Laura Gibbs

10. Searching for Search

Blog Post created by Laura Gibbs on Aug 8, 2018

Update: I submitted an idea for Instructure to hire an OER STRATEGIST to work on problems like these and strategicially develop the Canvas OER ecosystem in general. Please vote!

Instructure needs an OER STRATEGIST 


 

After getting off to a great start with the OER session on Wednesday, I then rushed off to the Engineering Ask-Me-Anything session. This was something I was excited about because there had been an opportunity to submit questions in advance both at the Community and at the InstCon Flipgrid. I shamelessly posted my question in both places: what's going to happen with global search? It was a big item from Khaki which was shelved without making any progress (details here), and I was really hoping to find out how the conversation was going to move forward and what we, as users, could do to help the engineering team sort through different use cases to rescope the project so that, if not global search, we could at least get improved search in different areas of Canvas and Canvas Commons.

 

So, the way the session worked was that there was a fantastic engineering team who was there to answer those questions plus other kinds of questions that regularly come up in these engineering encounters. I did take notes on the session (you can see my notes here), but this is not content I'm familiar with, so I don't trust myself to expand on my notes -- suffice to say, they covered a lot of really exciting topics, and the engineers are a very humorous bunch so it was a lot of fun too. Big focus on security and also on accessibility, which was great to hear about.

 

But here's the thing: as they worked their way through the questions, my question about Khaki and global search didn't get asked. And then they said they were all done with the questions from the Flipgrid and from the Community, but nobody had mentioned search. That felt weird: I made sure to ask my question in both places so that it would not get missed... but somehow it did.

 

So, I just went ahead and asked my question, and the answer was really disappointing: Chris Hunter explained that it was just too difficult to make search work with all the different levels of permissions. He didn't indicate that it would be rescoped. He didn't indicate that they would pursue search for content that was CC-licensed without limiting permissions. He didn't indicate that there would be search for Canvas Commons where permissions are presumably not a problem. No forward progress of any kind.

 

From the other discussion about the deferral of global search (including several comments from me, especially here, and with a really good comment from Joe Allen you can find here) , I learned about Atomic Jolt and how they do provide a search service, albeit for a very steep fee. If Atomic Jolt can get schools to pony up money for the search service, that proves in another way, beyond the Project Kahki vote, that this is something Canvas users find valuable. I strongly believe that if we are going to take Canvas seriously as a platform for developing and deploying content, we need to have search tools. Is anybody at Instructure listening? I hope they are...

 

Let's just take one example: Canvas Commons. Jared Stein shared in his keynote numbers about Canvas Commons usage, which was very encouraging, but the poor search features of Commons dramatically limit its usefulness. It does not search content right now, only titles and tags (and perhaps also the brief descriptions people include? I'm not sure), and, even worse, there is something very broken about the search that it does of title and tags right now: for some reason it seems to be just searching only on the first three characters of a search term. Maybe someone can figure out what's up with that. This has to be a bug and NOT a feature, right? Examples:

 

I search for elephant

https://lor.instructure.com/search?q=elephant 

and I get elephants and electricity and elections.

 

search for elephants

 

 

I search for calories

https://lor.instructure.com/search?q=calories 

and I get all the calculus materials:

 

search for calories

 

I search for ballet

https://lor.instructure.com/search?q=ballet 

and I get ballet and ballots and ballistics. And also "balance," which is what makes me think it is searching on just the first three characters. But who knows? It's just ... weird.

 

search for ballet

 

I am a huge fan of open content, and I would love to see Canvas being a force for good in the world of open content development (see my post here about open Canvas courses) ... but it is really hard to take content development seriously when Instructure is not going to provide content search features. I really could not recommend to any faculty member that they invest time in developing content on a platform that makes it hard, or even impossible, to discover that content in order to share and re-use it. 

 

And since Chris said that the problem with search is permissions, then maybe global search could be rescoped ONLY for content items that have some kind of CC license so that permissions are NOT a problem. Then we could get real search, including content search, for the Commons, plus a search feature for anyone who opens up their course content with a CC license. That sure might be an incentive to get people to think about the advantages of open course content!

 

So, I really enjoyed the engineering presentation, and I am impressed by all the great work they are going on security and accessibility in particular (you can watch the video when it comes out to learn more about that). I definitely understand the big projects going on, and how important they are... but I am not going to stop asking about search. The vote at Khaki showed that users put a high priority on search, willing to dedicate serious resources to it compared to other ways of spending those resources. I will keep hoping for more/better conversations about content development generally in the Instructure ecosystem (it needs some love, people! learning requires more than just quizzes, gradebooks, and data tracking), and I will also keep hoping about more/better conversations about search in particular.

Outcomes