rdavidr

I don't want to claim that Linkidex is better than Raindrop. Raindrop and Linkidex will serve different audiences. That said, we aren't quite ready to share our long term roadmap.

Thanks for your patience on this! This issue has been fixed.

Raindrop is great! Its written by one guy and the polish / quality of his work is top notch.

Our roadmap is going to bring us in a different direction than Raindrop, but more details on that later.

1) sharing is on our roadmap and core to our long term value proposition.

2) That is a great idea. Right now we default the title to the URL if a user doesn't input a title. Your suggestion is superior, will look into this.

3) This is something we have considered, but isn't currently on our road map. Right now we are looking at making Linkidex a great productivity tool as opposed to yet another social network.

I don't think a web app can do this, but a browser extension, which is on our roadmap could!

Regarding auto-categorize, how would you want that to work? I imagine this could be a machine learning problem, but I wonder if there is a naive solution that solves most of the problem instead of just throwing ML at it?

Text search across pages has come up multiple times in this thread. I shall add it to our list of features to investigate.

100% agree. A browser extension is on our roadmap.

thanks for all of this info!

The front end app maintains its own state. The front end fetches all of a given users links when it loads and stores them in application state (memory within the front end application.) If you lose your internet connection, the front end will continue to use the links that it has in its application state.

This isn't how Linkidex works right now, but using Linkidex as an example, the front end could queue up tasks such as "create link abc", "delete link "bcd", and "edit link cde" when offline. Then when your device regains an internet connection, it could send all of its queued up tasks to the back end and reconcile the changes with the database. PWAs make this kind of thing possible, and fairly simple to implement.

I am a huge fan of "The net ninja" on YouTube, and he has a great tutorial on PWAs: https://www.youtube.com/watch?v=4XT23X0Fjfk Note this is 3 years old and front end stuff changes quickly so this may be slightly out of date, but I still recommend this video series. PWAs are awesome.

Edit:

To clarify about linkidex working offline, you can still search your links when offline, assuming you loaded your links before going offline. You can't create, edit, or delete anything though, and if you refresh the app it will fail to reload just like any other web page. Maybe its unfair for me to say it 'mostly' works offline. I'll do a better job communicating this in the future.

The database is PostgreSQL.

We are using Material UI (https://mui.com/material-ui/getting-started/overview/) on the front end which is what makes the website responsive and mobile friendly. MUI has been great to work with, I highly recommend it.

The Progressive Web App part of things (https://developer.mozilla.org/en-US/docs/Web/Progressive_web...) is what makes Linkidex still mostly work if your internet goes down, and lets users "download" the front end web application to their phone so it behaves like a native mobile app would. For example, users who download Linkidex get an icon on their home screen that opens Linkidex, and the url bar isn't visible when the app is open, despite the downloaded app just being a web page thats opened with your phone's default browser.

archivebox is neat. Also it looks like they have the option to backup links to Archive.org, which was mentioned in another comment too.

I hope it does!

Making the error reporting of import / export is very high on my todo list. Right now the user just has to wait and doesn't get feedback, which is a bad experience.

I see one import failed due to something on my side. I've identified the issue and have a fix, I'll follow up as soon as its deployed. Sorry for the inconvenience!

We haven't plugged into a phone's "share" functionality yet, great suggestion!

Using web.archive.org is a fantastic idea I had not considered. I wonder if its ok to use their API to serve 'saved' versions of a given site instead of storing it myself. I shall look into this!

by local storage / export, do you mean storing your URLs, or downloading the archived text versions of pages?

To clarify, the suggestion here is to search the contents of a given Link, not just the link's title or url?

For example, if I have a webpage "www.linkidex.com/test" which displays a bunch of text, including "HackerNews2022", you need the ability to find this link by searching for `HackerNews2022`, assuming that you haven't named the link `HackerNews2022` or tagged it or categorized it as such?

Some of our competitors have this feature, but this feature comes at a cost. What happens if the given page you have saved changes? What if you save a page "New Laptop setup instructions" and those instructions change? It may be harmful to follow laptop setup instructions that are out of date.

To clarify, are you suggesting Linkidex automatically sync with whatever the user's bookmarks are on the user's current browser?

A naive solution could be to just display the first letter of whatever the user has named the given link. Not as helpful as a favicon, but it will prevent the page from looking broken if a bunch of favicons don't exist.

This is a great point. The chrome extension I was using before building Linkidex had a fixed layout, and I was often times able to quickly navigate it without reading what I was clicking on because I had memorized where certain things were.

the first UI I built with Linkidex was a grid instead of rows. I moved to rows because the grid became tedious for users to parse when they began to search or filter and grid elements started shuffling.

Linkidex does not currently get a sites Favicons, and currently doesn't have a way to fix where a given Link is on the page. I have catalogued this feedback thank you.

https://github.com/go-shiori/shiori I haven't heard of this one before, thanks for the info.

May I ask what made you decide to use Shiori instead of pocket or something else? It looks like Shiori is marketed as a "simple clone of pocket."

Notion Databases are neat! I’m not familiar with them, I just glanced over some of the documentation. I think its safe to say Notion Databases would let you create your own Linkidex (or your own todo list app, or your own food macro tracking app, etc) as long as you are willing to spend the time building and maintaining it, which you must if you have already put in the work!

May I ask what made you choose notion vs any alternative? Do you like having complete control over the shape of your data, do you like the UI, or is there something else about it?

Not a problem, thanks for checking it out!

“Test account for HN” registers an account with an email I’ll purge later, so people don’t need to give me their email address if they don’t want to.

1) I think the biggest difference is that Pocket (and some other competitors) try to keep you inside pocket. You can get to the original URL of something in your list, but it takes an extra step each time. Linkidex doesn’t try to keep you inside Linkidex, we just try to help you find the thing you are looking for as quickly as possible so you can move on.

2) Yep! Our roadmap includes browser extensions and native phone applications.