Hacker News new | past | comments | ask | show | jobs | submit login
Show HN: Mashups – Resurrecting Yahoo Pipes, my side project (mashups.io)
166 points by pyromaker 16 days ago | hide | past | favorite | 39 comments
Hey everyone.

For those who remember, Yahoo Pipes was a tool to mashup RSS feeds back in the good ole' days. :)

I really loved that tool, but of course, it was shut down.

Since then I know there's been a few tools and attempts at bringing it back.

I always wanted to create Yahoo Pipes clone myself.

So here is my small side project - Mashups.io

https://www.mashups.io

It's an MVP at the moment and well, let's see where it goes.

Thanks all!




Also check out https://nodered.org/ and https://github.com/huginn/huginn if you're interested in free and open-source software you can run yourself.


Is there a consolidated comparison website for these? There is also n8n that is on the rise now.


Not to mention Zapier and Make, which appear to dominate the same "proprietary solutions" space where this tool resides.


Also ActivePieces from YC



No, that's just a well-curated list. I am thinking along the lines of https://diskprices.com/ where you have all the characteristics/features mapped and can filter down using a faceted search (e.g. implemented using datasette) by what you need. For example, I know that if I need iOS Reminders integration, IFTTT has support for that but not n8n/make/zapier. And I think Make does not support Github integration with webhooks, only API calls. If you have a set of things in mind you'd like to integrate, picking the tool that supports a full combo OOB is a pain in the spreadsheet.


Windmill.dev too!

Pipedream!


I have been running Huginn on an Ubuntu VM since Yahoo Pipes shut down almost 10 years ago.

I've tried out NodeRed and n8n over the years but they've been a huge let down in ease of manipulating multiple RSS feeds and web scraping -- my primary use case for Huginn. I tried Zapier and IFTTT as well, but the free tier didn't allow for what I wanted to do. Since I could do it just as well with Huginn, I decided to stick the more complex flows there also.

Once you figure out the workflow (setting up agents, and which agents suit which task), then it's fairly straightforward.


how do these tools manage your credentials / tokens / cookies?


Is it just me but when I see a Rails app, I just walk away. I don't feel like setting that all up.


I played with pipes for a minute to manage an RSS ecosystem, but didn't build anything that I kept around.

The site seems to be down.

Let's say we all build wonderful apps on this, and then you run out of money or legal liability for hosting.

Is there a way to construct this so that we can export our Pipes to some standard format that the next person to pick up the torch could activate, and migrate to that?


I used Yahoo! Pipes to make a consolidated social feed some years ago (https://github.com/Xeoncross/MicroStream). It actually placed as an honorable mention in AListAparts 10k challenge.

It was really fun to have a service that could handle all the backend requests for you before sites like Zapier existed.


I miss yahoo pipes! I used to use it to aggregate movie reviews for some project I was working on. Brings me back.


I remember using it to read webcomics directly in my rss reader. It got me to really learn regular expressions.


Congrats on the launch! It's awesome to see someone resurrect the best parts of Yahoo Pipes, I really miss the simplicity of those days. Have you thought about supporting other sources, such as websites with tables or social media feeds?

I've been thinking a lot about how LLM agents need some kind of glue to connect them to everyday websites at a level higher than raw HTML. What you’re building here really reminds me of that idea.


My biggest use for Yahoo Pipes was filtering podcast feeds - namely, filtering out episodes I don't care about. For instance, a podcast I listen to has episodes where they just talk shop and others where they play Dungeons & Dragons. I really cannot be bothered to care about the latter type.

I actually wrote a little, easily modifiable Digital Ocean function I've been customizing and deploying per feed I want to filter. It feels like a hack, and it is, but it's "good enough". It's been working reasonably well and runs rarely enough that it falls into Digital Ocean’s free tier.


Brings me back too. But what stayed stuck on my memory, was something i read here on HN: at some point, one yahoo dev added his affiliate code to the links !!


Cool project! Yahoo Pipes was before my time, but I remember a Tom Scott video that mentioned it. Imagine having a service like that for free today…

This video has … views by Tom Scott: https://youtu.be/BxV14h0kFs0?si=J3sVKEJhkABUSjN_


Along the same lines, I recommend Retool's retrospective article [0] about pipes, for some of the perspective on how it looked from inside yahoo. I'll grant that, towards the end, it is an ad for their service.

[0] https://retool.com/pipes


Thanks for sharing this, so cool!


Is anyone able to scrape Bloomberg news? For years now they have no RSS and have made it very difficult to create one.



You will need to go through archive.is if you want to bypass the paywall and get the full article, not just the summary.


Typing this in the console:

  $$("div[data-component='headline']")
would select all titles in the home page. It shouldn't be super hard to build an RSS file using beautifulsoup for example. Can you be more specific about what you would need?


Great product, is awesome play with RSS feeds. Will check it out.

Want to add for anyone wanting to quickly apply a filter to one RSS feed I recommend https://www.siftrss.com/, it doesn't require create an account and it works perfectly.


I had a lot of fun wiring and rewriting an RSS feed using this. Super simple and worked well. RIP Yahoo Pipes, but this looks like a worthy successor.

And, I can't help but think there is gold to be found mining the graveyards of tech that people loved but was never a billion dollar revenue source...


Thanks for putting the work into this.

These kinds of software pipelines can be strangely satisfying to implement.

The development style of thinking of stream processing and online algorithms

It's also inspired by mapping and filtering and functional programming with flatmap.

It reminds me of factorio


Oh, this is really cool, and exactly what I'm looking for, but uh, I really wish I could test it! Thing is, I probably can't. I use a screen reader, so I doubt the interface would work for me, but maybe I'll try it, just for fun.


Wow interesting, I think I'm too young to remember yahoo pipes


Not gonna lie. I've been looking for something like this. It tried it, works really well. I'm gonna build a daily news pipeline for myself.


This is nice. Over the years I moved to Noda-RED for this kind of thing, but Pipes was just so much easier to use for quick queries and filtering.


I've heard many people here over the years mention it fondly. You must have left quite the impression on them.


Hit me up if you want to compare notes about Pipes successor implementations ;) And best of luck.


For some reason, on mobile, the demo and pricing links in the navbar get me a 404. Works fine on desktop.


Apologies - that's been fixed. Though I haven't really optimized the demo for the mobile view I think.


The demo was intuitive, congrats on the launch! Can I ask what you used to build the canvas UI?


most likely xyflow


can you add a JSON export option?


Good idea!




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: