Roam Research, new web-based outliner that supports transclusion & wiki features - thoughts?

Dynalist already does this :slight_smile:

Edit: however I agree it probably does solve the problem

A lot of the difference between Roam and Dynalist comes down to what you do with information after it gets entered into your Inbox. Dynalist wants you to file it away using the ā€œmoveā€ command, while Roam lets you leave it in place and file it by simply adding tags or turning the text itself into tags. (Roam also shows you ā€œunlinkedā€ items that are related by search terms even if they arenā€™t tagged as such.) This is why Roam can give you a new inbox every day. I donā€™t think the same thing would work in Dynalist, because you really need to file things away into the appropriate location in order to find them again. I have to say, the Roam approach is really appealing! (Which is why Iā€™ve already written three posts about it).

2 Likes

One thing I just want to add to this is that over the past few years of trying every digital system going Iā€™ve really come to see that there are psychological considerations to what makes a good system over and above how well a system might work for e.g. some perfectly rational, cognitively unlimited android.

As a small example I have found that one advantage of concrete ā€˜placesā€™ where I make a conscious decision to ā€˜sendā€™ things over simple tagging is the imprint in my memory. With the former, am I much more more likely to (1) know what lists I have and (2) remember where I sent a given item - itā€™s a more active process which yes adds a bit of friction, but that friction in itself can be a positive as it creates a stronger memory trace in my mind. The GTD guy has similar concerns about this - in one of his books he talks of the problem of ā€˜invisibleā€™ information in digital systems - you can throw everything in there, but will you ever see it again (even if its tagged, will you remember to search for that tag)? If you do have great systems for retrieval, and good behaviour around this, wonderful, but Iā€™ve personally found these behaviours harder to set up and maintain with ā€˜all in one box / tagā€™ systems vs concrete separate lists than I can ā€˜visualiseā€™ as separate spaces and furthermore are shoved in front of my face as I move around my lists. I can really see a poorly managed Roam system quickly falling into this problem.

Not saying whatā€™s best, just my experience (I have a bad memory generally so might be a stronger effect for me)

7 Likes

You say DynaList wants you to move stuff and Roam wants you to tag. I donā€™t see why DynaList couldnā€™t be used without moving.

My DynaList I have files wherein every day I create a new entry thatā€™s just the date. (What Roam does for you, but itā€™s literally 10 seconds a day, and I have full control.). Under that I put my notes for the day. I could tag if I choose, and use those tags to view a themed history of my work.

Now if we had cloning I could make topical documents and reference these inline in the dailies but thatā€™s generally not necessary. If I wanted to write an essay I could make a new note elsewhere and my daily entry would be scratch pad to help draft. I think I might even prefer this as it keeps context clear whereas a daily entry containing a cloned item would become anachronistic.

3 Likes

The main difference is how tags work in Roam. As I said above they are actually pages, not just search results - though they are that too. But even when they are search results you can still edit them in place, without going back to the original context to edit. (You can edit in place when search local tags in Dynalist, but not when searching across documents.) This is completely different workflow which obviates the need to move, or even clone, an item. (Though I have found cloning to be useful as well, but for different reasons.)

The Sidebar is a killer feature thatā€™s largely overlooked in favor of fancier features like bi-directional links.

While the ability to quickly cross-referencing notes in Roam is impressive, I am surprised by how useful it is to simply be able to work with multiple notes at the same time! You can open 1, 2, ā€¦multiple notes in the sidebar and work with all of them real time cut/pasting/re-arranging ā€“ and most importantly ā€“ discovering common concepts between them!

This ā€œmultiple notes open at onceā€ seems like such a low tech feature I am surprised that it might just be the most powerful feature I use on Roam!

dgg

3 Likes

Thatā€™s a good point David. Dynalist works well in two browsers side-by-side, but the ability to e.g. drag and drop from one window to another would be great.

3 Likes

Iā€™ve been thinking a bit about what Roam lacks that Dynalist does well:

  • API
  • Date range searches
  • Mobile app
  • Desktop app
  • Notes
  • Easy navigation and search between separate documents
  • Send to inbox from browser, mobile extension, API, or email
  • Custom CSS

And there are a bunch of features Dynalist is working on which Iā€™m not sure Roam will support? For instance some kind of end-to-end encryption, mobile URL scheme/deep-linking, true-offline, etc.

Did I miss anything?

They are very different apps so some features (such as separate documents) might not really be ā€œmissingā€ from Roam, but I am personally still not sure I want all my data in one single databaseā€¦

UPDATE 2020-06-30: Since I wrote this Roam has added ā€œDate range searchesā€ and ā€œCustom CSS.ā€

3 Likes

Yea now that we have archiving, separate files has really come into its own for me

Hi Erica, if the Dynalist team were to create another app with Roam type powers, what would it take to get the process started?

I would not mind putting $100.00 towards the project.

What if 50 individuals contributed in this manner - does this help the project get started?

6 Likes

That is an interesting idea. But I am not sure if the Dynalist Team has enough resources to create a whole new app while still maintaining Dynalist.

1 Like

Hereā€™s why I am askingā€¦they might be able to do it!

1 Like

I really think the user base would support @Erica doing this. Iā€™ll put my money where my mouth is right now. :smile:

2 Likes

Okay I am reading that the first time and I am a bit surprised.

1 Like

I never want to add more work to anyoneā€¦but I think there is an amazing opportunity to borrow from the Dynalist history (great features, great community and great support). The world needs a competitor to Roam Research and Iā€™d rather give @erica and @Shida my money than anyone else.

5 Likes

I still prefer growing DynaList where it makes sense. My thought is if we get Clone, then having a command called Clone Toā€¦ which gives a search dialog gets us pretty far towards a great knowledge management app within DynaList.

1 Like

I really donā€™t know what approach would make more sense (adding Roam features or starting over), but it is clear to me that Roam is on to something profound. The learning curve for Roam is much higher than for Dynalist, but as I delve into it I find myself using it more and more. One way to put it is that as you add more info to Roam it becomes more useful. While Dynalist can handle a lot of information elegantly, I feel like I have to spend more time re-organizing it to keep it under controlā€¦ It might be interesting for @Erica and @Shida to do both: create a kind of playground app where they can try new ideas without the burden of legacy code, and at the same time try to implement these experiments back in Dynalist? If that makes any sense, I would be happy to support it!

1 Like

The underlying theme here is if organization of knowledge in a network or in a hierarchy is better. In my opinion this is the wrong approach to think about it. These two structures are not mutually exclusive, they can coexist. Hierarchies are effective for large-scale, slow-moving efforts while networks are good in small-scale, quickly changing situations.

Dynalists approach is to apply a network (with tags and links) on an underlying hierarchy while roam does the opposite. I personally think that the first approach makes more sense and can be applied to more use cases.

The drawback people see in the Dynalist approach is (like you wrote too) that they need to re-organize more in Dynalist. I think this is a sign that the system people are using is not optimal, not the tool. With the right system you need just one shortcut (CTRL+Shift+M) to move everything to the right place. Keywords here are P.A.R.A or the Zettelkasten method.
The lack of a real hierarchy in Roam makes it easier to use not necessarily better in the long run.

10 Likes

Hierarchies are effective for large-scale, slow-moving efforts while networks are good in small-scale, quickly changing situations.

Yes, this gets at it perfectly. The world I work in requires me to constantly change and evolve my systems and I love that Roam has the flexibility to handle this. But I would also add that there is a lot of overlap and cross-fertilization between my various knowledge sets, and the Roam approach is useful for this as well. Thanks for helping clarify this.

3 Likes

Iā€™m of the same mind. Itā€™s true that DynaList forces some kind of hierarchy on our data, but Iā€™m not convinced thatā€™s a bad thing. In the past when Iā€™ve used sprawling wikis for my notes I often experience the nagging feeling of ā€œmissing somethingā€ ā€“ that there is information Iā€™ve forgotten about and canā€™t find again except by luck. Having pages organized in a loose hierarchy helps me feel like everything is ā€œin its placeā€.

ā€œCloningā€ would amp up DynaListā€™s capabilities tremendously. In my opinion, the main weakness of a hierarchy-based model is that often an item doesnā€™t fit in just one category, or have just one parent. Cloning would allow us to have multiple, curated, editable views of our data.

9 Likes