Khaki 2018 Update

33 37 4,124

At Khaki 2017, we committed to devoting 100 sprints of engineering capacity to the priorities selected by the group. We’re on track to accomplish that and have already delivered some of the projects. However, now that we’ve had an opportunity to do more detailed analysis, we know some of the projects will require more engineering effort than we originally estimated (based on the limited time and info we had during the 1-day Khaki event) and fall outside the 100 sprint capacity.


In this update, I have listed below the projects that will consume the 100 sprints of engineering work and will be delivered by InstructureCon 2018, as indicated by original prioritization. However, we’re not stopping at 100 sprints or InstCon 2018. Some projects could not completely fit within the 100 sprint scope, but we believe they are important enough to allocate additional resources for their completion.


Finally, there are some projects that have been deferred due to their size or have become unnecessary because of other projects. I’ve listed those as well.


100 Sprints Delivered by InstCon 2018

1. Admin Role Granular Permissions part 1
  (New user interface to understand and manage the assignment of roles)
July 2018
2. Canvas Content Duplication
  Module Items
  Quizzes.Next (accomplished through the copy assignment feature)


July 2017

July 2017

October 2017

November 2017

July 2018

June 2018

5. Large Course: Sections/Groups

  Section-Specific Announcements

  Section-Specific Discussions


March 2018

June 2018

6. Page View Mobile Activity June 2018
10. Mobile: Camera Access for Embedded Images




Bonus (Delivered after InstCon 2018)

1. Admin Role Granular Permissions part 2—breaking out create/edit/delete permissions (to be added permission by permission)
3. Student Feedback View
4. Advanced Reports
7. Rich Content Editor: File/Image Drag & Drop - Canvas Release Notes (2018-11-17)
9. Commons: Preview Activities and Content
12. SpeedGrader Screen Recorder



8. Canvas Course Global Search

11. Course Archive Process

13. Large Course: Grading

14. Roles/Permissions Masquerade

15. Elementary UX


As always, we welcome your continued input directly, through the product team, via your CSM, or in any way you can get it to us. I can be reached directly at

Community Coach
Community Coach

Wow, Mitch!  That's all I got! Wow!

Oh yeah, and a big round of applause for you and the product teams!


Community Coach
Community Coach

Thanks for the update  The changes that have been implemented have been a big improvement.  Content duplication has been a real time saver.  I'm looking forward to seeing more of these completed!  

Community Member

I appreciate the update,‌! The global search was the item of greatest interest to me, although even just in our discussion time at Khaki, I could tell that it was a huge undertaking... a REALLY huge undertaking...

Would there be some use in having a discussion here at the Community where people brought forward some more scenarios about how they would like to be able to use search so that something less ambitious might emerge that would still be of great use to admins and/or instructors and/or students...?

Adventurer II

Mitch, thank you so much for this update. I'm tickled to see the progress made in these various Khaki projects, and really looking forward to the functionality that's expected to come between April and July 2018.

However, I am disappointed to see that two projects -- global search and course archiving, which were part of the original dozen -- have been deferred. I do understand the need to prioritize and that there's a finite amount of development resources available.

Can you say more about what "deferred" means for the Project Khaki ideas? Should we expect to hear about about progress being made on these fronts after InstructureCon? Or are they effectively being shelved?



Community Team
Community Team


It's always beneficial to continue the conversation on the idea thread! 

Also, it might be worth getting the attention of some members of the‌ group that have special interest in this?  

Community Team
Community Team


Thank you for the question. Deferred does mean that it is not within the scope of Khaki. The specific project, or a related one may be picked up in the future, but it will not be completed as part of the Khaki commitment. 

Surveyor‌, Thank you for the update!

Now I'm going to go do a happy dance for the large course items and a cheer of woohoo for granular permissions!  Smiley Happy

Community Member

Ouch: I remember that the global search represented a really big allocation of resources, but it also was on the Khaki list because it had strong support. Is there no way that we could see some movement forward with limited search, even if global search has been deferred...? I can imagine lots of uses in which more specific forms of search could still be extremely useful. 

I can understand having to scale back from the very ambitious global search that people had brainstormed, but seeing no progress forward on search is very frustrating: one of the most powerful features of digital content is searchability, but we are not getting that benefit with our digital content.

Community Advocate
Community Advocate

Pleased to see recognition for the need to allow greater differentiation in Canvas courses with multiple sections. Discussions next on the agenda and patiently waiting for changes to pages and modules which will revolutionise the use of Canvas courses encouraging far greater teacher collaboration and sharing. 

Community Advocate
Community Advocate

Also, it would be good to see similar Khaki events taking place outside of USA. European audience (and appetite) for Canvas growing significantly! 


Hi Laura,

Chris on the Canvas product team here to help with questions (I presented at Khaki about our product development process). I agree, search is becoming a common navigation approach for digital content/info. Paired with speech-to-text it makes our lives pretty amazing (now I wish I could ask Alexa where my keys are ;-). 

While I'd never say never, as you saw at Khaki, we plan our resources in 6-12 month intervals so that we can remain flexible to adjust to customer and market needs. With the combination of our published priorities and these items from Khaki, our capacity is full for the foreseeable future. We'll keep our eyes open for opportunities to add more specific forms of search but can't commit to delivering something this year.

Community Member

Thanks for your reply, Deactivated user‌, and I enjoyed your presentation at Khaki. It was really useful to have that overview!

I'm ready and waiting to be persuaded that doing content development is worthwhile inside Canvas, and the lack of search is one of the reasons why I cannot see Canvas as a real space for content development. But I do appreciate the different ways I can deploy content in other ways inside Canvas! And of course lack of native Canvas search is all the more reason for people to OPEN their courses so that they can let Google do the searching for them. 🙂

Here's a second version of my plea: if/when global search gets put into the Khaki process next time, maybe it would be worthwhile to think about a more modular approach rather than a total-global search if the resources are not available for the total effort. Being able to search Pages would be a big boost, for example, especially for faculty who do a lot of content development in pages -- and they do, as you can see in this request for ways to organize the Pages space: 

(In fact, you'll my now inaccurate post-Khaki comments there about how search is coming and will alleviate the frustrations people are having with their Page space.)

Community Member

Ditto for Euro.

Also, yay for drag'n'drop - might reconvert quizzing colleagues who left Canvas for Microsoft forms...

Community Advocate
Community Advocate

Am interested in your reply Laura and your talk of content development within Canvas.

I am a big fan of Canvas pages and develop lots of content through these. Admittedly much of the exciting content comes from the ability to embed features from outside of the platform using a range of Web2 tools  (and how much I wish for Office365 integration to allow embedding on to pages).

I try to encourage staff to make use of this too as creating long lists of file links in Modules or giving students access to course Files areas makes engagement and enthusiasm a little more challenging - especially at lower end of school.

Whilst I definitely see a benefit in a global search tool, I wonder if to make this more accurate, we would also have to ask staff to start tagging pages? Personally speaking and perhaps this is more relevant to UK curriculum models, I would like to see Canvas allowing content pages and Modules to be linked to sections. In that way we can have better teacher collaboration and sharing within the one course rather than the norm which is different teachers their own courses for different classes but for the same topics..

Perhaps what is good about this particular posting is a recognition that developments in the platform do need to happen much quicker and the product needs to be more agile and flexible for a range of users/systems etc...

Community Member‌, thank you for your comment, and it has produced a tl;dr response here ha ha. While I was a big enthusiast for global search at Khaki (just because it seems crazy to me to have all this digital content and not be able to search it; that is just how the digital world works and one of its greatest advantages IMO), but I think you are exactly right that we need a bigger discussion of content development for its own sake, both in terms of what software features support good content development, and also what kinds of individual practices (tagging, modularity, etc.) along with school culture (openness, sharing) that support good content development.

I would say the content development side of Canvas is the most neglected area. Just look at the unholy mismatch between the Files area and the Pages area and you can see that there was not really any kind of plan for content originally. I suspect it is going to be the most neglected area in the future too. At Project Khaki I asked and others what kind of long-term vision they had for content, and they had nothing to say about that; it's just not something that's been on the agenda compared to the intensive work they've been doing on quizzing, gradebook, analytics, etc. 

Perhaps they are right to stay out of the content business so that people can use other tools for that and integrate the content into Canvas as needed. At the same time, if they want to really make progress with analytics and also with individualized learning pathways, OER, etc., they are going to have to do a better job. Not having content search features already built into the architecture is a big problem. And there are many big questions we need to ask also, like why content is course-specific to begin with. That's just a fundamental design flaw that does not make sense, but which may be very difficult to address since it is built into Canvas at a deep level I suspect.

Like other LMSes, Canvas has defined "courses" as administrative entities, filling up courses with students and then leaving those courses behind at the end of the semester as if they had never happened. But for instructors, that's not how the world works: we teach courses that persist, semester after semester, and it is instead the students come and go. As far as I am concerned, I teach 2 courses, but as far as Canvas is concerned I teach 6 courses every year, which means I've got 12 courses in my Canvas dashboard (we switched to Canvas last year), next year I will have a total of 36, and so on. But I know there are ONLY two courses in there, and in terms of content development there are ONLY two courses (and I share a lot of content between those two classes also: writing support, student success, campus enrichment, etc.). It's just the admin-oriented design of Canvas that expects me  to copy the content over and over and over again as if I had lots and lots of courses. Copying like that is bad for content development, bad for search, etc. But good for administrative record-keeping purposes where student transcripts define the architecture. 

Anyway, if they are going to defer the search features that I thought they had committed to at Khaki, then I hope at least there will be a conversation about content development in Canvas that will be wide-ranging and in-depth so that the questions about search will be better framed with community input. Right now, you can look through all the discussions here at the Community and you will find precious little about content development. And that is because the Community largely reflects the concerns and the activities of instructional designers and technology staff who, for the most part, leave the content development to instructors. Instructor voices are less well represented here at the Community, and also at Khaki. Most of the people at Khaki who were instructors were instructors AND something else: instructor and sysadmin, instructor and ID, etc.; my impression is that there were only a few of us who were just full-time instructors without any tech dimension in our job description. So, I get that they ran out of Khaki resources and we are not going to get search anytime soon, but let's not defer the discussion, okay? Otherwise, we are going to have the same problem all over again next time without priorities to help make the resource allocation more realistic.

As always, I'm happy to keep going all my content development outside of Canvas (using iframe and javascripts to pull it into Canvas as needed), and I always urge faculty to do the same. But I'm the oddball: other faculty understandably expect to do their content development in Canvas, and they are not being well served by the tools available for that right now. I understand that the technical overhead of global search was too much for the resources available... but I sure would like to see an ongoing discussion of search and also of content development so that next time there is an opportunity, Instructure will have a clearer, more specific set of user-defined goals and priorities when they do decide to allocate some resources to content development features, including search.

I might even continue to harbor a small hope for the most important content for my classes: student-created content... 

Yeah, I just had to get that in at the end. 🙂

Community Advocate
Community Advocate

Awesome reply. I will do it justice later on in much more detail, but thanks so much for taking the time to reply and share some (slightly worrying?!) content history...!

Community Member

Thanks and thanks also for connecting at Twitter; yay! The great thing about Instructure is that I think the discussions and conversations really DO matter, and eventually we are going to get to even better places because of the value they place on that shared process. 🙂

Pinging‌ for insight on the big picture of content... are we missing something? I am guessing this must have come up in discussions about analytics.


Thanks for the update and all of the work and the team at Instructure put into making Canvas awesome!!


laurakgibbs, I know this might not be exactly what you are looking for with search, but have you looked into Atomic Search by Atomic Jolt? I believe they will give it to one course for free for life. May be worth the investment with this being deferred. 

Community Member

Hi‌, and thanks for the suggestion! I actually don't do content in Canvas (for the reasons mentioned above), but I am really intrigued that this option exists. Ingenious! I don't recall anybody mentioning Atomic Search during discussions at Khaki. Looks like for my school an Atomic subscription would run $15K per year. Personally, I think that would be well worth it... I wonder if they are getting a lot of customers! Thank you again for the tip; totally new to me!

Learner II‌  I appreciate the update, and am looking forward to seeing how the improved permissions granularity and interface emerge.  I'm wondering if "6. Page View Mobile Activity" will affect any other mobile metrics, such as instructor interactions, access reports, or last access.  Is there a timeline that provides an estimate of when all mobile activity will be fully accounted for in all analytic reports and displays?  


Hi Peter,

This is Chris on Mitch's team helping with questions/clarifications. The objective of the "Page View Mobile Activity" is just as you describe - for all mobile activity to be fully accounted for in all analytic reports and displays. We'll be accomplishing this by storing the mobile activity as page views with all the other page view activity. That way, no matter how you get page view activity from Canvas, mobile will be included. This is actively being developed and we're still on track to deliver per the June 2018 timeline listed above. 

Surveyor II

The admin roll changed will be the biggest impact to Canvas users. 

Great job !

Community Member

laurakgibbs & ,

I wanted to second your comments about Search and Course Archiving.   On search, we became aware of Atomic Jolt's solution several months ago.  It is really cool and it works, including showing you only content you have access to see!  Here is a little video I made of our own content while we still had the trial.  I believe they just index your Canvas sites (we gave them an authorization token), no tagging needed.  However, they wanted roughly $50k per year for our size institution.   We said, we'll pass, Canvas is already building this for us and we wouldn't want to change the way the function works if the Canvas version is different than the Atomic Jolt version.   Needless to say, we too are disappointed that this isn't being pursued for completion by InstructureCon 2018.  Searchable digital content is the defacto standard now.   I'm not a programmer so I can't say how difficult this really is, but if a 3rd party already did it, why not buy the code, or try to replicate what they did?   This would really provide a competitive advantage to Canvas.   We know through various channels that Google is actively working on evolving Classroom into a full blown LMS.   They are the masters of search.  Why not stay ahead of that curve?  Google has already beaten Apple in the hardware space by taking 60% of market share (and growing) with Chromebooks, they are on their way to beat Microsoft O365 through competitive pricing and ease of use with GSuite and the LMS + integration standards such as OneRoster are the last bastion to seal up the education market.  

On course archiving, we are going into year 4 with Canvas.  4 years * 60,000 students * 8~ courses per student = 2M courses.   Most of these are no longer needed, however you can't just delete them as they contain student histories, content, rosters etc.  Our SIS is the official historical record, but Canvas contains all the detail of what happened in these courses.    Archiving so we can declutter, go about our work more efficiently by seeing just the current courses is desperately needed.  

Community Member

Here here!   If pages and modules could be differentiated like discussions and announcements have been (or will be), then we could basically have one canvas course per teacher instead of doing all the SIS / programming jujitsu we do in K12.  That plus search would get this to a new world class level.  


Course Archiving would probably also need the ability to retain student grades and submissions (discussions posts, quizzes, assignments).


I appreciate the mention of Atomic Search. Because this is a publicly available forum and you listed pricing for our product, I want to clarify in case anyone runs across this. Our current pricing for Atomic Search is $0.30/student/year for K-12 which for your school district, Howard County, MD (52,700) students, the price would be $15,810/yr, significantly less than the $50k/yr you put in your post. Perhaps we talked when we first released our product and were still trying to figure out the market price for our product. In higher ed, our pricing is $0.50/FTE/yr.

Community Member

Any ideas on when the SpeedGrader Screen Recorder might be released? I am most excited about this new feature!

Community Team
Community Team

rachelkrist No ETA shared at this point.

Community Member

Thanks for the update. Looking forward to Bonus item #3 Student Feedback View if that means you will be implementing the ideas outlined in 

Keep up the great work!

Community Member

Just wanted to let you all know students can now across mobile and web browser.

Community Member

Wow, awilliams‌, that is REALLY good news! At last! There a lot of people who are going to be VERY happy about this one!!!

Community Member

Is there any news or update on the status of Screen Recording in the SpeedGrader? Is it still being worked on?

Community Coach
Community Coach, I haven't heard anything about this for Speedgrader, but they did implement a screen recorder for Arc in May/June 2018 - 


Community Member

Interesting read about Arc. It however just looks like they have implemented a button that has you download a screen recording application on your computer. This still makes it so you have to upload the video content up to Canvas.

It looks like referenced some sort of SpeedGrader Screen Recorder that was suppose to be delivered sometime after the InstCon 2018. It would be soooo amazing if it was more than just the Arc solution. Having something integrated directly into the SpeedGrader like the web camera and audio option would make giving feedback so much better.


Community Coach
Community Coach

You don't have to upload the video back to Canvas with this application, it automatically uploads into the Arc library in your Canvas course. But yes, there's a one time download of an application.

Community Member

Sadly, our University did not purchase the Arc option for Canvas. However, from that article it still looks like it is recording the video on my computer and then asking me to upload it once the video is done.

"You can use the screen capture tool to record media for any length of time, depending on your computer's memory."

"To upload the screen capture, click the Done button [1]."

"To upload the screen capture, enter a title [1], description [2], and then click the Upload button [3]. You can also edit the capture [4], redo the capture [5], or cancel the capture [6]."

If it is recording the video on the computer it will still take processing time to upload it to the Arc library. I was hoping to cut out the processing time on my computer.

About the Author
I've done a lot of things over the years but the common thread has always been helping people. It is what I find most rewarding in this world and community is non-stop people helping each other.