Submitted for Judging

Favorites (1)

An innovative bookmarking 2.0 service that lets you 'gather' links as you browse the web.

To try it out please do the following

  1. Go to Http://www.teragulp.com/account/login
  2. Use gather as the invite code and register an account
  3. You will now see two 'public/featured' linklists that you can browse
  4. Use the 'Gulp It' bookmarklet to add your own links.

We use

  • iron.io cache to keep list state
  • iron.io worker to process and save each bookmark.
  • parse.com for user and other db info

With time we will be
- adding phantomjs screenshots for each bookmark
- extending the gathering process to include other information that we find as we browse (search results, page snippets, etc.)

Reasons why I used iron.io cache

  • Using parse.com or other db would require that each entry be a row and each list load would require a read of all the rows potentially slowing the system. iron cache however allows me to read the whole list in one 'gulp'

Reasons why i used iron.io worker

  • parse.com has a time limit to each call, so long lived backend processes can potentially timeout. With iron.io worker I can be assured that parse will always complete and I can undertake complex backend processing on the worker.
  • For ongoing backend analysis whether it be checking if links remain alve, getting screenshots and/or performing big data analysis and other headless activity, iron worker can simply be activated via a chron/webhook and each additional piece of application complexity can be incrementally added as separate workers.

So, using openshift to host, parse.com for database & user info, iron cache for state and iron worker for server processing one can now build a fast and scaleable webapp on the free-tier :-) You only pay for the domain name.


Srinivasan Sriram