Required for the Stanford challenge

Change tracking has been implemented by #257 and there is a count displayed to admins only.

Enhancements to this functionality are:

• forward and back links to navigate through versions
• which account made the change
• highlight which fields were changed and mouse over for previous field versions
• click to revert to a previous version (assumes all future versions would be deleted)

• token based
• expire after 7 days
• take user to a unsubscribe page but they are not actually logged in

Give participants the ability to opt out of email notifications at a per-challenge or per-post level

See bottom of https://www.crowdai.org/challenges/learning-how-to-walk

When users download files, we can optionally stream via browser-based torrents. WebTorrent requires no installation and S3 can produce files in torrents.

This could dramatically reduce our download traffic costs.

https://github.com/feross/webtorrent

Consider this also: https://github.com/TTLabs/EvaporateJS

both on Mac mail and iOS mail

Send an email notification to admins for any submissions.

Some file injection rules for containers have configuration fields, but are not implemented in the container builds. These need to be implemented at some point:

• Grading only
• Overwritable
• Datasets directory

Optionally enable challenge organisers to render Submissions Guidelines, Dataset description, etc on the Challenge Overview page

The current nested form is too complex now ... it should be replaced by several form objects. These can be coded with reform / cells.

https://github.com/apotonick/reform

https://github.com/apotonick/cells

Support for Syntax Highlighting in the forums will be really helpful.
For reference, you can have a look at my post on this thread : https://www.crowdai.org/topics/images-appear-to-be-corrupted/posts/new

I think it would be good if we can tag users like "@spMohanty" which points to a users profile (and also sends him a notification if he has opted for it) to make the long discussions more digestible.

Various deprecations to be updated before upgrading to Rails 5.1

DEPRECATION WARNING: `config.serve_static_files` is deprecated and will be removed in Rails 5.1.
2017-03-06T15:08:36.736467+00:00 app[worker.1]: Please use `config.public_file_server.enabled = true` instead.

DEPRECATION WARNING: `:nothing` option is deprecated and will be removed in Rails 5.1. Use `head` method to respond with empty response body. 

Mar 06 06:46:37 crowdai-prd app/web.2:  DEPRECATION WARNING: Passing an argument to force an association to reload is now deprecated and will be removed in Rails 5.1. Please call `reload` on the result collection proxy instead. (called from new at /app/app/controllers/posts_controller.rb:9) 

Mar 06 05:49:05 crowdai-prd app/worker.1:  DEPRECATION WARNING: alias_method_chain is deprecated. Please, use Module#prepend instead. From module, you can access the original method using super. (called from included at /app/vendor/bundle/ruby/2.3.0/bundler/gems/activeadmin-xls-a5a70018c626/lib/active_admin/xls/resource_controller_extension.rb:6) 

Mar 06 05:49:05 crowdai-prd app/worker.1:  DEPRECATION WARNING: In Rails 5.1, The `:subdomains` option of HSTS config will be treated as true if 

Mar 06 05:49:05 crowdai-prd app/worker.2:  DEPRECATION WARNING: `skip_filter` is deprecated and will be removed in Rails 5.1. Use skip_before_action, skip_after_action or skip_around_action instead. (called from <class:BaseController> at /app/app/controllers/api/base_controller.rb:4) 

Mar 06 05:49:05 crowdai-prd app/worker.2:  DEPRECATION WARNING: `skip_action_callback` is deprecated and will be removed in Rails 5.1. Please use skip_before_action, skip_after_action or skip_around_action instead. (called from <class:BaseController> at /app/app/controllers/api/base_controller.rb:4) 

Mar 06 05:49:05 crowdai-prd app/worker.2:  DEPRECATION WARNING: before_filter is deprecated and will be removed in Rails 5.1. Use before_action instead. (called from <class:BaseController> at /app/app/controllers/api/base_controller.rb:5) 

I get SSL error message on both crowdAI.org and openfood.ch

Ideally this link shouldnt throw such detailed errors in production mode : https://www.crowdai.org/challenges/learning-how-to-walk/leaderboards/video_modal?participant=yieldthought&score=875.386657&submission_id=171

Challenges are currently manually ended by an administrator.

Implement a background process or on-page trigger that automatically ends challenges when the end datetime has been reached.

Challenges should be sorted with the news on top (sort by start date) - at the moment the new challenges are at the bottom of the list:

Verify emails are correctly being sent to the users as per email preferences.

Re message received about a host account

I've logged into the Host account but when I press "crowdAI" in the navigation I get the following error message: NoMethodError in Organizers#show

Getting this error:

• challenge should be ending automatically (not working)
• send an email to all participants that a new challenge has started and a challenge has ended (but we can still accept submissions).

• simplify timeline (move it into challenges)
• follow general in-place editing model of the tutorials
• move all header fields + status into header section
• add private challenge checkbox to header section
• create a new admin level for profs
• admin only features marked consistently with an icon in collapsable areas
• number and name of downloaders visible to admin directly on downloads page
• downloads license editable directly on downloads page
• evaluation rules embedded into challenge edit of evaluation view
• hide docker configuration completely for now
• cohort admin only view: allows invitation of participants to private challenges and also shows activity summary: downloads, submissions, forum posts. Build as partials and reuse on participant home page
• add view as participant / view as admin (edit) toggle for admins

I think it would really be convenient for participants if the grading of their submission is automatically triggered upon submission. Given how much work goes into a particular submission, its a pain to wait for a manual trigger :'(

There have been two instances where there has been a problem with the submission file, and the error message is 'ant_id'.

What does this error message mean ... can we provide some more detail to participants?

Email preferences can be accessed either from within the participant profile, when logged in, or else by following the "Unsubscribe" link at the bottom of emails.

Following the unsubscribe link is seeming to allow changing of email preferences, but in fact the settings are not modified when the participant next logs in.

Refer this discussion posting: https://www.crowdai.org/topics/observation-vector/posts/new

In my case, they end up in the "Promotions" section, while I do get a lot of emails which are actually promotions but they still end up in my Primary Inbox.
Would be a pity if people dont even get to notice our newletters, especially when its so low volume.

The submission made emails, and the forum notifications do end up in my primary inbox though.

• S3 uploads and image cropping
• Challenge progress bar

File uploads with S3 / React / Rails

http://blog.littleblimp.com/post/119230396893/direct-uploads-to-s3-with-react-rails-and

How is this happening?

http://www.rorvswild.com/applications/152/errors/1820990

Currently the user 'Online now' indicator is based on a DB update, and causes many additional SQLs. This is inefficient when the site is scaled.

This should be replaced by an ActionCable solution

Dependent on #4

As you can see in this image, the buttons are not visually distinguishable from the rest of the text in the forum posts.
I think it should look more like the red button (that I created by simply modifying the markup to add btn btn-primary as classes) to make it more intuitive to use.

Click on leaderboard entry and then see all entries for the participant for this challenge, including animations.

Create a new engaging landing page for crowdAI

Potentially three boxes:

• about crowdAI
• box for participants
• box for hosts

A Jupyter notebook will be able to be added anywhere the user can currently add markdown, ie:

forum posts
challenge description (admins)
knowledge base
Gitlab integration is not in scope.

https://github.com/jupyter/nbconvert

Tests that are passing locally fail on the CI.

/home/travis/.rvm/rubies/ruby-2.3.1/bin/ruby -I/home/travis/build/crowdAI/crowdai/vendor/bundle/ruby/2.3.0/gems/rspec-core-3.5.4/lib:/home/travis/build/crowdAI/crowdai/vendor/bundle/ruby/2.3.0/gems/rspec-support-3.5.0/lib /home/travis/build/crowdAI/crowdai/vendor/bundle/ruby/2.3.0/gems/rspec-core-3.5.4/exe/rspec --pattern spec/\*\*\{,/\*/\*\*\}/\*_spec.rb
Coverage report generated for RSpec to /home/travis/build/crowdAI/crowdai/coverage. 659 / 1016 LOC (64.86%) covered.
/home/travis/build/crowdAI/crowdai/app/models/dataset_file_download.rb:8:in `<class:DatasetFileDownload>': uninitialized constant DatasetFileDownload::Resolv (NameError)
	from /home/travis/build/crowdAI/crowdai/app/models/dataset_file_download.rb:1:in `<top (required)>'
	from /home/travis/build/crowdAI/crowdai/vendor/bundle/ruby/2.3.0/gems/activesupport-5.0.0.1/lib/active_support/dependencies/interlock.rb:12:in `block in loading'

Strange, as the code is it failing on is part of the Ruby standard lib

  validates :ip_address,
            presence: true,
            uniqueness: false,
            format: { with: Regexp.union(Resolv::IPv4::Regex, Resolv::IPv6::Regex) }

Friendly ID is not updated when the challenge title is changed. Add additional code to facilitate this.

http://stackoverflow.com/a/36236475/993592