Announcing Typesense v0.23 and Addressing Queries

🚀📢 @channel Announcing the general availability of v0.23 of Typesense! 🎉

Here are some key highlights:

✅ Phrase search: wrap keywords in a query with double quotes to search them as a phrase, e.g. "new york".

✅ In-place collection schema changes: we now support fields to be added or dropped from a collection schema in-place.

✅ Infix searching: find string within strings, which is useful for entities like model numbers or email address.

✅ Sorting on string fields: helps build interfaces where you need to sort A-Z or Z-A or lexicographically.

✅ Treat space as typo: search for "basket ball" if "basketball" is not found or vice-versa.

✅ Improved multi-field matching: much better performance and accuracy when query keywords have to be matched across multiple fields in a document. (~40-50ms improvement in our test data sets with tens of millions of records).

✅ Improved support for Cyrillic languages - better highlighting and fuzzy search

✅ ARM compatibility: an ARM build is now published for every release

✅ Homebrew: a new Homebrew installation method for Typesense on macOS

✅ Each multi-search query can have an independent x-typesense-api-key key. This is useful to specify different scoped search API keys for each search.

✅ ~10x performance improvement when updating/deleting numeric fields.

Here's the full changelog: https://typesense.org/docs/0.23.0/api/#what-s-new

Here's how to upgrade to this version: https://typesense.org/docs/0.23.0/api/#upgrading

A big THANK YOU to everyone who contributed to this release in the form of feature requests, ideas, feedback, bug reports and code! 🙏 🙏

nice timing, as i watch wwdc! super excited about the double quotes feature

Awesome update! 2 questions:
• for schema changes -- does that include modifying a field (e.g. int32 -> int64) without drop -> create? Or does it need to be dropped and re-added (and thus data re-indexed?)
• As a recent Cloud customer -- how do my cloud clusters receive this new update?

Ross

> does that include modifying a field (e.g. int32 -> int64) without drop -> create?
I assume you meant drop->create the collection. That's correct. Here's how: https://typesense.org/docs/0.23.0/api/collections.html#modifying-an-existing-field

So in your case you would do something like:

     {
         "fields": [
           {"name": "numeric_field", "drop": true }
           {"name": "numeric_field", "type": "int64" }   
         ]
       }

> As a recent Cloud customer -- how do my cloud clusters receive this new update?
You'll find a "Modify Configuration" option on your cluster dashboard. You can schedule an update from there. (Edit: Just added this info to the announcement).

Jason + better support for Cyrillic - non Latin languages 👍

Oh yes! Just added it

Curious, any idea of the price-performance improvement on ARM?

We haven't benchmarked Typesense on ARM yet, but these stats should be in the ballpark: https://www.scylladb.com/2021/09/16/aws-graviton2-arm-brings-better-price-performance-than-intel/

> Arm instances provide 15%-25% better price-performance

Hopefully even better on Graviton3

Amazing effort and work guys! Congratulations! 🎊

yay for exact phrase matching :partying_face: thanks for all the hard work!