My issue with all bookmark managers, including the one I currently use (Larder.io), is re-discoverability. I will happily bookmark this and that, but it's unlikely I'll ever actually go back to the bookmark. Every once in a while I'll go through the library when I feel like looking for random gems, but the first port of call when trying to solve a problem will always be a search engine.
I wish a non-Google search engine would offer a bookmark manager that integrates into its everyday search, giving preferential treatment to previously-bookmarked material. That would be the be-all end-all of bookmark managers.
Exactly!
That led me to develop an extension for quick bookmark capture into plaintext: https://beepb00p.xyz/grasp.html It makes it way easier to search and organise bookmarks, keep some context, etc.
I also describe my search setup here: https://beepb00p.xyz/pkm-search.html#other . Basically, I can search in all of my stuff with a single hotkey, whether it's pinboard, instapaper or just plaintext bookmarks.
What about a bookmark manager or solution that preserves your privacy. A specific list of bookmarks not only has an extreme degree of uniqueness, it's content (links) are highly revealing about the interests, problems and so forth of any one person. This then tied to someone's IP address is what makes me wonder what solutions exist that let you keep all of that private while being able to sort, save and transfer your bookmarks.
>I will happily bookmark this and that, but it's unlikely I'll ever actually go back to the bookmark.
That might be a problem with your process, not with your tools.
>I wish a non-Google search engine would offer a bookmark manager that integrates into its everyday search, giving preferential treatment to previously-bookmarked material.
Most browsers already do that with local bookmarks, to a degree. (they don't search the article contents, only the bookmark tags and titles).
> That might be a problem with your process, not with your tools.
"You're holding it wrong" is rarely a valid critique.
My process accumulates knowledge; the problem is that this knowledge is saved by bookmark managers as a set of dumb pointers. There must be a way to surface them, at a later date, in circumstances where I'm searching for the knowledge behind those pointers, rather than for the pointers themselves.
> Most browsers already do that with local bookmarks, to a degree.
The current stuff is just basic pattern matching. A real search engine would obviously work at a very different level.
I don't think it's a problem that browsers can solve alone, it's a feature that search engines should offer and currently don't (the only one I know that used to have a bookmark manager was Google, but I don't use it and I don't even know if it still has such a service). Or maybe bookmark managers should be smart enough to offer search capabilities so good that I can use them as the point of call for search (e.g. when I search among my bookmarks, as well as searching the actual pages behind those bookmarks, transparently integrate real search-engine results rather than saying "0 found").
> Most browsers already do that with local bookmarks, to a degree. (they don't search the article contents, only the bookmark tags and titles).
That’s a pretty big problem- the specific term I remember is rarely in the title or URL. Not to mention all the times I’ve returned to the same bookmark only to see the content have been deleted or the domain expired. Default bookmark managers in browsers are complete rubbish.
Local cache would definitely be the (easiest) way of solving it. Tools like Memex [0] are most of the way there.
But a text-only copy on my local device isn't great if the content had special formatting in presentation. Also, it misses out on images or embedded videos. That's where something like ArchiveBox [1] comes in.
> ArchiveBox takes a list of website URLs you want to archive, and creates a local, static, browsable HTML clone of the content from those websites (it saves HTML, JS, media files, PDFs, images and more).
But really what I'd like to see at some point is an opt-in community tool where every page I visit that fits a certain criteria (URL, topic, special mark by me, etc) is fully cloned and uploaded to IPFS [2] for anyone interested in that topic to find and use later - regardless of what happens to the source content. Definitely a host of legal issues around this, but not impossible.
Vimium sort of does this. If you press O and input a term, you get results from your history, but if you don't select any, it will google this term. You can also press B, which searches bookmarks specifically, and again if it doesn't find anything, it will search the term on Google.
Also I find this B often gives better results than Google's address bar autocompletion.
It's a pretty big shift from browsers like Chrome and Firefox, but I really like how Qutebrowser handles bookmarks (and "quickmarks"). It's very similar to what you're describing.
I think Safari has good bookmark integration; I frequently re-find bookmarks and items in my reading list just from the title. Of course searching the content as well would be a great feature.
Bookmark OS comes close for me, but the ugly UI ruins the experience. Nothing is more annoying than the Russian doll experience of seeing a hacky version of macOS doubled up inside of the chrome of your actual OS.
Love Raindrop in theory (design-wise they have the right idea), but I dumped it in practice due to the fact it's too bloated and buggy and runs much slower than just keeping my bookmarks on Safari.
Pinboard, while being minimalist which I love, is minimalist in the wrong way. The visual hierarchy is all over the place and impossible to decipher through. Scanning down a list of bookmarks is painful.
So I've largely just been using the standard bookmarks feature in safari--which is actually refreshing and awesome.
Except when I do that, I can't sync bookmarks to other browsers. I use Chrome on mobile and sometimes on desktop.
No p2p-distributed portable across all browsers with ability to do semantic topology tagging integrated with webarchive and some translation service with ability secure parts of subgraphs and forkshare other parts with all proper layers views of visualization - from fast tag and to publish and research. With all these rdf. Crdts, type theory of patching, fuzzy vertex merges, deduplication, collective intilligence and graph marketplace. And dozen of other things. When you, the developer, will write one for me? You do not earn on giant huge market of knowledge consumption and digestion. Until than, internet is useless place without new T B Lee and Ted Nelson. I cannot live in the world without such bookmark manager. Please save me from self destruction.
Neat! Thanks for putting my LinkAce on the list too, depite being in beta.
I find it very interesting, that the "market" for bookmark manager kinda exploded in the past years, at least from my perspective. Tons of new apps were developed, all of them with their own specific feature set. And yet, finding the right solution is very hard because a) projects are sometimes barely visible or b) you have very specific needs. I find the first point is one of the biggest problems about open source projects and stuff. There are so many great projects out there but marketing is hard and/or expensive so many of these projects can only be found on page 2 of the Google results. Overviews like these are a great help, especially if they are up-to-date and cover a lot of projects.
To be honest, there are some solutions out there looking really promising and if I would have found them earlier, I probably wouldn't have built LinkAce in the first place. Now it's a great side project to work on that I actively use.
You could make a program that would fetch the content of those sites and use those libraries that create word frequency clouds to get the most common word. Then this word would be the category.
An interesting weekend project idea.
One of the problems is that any two people likely disagree on what “categorized” means. Tags? Hierarchy? Tag hierarchy? What defines a hierarchy? Etc etc.
I've been thinking on this for some time. Quick sorts for my use might be by date for one, by url (base domain) for another, common found terms within urls (news, product, etc)
then start looking at other tags, terms within the pages, stuff like that. But just having the quick sorts and exports on those above would make it more manageable.
I'm still in the process to find the perfect one. So far I use Trello lists for my bookmarks but it lack of search and tags and various other stuff.
I have boards with over 5000 links, my github has over 3000 stars, it is hard to maintain and reuse the knowledge or find stuff you know you are going to look for in the future.
So I am on the works on creating my own, A Trello style boards but more suited to links and knowledge search. Far from complete but it is a start.
Tefter (https://tefter.io) co-founder here. I’m glad we made the list!
It’s under the visual-based section, but arguably meets the criteria to be in many more categories.
It’s list-based and tag-based, since you may organise your bookmarks using either lists or tags.
It’s also search based, since bookmarks have their contents indexed and ranked and there’s autocomplete. For read-later, we keep archives and you can mark bookmarks as “read later” to read them on the mobile app.
However our favourite feature is the team collaboration. You can create organisations, install the Slack app and add bookmarks and aliases (think golinks.io type of short links).
We use a combination of heuristics and integrations with the most popular websites. In most cases we'll suggest relevant tags. Give it a go and let us know if it works well for you. We're still tuning it.
Recently, I have introduced it at work. All project bookmarks are maintained as plain text (yaml) in the project's git repository, so they can be adjusted by anyone (with Git knowledge/access). They are deployed on every commit automatically, so all collegues can access the latest version. It is not as end-user-friendly as most tools, but it has advantages for tech-savy people (like the mentioned plain-text managing via git or similar).
It's instant search of my 1503 bookmarks. Just titles/tags/urls right now, but I plan to index the text on the pages.
Meant to replace my current bookmark solutions (neither of which are in the list). So there is a feature to import netscape-formatted bookmark files. I currently use:
Heh - they don't have the one I created, so it's missing at least one! Don't worry, though - I'm not proud of it, it was one of my first Perl scripting efforts, and interactive web-site efforts too (perl cgi - ugh). So it's best if its never looked at again...
after some research found that one to be close in some ways to what I wanted. Found a missing feature I think it was 'deleting chunks based on urls' not supported that I wanted, I wrote the developer and never heard back via email, so I gave up on it. Kept it installed as a reminder to make something similar one day.
It's been a few years and now I wonder if gmail or some similar anti-spam system didn't just blackhole the email from my custom domain and that was the issue, not a dev that doesn't answer. hmm.
I wish a non-Google search engine would offer a bookmark manager that integrates into its everyday search, giving preferential treatment to previously-bookmarked material. That would be the be-all end-all of bookmark managers.