Skip to content

Latest commit

 

History

History
703 lines (398 loc) · 66.1 KB

CHANGELOG.md

File metadata and controls

703 lines (398 loc) · 66.1 KB
Raw

Changelog

6.13.0 - ??

Improvements

  • We've moved away from the "master" terminology to "main" for the default code branch, and updated links for that.

6.12.1 - 2022-08-11

Bugs

  • "Show Only Rated" feature in Quepid busted for ES. Highlighting on _id after a terms match in ES causes a index out of bounds exception. Also, the API method in explainOther needed to have the right case, might make a constants file at some point. o19s/splainer-search#109 by @worleydl fixed the issue in Splainer. Bumped in Quepid via o19s#547 by @epugh.

6.12.0 - 2022-08-09

Features

  • Search box (to find queries). o19s#506 and o19s#512 by @epugh fixes o19s#493 by @theaaronposc. Thanks Aaron!

  • You can now generate a new Case through the thor command line tools, useful when setting up lots of cases or as part of a script. o19s#533 by @epugh.

  • Surface how many unrated documents you have for a query as an alert bubble. o19s#532 by @epugh.

  • The Frog Pond Report tells you about the state of your ratings. Understand the distribution of queries that need ratings and decide if you need to "hop to it" ;-). o19s#526 by @epugh.

  • Create missing queries when importing ratings. o19s#509 by @epugh.

Improvements

  • Gray out Queries with no ratings. Previously they showed up as Red with a 0.0 score, the same as if everything had bad ratings! o19s#516 by @worleydl and suggested by @renekrie.

  • Keycloak Version 17 changed the default url to not have the /auth in the namespaces. o19s#536 by @epugh fixes o19s#528. We now test Quepid with Keycloak 18.

  • Average Precision Scorer AP@10 wasn't implemented following the most common definition of AP. o19s#535 by @david-fisher fixes this! Thanks David!

  • Taking a Snapshot is intensive process that makes a unique query to the search engine. o19s#513 by @epugh provides nicer error message.

  • Remove some more unused AngularJS code. o19s#529 by @epugh.

  • 'Information need' import doesn't works for an empty case. o19s#508 by @epugh fixes o19s#507 by @atarora.

  • Bump to latest Rails to pick up CVE fixes, Ruby 2.7.6 recommended by Heroku, Bundler 2.3.18, plus general updates to gem dependencies.

Bugs

  • The detailed export of the case doesn't exports the 0 documents matching queries. o19s#504 by @atarora fixes o19s#501.

  • Move the logic for the default ES and Solr server definitions from server AND client side to just client side. o19s#505 by @epugh fixes o19s#503.

  • Limited encoding renders snapshots unusable. o19s#510 by @epugh fixes o19s#499 by @KennyLindahl. Thanks Kenny for the analysis!

  • Fixed missing query fields from cloning a case. o19s#520 by @epugh

6.11.0 - 2022-04-23

Features

  • Documenting the explicit Information Need of the user is a core part of running human judgements, and historically people used the Notes field for this. Now we have a Information Need attribute for every Query, and you can import/export that data via CSV, as well as edit it in the UI. Mouse over the query to see exactly what the user is trying to do! o19s#480 by @epugh.

  • The details of a specific query exposed by the Explain Query modal is in a Human readable format, but isn't in a JSON compliant format. Now we sort the keys alphabetically, which helps the user dig through the details, and have a Copy as JSON button so you can use other tools to look at the JSON information. o19s#497 by @epugh fixes o19s#492.

  • Delete All Queries for a Case was requested by @negyesi in o19s#475. At the case level added a Delete button that brings up a modal that lets you delete the case, archive the case, or delete all the queries. o19s#486 by @epugh.

  • Frustrated that your custom API that looks like Solr to Quepid has to deal with JSONP callbacks? Now you don't, as we have introduced the ability to swap from the default JSONP to GET api methods when you define your Solr end point. This is going to simplify the work that folks have to do. o19s#495 by @epugh. For more information refer to the docs at https://github.com/o19s/quepid/blob/main/docs/endpoints_solr.md#solr-endpoints-structure.

  • Do you have a relative image url in your search engine that you want to show via thumb: or image:? You can now provide some JSON in the field specification to provide that data: id,title:title,{"name": "image_url", "type":"image", "prefix": "https://www.example.org/images"}, description. This feature is experimental as we learn more about how to have richer field specifications, and may change! o19s#487 by @epugh.

Improvements

  • The link to view a single Solr document specifies XML, which I realize is a very OK boomer approach these days! o19s/splainer-search#107 by @begomeister fixes o19s#494 by swapping to JSON format instead.

  • github.com/teleborder/FileSaver.js repo disappeared, so migrate to an actual published node module file-saver instead. o19s#498 by @epugh.

  • We had a mish-mash of query and query_text as headers in our CSV files, so we've standardized on query to make export/import simpler. o19s#489 by @epugh.

  • We discovered that Redis was ONLY used in Quepid to support sending events to Google Analytics! If your QUEPID_GA isn't specified, then Redis is no longer required. Thanks to @rbednarzCBI for updating the production docker settings to remove Redis in o19s#488.

  • preserve search engine type when swapping protocol, don't default to Solr! o19s#459 by @epugh.

Bugs

  • The various CSV upload and export features didn't really handle , in the text properly, so we gave that a good going over. o19s#482 by @atarora fixed by o19s#483 and o19s#485 by @epugh

  • Tracking the history of your tries, we had a limit in the database that prevented you from having lots of tries. o19s#478 by @KennyLindahl and o19s#470 by @korlowski. Fixed by o19s#479 by @epugh. Thanks @KennyLindahl for the additional testing work you did to help me fix this!

  • Fixed the ability to peek at the next page of result when you have less then 10 results left. Thanks @anegyesi for submitting o19s#473 and fixed in o19s#496.

  • Noticed the case level graph isn't useful? PR o19s#491 by @worleydl, Save scores less often, fixes stretching of graph, fixes this.

  • @thePanz opened a ticket a year ago (o19s#330) about ratings getting progressively slower in the UI, and a huge amount of updates to the backend. Thanks to @worleydl and o19s#490 this should be fixed.

6.10.1 - 2022-02-28

Bugs

  • Updated Splainer-Search to https://github.com/o19s/splainer-search/releases/tag/v2.13.0 to enhance snapshots with ES, and make our parsing of ES.

  • When Quepid switches between HTTP and HTTPS protocols, the flash message "Reloaded on HTTP protocol for Quepid app to match search engine URL" causes some CSS layout issues which led to the "Run my Queries" button to slide off the page! Darn you CSS. First spotted by @nathancday, while giving a demo of course! Fixed by o19s#463 by @epugh.

  • When reloading from HTTP to HTTPS Elasticsearch during the case wizard, we lose the default query params for ES, we end up with the default Solr query params! o19s#457 by @epugh and @wrigleyDan. Fixed by o19s#464 by @epugh.

  • Build was failing for Apple Silicon M1 users! o19s#430 by @llermaly and fixed by @wrigleyDan via o19s#460.

  • Missing doc" error when comparing with snapshot in Elastic when using non standard Id, o19s#466 by @KennyLindahl fixed by splainer-search 2.13.0.

6.10.0 - 2022-01-03

We've taken a rag and polished up the Find and Rate Missing Documents modal box in this release. We've also completely redone how we attempted to handle http and https connections from Quepid to http and https search engines that was first introduced into 6.9.0 based on what we've learned in the real world.

Improvements

  • Reworked the handing of http and https connections to Quepid when connecting to http and https search engines. Added in Development mode a nginx based proxy to make http://localhost/ and https://localhost/ work, which is also how you would deploy Quepid in a Production set up. Still works fine on http://localhost:3000 if you only use http based search engine connection. Big thanks to @jzonthemtn for o19s#451 which fixes o19s#444 and o19s#438.

  • The Find and Rate Missing Documents query interface assumes you use the Lucene query syntax. Turns out we have a highlighter built in, so enable that for Lucene syntax. o19s#453 by @epugh.

  • Writing your own scorer? The modal popup window is rather cramped, so let's give the editor room to breathe by making them larger! o19s#452 by @epugh.

  • In the early days of Quepid, looking up a single Solr doc would trigger a facet.field on every field you display in Quepid. For some Solr's, this can turn a quick lookup for a single document to a 30 second or more ordeal because of unexpected faceting! Thanks to @jeffryedvm for opening o19s#442. Fixed by o19s#456 by @epugh.

Bugs

  • On the Find and Rate Missing Documents screen the ability to show just the rated documents had an issue that you had to click the button twice, making you think it was broken. o19s#454 and o19s#423 by @epugh are fixed in o19s#455 by @epugh.

6.9.1 - 2021-10-27

Improvements

  • Splainer-search 2.11.0 handles nulls the same as empty arrays ({ }) in Solr output. Nulls might happen when a Proxy is used. o19s#440 by @epugh fixes issue discovered by @mkr in o19s/splainer-search#97.

  • Node 15 is not a LTS version, which leads to a warning message! o19s#437 by @epugh bumps this.

6.9.0 - 2021-10-27

Features

  • Chrome 94 forces you to be on https URL to access HTTPS secured search engines. Quepid now checks to see whether it is running on the same protocol, HTTPS or HTTP that the search engine is running on, and if not, prompts you to reload the app on the same protocol as the search engine. This required updating the messenging in the Case Wizard and the Tune Relevance screens. We also simplified the logic in the home_controller.rb around bootstrapping a Case and Try. Thanks to @tboeghk for pairing with @epugh on how to solve this problem. o19s#431 by @epugh fixes o19s#426 by @DmitryKey and o19s#432 by @bbirrell.

Improvements

  • Remove remnants of "SecureApp" angular1 app that was refactored a while ago into regular Rails app for login screens. o19s#434 by @epugh fixes o19s#433.

  • Logic to enable Google Analytics tracking was crufty and caused classloader issues in development mode. o19s#419 by @epugh fixes o19s#418.

6.8.0 - 2021-09-28

Features

  • Understand the evolution of your various Tries by visualizing the history as a tree data structure. Understand where you went back to an earlier Try and branched off in new directions. o19s#412 by @epugh fixes o19s#412 by @DmitryKey.

  • Deploying Quepid in a cloud native way requires having a healthcheck end point to confirm Quepid is available and happy. o19s#404 by @epugh fixes o19s#404 by @mkr. Thanks to @xaviersanchez for the inspiration to do this.

  • Now sort queries by most recently modified. o19s#334 by @DmitryKey and @epugh fixes o19s#307.

  • Excited to share the improvements you made to your Case? You can now drop your link to the Case into Slack or other messaging tools, and it will unfurl, i.e show the Case Name and the most recent Try number. You will still need to log into Quepid and have access to that Case. o19s#417 by @epugh fixes o19s#416.

Improvements

  • The messaging around how many documents are being scored, i.e @10 or @40 is now driven by the variable k defined in various Scorers. Commit 73a378f9bae777cb504618cdc3361bfec52ff82f by @epugh fixes o19s#413 by @DmitryKey.

  • We audited our Docker builds, and _shrank the images from 2.42GB to 1.44GB! Bumping Node version to 15, be smarter about what folders we copy into our production image, and some general fixups. o19s#409 by @dworley and @epugh.

  • The logic to check if Google Analytics is enabled via ENV variable is different then the normal Quepid ENV handling. o19s#419 by @epugh cleaned this up, and solves a class loading issue in the DEV mode Quepid, increasing developer happiness!

  • Trying to find a doc that SHOULD match and be rated, but can't find it? Using a global scorer like NDCG and curious what all the ratings are? We have reworked the Explain Missing Documents UI into a Find and Rate Missing Documents that lets you see all your ratings associated with a query, and then search and find other documents that are missing. Preview video here https://share.descript.com/view/IkxWkjSzzKF. o19s#407 by @epugh fixes o19s#406 by @mkr.

6.7.0 - 2021-08-26

Features

  • Elasticsearch fields computed using script query now work! Back in February @CGamesPlay contributed a PR (o19s/splainer-search#90) to splainer-search to support this, and now it automagically works in Quepid! Thanks @DmitryKey for opening o19s#327 and supporting this feature.

  • Elasticsearch templates can now be used in Quepid! o19s#400 by @epugh exposes the underlying splainer-search functionality in Quepid. Thanks @danielibars for opening o19s#274 and supporting this feature.

  • Curious what parameters are being used by Solr to construct the query? We're now adding a echoParams=all to queries and surfacing them in the Explain Query modal at the query level in Quepid. Currently a Solr only feature. o19s#397 by @epugh with inspiration from @mkr.

Improvements

  • Continue cleanup of JavaScript, eliminated last remnants of "unit test" style scorers, and renamed customScorer to just scorer in front end app.

6.6.1 - 2021-08-11

  • In testing the Keycloak oAuth integration via the Chorus project, identified that the name of the Client that is passed to Keycloak was tied to the KEYCLOAK_REALM env setting. Changed this to be hardcoded to quepid so it can be different then the Realm name (i.e the Realm is Chorus in the Chorus project, and the Client is quepid). Commmit de9e0e645ddb4cb02644680ed8af456a167023dd by @epugh.

6.6.0 - 2021-08-09

Features

  • OpenID Support! Historically you had to create your own account on Quepid, but that is another barrier to entry, and something else to keep track of. Inspired by the issue Add OAuth Providers by @gregoryduckworth o19s#188, o19s#280 and o19s#389 by @epugh implements the first two providers, Google and Keycloak.

  • Explain Query Parsing! We often want to know what the search engine is doing to our query. Now you can click the "Explain Query" popup and see what is happening to your query and the various filter queries. This works with both Solr and Elasticsearch. https://twitter.com/dep4b/status/1420480501710827520. o19s#391 by @epugh.

  • Frequently type deftype instead of defType in Solr? I found an old, but broken feature that provided a popup warning on common Solr query syntax errors, which has now been restored. o19s#394 by @epugh fixes o19s#390.

Improvements

  • Quepid is running on Rails 6! Rails 6 was released ~26 months ago, and represents the future of Rails. I'm excited that this push initiated by @DmitryKey is going to bring us some great new features like: better developer experience with Webpack for JavaScript, ActionText to handle better text formatting of notes and messages about Cases and Queries, ActionCable which will let us notify all the users of the same case that data has changed when it changes. o19s#381 by @DmitryKey with assist from @epugh.

  • Make our ActiveRecord modeling for ownership the same. Teams have an owner, Scorers have an owner, but Cases have a user. Now we have case.owner relationship. o19s#359 by @epugh.

  • The nav bar link for Scorers is labeled "Custom Scorers", and takes you to an page called /advanced. Both of the names are rather offputting. Creating Scorers is a normal common thing to do in Quepid, so lets label it Scorers. o19s#392 by @epugh fixes o19s#351.

  • We had, as someone put it "dinosaur" age CSS tags for mozilla and webkit browsers that triggered lots of browser warnings in the console. o19s#380 by @epugh cleaned up many of them. It has also prompted some thinking about migrating to Bootstrap 5 for the non core pages. If anyone has some CSS skills, please contact @epugh, to help refine the header menu in Bootstrap 5 CSS.

Bugs

6.5.5 - 2021-06-30

Features

  • You can now tag a field with translate:, as in translate:content and you will get an icon to pop the text open in a Google Translate in a new browser window. o19s#386 by @epugh.

Improvements

  • You can now export fields that have the formatting modifiers thumb and image using the detail format. Also improved the handling of the General and Detail export from the Case Listing page. o19s#383 by @epugh fixes o19s#382. Thanks @DmitryKey for the improvement assist and spotting the Case Listing export issue.

  • Admin user can now reset a users password with a new password. o19s#385 by @epugh to fix issue identified by @michaelcizmar. Thanks Michael!

  • Trying to communicate about HTTPS better when you set up a case. o19s#384 by @epugh inspired by o19s#279 by @arafalov.

6.5.4 - 2021-06-16

Features

  • You can now Clone an existing Scorer!. Super useful when you want to start with one of the communal scorers that only go to 10 deep, like P@10, and you want it to be P@40. Introduced as part of o19s#379 by @epugh.

Improvements

  • Enable TLS for Redis 6 on Sidekiq on Heroku. o19s#370 by @michaelcizmar fixes o19s#271 by @michaelcizmar. Thanks Michael for your first PR to Quepid!

  • During the Case Wizard setup, allow space delimited list of fields, like title overview to be pasted in, just like comma delimited lists. Thanks @peterdm for the assist on the Regex! o19s#378 by @epugh.

  • Revamped the layout of the Scorer creation and editing screens to be visually cleaner. Retired the old Fibonnaci scale, and renamed Default to Detail, as well as Short to Graded scales. Introduced Binary as a new, default scale. @DmitryKey and @epugh paired on this during Quepid Qommunity Qoding hour, resulting in o19s#379.

  • Shrink production Docker image of Quepid from 2.19GB to 2.17GB by not installing development and test Gems. Commit 426d2677f6c4a8380971ddc1b0faa42a53a48879 by @epugh.

Bugs

  • Preserve the chosen scorer when cloning a case, we were defaulting to the users default scorer. o19s#372 by @epugh fixes o19s#273. Thanks @binarymax and @nathancday for spotting this!

  • Ensure that CSV export always has all the columns expected. Queries that aren't fully rated and have blank values didn't put out complete set of , separators. o19s#371 by @epugh fixes o19s#364 by @sstults. Spotted while working with Quaerite.

  • If you don't have a Terms & Conditions url, then your users don't have an agreed_time, so the activity pulse chart in the admin errors out. o19s#373 by @epugh sorts this out.

  • In the dawn of Quepid, we had agreed_time but not a boolean agreed. A user let me know he couldn't update his email address (from an account in 2015!) because he didn't have agreed=true set. o19s#374 by @epugh cleans up some old user data.

  • Fixed the labeling behavior for a Scorer. @DmitryKey and @epugh paired on this during Quepid Qommunity Qoding hour, resulting in o19s#379, fixing o19s#376.

6.5.3 - 2021-06-01

Bugs

  • In 6.5.2 we introduced the image:image_path to show larger images, however there is odd wrapping if you don't have many fields listed for a doc. o19s#365 by @epugh fixes this.

  • If you didn't have email set up, then sending invites wouldn't work. o19s#369 by @epugh adds better handling for when email hasn't been set up for Quepid.

6.5.2 - 2021-05-27

Improvements

  • We have a new Favicon, but didn't use it everywhere, so o19s#363 by @epugh displays the new favicon everywhere.

Features

  • Rating documents where the image carries most of the information? The existing thumbnail:image_path limits you to a 75 pixel wide image. We now have image:image_path which displays a 200 pixel wide image. This is an alternative to the media:image_path approach, as this shows up next to your fields versus inline with your fields. o19s#362 by @epugh fixes o19s#360.

6.5.1 - 2021-05-25

Bugs

  • Now that we've had the team invitations feature out for a while, we've found places where we make assumptions that aren't correct. This fixes the Admin panel's user page, and makes on the Teams page the messaging around pending invites. o19s#339 by @epugh.

  • We discovered that nDCG@10 score for products with a hitCount below 10 (K) was always 0. o19s#336 by @rudibatt fixes this. Thanks @rudibatt for this contribution!

  • Passing in quote delimited parameters to bin/docker r wasn't preserving or quotes or escaping the spaces, causing the thor user:create [email protected] "Eric Pugh" "mysuperstrongpassword" example to fail. o19s#340 by @epugh fixes o19s#338.

  • Thanks to Liz Novak at VIN for discovering that a non lowercase email address like [email protected] wouldn't be considered a valid email address, so you couldn't send a invite to join Quepid. o19s#342 fixed by o19s#346 by @epugh with assist from @eribeiro.

Improvements

  • Uglifier for JavaScript appears to not work with ES6, and so using Terser instead. o19s#329 by @epugh fixes this.

  • Add to the Team page the ability to copy the team invitation url to send via other means then Quepid's own email setup. This is helpful if you don't have email set up, or if you want to share invites with team members via chat tools like Slack. Also fixed the user flow of registering after someone has sent out an invite for someone. o19s#337 by @epugh fixes o19s#335 and o19s#326.

  • We've added linting for our Dockerfiles via Deepsource.io. o19s#343 by @epugh fixes o19s#341.

  • Our Docker images are getting bloated by including log files from dev and test, leading to almost a 3 GB image. Adding a .dockerignore has got us back to around 2.1 GB image. o19s#358 by @epugh fixes o19s#353.

  • Remove the remnants of the query test (unit test style custom scorers) and the default scorer table that are still in the source code. o19s#357 by @epugh fixes o19s#349.

  • Sometimes you want to create a Case even when you can't validate the connection. Introduce a option to bypass the validation. o19s#356 by @epugh.

  • When we export basic rating data there is an extra LF character at the end of the CSV file. This was causing issues when importing ratings into the Quaerite project (tballison/quaerite#14). o19s#355 by @epugh with assist from @sstults fixes o19s#354 by @sstults.

  • Quepid 6.5.0 started failing on Kubernetes clusters due to the need to be able to write to the filesystem. Bootsnap dependency upgrade avoids this. o19s#344 by @epugh. Thanks to @LiuCao0614 and @tonomonic for the assist on this.

  • Revamp of how all the permissions for doing actions related to Teams, Scorers, and Cases to use the Permissions data we aleady have for a user. Lots of cleanups on those functions. Finally dealt with the left over need to have a Case per User, or the UI blew up. Now you get a nice message, and the UI doesn't blow up. o19s#347 by @epugh fixes o19s#352, o19s#331, o19s#278.

6.5.0 - 2021-04-22

favicon

I can sum up this release of Quepid as either the Hey friend, come join me in improving search release or the so much technical debt has been paid down release.

For those of you upgrading your Docker based install, there are two new ENV variables you need to add to your docker-compose.yml file:

- RAILS_LOG_TO_STDOUT=true
- RAILS_SERVE_STATIC_FILES=true

Hey friend, come join me in improving search

We have added features that make it easier for you to invite your colleagues to come join your team and start rating documents. We have reworked the initial case creation wizard to be smarter about popping up only the first time you decide to create your own case, instead of having it jump into the user flow when you join an existing team with existing cases. This should make it easier to bring folks interested only in rating docs into Quepid. As part of this, we've also added support for deleting individual Cases and User accounts. So if you invite someone, and change your mind, you can clean up after yourself.

so much technical debt has been paid down

We are finally off Rails 4.2 and are now on the latest Rails 5 release, and we have a shiny new favicon to go with it! We've updated all of our frontend dependencies to the latest versions possible, congruent with the core app still being based on Angular1. The lift from Rails 4 to Rails 5 was a massive effort, and a huge round of thanks goes out to everyone who reported bugs. We've shrunk our list of dependencies by 10% over the v6.4.1 release, and 29% compared to v6.0.1 release, and we're set up now to look at new front end technologies!

We also did some housecleaning by ripping out some features that hadn't seen adoption by users, including "soft delete" of queries and the "unit test" style of custom scorers.

Features

  • Invite your friends to join your team on Quepid! You can now send an email with an invite code to join Quepid and your specific team. o19s#259 by @epugh. o19s#305 by @epugh introduced better support for tracking of if someone had gone through the create case wizard or not, extending this feature.

  • Add support for sending emails via SMTP, or use Postmark, or don't send emails. o19s#276 by @gabauer fixes o19s#275.

  • Let a user (or an Administrator) delete their account from Quepid, handling their cases, scorers, and team memberships. o19s#315 by @epugh fixes o19s#311.

  • You can now Delete a case altogether! Historically we had an Archive function, so that you could restore an old case. However, if you are like me, you create lots and lots of throwaway cases, so this allows you to clean up your Quepid setup. This PR also fixed some data modeling issues, and the problem of sometimes have a Try of zero, instead of the default first try of One! We also always include the Try number even if you have named the Try, since that is a big deal. o19s#288 by @epugh fixes o19s#250. Thanks @DmitryKey for help QA'ing this code.

Improvements

  • We now have a favicon for Quepid! o19s#317 by @DmitryKey.

  • Upgrade to Rails 5 and Ruby 2.7.2! We have been stuck on Rails 4.2 for years, and this unlocks a lot of new capabilities. o19s#256 by @epugh with inspiration from @worleydl.

  • Renamed "Export Case" and "Import Ratings" to just "Export" and "Import". o19s#270 by @abhishekChoudhary93 fixes o19s#195.

  • Pay down some tech debt by removing our use of Coffee Script in Quepid. o19s#283 by @epugh.

  • Build time on CircleCI dropped from ~135 seconds down to ~105 seconds by removing sass gem in favour of sassc. o19s#285 by @epugh.

  • We have simplified our queries that involve cases and users that are part of teams by making the simplifying assumption that all owners of teams are also members, which turns out to be true in practice. We also have audited our use of .includes to load child objects (avoiding the N+1 problem), and used more .preload to try and deal with running out of temp space in our database. o19s#286 by @epugh.

  • Turns out we had a ERD diagram all along, but it was hidden. Now you can see it on our Data Mapping page, plus we have how to recreate it documented and integrated. o19s#287 by @epugh.

  • Remove obscure quepidIf.js file that doesn't seem to do anything. o19s#293 by @worleydl.

  • The Export All Cases feature only supported the old "Detail" format, and none of the other export formats, like LTR, Basic, or RRE. Plus the filtering options of "All, Owned, Shared" was pretty simplistic. It wasn't being used, and it added some complexity and performance issues to List All Cases page. So we removed it. o19s#295 by @epugh fixes o19s#294.

  • Unit Test style custom scorers let you run a unit test that asserted specific things about specific docs at specific ranks in your search results. This logic however was always convoluted, and only 15 people since 2019 have used it, and I suspect by accident ;-) We want better ways of handling this type of function, so removing this to pay down some tech debt, simplify the database architecture, and open the door to new approach. o19s#296 by @epugh fixes o19s#290. o19s#313 also dealt with this.

  • We have removed the Soft Delete for Queries to simplify how Quepid works. If you delete a query in Quepid it is now fully deleted from the database! This is a nice bit of paying down tech debt. Huge thanks to @DmitryKey for testing this PR. o19s#299 by @epugh fixes o19s#298 by @DmitryKey.

  • The /case url is one we frequently end up with, and it generates a 404. Thanks @DmitryKey for suggesting we have it route to the case listing page /cases instead. o19s#304 by @epugh.

  • The logic around when we popped open the "Create a Case" wizard for a brand new user was somewhat split between the backend home_controller.rb and the front end. It also made anyone who was invited to a team just for rating purposes go through the Create a Case Wizard on their first login, which was awkward. So, converted the concept of a "first_login" for the wizard to just a boolean "completed_case_wizard", and now it it checked if you have NO cases, and popped up, or if you click "Add a Case". o19s#305 by @epugh fixes o19s#281.

  • Upgraded to the latest version of Angular 1, 1.8.2, and many of the other front end dependencies. o19s#308 and o19s#324 by @epugh and o19s#320 by @worleydl deals with this.

  • You can now override the from email address from the default [email protected] to your own email address. o19s#322 by @slawmac. Thanks @slawmac for this improvement!

Bugs

  • You can export a rating that has no actual rating value chosen! o19s#266 by @epugh fixes o19s#265.

  • Fixing typo in defining aria-hidden html tag. o19s#269 by @abhishekChoudhary93 fixes o19s#268.

  • Rails 5 Upgrade Hangover: fixed strong parameters validation issues, fixed wonky left outer joins on cases showing duplicate cases (and killed performance), eliminated the user_case_finder.rb class in favour of simpler scopes. o19s#292 by @epugh w/ assist from @worleydl. o19s#314 by @worleydl is related. o19s#321 by @janhoy fixed by o19s#323 by @epugh.

  • Thanks to @DmitryKey for spotting that we were not tracking scorers for cases (bug introduced when we added the Show Only Ratings feature). o19s#303 by @epugh and @worleydl fixes both o19s#306 AND o19s#289. A twofer!

  • Thanks to @DmitryKey for spotting that we were not properly storing the Elasticsearch choice. o19s#310 by @epugh fixes o19s#309.

  • There is a known issue where expanding/collapsing queries on a larger case, the UI can become completely locked in Chrome, forcing the user to close out the tab. Thanks to some next level debugging by @LiuCao0614 we have a workaround that disables the JQuery based drag and drop sorting feature that appears to avoid this bug. This is not a fix! This is a workaround! Set QUERY_LIST_SORTABLE=false in your environment to disable the sorting of queries feature. o19s#272 tracks this ongoing issue.

6.4.1 - 2021-01-14

In the 6.4.0 release, the correct splainer-search NPM package was missed in the production Docker image.

This release resolves that oversight. Thanks @LiuCao0614 for testing the upgrade process and reporting the issue.

Improvements

  • Small housekeeping update for the hosted app version running on Heroku. o19s#267 by @dworley.

  • Clean up last scoring run details on the Cases Listing page. Now better iconography to call out cases that have queries that need rating. o19s#261 by @epugh fixes o19s#219.

Bugs

  • Don't export ratings that don't have a rating set for RRE and Basic export formats. o19s#266 by @epugh fixes o19s#265.

6.4.0 - 2020-12-18

Only Rated toggle is a powerful new feature added to Quepid, our first relevancy centric feature in a long while.

In certain relevancy cases, you may only care about the ordering of a subset of documents within the entire result set. One particular use case may be in e-commerce where you want certain products to be boosted and others to be demoted in the ranking algorithm. While this was doable in previous versions of Quepid, it could get difficult to manage the list of rated documents, especially if the list went many pages deep into the results.

To help solve this problem we worked with Kobler to introduce several new features to Quepid. We strived to make it easier to work with your corpus of rated documents in the main case view, inside of Explain Missing Documents and within the scorers themselves.

This feature was developed by @worleydl in o19s#229, with an initial POC by @janhoy in o19s#220, and resolves issue o19s#163. Much thanks to @janhoy and the folks at Kobler for this feature.

Features

  • Added "Show Only Rated" toggle to the main searchResult listing
  • Explain Missing Documents modal now defaults to showing all of your rated documents when you first open it up
  • Added eachRatedDoc function to ScorerFactory to iterate over rated documents in scoring
  • Provided refreshRatedDocs(k) in ScorerFactory for loading a custom amount of rated documents up to k

Improvements

  • Migrated scoring to an asynchronous pipeline
  • Fixed stuck "Updating queries" message when creating a new case

6.3.2 - 2020-12-08

Features

  • Round trip export and import LTR ranking files! o19s#204 by @epugh fixes o19s#162 by @DmitryKey.

  • Disable web signups by setting SIGNUP_ENABLED=false. o19s#238 by @tonomonic.

Improvements

  • Easy in-place editing of case name and the try name to encourage using those features! Microinteration FTW! o19s#242 by @epugh.

  • Demonstrate richness of queries in Quepid when you use the TMDB dataset. o19s#236 by @epugh fixes o19s#224.

  • Update Javascript references to application/javascript. Pay down some tech debt! o19s#223 by @epugh

  • Simplify handling doc id's that have periods or slashes in then, and avoid base64 issues by passing that in the JSON payload. o19s#233 by @epugh fixes o19s#228 and o19s#221.

  • Some Solr collections need manual setup of the wt=json mimetype. Add better text message for users. o19s#235 by @epugh fixes o19s#178.

  • The CSV export format has a CSV injection vulnerability that is now fixed. o19s#245 by @nicholaskwan fixes o19s#231.

  • The Javascript based scorers could be used maliciously, so we've added an environment variable COMMUNAL_SCORERS_ONLY that controls if only Admins can create communal scorers for use by users. o19s#246 by @jacobgraves fixes o19s#232.

  • Don't show the Sharing icon for communal scorers, since they are implicitly shared globally! o19s#251 by @epugh fixes o19s#247.

  • Allow the export and import cycle of ratings using both CSV and RRE formats to include queries with no rated documents. Don't create a partial rating on import where we only have a doc id but no rating. o19s#252 by @epugh fixes o19s#244 by @LiuCao0614

  • Make it easier for scrapers and automated test tools to log into Quepid by adding some id and class tags to the login forms. o19s#257 by @epugh.

Bugs

  • Discovered that the migrations from communal scorers being @5 to @10 didn't always run cleanly. Commits 94dd23990422901082d79b121c1ca86a76907dc3, 8317b543530cc387d5cb89b4942acea5da57ce23, and 19b046485db530162c213a593e5b2e9df8fbbf07 to deal with this.

  • Discovered that DELETE of ratings didn't work, and had to work around that. Commit 153047cd4b75d626695f5fc38832f6202eed9007.

  • Missing authorization check for Team Owner. o19s#230 by @jacobgraves fixes o19s#230 by @testerTester0123456789.

  • Can't rename a case on the Teams page. o19s#240 by @epugh fixes o19s#213

  • Fixed scoring of AP@10 and NDCG@10 when you have fewer then 10 results. o19s#253 by @nathancday fixes o19s#225 by @epugh.

6.3.1.2 - 2020-09-16

  • Silly cut'n'paste error that should have been caught with more testing before the 6.3.1.1 release, not the day after. Fixed in commit 2e322b337cc62895847df0ed95ba6a68683dad5f by @epugh.

6.3.1.1 - 2020-09-15

  • Default communal scorer was set to AP@5, however in release 6.3.1 we renamed it to AP@10, so doing a quick release. Fixed in commit 182f14d8a667759cdda559fd0ed5e063167b6914 and ad29ad199dcd25231d364e2ca95d2c09cac195ac by @epugh.
  • Also found some issues with the seeds.db names being used. Commit 26edccfa407119b46b6f4316f44f34d0e3e87a9f by @epugh.

6.3.1 - 2020-09-14

Features

Improvements

  • When we rolled out classic graded scorers (CG, DCT, NDCG) we used a 5 point scale, that allows you to pick a "no choice" middle option. However industry standard (for example RRE) is 4 point scale, so lets use 0,1,2,3 as our choices. New Best Practices document https://github.com/o19s/quepid/wiki/Judgement-Rating-Best-Practices by @binarymax is availble to help educate you. o19s#206 by @epugh fixes o19s#205.

  • Closely related to the scale change, all the classic scorers used a @5 depth of scoring, however the other industry standard in rating is to score ten deep documents, @10, so o19s#209 by @epugh makes this change as well.

  • Encourage use of profile pics by adding help text in Profile page that they come from Gravatar.com. o19s#202 by @epugh fixes o19s#201.

  • Use the same header navigation bar in both the main application and the Admin screens. o19s#203 by @epugh refactors this logic.

  • Added bin/docker c option to jump into the Rails Console during development, and cleaned up the other options. o19s#200 by @epugh.

Bugs

  • Running bin/setup_docker more than once always had hiccups on dropping MySQL, and needed a docker-compose -v, so just make that part of the script. o19s#208 by @epugh fixes this.

  • Making HTTP links clickable wasn't working in some cases. o19s#211 by @e-budur fixes o19s#183.

6.3.0 - 2020-09-01

Features

  • We want to export unrated query/doc pairs, which is only supported via exporting a previously created Snapshot. Add support for exporting a Snapshot to the Basic format on the Export screen. o19s#191 by @epugh fixes o19s#185.

Improvements

  • When importing queries, you can have it clear all existing queries. This feature only worked for CSV files, not other formats like RRE. o19s#193 by @epugh solves this.

  • You can now import queries using the Import Ratings screen. Most of this was already supported, however now the modal dialog has better instructions to the user, and nicer validation of CSV formatting. o19s#192 by @epugh solves this.

  • When sharing Cases or Scorers, the lookup for what Teams you are part of created a really large JSON response (@epugh would get back a 1.4 MB payload!). We put this API response on a diet! o19s#197 by @epugh solves this.

Bugs

6.2.2 - 2020-07-09

Improvements

  • When exporting for RRE, we need the ES or Solr index name. Extract this from the url for the most recent try and save a step! o19s#167 by @epugh fixes o19s#159.
  • If you link to either a case that doesn't exist, or you don't have permission for, or a try that doesn't exist for a case, then provide messaging back in the UI! Let's share some Quepid Cases! 933ed257198ebe21ff86b7e35573d3172cc2e593, 99ac27c1f8698ed726580a4c46eaf6810a4372d2, and 37b95b89fa848b0af2bae3d5a9541141e5d80d62 by @epugh to main branch fixes o19s#158.
  • getCaseByNo only used in tests. o19s#173 by @epugh removes code.
  • There was a partially working attempt at a result grid view instead of list view. It wasn't rendering in the UI, and we want to have a more general solution in the future, so removing the code to simplify Quepid. o19s#174 by @epugh fixes o19s#171.

Bugs

  • When using the case wizard, you couldn't cut'n'paste in a long list of fields like overview_en, overview_idioms as they became a single tag. o19s#166 by @epugh fixes o19s#165.
  • A : in the case name was converted to a when exporting a case. o19s#169 by @epugh fixes o19s#168.
  • The implementation of autosaving your notes per query didn't work well. Going back to an explicit save button. o19s#170 by @epugh fixes o19s#164. Thanks @DmitryKey for the issue!
  • Remove warning in server log on parameters. o19s#182 by @epugh fixes o19s#180.
  • Community member spotted that doc id's with blah-http-blah triggered base64 encoding. Simplified handling docs with an id with a . or / in them. o19s#179 by @worleydl and @epugh fixes o19s#175.

6.2.1 - 2020-06-18

Thanks to some feedback from the community, we figured out that the SQL script for migrating data from the DefaultScorer table to the Scorers table (and being tagged as communal) didn't run reliably. We've pulled it out as db/release_6_2_0_merge_default_scorer_into_scorer.sql for folks to look at.

  • o19s#157 identified the issue and fixed by @worleydl.
  • Commits a1fc942d32e3d524836492f745735ce4ec4972ff and 029dd0cafe8caa492095c9483617b623a6a4e437 and 1a3c997f59b144a1cbffa59a04c67cb3e051b32d cover the migration fixes.

6.2.0 - 2020-06-11

This release of Quepid finally introduces the classical search geek scorers like NDCG, DCG, Average Precision, and friends by default. We've also fixed a large number of bugs introduced by changes in Firefox and Chrome browser security models that happened in the first part of 2020, as well as recent updates to Elasticsearch and Solr.

The burgeoning suite of open source tools for relevancy all require judgement lists. We revamped Quepid to handle exporting and importing from your favorite tools like Quaerite and RRE, as well as a Learning to Rank formatted export. You can see this in action in Chorus, an ecommerce search focused stack that includes Quepid.

Ease of development has long been an issue with Quepid. In this release we've worked on developer happiness by auditing all of the dependencies we include. We're removed quite a few Ruby gems and Node packages that weren't used, and upgraded the rest (replacing PhantomJS with Puppeteer for headless testing). yarn.lock dropped from 6706 lines to 1525 and Gemfile.lock from 448 to 330! This allowed us to finally move to ECMAScript 6 and Node version 10. Lastly we have cut the production Quepid Docker image from almost 3 GB down to 2.3 GB.

Upgrade Notes

Follow the steps outlined at https://github.com/o19s/quepid/wiki/Installation-Guide#updating-quepid.

You will need to run some database migrations to update the database and insert the new classical scorers (NDCG, AP, etc).

There is a sql migration script in ./db/release_6_2_0_merge_default_scorer_into_scorer.sql that you might want to look if you have been using Quepid locally for a while. This script is primarily meant to support updating http://app.quepid.com database for the 6.2.0 schema.

QUEPID_DEFAULT_SCORER is a new environment variable specifying the name of the default scorer for users when they create a new case. QUEPID_DEFAULT_SCORER=AP@10 is what app.quepid.com uses.

Features

  • Quepid Default Scorer Admin screen is buried in UI. Add it to the dropdown menu. o19s#96 by @epugh.
  • Support Basic Auth in Quepid with Elasticsearch. o19s#114 and o19s/splainer-search#79 by @CGamesPlay fixes o19s#109.
  • When installing Quepid, use thor user:create to create an administrator! o19s#112 by @epugh fixes o19s#107.
  • Enhanced export ratings function that follows the standard query,docid,rating supports full export/import lifecycle and easier integration with other tools. o19s#128 by @epugh fixes o19s#126.
  • Export ratings in RRE and LTR file formats. Import ratings from RRE Judgement JSON file. o19s#137 and o19s#139 and o19s#152 by @epugh fixes o19s#133
  • Classical scorers like AP, DCG, and NDCG are now shipping by default. All new users who sign up on Quepid will start with AP@5. This effort is still in progress, for example right now the scorers are named AP@5, DCG@5, NDCG@5 because they are hard coded to only look at the first five ranks. We also need to work on the Q Score Graph's vertical axis handling. Huge effort in o19s#100 by @nathancday and @epugh to get this done. This change also removes the individual query level scorer being assignable, which was a bad idea (o19s#132) and restores the ability to write a custom unit test for a query in Javascript (o19s#120) which is an interesting idea! Thanks to @janhoy for being the impetuous to get this work done!

Improvements

  • Community member reported the default demo TMDB Elasticsearch server on port 9200 had no data. Having seem random "drive by vandalism" of our demo Solr and ES servers that run on default ports, we want to move away. Yes, security by obscurity. o19s#103 by @epugh changes default used in Wizard to port 8985 and 9206. o19s#125 and o19s#104 migrate existing URLs to the new ones.
  • Increase Developer Happiness by speeding up the bin/setup_docker script by caching RubyGems. 50% speed up! o19s#105 by @nathancday.
  • Remove the concept of "communal" i.e shared with everyone scorers. This feature has been in Quepid for years, yet even @softwaredoug didn't remember it! o19s#99 by @epugh fixes o19s#98.
  • In the beginning of Quepid, before there was a Search Relevancy community, there was just one, slightly janky scorer, that wasn't like any of the standard search geek scorers (looking at you NDCG) that we use today. We are building towards supporting many scorers, so it's time to remove the DEFAULTS definition of a scorer, and the related scope in ActiveRecord. o19s#97 by @epugh.
  • Solr 8.2 tightened up the security profile for accessing it that we depend on via JSONP to have Quepid work. Now the wizard provides you the command to run on your Solr if it can't connect. o19s#95 by @epugh fixes o19s#92.
  • Using the default tmdb dataset? Demonstrate the thumb:poster_path feature. o19s#94 by @epugh fixed o19s#72.
  • Rename user.username in database to user.email since that is what we use. Clean up API. o19s#113 by @epugh fixes o19s#111. You will need to run database migration for this release!
  • Audited code base to prune dependencies and remove unused code. o19s#121, o19s#119, o19s#118, o19s#116 all worked towards this goal.
  • Turns out Quepid has a Curate interface. It isn't supported, so in the interest of reducing our codebase, removing it. Points the way for the future however! o19s#123.
  • Deal with sameSite cookie warnings from Firefox. o19s#131 by @epugh fixes o19s#130.
  • curator variables (i.e knobs and dials) that aren't used in query cause weird UI. o19s#135 by @epugh fixes o19s#64.
  • Saving two annotations in a row doesn't work, you need to rescore per annotation. Fixed in o19s#136 by @epugh.
  • Users can opt out of community marketing emails. Add a /admin/users.csv export to make keeping track of that easier. Thanks @flaxsearch for the suggestion.
  • Inconsistent use of X icon in modal popups is frustrating. o19s#148 and o19s#149 by @worleydl fixes o19s#146 and o19s#145.
  • Swap the sorting of tries in the history tab to the newest first, going back in time. o19s#151 by @epugh fixes o19s#143 by @renekrie.
  • Be clearer that MySQL is the only supported database platform. o19s#156 by @epugh fixes o19s#155 by @janhoy.

Bugs

  • Wizard Autocomplete Didn't work well with Keyboard. Autocomplete suggestion had to be clicked with a Mouse. o19s#94 by @epugh fixes this by upgrading package.
  • Multivalued and nest JSON fields didn't display well, you would get [object Object] instead. Now we display arrays and Json properly. o19s#117 by @CGamesPlay fixes o19s#52.
  • fixed highlighting throwing an error on Solr date fields by using hl.method=unified in Splainer-Search v2.5.9. o19s#84 created by @janhoy.
  • fixed fields with a . like foo.bar failing to be rendered in UI in Splainer-Search v2.5.9. o19s#106 created by @rjurney.
  • When a try number is not specified for a case, automatically redirect to the most recent try. o19s#122 by @epugh fixes o19s#110 by @binarymax.
  • Case shared via Team with me has NaN for the Try. o19s#134 by @epugh fixes o19s#83.
  • Not able to export cases from a Team page. Commit quepid@021bda8627463986f1ff04ffb1c290195c428c5b by @epugh fixed this.
  • Race condition in laying out the main Quepid Screen and the Tune Relevance screen. o19s#147 by @worleydl fixes o19s#144 by @epugh.
  • Using the "peek" function to page to deeper search results would lose your escapeQueries choice. o19s/splainer-search#87 by @jorgelbg fies this. Bump to splainer search 2.6.0 to include the fix.

6.1.1 - 2020-07-03

  • Community member reported race condition in standing up Rails and MySQL and issues with PhantomJS install in the developer docker-compose.yml and Dockerfile.dev setups. o19s#75 by @epugh fixes o19s#76 and o19s#73.
  • Add .dockerignore file to prevent unrelated changes from breaking Docker layer cache fixes by @TheSench o19s#80
  • Fix issue where you couldn't clone a case without including the full history. o19s#89 by @worleydl fixes o19s#37 Thanks @janhoy for submitting this bug.
  • Fixed display of notes for query disappearing after collapse and then expand of query window. o19s#88 by @dworley fixes o19s#87
  • NDCG@10 doesn't include documents that are rated via Explain Other, it only looks at the documents returned by the search engine. We want to be able to have NDCG look globally at all rated documents. o19s#90 by @nathancday and @worleydl fixes o19s#78. Note: We currently have the NDCG@10 scorer in app.quepid.com, however it hasn't been backported to the Docker image or the dev setup in Quepid. See o19s#91.

6.1.0 - 2020-02-01

This release changes the database schema, so you will need to run docker-compose run --rm app bin/rake db:migrate if you have an existing Quepid.

  • Cloning cases now carries any magic variables and their values along to the new case o19s#55 by @epugh fixes o19s#37
  • Only treat field content that begins with "http" as a link. o19s#35 by @pfries fixes o19s#34
  • Update Elasticsearch logo. o19s#38 by @moshebla
  • Remove sqlite from gem, no longer used. o19s#41 by @epugh fixes o19s#40
  • Better look and UI experience for the dev panel. o19s#39 by @moshebla
  • Show or don't show the T&C's link based on the Quepid configuration. o19s#42 by @epugh fixes o19s#44 by @flaxsearch.
  • Add Query button activates in response to query text entered. o19s#43 by @moshebla
  • Prevent duplicate queries when using bulk query importer. o19s#49 by @epugh fixes o19s#48
  • Production oriented Docker Compose setup for Quepid is ready! Install guide at https://github.com/o19s/quepid/wiki/Installation-Guide. o19s#36 by @epugh was based on original work in o19s#33 by @synhershko. Thanks!
  • Update the sample TMDB dataset for Elasticsearch to support thumb:poster_path in the field listing. o19s#53 by @epugh
  • Collapse query results well from bottom (in addition to the top) o19s#28 by @epugh fixes o19s#20 by @peterdm. Thanks @ychaker for reviewing PR.
  • Explain Other on ES 6 and 7 Broken. o19s/splainer-search#74 by @worleydl fixes o19s#25.
  • Support for embeds of audio/image/video via media:. o19s#62 by @worleydl fixes o19s#56 by @flaxsearch
  • Autocomplete in Case Setup Wizard properly handles media: and thumb: prefixes. Nice refactoring for more prefixes. o19s#46 by @moshebla with refactor by @worleydl.
  • Post the onboarding wizard you would often see your queries being stalled in loading. Only workaround was to reload the Javascript app. o19s#69 by @worleydl fixes o19s#66
  • app.quepid.com needs to get explicit consent from users to receive emails related to Quepid and related features. Introducing a new customize_quepid.rb file in initializers to start supporting more customizations of Quepid. o19s#68 by @worleydl.
  • Only show cookie acceptance popup if COOKIES_URL is set. o19s#71 by @epugh fixes o19s#70 by @epugh.

6.0.2 - 2019-11-29

6.0.1 - 2019-11-05

  • Update to splainer-search 2.5.0.
  • Fix setting the default case for exporting when you first load Quepid. o19s#15 by @worleydl fixes o19s#12
  • Allow a document to be rated to have an id with a period in it like mydoc.pdf. o19s#6 by @epugh fixes o19s#5
  • Default query when setting up a new case works across Elasticsearch versions 5, 6, and 7. o19s#3 by epugh fixes o19s#2
  • Updates to CircleCI and update Ruby 2.5.7 by @ychaker

6.0.0 - 2019-07-25

  • Removes everything related to payments, and makes Quepid "free"
  • Adds support for using a CORS proxy for Solr instances that are not configured to allow connections from Quepid
  • Replaces use of Vagrant in development in favor of Docker
  • Adds support for GDPR
  • Adds support for deploying Quepid on Heroku
  • Upgrades Ruby version
  • Upgrades Angular from 1.4.x to 1.7.x
  • A bunch of bug fixes

5.0.2

  • bugfix-888 Fixes #886: Locks version number for CSV module.

5.0.1

  • bugfix: ES5 displaying attributes - fields are not automatically stored so stored_fields will not return anything, instead get the data from the _source field

5.0.0

  • feature: q-score - Added graph for score history
  • feature: annotations - User can now take a snapshot of the score with an annotation to describe that point in time
  • bugfix-879-880-heatmap-safari Fixes heatmap in Safari

4.2.0

  • add support for ES 5+ with stored_fields vs fields

4.1.1

  • bugfix: Prevent re-escaping % when it's part of an escape char

4.1.0

  • bugfix-873-change-try-name Fixes renaming try and reloading name in list
  • bugfix-872-duplicate-try Fixes #872: Duplicating a try
  • bugfix-852-redraw-results-with-snapshot-open Fixes updating results in open query while comparing with a snapshot
  • bugfix-848-close-modal-after-deleting-try Closes modal after deleting try
  • enhancement-862-document-qa-workflow Adding documentation for QA flow
  • feature-863-explain-disabled-features-for-trial-user Enable popups that explain why trial user cannot perform action
  • enhancement-document-magic-variables Adds documentation for Magic Variables
  • bugfix-844-add-non-existing-user-to-enterprise Fixes feedback when adding non existent user to enterprise
  • bugfix-857-case-insensitive-emails Fixes logging in with emails (making them case insensitive)
  • bugfix-845-add-team-member Fixes feedback when adding a team member
  • bugfix-843-user-autocomplete Removing entire list of users from autocomplete when adding team member
  • feature-use-n-docs-in-scoring Allow users to set how many results to display by default instead of the default 10 results
  • feature-company-name-in-profile Adds ability to specify company name in user profile
  • feature-120-clone-case Adds ability to clone a case
  • bug-796-mass-import-ratings Fixes bug in mass import of ratings
  • bugfix-767-viewing-snapshots Fixes bug while views snapshots
  • bugfix-big-snapshot Fixes bug while viewing large snapshots