source: https://retool.com/pipes
A tool for connecting data across the web that never came to be
- started off as a “skunkworks” effort within Yahoo as part of cutting edge ideas
- finished version was intended for end users to help program websites
- Final product was a cloud-hosted service where users could drag and drop elements to “roll their own information feeds”
- these feeds could take data directly from websites, be filtered, merged, and processed and produce a final stream that could be subscribed to
- Pasha Sadri lead this effort
- The problem emerged when they launched they got 100,000s of users rather than thousands and each pipe in aggregate pulled a ton of data and required a lot of processing
-
- looks a lot like natto.dev
- Supported RSS, Atom, CSS, RDF, etc. and eventually YQL
- and very powerful output formatting:
-
Pipes output could be embedded into a page or formatted as HTML, JSON, KML, RSS, or XML. There was even a way for a non-Yahoo website to include JavaScript that queried a Pipe workflow and received the results back without any front-end browser or back-end server involvement. Pipes, in effect, could power other sites.
-
- But other websites also started blocking pipes
- ex: craigslist, supposedly because of resource hogging after someone created a map of craigslist listings (source)
- lol they literally hardcoded “pipes.yahoo.com” and blocked it
- but they brought it back 2 weeks later
- ex: craigslist, supposedly because of resource hogging after someone created a map of craigslist listings (source)
- HN discussion on the article
- effectively allowed you to bypass cors to build clients that used data from other sites
-
A youtube video with title that changes with the view count: https://www.youtube.com/watch?v=BxV14h0kFs0