * {
	display: inline;
}

.markdown-embed-link {
	position: static;
	display: inline;
}

[argentum]

Autohotkey script to create new notes based on selections:

@_ph

I’ve made a crude autohotkey script to create .md files out of text selections, then it puts the transclusion link in the clipboard ready for pasting… maybe someone can use it as a poor man’s ersatz of block reference.

details :
. choose a folder to store the snippets (edit the script to configure the folder’s location first). the script will create folder by month and store all snippers here
. select some text (from an obsidian doc, browser, etc…)
. CTRL+SHIFT+C (copy) or CTRL+SHIFT+X (cut) will create a new .md file in /snippets/yyyyMM/ :

vault/snippets/yyyyMM/xx_yyyyMMhhmmss-msec_%sourcewindowtitle%.md

. the link to the .md file is automatically copied in the clipboard and ready to be pasted in another obsidian doc as a transclusion (the script generates ![[linktosnippet]] and puts it in the clipboard).

my current css makes inline transclusion visible with a blue border on the left + red icon on the upper left side :

my use case is to split long files into smaller snippets that can be referenced elsewhere (poor-man’s text blocks in a way). So I select a piece of text, CTRL+SHIFT+X to cut and immediately CTRL+V at the same place : the original block of text has been replaced by a transclusion. The transcluded file is located in /snippets (at the very bottom of the list), I can then rename/move it elsewhere for storage (links will be updated as long as the file opration is made in obsidian).
Alternatively, I use CTRL+SHIFT+C to make “children blocks” that can be transcluded/edited without modifying the original reference.

DISCLAIMER :
The script is very limited, it’s text only, no images, no formatting. I’m not a coder, so it’s use at your own risk. backup/clone your docs before testing… It only work with AutoHotkey ANSI (autoHotkeyA32.exe) because it’s easier to deal with accents.

odsbn-autosnipp.zip (1.2 KB)

a small gif to see what it does

[argentum]

How to add “naked” embeds

"Naked" Embeds (CSS tweak)

As per a request on the Obsidian Discord (shoutout to @scmwiz, @spacejunky, @kard32 and others) this allows note embeds to look as if they were part of the note in preview mode. I’m not a huge fan of this approach, myself. I prefer to see the embedded note look a little different. But that’s just my personal taste. Screenshot: [20200520113915] Instructions: Copy the snippet into your vault’s obsidian.css file. It should be compatible with whatever other theme you have in there at the time. …

[Dor]

My current usage:
I have a text expander set up with a zettelkasten type date/time and an h6 followed by the date time.
I add text as desired either before the date/time or after - depending on how I envisage searching and preferred label.
I tend to use the pure date/time for new notes and the h6 one for blocks in a note.

This is quite sufficient for me. Unique names and links guaranteed. Most of the time I have no need for blocks, so I may not be typical of those who are interested in them.