Ask HN: Do you curate links/bookmarks? Would like to know how the tech community deals with bookmarks and links. They might just be read-later entries that you want to visit later on. Or could the links you have been collecting be a valuable resource? How would you share them? |
It's a primary reason I continue to run my big fat mouth on blogs the rest of the world largely seems to have a yawning disinterest in most of the time.
If the website contains an important bit of info, I try to send the page to the Wayback Machine so that it's not lost to time when the website eventually dies, and I also copy that info in my notes (currently using Joplin)
The challenging bit is providing a sound enough structure so every item finds a place almost organically. If it’s a simple list then it can take a while to find a specific entry and if it’s too categorised you’ll run into a similar problem. I like to use trees with symlinks.
If something is interesting enough to keep I clip in into Evernote and tag it, then I can be confident it will be there when I go looking for it at some undefined point in the future.
These days I just save them to Pinboard, like I used to, and sometimes go back and find something I vaguely remember...
Bookmark wise I mostly just use the standard bookmark manager. I have considered a self-hosted syncing mechanism but I only have a few hundred bookmarks, it is stuff I use.
I put every blog, site I want to follow on my freshrss instance(rss.ishanjain.me).
I use Reeder(on iOS) and FeedMe(on Android) to read those posts. When I like something, I add it to Pocket where it gets archived and stuff and I can revisit it later at any point.
I use Browser bookmarks to store interesting websites, Like tools, Stuff other than articles basically.
Worth a try I would say, but I won't make a definitive recommendation for now.
I do a fair bit of reading in Feedbin. Anything I read and want to save, I tag and send to Pinboard. If I want to read it later, I mark it as such and also send it to Pinboard.
Maybe add a 'ril' tag for every thing you want to read later, no tags for generic queue of things to look at, 'archive' for things you want to keep, etc.
Articles I plan to read later are saved in Pocket and archived after I read them.
How it differentiates: single click save - keeps track of the referer if possible - show a QR code - keeps a cached copy
The main reason behind my building Histre is the idea that we throw away a lot of the signal we generate while doing things online and this can be put to good use for ourselves.
Bookmark management is just a special case of knowledge management. What you really need is a knowledge management tool that is easy to use. You'll get a ton of other benefits too.
As it is right now, Histre aids the casual online research we all do (ie the explore -> filter -> decide loop). For example, it removes friction in taking notes on links you're looking at, with free-form tags that you don't have to create first and other such niceties that add up. And it easy to group notes into notebooks and share. In short, when you have to look at a bunch of links for something (decide on AirBnB, people to hire, material for your next blog post, etc) Histre makes your life easier. But this is just the starting point for what Histre intends to do.
IMHO the biggest problem with apps like Evernote, Notion etc is that it becomes digital hoarding, and not a knowledge base. And the knowledge base focused apps out there involve a lot of manual upkeep, which almost never happens, especially at work. Things start out okay and quickly fall into disrepair. I'm differentiating from the other note taking apps by automatically putting together a knowledge base (grouped by topic etc).
One idea I'm excited about and I'm working on right now is: Histre automatically fetches updates from the websites you visit, ranks the websites with things like lack of ads / referral links and ranks the new posts with your 'revealed preferences' of what you tend to actually read from the list of updates previously shown etc. Personally I think this is will be what replaces social news sites, which are too sensitive to people who bother to go upvote on /new and has to cater to the lowest common denominator.
Automatic Upkeep (WIP): Histre detects links/notes related to your existing notebooks and offers to update those notebooks with the new links and notes. This is similar to how Google Photos suggests new photos for your existing albums. This solves the upkeep problem. Currently people create knowledge bases with good intentions and it becomes stale and useless quite fast.
Exports: I'm working on org-mode exports, as I'm an Emacs user myself. Other export formats coming soon.
Integrations: There is Hacker News (import and optionally share your upvotes) and Telegram (take notes using the Telegram app) integrations now. I'm working on Twitter next (lists of Twitter profiles -> follow all / block all).
1. Every article I will read, I'll drop straight into Pocket, to highlight/read. I will never rea the article outright - this is important, because some clickbait can find me in a "vulnerable" state. After this time has passed, I'd delete it out of my Pocket without reading if I determine it's not worth it. I delete about 40% of the stuff I intended to read. In pocket, I would highlight heavily (I have pocket premium)
2. All my highlights, notes from the books I read, my Tweets, articles that I write and newsletters I send end up in my Evernote. This gives me the single source of truth for the knowledge I am gathering. I can instantly search for a figure from an article I've read 3 years ago. I organize it by topic and treat it as a "Commonplace Book".
As such, this is not for organizing "Links", because I dont care about links. I care about quickly finding the piece of information I've seen on the web, and -of course - proper attribution.
This is like reason #17 why I wish we had a hypertext system (like we used to have) that was separate from the application deployment system (that’s been given to us).
I capture a lot of bookmarks into my org-mode and I don't spend much time deliberately curating. Once in few weeks I go through captured bookmarks and quickly assign a priority to each one (just a matter of pressing hotkey). Then I sort everything by priority, typically ending up with 10-15 higher priority ones that I would tag, put into my reading queue, refile etc.
Rest of it isn't curated and serves as my personal search engine [0]. Often instead of searching in google I'd first search in my emacs and find some relevant information in my knowledge base.
In addition, I'm working on promnesia [1], a browser addon that integrates links in my org-mode files with my browser. E.g. - when I visit some blog, it would show me that I've got few blog posts from that blog bookmarked (along with my private notes and annotations), which typically means that the blog is worth exploring more - when I visit someone's twitter profile it might prompt that I've retweeted/favorited some of that person's tweets
[0] https://beepb00p.xyz/pkm-search.html#personal_information
Every few weeks I conduct housekeeping on my email. This moves the good bookmarks over to my personal git wiki or deleted. This ensures that every bookmark I have gets visited at least twice. This also allows me to not overthink before adding a bookmark without ending up with incomprehensible huge lists of bookmarks that I never end up re-visiting.
I've learned over the years to keep this system simple. Understand and realize that whatever bookmark system you use will eventually go away. If its good enough, they will try to monetize, if not they will be forced to close shop. Either way it gets killed, at least for me.
Another side note: not all bookmarks are equal. Some are just references you want handy when you want to point someone to them, usually in an online argument where response time matters. Some links have golden content on the other side. You want to read, re-read, internalize and remember (spaced repetition) this. Some are work related. Some are aspirational stuff you want to get to in the future. So knowing what and why I am bookmarking something was key to keeping it real.
I have made 95%(2000+) of my bookmarks public on GitHub[0]. I have categorized them and host them on a public repository(110+ forks, ~1800 stars) and push every week. I have also made the entire repo available as a GitBook with search and export as PDF functionality.
Search (especially incremental) is very important, there's no way I'm going to read through someone else's 2000+ bookmarks, whereas searching for topics that interest me is feasible.
Wondering why you publish both gitbook and mdbook? I was planning to use mdbook to release my bookmarks to public; gitbook development has stalled as far as I understand.
As a product designer I'd love to sink my teeth into this problem but unfortunately it's a dev-heavy project and that's outside my comfort-zone.
For now I'm sticking with my fallback solution — dozens of 'link dump' notes in the Mac Notes app. It's searchable, lightweight, and flexible but a nightmare to manage. There must be something better out there.
I’ve been using Pinboard for the last 10 years or so and I’m super happy with it. Love the speed, the text-heavy interface and the high information density. I love that it never changes. I use a separate app my phone (the website is unusable on mobile) and it works great. I don’t need anything more.
What I'm doing now instead, is curating a set of links around some topics, and sharing them publicly. The most portable way of storing and presenting them is either an HTML page on one of my websites or a markdown file mirrored to Github. That way I can be sure that my data isn't locked into some proprietary format by an entity who might get Incredible Journey-ed for whatever reasons.
Here are my personal favorites: https://jmstfv.com/bookmarks
And here's the curated list of businesses publicly sharing their expenses: https://github.com/jmstfv/open-expenses
In order to view all these saved links as well as share with anyone interested, I created a website to republish them all on one page:
There's more info about the philosophy and implementation on GitHub, as well as all the source: https://github.com/jacobwgillespie/saved-for-later.
On a page that I think may be useful I click a menu item in firefox which activates org-protocol sending the URL to an Emacs server, starting Emacs if needed. Emacs prompts me for some tags and then makes an entry in an org file which I sync to all my computers, home and work, via syncthing. I'll then either simply close the Emacs frame or if appropriate leave it open while I read through the page and add notes to the org entry. Later, if I have reason to revisit then I'll search on terms my brain is able to associate and that hits on the tags, text or page title that were stored.
Scripts and configuration are propagated to different computers using Git via a personal Gitea instance and vcsh/myrepo tools.
You can save links in the web app, via email, browser extension or pull automatically by integrating with a Medium blog or any RSS feed. Link collections get automatically turned into an email newsletter that can be personal, public or shared within a group.
We recently launched 'Picked for you' which suggests links you might find interesting. Also, we launch new features daily so feedback and requests are more than welcome!
The Alfred workflow makes capture and tagging pretty seemless.
Edit: just read the op properly, not after pinboard. I also would like a visual manager mainly for design refs. I don’t personally use it but Are.na is a cool platform for visual bookmarking of media alongside articles / sites.
I have the custom domain feature configured so all my links are here:
Full disclosure - I’m the founder and lead developer, I haven’t launched yet on HN but if anyone wants to try it out feel free to setup an account.
I tend to organise links in lists. Some examples:
* https://tefter.io/zorbash/lists/ruby-rails * https://tefter.io/zorbash/lists/devops * https://tefter.io/zorbash/lists/elixir-resources * https://tefter.io/zorbash/lists/web-development
I would like to use bookmarks more, assigning tags, using them in my planning, but Firefox bookmarks just suck and I end up spending too much time without any benefits. There were some external tools for that, but they aren't better than a plain text file because of poor browser integration.
Anything that can be annotated and put into context ends up somewhere in my *.md notes.
So I started sending them out as a curated list of a handful of links per week, alongside a short explanation of why it mattered or a quote.
8 years later, I still sort of remember if I sent something, and I no longer use bookmarks.
apologies for site not being done yet and article not fully fleshed out. but the topic has come back many times now that i think i needed to share what i thought about it.
to answer your question: we need to share and distribute our bookmarks in GitHub. it works for developers because our search queries are keyword-based.
pocket, pinboard.in, etc. don't work because they feel like "todo" lists.
I use browser bookmark folders only for links for private stuff e.g. fun+memes, cooking, curated lists of <FOO>, ... Also I maintain markdown files for documenting changes/tweaks and modifications made to various machines on my network.
As a VI user I never looked at org-mode before. I'll probably do so in 2020 (https://github.com/jceb/vim-orgmode).
It is intended to replace the “new tab” page of the browser and works really well!
When I first “launched” it had a minimal set of features and scratched my own itch. I wanted to charge $12/year. However, I discovered that nobody even wanted to pay $1 per month!
I’ve since created designs for a new UX, which is what you see in the screenshots, and put up a landing page. All out of the startup playbook. However, nobody has expressed an interest in the product, so I haven’t opened up access for others.
This makes them easy to find again at any later time due to mediawiki's excellent built-in search facilities.