Foam Graph: What's next? #1199
Replies: 8 comments
-
|
What would it take to add linking via tags? Does it suffer from the same hurdle/issue as grouping by tags? |
Beta Was this translation helpful? Give feedback.
-
|
If you want to have a data dump, shouldn't you define for what use? To have a interchangeable format with a research and analyzing tool like any of those would be a great feature for many researchers that what to use markdown as their research log formats (me included)... I watched the youtube chat you had in March and there you mentioned labels on edges... and something about talk to others to maybe create a "standard" for networked graphs... have any of you looked at https://github.com/HEmile/juggl . |
Beta Was this translation helpful? Give feedback.
-
|
With regards to filtering, if you can allow user's fuzzy search into a text box with display a graph with responsive filtering based on what's being input, that would be amazing, kind of like |
Beta Was this translation helpful? Give feedback.
-
|
Would also be nice to filter/distinguish edges of the graph (e.g. set "colors" of edges). Sometimes I want to have the link to another node, but don't have to complicate the structure (which may be a good thing, but still). No idea if that is feasible though. |
Beta Was this translation helpful? Give feedback.
-
|
I just add the comment about JUGGL here to... JUGGL for Obsidian has most of what you talk about, and he is open to help to implement it for other markdowns editors... I am not a developer, so I can't really help much more than give you som tips and comments... |
Beta Was this translation helpful? Give feedback.
-
|
Would be great to have an option to output the data powering the graph as JSON locally, can then easily be consumed by something like Gatsby 👍 |
Beta Was this translation helpful? Give feedback.
-
|
Github & VSCode have been working together recently on new VS Code API's and libraries to help facilitate more dynamic and custom UI experiences in VS Code that could help with the new Foam graph experience's I've started a discussion in the link below with more details & would llove to hear the Foam community's thoughts 💡 🙏 |
Beta Was this translation helpful? Give feedback.
-
|
Alternative approach for "Data Dump". There is a project BrainDB which converts folder of markdown files to "graph" database (graph in quotes, because it is rather Relational database, but you can get graph anyway). Then you can write custom script to dump data as any format you like. For example, as Graphology. |
Beta Was this translation helpful? Give feedback.
-
Foam Graph
This is a mixture between current state of affairs and a call to action.
That is, every issue and possible improvement point you see here can be thought as a place where Foam can be improved.
As improvements will be required to be PR'd into the repository, discussion is also open
Current State
Customization
This year we added some customization settings to the graph in: #438
We hope to add many more during 2021.
Coloring by Date
Recently, I saw the following issue: #485
In summary, the issue suggests nodes to be colored according to "Last Modification Date".
While I do not think it is a feature that should get into the core,
I can understand why it would be an awesome feature and I have some ideas.
If you use
gityou can highlight nodes according to their status, or track the knowledge-base progress through time.This of course may not even be possible, but would be cool nonetheless.
Coloring by Attributes
I consider attributes to be embedded metadata on the file.
These can be tags, or other "keys" defined in the YAML frontmatter (see #388).
This is inspired by #2.
Filtering
Filtering should not be too hard to implement, however we are still thinking of ways to filter nodes.
Existing Requests
Text Filtering
Filtering would not be complete without filtering by text (even though I forgot about it).
If the search feature is available in the VSCode API, we can take advantage of that too.
Filtering by date could also be a thing.
Being honest, if the filtering is flexible enough we can filter on a ton of things, but do not get your hopes up just yet.
Regardless, we are open to suggestions!
Grouping
I will preface this part by saying that grouping is (almost) entirely dependent on the graph library.
Why almost you ask? Because we can always add some fake nodes and hack our way into have a semi supported version of grouping.
Grouping by folder would actually work that way by nature, but tags would require a second graph to be built.
The ideas around grouping are the same as for filtering:
Data Dump
This is useful for many reasons, mostly for support on the web and other visualization tools. #308
While it is not supported yet, adding it in is just a matter of deciding on the graph structure and commit to it.
There are several formats we can pick from, for example Turtle (https://en.wikipedia.org/wiki/Turtle_(syntax)).
Performance
Performance is not perfect yet. In this field we are currently dependent on the graph library we use https://github.com/vasturiano/force-graph.
I encourage anyone with JavaScript experience to take a look into the issues below.
Relevant Issues
Other Thoughts
Most of the ideas in this document are dependent on some previous work over the graph.
The API needs to be stabilized, a work that can only be done by using it for more features and ironing out the kinks.
Beta Was this translation helpful? Give feedback.
All reactions