Ask HN: How do you manage your bookmarks?
Every time I saw something cool on the internet, I add it to my bookmarks.
Result: I have more than 5000 bookmarks on firefox.
I tried to add keywords and descriptions, but it takes me too much time to do that every time.
I also tried to use notion.so (an amazing tool !) to manage my bookmarks, but the plugin "Notion Web Clipper" does not allow to add tags when you save a link.
And you, how do you deal with your bookmarks? Any good tips to keep interesting resources close to you?
271 comments
[ 4.4 ms ] story [ 266 ms ] threadI have 10/15 bookmarks in the top toolbar that are releant to my work. All the rest goes into limbo, and I don't even bookmark much anymore.
When needed, I write notes (wiki, or paper) with links contextualized by topic/need.
The problem with bookmarks is those 3 together (that are required for bookmarks to be of any usefulness - to me anyway):
1) it's easy to create one, there should be more friction;
2) it's hard to have a long-term storage/indexing system (Delicious & equivalents were a good direction);
3) it requires a custom setup for bookmark search to be integrated.
Also, in the mid-term, most URLs are broken.
A paper notebook, on the other hand:
1) requires a conscious effort to note something done (and to have around also);
2) is relatively a secure & long term storage; it's discoverable;
3) the effort marks a memory inprint that helps indexing in long term memory, just for the "I remember I noted that down somewhere when I was in...".
My offline bookmarks/notes/snippets/activities feel way more productive and fulfilling than the computer-based ones in the long-term.
And then it happened again with Reddit. Same thing: I save an interesting comment or post, and then never see it again. I sometimes go back to get a specific comment, but it takes forever to find it. So my solution was to make it some type of game. I find time every month to just attack those saved posts and comments, and check them out 1 by 1. Some of them are either videos that I couldn't watch at the time, or articles I thought I should read later. I watched the videos, read the articles, and then unsaved the posts. If their is a post I think there is a big chance I'll go back to again, I'll keep it saved.
After took me a couple of sessions to go through all my backlog, but after every session, I learn so much, probably laugh a lot as well, and my bookmarks are kept to an absolute minimum.
So yeah, no extension will help you, I'm sure more than half of your bookmarks are a one time read, many are also not interesting to you anymore, and those you really want to keep are less than a hundred.
It's great (again).
I have around 3.5k bookmarks, sloppily tagged, and it's such a great resource to have.
Now that he's again paying attention to his service, I haven't had issues again.
Please reach out to me (contact info in my profile) as I have a side project I would like to resurrect that tackles these same issues.
The benefit of org-mode (or any plaintext bookmarking) is that it makes it instantly searchable in my emacs. I accept that I probably will never have time to properly read through everything I clipped. But it works as a personal search engine: when I search for some topic/tool/etc., I run into some related stuff that I already clipped. It has higher information quality than googling it because my past self already found it interesting and curated these pieces of information. I describe my setup for searching in emacs here: [0].
Another benefit of org-mode is that I can add tags, notes and basically treat bookmarks as any other piece of knowledge I keep in my org-mode files.
Basically, I only use browser bookmarks for services (i.e. social networks/dashboards/etc) now. If it's some sort of knowledge or anything interesting, it gets clipped into org-mode.
For clipping pages into org-mode I tried using org-protocol [1] for bookmarking straight into org-mode; it was unreliable for me, so I wrote my own extension [2]
And finally, I'm working on a browser extension [3], that would unify 'bookmarks' and browsing history from anywhere, whether they are youtube lists/github stars/twitter likes/IM messages or even plaintext files on your filesystem
[0] https://beepb00p.xyz/pkm-search.html#personal_information
[1] https://github.com/sprig/org-capture-extension
[2] https://beepb00p.xyz/grasp.html
[3] https://github.com/karlicoss/promnesia#demo
https://wiki.nikitavoloboev.xyz/
Searched and parsed with Alfred workflow.
https://github.com/nikitavoloboev/alfred-my-mind
The bookmarks are links under ## Links heading of any markdown file. Here are few with some links
https://wiki.nikitavoloboev.xyz/programming-languages/go
https://wiki.nikitavoloboev.xyz/programming-languages/go/go-...
https://wiki.nikitavoloboev.xyz/machine-learning
There is lots more. Around 16,000 lines of markdown in the wiki now. Writing the code to meaningfully parse it now.
https://mobile.twitter.com/nikitavoloboev/status/12165452379...
Here are all the topics included in the wiki. Enough to learn for a life time.
https://github.com/nikitavoloboev/knowledge/blob/master/SUMM...
Oh and the wiki of course does not include private links. Those live in my main browser (Safari) and are searched with workflow too.
https://github.com/deanishe/alfred-safari-assistant
And I optimized the top sites to be most popular pages I visit to get my news on mobile. Accessed by opening new tab.
https://github.com/nikitavoloboev/my-ios/blob/master/README....
On mac I go as far as binding certain sites to open with a single key using Karabiner. ie pressing b + n opens https://hckrnews.com
https://github.com/nikitavoloboev/dotfiles/blob/master/karab...
Although now I do a more low-key approach, just storing them in a flat list and for current topics that interest me I create folders. My "focus interests" change quite rapidly so I don't care about archiving anymore.
I currently have 4319 links, collected over 13 years and 8 days (started with Delicious, then Pinboard).
Link rot and relevance is an issue, and I rarely go back to links that are more than a year old.
I don't delete broken links though, I like to think my Shaarli is a fair representation of my interests over the years.
I recently developed a Shaarli extension for Ulauncher. [0]
[0] https://ext.ulauncher.io/-/github-sebw-ulauncher-shaarli
With the risk of sounding like someone on StackOverflow, do you even need to do this?
I have <100 bookmarks at any time and regular bookmark directories with browser sync work great. I don't have a good solution for efficiently managing thousands. But I also can't imagine that you'll ever actually read all of them, or need to re-read most of them. It's okay to see something on the internet and not save it.
Some long time ago i started adopting the habit of keeping my tab-count low to get more focus on things i need in this exact moment. I only work session-based and like to start things of clean every day. This then resulted in (obviously) lots and lots more bookmarks. But soon i realized that this is headache inducing too. Since then I try to think about if i really need this bookmark and also clean my bookmarks out every few months.
The few bookmarks i have get organized in simple folders like news, programming, books, <insert-current-project>, music. That's it. Done. Try to be minimalistic my friends! It's awesome!
Yeah, just save every page you ever open to disk. Just in case. Right? Also sync all of that data on multiple machines. But what if the syncing service disappears some day? Better ask for a self-hosted solution. But a free one. Which saves everything. Just in case anything ever gets lost. How much better life would be then.
Your browser also caches aggressively.
If targeted to specific high-value sites, or setting retention based on site / content value (some automatic, some less so, some short-lived, some logner), you'll end up with a useful and usable local archive with what is today very small amounts of storage -- even a few GB of text out of a TB or more, isn't much, and that would be a pretty extensive collection.
If the content can be reduced such that it's just necessary text (excluding web crud and more), the end result is likely much smaller still. I've experimented with reducing Washington Post articles and homepage to a simplified view, by selecting specific HTML elements, and the result weighs in at about 3-10% of the source page.
A typical online article likely runs about 800 words. If you read (or save) 20 articles a day for a year, thats about 300 MB.
You would eventually fill a 1 TB hard drive with text at that point. In about 3,400 years.
Every bookmark you create is a time commitment at a later date. The more you have, the less time you have available at any later point in time.
[1] https://docs.microsoft.com/en-us/previous-versions/windows/i...
Linkdrop doesn't have any unread article lists. When you "save" an article, it queues it up to send it to you in an email the next day.
Since I've been using it I find that my actual bookmarks are things I want to save instead of just random articles. Makes things a lot cleaner :)
It's got a good number of users atm and things are running smoothly, it's also free. If you do use it, feel free to send me any feedback you have. I've been trying to find more time to hack on it and would love some more direction.
Hope it serves you well
My main use case was to have a single place for all my bookmarks across Android, iOS, OS X and Linux, and across Firefox and Chrome.
Among all the solutions I evaluated, Raindrop.io had the best UX at the time.
https://github.com/calpaterson/quarchive
It's quite early but I'm able to use it as a near replacement for pinboard (including search but not including tags). I have circa 6k bookmarks so I feel your pain.
It's based on sync of your browser bookmarks via an extension. This works well because it allows you to bookmark on your phone.
Here are some features I have planned:
1. Show what pages link to a bookmark
2. Show discussions (HN/Reddit/Lobsters) about a bookmark
3. Full text search (including of PDFs), perhaps including the above discussions
4. The ability to additionally just record everything you browse (including full text)
Right now it works for me but looks a bit basic - I plan to soup up the graphic design once I have the MVP working:
https://i.imgur.com/5VdgGU7.png
I'm hoping to make it available as a hosted service at some point
If you're interested in this topic, have feature ideas or are even interested in working on it with me either comment here or please send me email - cal@calpaterson.com . I know social bookmarking is not trendy any more but I still use it a lot and pinboard is a bit limited.
i derive _so_ much value from having well tagged items in Pinboard.
I think this offers the exact same thing as the popular pinboard but it's free !
http://www.gettoby.com/
I find it has some tiny glitches with drag and dropping to rearrange, but otherwise it's free and works pretty well.
If you're not then don't store them, just google them when you need them. That will either help you find the link you wanted, or it will find you something better, because the world has moved on since then.
EDIT: Looks like Pinboard does this already! Mea culpa.
the whole point of bookmarks is to be able to retrieve the info later. that archive has helped me SO many times.
it doesn't suggest links to you and the visual style is very different.
It's run by a single developer who responds quickly to feedback.
IMNSHO it's worth every penny.
So I am now building Liste[1] thinking of it as a more power-user focused tool.
Aim is to make Liste a powerful read/watch/listen-it-later + a personal knowledge base tool. In the long run I aim to replace goodreads/imdb et al with only source of truth in my life - all stored in Liste.
Beta launch is targeted in Q1 2020. Probably by mid-feb.
[1] https://getliste.com
https://LearnAwesome.org
It's FOSS, of-course.