# Sphinx search for Discourse

**URL:** <https://meta.discourse.org/t/sphinx-search-for-discourse/12548>\
**Category:** Development\
**Created:** [2014年二月7日 23:29 UTC](https://meta.discourse.org/t/sphinx-search-for-discourse/12548 "2014-02-07T23:29:12Z")\
**Posts on this page:** 8\
**Page:** 2

<div class="post-metadata">

**Author:** ![katie](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/katie/32/114266_2.png) [@katie](https://meta.discourse.org/u/katie)\
**Post date:** [2014年三月16日 23:24 UTC](https://meta.discourse.org/t/sphinx-search-for-discourse/12548/21 "2014-03-16T23:24:55Z")

</div>

> [@neil](#):
>
> But why? How is postgresql’s search failing us?

It is not failing for now, they’re just a better option to consider in the future.

Is discourse being built to be a unique forum experience with innovation and great features that other forum software might lack or is it built just to handle sharing, communication, and discussion in a different way?

I’ve created several topics about Discourse on VB and XF because i was impressed with its simplicity and its strive for innovation which is something that drove me toward looking at. Vb has many features and probably that is why many were like “it looks nice and ok.. well.. but not for big discussions” - that was many people first impression.

I still like Discourse for its simplicity and I want it to see it better and unique : ) i believe it has a great future.

> [@michaeld](#):
>
> You can’t compare MySQL based boards where search is slow and painful to Discourse. One of the reasons it uses Postgresql is to work around those limitations.

Well yes, certainly not for VB 4.x because that is what they used (horrible approach) but with Vb 5.x where they started to use Sphnix, it was a better move.

---

<div class="post-metadata">

**Author:** ![sam](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/sam/32/102149_2.png) [@sam](https://meta.discourse.org/u/sam)\
**Post date:** [2014年三月16日 23:41 UTC](https://meta.discourse.org/t/sphinx-search-for-discourse/12548/22 "2014-03-16T23:41:02Z")

</div>

In an interesting turn of events, adding a dependency to something like lucene or sphinx is now not really a huge deal - now that we use Docker for deployment.

That said, full page results must come first, we need to continue refining pg, till we strongly believe it is just too limited for our purposes.

There is a whole unexplored vector (pg\_trgm) that needs testing first. I don’t see any reason to jump off the bus before we

1. Come to a strong understanding of the limitations of the bus
2. Agree that the bus is too limited

---

<div class="post-metadata">

**Author:** ![michaeld](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/michaeld/32/1594_2.png) [@michaeld](https://meta.discourse.org/u/michaeld)\
**Post date:** [2014年三月17日 08:14 UTC](https://meta.discourse.org/t/sphinx-search-for-discourse/12548/23 "2014-03-17T08:14:33Z")

</div>

> [@katie](#):
>
> It is not failing for now, they’re just a better option to consider in the future.

This is technology push that’s (currently) without benefits. It adds complexity and will be limiting to adoption.  
First identity the problem, then look for tools to solve it. Not the other way around.

Don’t get me wrong, I really love Sphinx, and use it in several large projects.

---

<div class="post-metadata">

**Author:** ![anon21958287](https://avatars.discourse-cdn.com/v4/letter/a/ee59a6/32.png) [@anon21958287](https://meta.discourse.org/u/anon21958287)\
**Post date:** [2014年八月22日 21:00 UTC](https://meta.discourse.org/t/sphinx-search-for-discourse/12548/24 "2014-08-22T21:00:23Z")

</div>

> [@sam](#):
>
> That said, full page results must come first, we need to continue refining pg, till we strongly believe it is just too limited for our purposes.

Full page search is really needed. I was looking for something here and I couldn’t even tell when a topic / post was made. Almost everything was pretty old, and I suspect there’s more out there, but the limited results and information about the search was very frustrating.

---

<div class="post-metadata">

**Author:** ![codinghorror](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/codinghorror/32/110067_2.png) [@codinghorror](https://meta.discourse.org/u/codinghorror)\
**Post date:** [2014年八月22日 21:03 UTC](https://meta.discourse.org/t/sphinx-search-for-discourse/12548/25 "2014-08-22T21:03:03Z")

</div>

Until we have that, try Sam’s little hack for now

[https://meta.discourse.org/?search=canary](https://meta.discourse.org/?search=canary)

---

<div class="post-metadata">

**Author:** ![jeans](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/jeans/32/104910_2.png) [@jeans](https://meta.discourse.org/u/jeans)\
**Post date:** [2014年八月26日 11:31 UTC](https://meta.discourse.org/t/sphinx-search-for-discourse/12548/26 "2014-08-26T11:31:48Z")

</div>

We created a plugin for full page search results: [GitHub - www2014/discourse\_advanced\_search · GitHub](https://github.com/www2014/discourse_advanced_search) - you’re welcome to contribute

---

<div class="post-metadata">

**Author:** ![anon21958287](https://avatars.discourse-cdn.com/v4/letter/a/ee59a6/32.png) [@anon21958287](https://meta.discourse.org/u/anon21958287)\
**Post date:** [2014年八月26日 13:02 UTC](https://meta.discourse.org/t/sphinx-search-for-discourse/12548/27 "2014-08-26T13:02:21Z")

</div>

Thanks. I noticed this the other day, and suggested that we try it out. I’m not an admin, so that’s as far as I can go.

---

<div class="post-metadata">

**Author:** ![Webinsane](https://sea3.discourse-cdn.com/meta/user_avatar/meta.discourse.org/webinsane/32/160268_2.png) [@Webinsane](https://meta.discourse.org/u/Webinsane)\
**Post date:** [2014年十月12日 06:05 UTC](https://meta.discourse.org/t/sphinx-search-for-discourse/12548/28 "2014-10-12T06:05:30Z")

</div>

Now my discourse bookmark is:

[https://meta.discourse.org/?search=webinsane](https://meta.discourse.org/?search=webinsane)

Because no one is as cool as me 😄

[上一頁](https://meta.discourse.org/t/sphinx-search-for-discourse/12548.md?page=1)
