Hacker News Reader

Comment by iansinnott | Hacker News Reader

I have something of an obsession with being able to search my digital history. To this end I've created a few projects which I still use almost daily:

- full text tabs forever [1]: chrome extension to provide full text search over web pages I've visited

- browser-gopher [2] (formerly browser parrot): collect browsing history from all browsers

- prompta [3]: a ChatGPT UI with full text search and some other niceties

[1]: https://chromewebstore.google.com/detail/full-text-tabs-fore...

[2]: https://github.com/iansinnott/browser-gopher

[3]: https://github.com/iansinnott/prompta

edit: formatting

View on HN

> full text tabs forever [1]: chrome extension to provide full text search over web pages I've visited

This is amazing.

But wouldn't this extension end up taking up a massive amount of space? Does it index text which is fetched and hydrated after page load?

Also, does it strip HTML from the text so that text <strong>like this</strong> can be searched with exact phrase matching?

Is the source code for fttf available? I'm sure you're lovely, but I'm nevertheless reluctant to install an extension which has access to all pages

It doesn't take up too much space. Currently about 70mb on my system. It will grow over time of course.

It used a distilled version of the web page, i.e. "reader mode" and indexes that rather than the full HTML. So yes it indexes plain text only and, in theory, ignores headers footers and other non-interesting parts of web pages.

Anecdotally I find it invaluable for finding a web page I know i've seen but can't remember.

It's not open source as of this comment, but the plan is to open source it. Agreed that OSS for something like this is important. It will index all your auth-ed pages too.

It's also not minified or obfuscated, so the source is "available" in that sense.

> It doesn't take up too much space. Currently about 70mb on my system. It will grow over time of course.

If the publication date (Nov 30) is an indication of how long you've been using it, it's not exactly lightweight :P

> It's not open source as of this comment, but the plan is to open source it

Might be really useful with some configuration around which domains to archive; I may contribute (or fork if you're not accepting contributions) if it is released under an OSS license

> It's also not minified or obfuscated, so the source is "available" in that sense.

Amazing!

Cheers! would be happy to have contribution.

I've been using it for longer than the publish date. A few months at this point. It's also roughly twice the size it needs to be I think, as the FTS index could be optimized. So there's room for improvement.

https://github.com/iansinnott/full-text-tabs-forever

I really love full text tabs forever, but as another commenter noted, it’d be good if it was open sourced to make sure it’s not sending the data to the devil.

Thanks for sharing!

A quick DDG shows this github repo - same username on HN and GH, so one can assume this is the source: https://github.com/iansinnott/full-text-tabs-forever

Standard disclaimer: the version on Chrome Web Store could be different than the Github Repo.

Edit: ah, it's 1 hour old, I guess OP just made the repo public.

Full text tabs looks great! Would the code work on Firefox?

Not currently, but it was initially built with that in mind. Namely, i used web extension polyfills to unify the extension API of Chrome and FF. However, while trying to achieve MV3 compatibility I used chrome-only APIs because the polyfills made debugging more difficult.

It will take a bit of work to support FF, but not too much work

>full text tabs forever

Thanks, this is awesome!

Reply on news.ycombinator.com