This was perfect for me, but I missed the timestamps. So here are some: 00:00 Capture vs recall/search 00:31 Obsidian six vaults, friction 01:11 Reference system, commonplace notes 01:25 Slow down to understand better. 01:56 Example: Learning Go 02:14 Dynamic hierarchy of tags 02:52 Literature note, reference system 03:38 Tags for the dynamic hierarchy 04:05 Normal notes from book or video 04:28 Find/search for this knowledge 05:01 Keywords, templates 05:47 Hierarchy perspective 05:56 Dynamically generated 06:12 Filter template 06:29 Slashes represented in hierarchy 06:56 Templates, sub-hierarchy 07:27 GitHub repository, public 07:56 Tag at the end draws attention 08:09 Square brackets or hashtag links same place 08:34 Easier to read 08:50 Cleaner look 09:48 Search for code snippets 10:18 Block level indentation 10:46 Zoom into a note 10:58 Bullets as narrative for future notes 11:19 Summarize 11:41 Another benefit 11:53 Tag wrangler obsidian plugin I prefer Josh keeps helping us with more Logseq notes about his working systems and current thinking/reading (capture.) We can then help with the timestamps (recall) and links to the other videos he references.
Thank you for taking the time to put all those timestamps together! I'll do better in future and make sure to include them. I also really enjoy when someone taken the time to include the. :)
Nice approach! I use something similar, but instead of relying on tags, I'd use slashes in the names of particular pages themselves. Logseq allows for that too. Let's say, I want to talk about Spring Boot (the Java framework). I'd create a page titled "Java/Frameworks/Spring Boot" and as one of its first proeprties, add an alias to Spring Boot itself (alias:: [[Spring Boot]]). One thing I really dislike about page layouts in Logseq, is that the hierarchy listing is at the very bottom. Imagine a page with hundreds and hundreds of backlinks where you have to scroll a ton to get to the hierarchy. It is really inconvenient. I'd much rather have the hierarchy above the backlinks.
good thinking! but careful when you do that, you won’t be able to use the namespaces as filters in your linked references because the alias is all Logseq is seeing now.
Great video Josh! Love the dynamic hierarchies and preservation of context. Though seems at the cost of fragmentation. Imagine you learn 10 languages. You may have 10 regex tags golang/regex, javascript/regex, etc. Some of the regex things you learn may be more generic, so you may have a top level #regex tag as well. How do you think about the 11 regex tags you'll have then? How to decide at which level to tag. One solution would be to tag both, but will lead to clutter. The ultimate solution would be if you could have the benefit of the tag hierarchy, and click the subtag either as part of the hierarchy (which would lead you to #golang/regex) or just as #regex, which would then combine the #golang/regex and #python/regex, etc. Almost the opposite as how the filter works.That would give both the dynamic hierarchy, which helps with discovery and context, and give the flexibility of crossing the hierarchy. Looks like that isn't possible at the moment afaik.
Already I've discovered that flexible is key with this dynamic structure. In your example with regex, my solution to the problem would be to distinguish between the syntax of regex and the implemenation of a regex engine within a given language. Blocks associated with the syntax would begin to form a new "regex" namespace and document the use of regex itself, while the remaining sub-topics that demonstrate how to use regex within a language would be kept within the language namespace, for example. golang/regex. As you mentioned you could also use an additional tag, creating multiple paths to a particular block. That allows a single note to become part of multiple sequences, much like How to Take Smart Notes would approach it. All that said, it's still a sandbox for me atm. lol subject to change.
@@joshuaduffney Great advice Josh! But don't you think that would lead us down a rabbit hole? To be consistent, you'd need to tag the block with both #regex/golang and #golang/regex because you don't know which angle you'll take in recovering this information which is relevant to both topics! In set theory terms, the block lives in the intersection of #golang and #regex. Doesn't matter which came first conceptually. As @Arthur Kosten mentionned, the ideal solution is to tag both #golex #regex as they are both potentially generic concepts but I can see a way in which Logseq implements polyhierarchy of tags so you'd set it up once and not at the block level everytime.
@@matthieub3973 that’s a good point. I chose to use Golang/regex because I’d look for that snippet only when interested in how to use regex in golang code. I might later tag blocks with #regex that contain patterns or concepts specific to regex not tied to a programming language. But that’s how my brain works and I do agree just using two tags golang and regex is the simplest option and removes the need to maintain a hierarchy. :)
Great stuff. I've been thinking about this every day for the past week. I think the more flexible the structure, the more complex, and the more important the use of queries to keep the friction low. A three-level progression of understanding: simple, advanced with simple query extensions, and full datalog style queries. Even before starting with simple queries, there are plenty of special cases to become aware of. While running tests to see what works I keep forgetting to re-index. If I had any hair I'd have pulled it out. There are gems of knowledge spread around the many online resources but I have yet to find a definitive guide.
Queries isn't something I've worked within Logseq. I find the syntax extremely frustrating, and that's saying something I've done a fair amount of database administration. At any rate, I can relate to the need to re-index. And I wish these tools made it easier for us. Thank you for your kind words and thoughtful comment.
That was incredibly helpful and will implement immediately. New subscriber! Can you provide a brief tutorial on how you set up your page properties (content in gray boxes)? I know this is supposed to be easy, but for some reason I can't get it 100 percent how I want it. Thanks!
Glad you enjoyed the video Cami, thanks for subscribing. :) Are you referring to the bibliographical data I collected for the sources? If that's the case I'm using the being_quote block. When you type < it will pop up with all the available admonitions like Tip, Warning, etc.. below is an example. #+BEGIN_QUOTE Author(s): [[Todd McLeod]] Url: www.udemy.com/course/go-programming-language/learn/lecture/5988980 Source code: github.com/GoesToEleven/golang-web-dev tags:: #course #programming #[[software development]] #+END_QUOTE
I didn’t see your explanation of how to create the filter that shows you all the hashtags and their hierarchy. Can you tell us how do you set up the Logseq query to provide that result?
Great contribution Josh! Thank you for making it! I'm in the thick of trying to figure out how to structure my notes within Logseq and often find myself envious of a friend with a great folder hierarchy in Obsidian. Using namespaces along the lines of what you described here seems like it has the potential to help me bridge the gap!
Glad it was helpful! One thing I'll mention is that you're namespaces will naturally need to change as your understanding of an area evolves. So don't be afraid to rename different sub-topics within a namespace to better organize it. But don't front load those decisions, let them evolve. :) You can use editors like VScode to search and replace strings of text which makes renaming the namespaces pretty painless.
@@joshuaduffney Thank you!! That makes sense, yeah! I'd rather let it be bottom-up like that. But I had indeed been starting something top-down with my initial foray into this. So I'll course correct after hearing this.
@@DavidManuel I completely understand, top-down is really difficult to avoid. I've probably made that mistake 3-4 times now with Obsidian and finally breaking free of it with logseq.
What about opening your *.md Logseq files from Obsidian and use the Tag Wrangler plugin if you want to change some particular Tag? Or just changing the [[link tittle]] in Obsidian, which changes it in all the pages where its placed. Isn't it?
On your last point about plugin to rename tags - isnt it sufficient to go to the tag's page and rename its title? AFAIR it updates all the linked references as well.
Like tagging, but realised we have to work to build and maintain them. Folders are static, but with a simple structure the system does the work for you. This point on effort isn't stressed enough.
Tagging takes a tremendous amount of effort, I totally agree. You'll want to make sure you have a good reason for making such an investment. Luckily for me I think it's starting to pay off. :)
Great question, but I'm not sure. Logseq has an active discord and forms site that I'd ask. There are a bunch of Logseq wizards that hang out there and I'm certain they can help. :)
@@DaniloFaria7 yeah, would not feel comfortable with this. I used "-" instead of "/" in Obsidian and was not happy. In Roam it works well because namespaces are not yet supported.
Hey Jason, I am still using dendron my 30 day experiment turned into a 90 day one. lol It's been working really well for me to manage my work life, since most of my is spent in VSCode anyway. Any interest in a video covering my P.A.R.A method setup in Dendron?
@@joshuaduffney Whoa! First of all Thank you for the reply. I have watched all of your videos and after finishing them I wanted more. You should make more videos. I am also learning web development and for that I am trying to use PARA method where all my project related resources stay in the project file or folder. You get the idea. After trying Notion, Obsidian and Logseq. Ive started using Dendron for the same, and I kind of leaning towards it more.I would really love to see your workflow and setup in dendron. Please make one.
This was perfect for me, but I missed the timestamps. So here are some:
00:00 Capture vs recall/search
00:31 Obsidian six vaults, friction
01:11 Reference system, commonplace notes
01:25 Slow down to understand better.
01:56 Example: Learning Go
02:14 Dynamic hierarchy of tags
02:52 Literature note, reference system
03:38 Tags for the dynamic hierarchy
04:05 Normal notes from book or video
04:28 Find/search for this knowledge
05:01 Keywords, templates
05:47 Hierarchy perspective
05:56 Dynamically generated
06:12 Filter template
06:29 Slashes represented in hierarchy
06:56 Templates, sub-hierarchy
07:27 GitHub repository, public
07:56 Tag at the end draws attention
08:09 Square brackets or hashtag links same place
08:34 Easier to read
08:50 Cleaner look
09:48 Search for code snippets
10:18 Block level indentation
10:46 Zoom into a note
10:58 Bullets as narrative for future notes
11:19 Summarize
11:41 Another benefit
11:53 Tag wrangler obsidian plugin
I prefer Josh keeps helping us with more Logseq notes about his working systems and current thinking/reading (capture.) We can then help with the timestamps (recall) and links to the other videos he references.
Thank you for taking the time to put all those timestamps together! I'll do better in future and make sure to include them. I also really enjoy when someone taken the time to include the. :)
Nice approach! I use something similar, but instead of relying on tags, I'd use slashes in the names of particular pages themselves. Logseq allows for that too. Let's say, I want to talk about Spring Boot (the Java framework). I'd create a page titled "Java/Frameworks/Spring Boot" and as one of its first proeprties, add an alias to Spring Boot itself (alias:: [[Spring Boot]]).
One thing I really dislike about page layouts in Logseq, is that the hierarchy listing is at the very bottom. Imagine a page with hundreds and hundreds of backlinks where you have to scroll a ton to get to the hierarchy. It is really inconvenient. I'd much rather have the hierarchy above the backlinks.
good thinking! but careful when you do that, you won’t be able to use the namespaces as filters in your linked references because the alias is all Logseq is seeing now.
Bro, Do you use logseq in ipad ? or any other tablet ? . How is the user experience ?
Great video Josh! Love the dynamic hierarchies and preservation of context. Though seems at the cost of fragmentation. Imagine you learn 10 languages. You may have 10 regex tags golang/regex, javascript/regex, etc. Some of the regex things you learn may be more generic, so you may have a top level #regex tag as well. How do you think about the 11 regex tags you'll have then? How to decide at which level to tag. One solution would be to tag both, but will lead to clutter. The ultimate solution would be if you could have the benefit of the tag hierarchy, and click the subtag either as part of the hierarchy (which would lead you to #golang/regex) or just as #regex, which would then combine the #golang/regex and #python/regex, etc. Almost the opposite as how the filter works.That would give both the dynamic hierarchy, which helps with discovery and context, and give the flexibility of crossing the hierarchy. Looks like that isn't possible at the moment afaik.
Already I've discovered that flexible is key with this dynamic structure. In your example with regex, my solution to the problem would be to distinguish between the syntax of regex and the implemenation of a regex engine within a given language. Blocks associated with the syntax would begin to form a new "regex" namespace and document the use of regex itself, while the remaining sub-topics that demonstrate how to use regex within a language would be kept within the language namespace, for example. golang/regex.
As you mentioned you could also use an additional tag, creating multiple paths to a particular block. That allows a single note to become part of multiple sequences, much like How to Take Smart Notes would approach it.
All that said, it's still a sandbox for me atm. lol subject to change.
.
@@joshuaduffney Great advice Josh! But don't you think that would lead us down a rabbit hole? To be consistent, you'd need to tag the block with both #regex/golang and #golang/regex because you don't know which angle you'll take in recovering this information which is relevant to both topics! In set theory terms, the block lives in the intersection of #golang and #regex. Doesn't matter which came first conceptually. As @Arthur Kosten mentionned, the ideal solution is to tag both #golex #regex as they are both potentially generic concepts but I can see a way in which Logseq implements polyhierarchy of tags so you'd set it up once and not at the block level everytime.
@@matthieub3973 that’s a good point. I chose to use Golang/regex because I’d look for that snippet only when interested in how to use regex in golang code. I might later tag blocks with #regex that contain patterns or concepts specific to regex not tied to a programming language. But that’s how my brain works and I do agree just using two tags golang and regex is the simplest option and removes the need to maintain a hierarchy. :)
Great stuff. I've been thinking about this every day for the past week. I think the more flexible the structure, the more complex, and the more important the use of queries to keep the friction low. A three-level progression of understanding: simple, advanced with simple query extensions, and full datalog style queries. Even before starting with simple queries, there are plenty of special cases to become aware of. While running tests to see what works I keep forgetting to re-index. If I had any hair I'd have pulled it out. There are gems of knowledge spread around the many online resources but I have yet to find a definitive guide.
Queries isn't something I've worked within Logseq. I find the syntax extremely frustrating, and that's saying something I've done a fair amount of database administration. At any rate, I can relate to the need to re-index. And I wish these tools made it easier for us. Thank you for your kind words and thoughtful comment.
That was incredibly helpful and will implement immediately. New subscriber! Can you provide a brief tutorial on how you set up your page properties (content in gray boxes)? I know this is supposed to be easy, but for some reason I can't get it 100 percent how I want it. Thanks!
Glad you enjoyed the video Cami, thanks for subscribing. :)
Are you referring to the bibliographical data I collected for the sources? If that's the case I'm using the being_quote block. When you type < it will pop up with all the available admonitions like Tip, Warning, etc.. below is an example.
#+BEGIN_QUOTE
Author(s): [[Todd McLeod]]
Url: www.udemy.com/course/go-programming-language/learn/lecture/5988980
Source code: github.com/GoesToEleven/golang-web-dev
tags:: #course #programming #[[software development]]
#+END_QUOTE
@@joshuaduffney thank you! I will give this a try.
I didn’t see your explanation of how to create the filter that shows you all the hashtags and their hierarchy. Can you tell us how do you set up the Logseq query to provide that result?
Great video Josh… very helpful. Particularly like the way you summarise at the end, which helps with reinforcing the message. 😊
Great contribution Josh! Thank you for making it! I'm in the thick of trying to figure out how to structure my notes within Logseq and often find myself envious of a friend with a great folder hierarchy in Obsidian. Using namespaces along the lines of what you described here seems like it has the potential to help me bridge the gap!
Glad it was helpful! One thing I'll mention is that you're namespaces will naturally need to change as your understanding of an area evolves. So don't be afraid to rename different sub-topics within a namespace to better organize it. But don't front load those decisions, let them evolve. :) You can use editors like VScode to search and replace strings of text which makes renaming the namespaces pretty painless.
@@joshuaduffney Thank you!! That makes sense, yeah! I'd rather let it be bottom-up like that. But I had indeed been starting something top-down with my initial foray into this. So I'll course correct after hearing this.
@@DavidManuel I completely understand, top-down is really difficult to avoid. I've probably made that mistake 3-4 times now with Obsidian and finally breaking free of it with logseq.
What about opening your *.md Logseq files from Obsidian and use the Tag Wrangler plugin if you want to change some particular Tag? Or just changing the [[link tittle]] in Obsidian, which changes it in all the pages where its placed. Isn't it?
chaging the title in Logseq also changes it everywhere where it is used
On your last point about plugin to rename tags - isnt it sufficient to go to the tag's page and rename its title? AFAIR it updates all the linked references as well.
You could use alias:: --thus tag1 and aliased tag2 lead to the same page/block
You might also find the query {{namespace YOURNAMESPACE}} useful eg {{namespace golang}} or {{namespace golang/template}}
Oh, very nice! That makes quite a nice index page. :) You just leveled up the dynamicness of the this method, thank you!
Like tagging, but realised we have to work to build and maintain them. Folders are static, but with a simple structure the system does the work for you. This point on effort isn't stressed enough.
Tagging takes a tremendous amount of effort, I totally agree. You'll want to make sure you have a good reason for making such an investment. Luckily for me I think it's starting to pay off. :)
Thx Josh a new angle appreciated
Great video Josh 🙏🏽
Appreciate it! Just bought your course and looking forward to leveling up my logseq skills. :)
Very helpful!
Very helpful!
Is there a way to disable namespaces? I’m using a lot of legal texts that usually contain „/„
Great question, but I'm not sure. Logseq has an active discord and forms site that I'd ask. There are a bunch of Logseq wizards that hang out there and I'm certain they can help. :)
You could use the opposite symbom \ which doesnt create a namespace
@@DaniloFaria7 thanks for answer, but when Legislative Acts use "/" I can't just other symbols. I need to use correct names.
@@ToriZealot when sharing the notes, you could replace the characters with some automated bot. But i guess that might be tricky.
@@DaniloFaria7 yeah, would not feel comfortable with this. I used "-" instead of "/" in Obsidian and was not happy. In Roam it works well because namespaces are not yet supported.
Very interesting!
I thought so too. :) Thanks for watching.
Do you use still use dendron?
Hey Jason, I am still using dendron my 30 day experiment turned into a 90 day one. lol It's been working really well for me to manage my work life, since most of my is spent in VSCode anyway. Any interest in a video covering my P.A.R.A method setup in Dendron?
@@joshuaduffney Whoa! First of all Thank you for the reply. I have watched all of your videos and after finishing them I wanted more. You should make more videos. I am also learning web development and for that I am trying to use PARA method where all my project related resources stay in the project file or folder. You get the idea. After trying Notion, Obsidian and Logseq. Ive started using Dendron for the same, and I kind of leaning towards it more.I would really love to see your workflow and setup in dendron. Please make one.
@@joshuaduffney I'm so excited to see your PARA setup, how long do I have to wait?
@@jasonbourne757 I just got my new recording gig setup today, so hopefully I'll have it posted by EOW. :)
@@joshuaduffney That's great news! 👍