8 ms·

The critical issue with this so-called stack is the effort to set this up. It seems simple in words but consider how many steps, tools, plugins, devices, human

The critical issue with this so-called stack is the effort to set this up. It seems simple in words but consider how many steps, tools, plugins, devices, human intervention are required to achieve it. Is it reliable? Nope. Is it better? Nope. You still have to remember those keywords you added so you haven't gotten rid of the memory problem. If even one moving part fails to function properly, the whole system will break so you haven't solved the maintainability problem. And clearly, this doesn't solve the time problem either. When I bookmark something on HN, it is usually because I don't have enough time to read it or I want to read it again later on. I don't bookmark comments or save them. Not because they aren't interesting but because informal content like comments are really easy to understand. They are short and to the point. So I take what information I can and move on. If I forget...well, maybe it wasn't that important to remember in the first place. If I forget but want to remember, which is rare, I can just go through the bookmarks, open the relevant link, and go through the comments again or Ctrl + F if I remember something. This is rare. So rare that I can afford to do it manually. The problem with "knowledge assimilation" aka note taking is that after a certain point, you have more "knowledge" than you even want. The extra becomes noise. At first, the idea to clip and save anything interesting seems so fascinating but soon you realize that you have saved so much that it's become a problem. That's why using the right tool for the right job is important. Bookmarks are perfect for things you want to save without them getting in the way of more important things. They are just a list. Nothing overkill. No complex organization. No complicated processes. Just press a button and viola! I do, however, commend the ingenuity of the OP in making such a system. I wonder how long it took to set this all up.

BoysenberryPi 5y ago
I agree with you for the most part but >The problem with "knowledge assimilation" aka note taking is that after a certain point, you have more "knowledge" than you even want. The extra becomes noise. At first, the idea to clip and save anything interesting seems so fascinating but soon you realize that you have saved so much that it's become a problem. This doesn't matter at all. The thing with Obsidian and other bidirectional linking tools is that if you never see a note again, if you never link to it again then it doesn't matter. There is no cluttering the system.
shane_b 5y ago
I do end up with more knowledge than I want when I add links to a long file. It’s important to look a second time to remove the noise. For me, I prune every other week and usually takes 10 mins tops. If nothing else, my brain gets a second chance to add the idea to memory.
Abishek_Muthian 5y ago
You raise some pertinent questions, I'll try to answer them. > It seems simple in words but consider how many steps, tools, plugins, devices, human intervention are required to achieve it. Is it reliable? Nope. Is it better? Nope. I had to create this setup because 'I had the need', Browser bookmarks alone just didn't cut it, I have at least a decade of bookmarks, For old bookmarks I have to check every page with a particular tag before I can get some value out of them. In spite of trying numerous browser bookmark-organizing tools[1], My need wasn't addressed completely. > You still have to remember those keywords you added so you haven't gotten rid of the memory problem. Knowledge management suites specifically address this with 'knowledge-graphs'. All I have to do this put some thought into naming the tags when saving(highlighting in my case), link to relevant notes as and when I write them . Although I wish I could address the memory problem itself[2], I can now also use spaced-repetition for the content I'd like to remember(Even Kindle apps have built-in flash cards feature). > The problem with "knowledge assimilation" aka note taking is that after a certain point, you have more "knowledge" than you even want. The extra becomes noise. I don't think I'll ever have more knowledge than I ever need, Intellectual curiosity is what makes life less painful to me; literally and metaphorically. > I do, however, commend the ingenuity of the OP in making such a system. Thanks, I like to reiterate that 'I don't claim my-setup to be the best setup for the task', It's evident from this thread that many others use what fits them and 'HN -> Kindle -> Obsidian' happens to be mine. > I wonder how long it took to set this all up. As a developer, HN to Kindle: ~14 hours for the first release, Further development/bug fixes took few minutes to an hour every now and then(Check change log). Not counting the go based web framework I re-use for web application (or) using existing ML models for classification of HN stories. Kindle to Obsidian: Couple of minutes to find the plugin, Few more minutes to find what I need from Nunjucks template, A week to work with the developer to fix the feature parity with Nunjucks on the plugin(Effort here completely is from the developer). Obisidan: Since I already use Markdown/ YAML for all my writing, There's was not much of a learning curve here. I carry on incremental development as I have finite time(due to health reasons) to work on something each day. Someone with lesser constraints could do all the above development within a day. As for an user, This complete stack could be setup in 10-15 minutes. Disclosure: The links below are from my problem validation platform, I've linked them to show that I've explored some of the problems stated by the parent. [1] https://needgap.com/problems/57-i-forget-my-web-bookmarks-quite-often.-internet-organizing https://needgap.com/problems/57-i-forget-my-web-bookmarks-qu... [2] https://needgap.com/problems/41-human-memory-lack-of-thereof-psychology-neuroscience https://needgap.com/problems/41-human-memory-lack-of-thereof...
throw10920 5y ago
Your comment indicates that you haven't done any serious knowledge management, so I'll explain some of it for you. > The critical issue with this so-called stack is the effort to set this up. It seems simple in words but consider how many steps, tools, plugins, devices, human intervention are required to achieve it. This makes it pretty clear that you don't actually understand the system very well, because it is not very complex. Let's count the tools: HN to Kindle, Kindle, obsidian-kindle-plugin, Obsidian. That's four pieces of software (five, if you count HN), and you only need to set-up/manage three (two? Obsidian works really well out-of-the-box) of them. This could be done in, what, two hours? > Is it reliable? Nope. Incorrect. Software doesn't just break - people change it. The only piece in this stack that the user can't control the updates of (which is the only way they could reasonably break) is HN - and HN has a very stable "API" (I assume he's scraping HTML, and HN HTML rarely changes). > Is it better? Nope. Also incorrect, and this is the thing that gives away the fact that you have no experience with knowledge management - or even serious reading, for that matter. Almost anyone who reads a lot (or has studied human learning) can tell you that retention is lower when reading from a screen than from paper (and a Kindle is closer to a piece of paper than to a screen, from a cognitive perspective), and anyone who's done knowledge management can tell you that tags and highlights are incredibly valuable tools for learning and KM. > You still have to remember those keywords you added so you haven't gotten rid of the memory problem. Again - a naive statement that wouldn't be uttered by anyone who does KM. You learn frequently-used tags very quickly, and can always refer to the other ones (or just adopt a "create a new tag if it doesn't exist, and then rewrite later" strategy) if you don't remember a name. I've been using a KM system that is tag-like for several years now, and I learned all of my frequent tags within a few hours of continuous use. Memory of the tags just isn't a problem at all - and they make it significantly easier to find content, so they do solve the memory problem. > the whole system will break so you haven't solved the maintainability problem. See previous comment about software stability - it won't break unless you want it to, or unless the HN API changes. > And clearly, this doesn't solve the time problem either. Yes, it does. After a fixed up-front cost of setting up the system, it makes it easier to capture and retain knowledge, and learn/take notes, for the rest of the life of the system. You should look up big-O notation - this system has a setup cost of O(1). > When I bookmark something on HN, it is usually because I don't have enough time to read it or I want to read it again later on. I don't bookmark comments or save them. Not because they aren't interesting but because informal content like comments are really easy to understand. They are short and to the point. You must not read some of the more interesting comments, then, because I have dozens that are multiple long paragraphs long, many of which are difficult to find again without explicit management (e.g. https://news.ycombinator.com/item?id=27017072 https://news.ycombinator.com/item?id=27017072, which I've had to look up on five separate occasions now, and is difficult every time because it's not tagged properly). > So I take what information I can and move on. If I forget...well, maybe it wasn't that important to remember in the first place. Try applying that strategy to getting a degree in college and see how far it gets you. (there's a tiny chance that it worked for you - but I can assure you that the vast majority of college students have to study in order to pass their classes) > If I forget but want to remember, which is rare, I can just go through the bookmarks, open the relevant link, and go through the comments again or Ctrl + F if I remember something. This is rare. So rare that I can afford to do it manually. Again, it seems like you're not reading the more interesting comments. Or, perhaps, you're not reading very much - I have over 6800 bookmarked HN comments and comment threads (derived from reading dozens of comments a day), which also happens to be impossible to "just go through the bookmarks" bookmarks are simply inadequate for that scale of items. > The problem with "knowledge assimilation" aka note taking is that after a certain point, you have more "knowledge" than you even want. The extra becomes noise. At first, the idea to clip and save anything interesting seems so fascinating but soon you realize that you have saved so much that it's become a problem. The "extra" only "becomes noise" if you don't have a good system to store it (which, as we've established, you don't). Folders aren't adequate. Tags are slightly better. A Gmail-label-based system could probably handle hundreds of thousands of items, at least - and I've never found a piece of information in my system that would have taken longer than 20 seconds to find if I had labeled it properly. > That's why using the right tool for the right job is important. Bookmarks are perfect for things you want to save without them getting in the way of more important things. From an organizational point of view, bookmarks are just bookmarks. They're not an organizational tool. However, if you meant "bookmarks plus folders and tags" then you're a little more on-track - such a system seems to scale into the tens of thousands of items, at least based on my own collection of that size - but all existing browser bookmarking systems are still woefully inadequate for someone serious about learning on the Web. From a note-taking and implementation point of view, bookmarks are even worse - browsers are not good note-taking tools, and running a full browser on an e-paper device like a Kindle is a bad idea (if it's possible at all). > They are just a list. Nothing overkill. No complex organization. No complicated processes. Just press a button and viola! The listed process is only "overkill" if you either (a) have something approximating a perfect memory (which means that you don't need to do knowledge management in the first place - lucky you!) or (b) aren't really interested in reading or learning that much from the internet. As someone who is interested in learning gobs of stuff from the internet, I can assure you that not only is this system not "overkill", but it's inadequate - I would need better tools than it (although mad props to the author for taking many steps in the right direction relative to other currently available tools). Neither is it "complex organization" - it's not powerful enough. And it's definitely not "complicated" by any stretch of the word, certainly not compared to the complexity of the platforms its running on, nor is its difficulty of use comparable to that of using those platforms in the first place. > I wonder how long it took to set this all up. Your subtle jab against the author putting in quite a reasonable amount of work to satisfy a real need isn't lost on me. This seems like a perfect Chesterton's Fence moment - you don't understand why people want to do knowledge management, so you don't get the point of the system. There is a need for it, though - a need that's shared by the author, me, half of my friends, and all of my co-workers at my day job. You seem to be an outlier in that you don't understand that need (especially as a HN user), so I suggest you learn why people do it before making comments about it. Moreover, you seem to have the mindset that it's a good trade to accept simple, inefficient tools that don't provide much value to your workflow as long as they're easy to learn. It's a popular attitude, but almost always misguided - I strongly encourage you to re-think it. The definition of a tool is a device that allows you to either accomplish a task that you couldn't in the first place, or makes it easier to accomplish a task that you could already do. Tools are meant to increase efficiency, and computers are very good at automation. Throwing away useful tools that can make you more productive for the rest of your life just because they require more upfront work is the epitome of foolishness.

BoysenberryPi 5y ago

I agree with you for the most part but >The problem with "knowledge assimilation" aka note taking is that after a certain point, you have more "knowledge" than you even want. The extra becomes noise. At first, the idea to clip and save anything interesting seems so fascinating but soon you realize that you have saved so much that it's become a problem. This doesn't matter at all. The thing with Obsidian and other bidirectional linking tools is that if you never see a note again, if you never link to it again then it doesn't matter. There is no cluttering the system.

shane_b 5y ago

I do end up with more knowledge than I want when I add links to a long file. It’s important to look a second time to remove the noise. For me, I prune every other week and usually takes 10 mins tops. If nothing else, my brain gets a second chance to add the idea to memory.

Abishek_Muthian 5y ago

You raise some pertinent questions, I'll try to answer them. > It seems simple in words but consider how many steps, tools, plugins, devices, human intervention are required to achieve it. Is it reliable? Nope. Is it better? Nope. I had to create this setup because 'I had the need', Browser bookmarks alone just didn't cut it, I have at least a decade of bookmarks, For old bookmarks I have to check every page with a particular tag before I can get some value out of them. In spite of trying numerous browser bookmark-organizing tools[1], My need wasn't addressed completely. > You still have to remember those keywords you added so you haven't gotten rid of the memory problem. Knowledge management suites specifically address this with 'knowledge-graphs'. All I have to do this put some thought into naming the tags when saving(highlighting in my case), link to relevant notes as and when I write them . Although I wish I could address the memory problem itself[2], I can now also use spaced-repetition for the content I'd like to remember(Even Kindle apps have built-in flash cards feature). > The problem with "knowledge assimilation" aka note taking is that after a certain point, you have more "knowledge" than you even want. The extra becomes noise. I don't think I'll ever have more knowledge than I ever need, Intellectual curiosity is what makes life less painful to me; literally and metaphorically. > I do, however, commend the ingenuity of the OP in making such a system. Thanks, I like to reiterate that 'I don't claim my-setup to be the best setup for the task', It's evident from this thread that many others use what fits them and 'HN -> Kindle -> Obsidian' happens to be mine. > I wonder how long it took to set this all up. As a developer, HN to Kindle: ~14 hours for the first release, Further development/bug fixes took few minutes to an hour every now and then(Check change log). Not counting the go based web framework I re-use for web application (or) using existing ML models for classification of HN stories. Kindle to Obsidian: Couple of minutes to find the plugin, Few more minutes to find what I need from Nunjucks template, A week to work with the developer to fix the feature parity with Nunjucks on the plugin(Effort here completely is from the developer). Obisidan: Since I already use Markdown/ YAML for all my writing, There's was not much of a learning curve here. I carry on incremental development as I have finite time(due to health reasons) to work on something each day. Someone with lesser constraints could do all the above development within a day. As for an user, This complete stack could be setup in 10-15 minutes. Disclosure: The links below are from my problem validation platform, I've linked them to show that I've explored some of the problems stated by the parent. [1] https://needgap.com/problems/57-i-forget-my-web-bookmarks-quite-often.-internet-organizing https://needgap.com/problems/57-i-forget-my-web-bookmarks-qu... [2] https://needgap.com/problems/41-human-memory-lack-of-thereof-psychology-neuroscience https://needgap.com/problems/41-human-memory-lack-of-thereof...

throw10920 5y ago

Your comment indicates that you haven't done any serious knowledge management, so I'll explain some of it for you. > The critical issue with this so-called stack is the effort to set this up. It seems simple in words but consider how many steps, tools, plugins, devices, human intervention are required to achieve it. This makes it pretty clear that you don't actually understand the system very well, because it is not very complex. Let's count the tools: HN to Kindle, Kindle, obsidian-kindle-plugin, Obsidian. That's four pieces of software (five, if you count HN), and you only need to set-up/manage three (two? Obsidian works really well out-of-the-box) of them. This could be done in, what, two hours? > Is it reliable? Nope. Incorrect. Software doesn't just break - people change it. The only piece in this stack that the user can't control the updates of (which is the only way they could reasonably break) is HN - and HN has a very stable "API" (I assume he's scraping HTML, and HN HTML rarely changes). > Is it better? Nope. Also incorrect, and this is the thing that gives away the fact that you have no experience with knowledge management - or even serious reading, for that matter. Almost anyone who reads a lot (or has studied human learning) can tell you that retention is lower when reading from a screen than from paper (and a Kindle is closer to a piece of paper than to a screen, from a cognitive perspective), and anyone who's done knowledge management can tell you that tags and highlights are incredibly valuable tools for learning and KM. > You still have to remember those keywords you added so you haven't gotten rid of the memory problem. Again - a naive statement that wouldn't be uttered by anyone who does KM. You learn frequently-used tags very quickly, and can always refer to the other ones (or just adopt a "create a new tag if it doesn't exist, and then rewrite later" strategy) if you don't remember a name. I've been using a KM system that is tag-like for several years now, and I learned all of my frequent tags within a few hours of continuous use. Memory of the tags just isn't a problem at all - and they make it significantly easier to find content, so they do solve the memory problem. > the whole system will break so you haven't solved the maintainability problem. See previous comment about software stability - it won't break unless you want it to, or unless the HN API changes. > And clearly, this doesn't solve the time problem either. Yes, it does. After a fixed up-front cost of setting up the system, it makes it easier to capture and retain knowledge, and learn/take notes, for the rest of the life of the system. You should look up big-O notation - this system has a setup cost of O(1). > When I bookmark something on HN, it is usually because I don't have enough time to read it or I want to read it again later on. I don't bookmark comments or save them. Not because they aren't interesting but because informal content like comments are really easy to understand. They are short and to the point. You must not read some of the more interesting comments, then, because I have dozens that are multiple long paragraphs long, many of which are difficult to find again without explicit management (e.g. https://news.ycombinator.com/item?id=27017072 https://news.ycombinator.com/item?id=27017072, which I've had to look up on five separate occasions now, and is difficult every time because it's not tagged properly). > So I take what information I can and move on. If I forget...well, maybe it wasn't that important to remember in the first place. Try applying that strategy to getting a degree in college and see how far it gets you. (there's a tiny chance that it worked for you - but I can assure you that the vast majority of college students have to study in order to pass their classes) > If I forget but want to remember, which is rare, I can just go through the bookmarks, open the relevant link, and go through the comments again or Ctrl + F if I remember something. This is rare. So rare that I can afford to do it manually. Again, it seems like you're not reading the more interesting comments. Or, perhaps, you're not reading very much - I have over 6800 bookmarked HN comments and comment threads (derived from reading dozens of comments a day), which also happens to be impossible to "just go through the bookmarks" bookmarks are simply inadequate for that scale of items. > The problem with "knowledge assimilation" aka note taking is that after a certain point, you have more "knowledge" than you even want. The extra becomes noise. At first, the idea to clip and save anything interesting seems so fascinating but soon you realize that you have saved so much that it's become a problem. The "extra" only "becomes noise" if you don't have a good system to store it (which, as we've established, you don't). Folders aren't adequate. Tags are slightly better. A Gmail-label-based system could probably handle hundreds of thousands of items, at least - and I've never found a piece of information in my system that would have taken longer than 20 seconds to find if I had labeled it properly. > That's why using the right tool for the right job is important. Bookmarks are perfect for things you want to save without them getting in the way of more important things. From an organizational point of view, bookmarks are just bookmarks. They're not an organizational tool. However, if you meant "bookmarks plus folders and tags" then you're a little more on-track - such a system seems to scale into the tens of thousands of items, at least based on my own collection of that size - but all existing browser bookmarking systems are still woefully inadequate for someone serious about learning on the Web. From a note-taking and implementation point of view, bookmarks are even worse - browsers are not good note-taking tools, and running a full browser on an e-paper device like a Kindle is a bad idea (if it's possible at all). > They are just a list. Nothing overkill. No complex organization. No complicated processes. Just press a button and viola! The listed process is only "overkill" if you either (a) have something approximating a perfect memory (which means that you don't need to do knowledge management in the first place - lucky you!) or (b) aren't really interested in reading or learning that much from the internet. As someone who is interested in learning gobs of stuff from the internet, I can assure you that not only is this system not "overkill", but it's inadequate - I would need better tools than it (although mad props to the author for taking many steps in the right direction relative to other currently available tools). Neither is it "complex organization" - it's not powerful enough. And it's definitely not "complicated" by any stretch of the word, certainly not compared to the complexity of the platforms its running on, nor is its difficulty of use comparable to that of using those platforms in the first place. > I wonder how long it took to set this all up. Your subtle jab against the author putting in quite a reasonable amount of work to satisfy a real need isn't lost on me. This seems like a perfect Chesterton's Fence moment - you don't understand why people want to do knowledge management, so you don't get the point of the system. There is a need for it, though - a need that's shared by the author, me, half of my friends, and all of my co-workers at my day job. You seem to be an outlier in that you don't understand that need (especially as a HN user), so I suggest you learn why people do it before making comments about it. Moreover, you seem to have the mindset that it's a good trade to accept simple, inefficient tools that don't provide much value to your workflow as long as they're easy to learn. It's a popular attitude, but almost always misguided - I strongly encourage you to re-think it. The definition of a tool is a device that allows you to either accomplish a task that you couldn't in the first place, or makes it easier to accomplish a task that you could already do. Tools are meant to increase efficiency, and computers are very good at automation. Throwing away useful tools that can make you more productive for the rest of your life just because they require more upfront work is the epitome of foolishness.