Glad to hear that. This is also a big pain point for me.
And @Oshyan presumably you’re aware a big hole in Notion is they don’t index comments in their “new” search from earlier this year? So big opportunity for Michael and co. if the comments also come through as part of the search. I would really like to petition to include that, as my team endeavors to keep as much communication in our tool as we can, and not outside in Email, Slack, etc.
Ah, interesting, I didn’t actually notice that as I use Notion solo and don’t use Comments that much lately. But I do use Comments in other systems for my own reference, and ultimately would prefer Notion to have that too. So it’s good to know that got removed, and hopefully will be coming back. Their search update was overall a good improvement from before (though what they had before was… terrible, hah). But it was not a total fix, clearly…
Hey, just a point of clarification, although may not be necessary: Notion in the previous iteration of search, which you pointed out was not good, did not actually index comments at that time, either. So they keep missing that big piece. And I’d say Fibery with the “most recently viewed” at the top of the limited search we do have right now, is superior even to Notion’s old “original” search.
Michael and @Oshyan (and @Chr1sG as I assume as you increase use this may be of need to you) - I was thinking about the fact that Notion doesn’t include comments in search and we have an opportunity to beat them to that function. So when I saw the huge AirTable release today of synced bases and Automations, I thought about do they have that? Well, they don’t - first request in 2015!!
While I’m at it, I’ll close the loop on the “big 3” and add that Coda doesn’t do this, either.
So real chance for you guys. As. I mentioned in your post @Oshyan about use cases, I am really big on Fibery becoming as much of a central communication hub around the work my team does in here, so a good search function is absolutely key to that long-term goal. I assume all reading this may know that Slack really has pushed hard for its search functionality, and they have big benefit from it. If you think about it, aren’t comments in Work Mgmt tools like Fibery, Notion, etc. essentially just as important to a team? So it seems really a travesty that when you’re in those tools, you can’t search through what was discussed in comments!!
If you have a moment Michael, would love to hear from you as to whether you guys plan to include comments in search results. Thanks!
Agree, and I’ll just add here that for me personally, when a feature cannot be added, or is more difficult to implement, I would appreciate knowing that fact, even if there can be no indication of whether or when it might be added.
So in the case of Search in Comments, let’s say that extending search to index comments would make it 10x slower for all queries and/or make search indexes 10x bigger, both of which would cause problems for Fibery dev/hosting/etc. It would still be really good to know from @mdubakov or others on the team that they have looked at the request and unfortunately have identified significant blocking challenges like this. Or if they want to do it but identified UI issues that need to be solved first, such as search filtering.
Great points! Indeed, if it turns out this particular request, or others, can’t be done for a certain reason, I feel like most of the user base will be understanding. So I don’t see why it’s that risky to explain and effectively “reveal” a part of the roadmap in this case, although I realize it is a very thin line to promise too much roadmap to users as that can be hard to deliver.
How about just an option to select which fields to include in the search query? Depending on the type for our case, different fields of information would be useful to display.
In our case, it’s mostly just 1–2 fields like a customer ID or a product state, which would be huge time savers so we didn’t have to investigate a lot of the results (thank God CTRL+Enter is possible!).
If this or something like it would be implemented, I am fairly certain that would be my main method of navigating around our workspace. Just CTRL+K, write, find, enter. Boom, I am where I want.
I still use search a lot to navigate around, but problem is that it shows only name fields. In our “main” place of interest (table) the name of the row not important at all. The only purpose it has at the moment is being identifiable in searches and references. We generate the name though a button by combining pieces of information from other fields and entities of the parent.
Hey Michael, just curious if you have any further update on this, in light of the [In dev] label. I know this is obvious to you, but another big basic piece of functionality in Fibery, so very curious for any update you can provide…also would be really curious how much you plan to index, stuff like comments (not indexed by Notion). And do you plan things like showing snippets of Rich Text in the results preview, would also be very useful…
Finally, I’d love to be able to navigate to a top-level App “home” - which I know doesn’t really exist yet - via the search dialog. I can get just about anywhere else quickly in Fibery with “ctrl” + “K,” but trying to find apps via scrolling in the Left Nav has been cumbersome for my team.
Yeah, I had a big and unexpected need for it just a couple days ago. I only could have worked around it if I’d anticipated the need and somehow copied all the data to other tools along the way, i.e. a total pain in the butt. So my need for full text search is not constant, but when it arises, it’s critical.
I agree. Since the team considers this one done, and the title is a bit unclear (“entity contents and docs”), maybe we need another feature request to index all fields?
I had the thought that if it would help the Fibery team to implement indexing of more fields while minimizing impact on performance, index size, etc., it could be interesting to have a per-database selection for which fields get indexed. There are quite a few fields on many of my DBs that I wouldn’t need or necessarily even want to be used for search.
Good to see the topic of search getting some activity here. Another thing I’ve noticed of late (probably could use yet another request) is that the order of search terms seems to dictate results more than it should - in other search implementations you can flip-flop the terms and the results stay the same. And generally just want to support again the indexing of more in Fibery, most notably for my need Index comments in search.
Hoping all this discussion will lead to some increased push to move better search up on the priority list for development!
I just tried rolling my own indexable-fields-control, by adding a Rich Text field to contain all the text I want indexed for an entity.
– UPDATE – The following is inaccurate: Indexing IS working on all RTF fields –
And then I found out that “entity Description” is the ONLY Rich Text Field that gets indexed - not ALL Rich Text fields.
So perhaps the only way around it would be to re-purpose the entity Description field as the “Indexable content” field, and then create a new Rich Text to replace Description (if needed).
Then you need a Formula field that concatenates the contents of all non-RTF fields you want indexed (Texts, Multi-Selects, etc), plus a Rule that triggers when this Formula is updated, and copies it to the “Indexable Content” RTF (formerly entity Description).
If any RTFs also need to be indexed (e.g. the “replacement entity Description” field) , then you’d need to use an hourly-scheduled Rule instead to update the “Indexable Content” RTF - because Rules can’t trigger on changes to RTF’s, and Formulas can’t reference RTF contents.
Matt, I read this and thought you were saying that there is only one Rich Text field that will get indexed, the first one you create. I tested this and I have entities with multiple Rich Text fields and it seems like the content in the other Rich Text fields does show up in search. Am I misunderstanding the issue here? Definitely would like to see as much as possible indexed in Fibery!