Rollbar vs TrackJS JavaScript Error Logging

This feature is quite an interesting technical feat, considering that it shouldn't really be possible to get error information form 3rd party scripts. See this and this for discussions.

Free plan allows for up to 5000 events/month, customizable rate limits, integrations (e.g. with Slack), multiple users, and 30 day retention.

Rollbar can be integrated with Slack to send notifications of various event to a set Slack channel based on customizable rules.

Stack trace with function calls and line numbers.

At least as of 2015-July-14. Many other prominent loggers are blocked, e.g. Loggly and LogEntries.

Rollbar allows sending any log messages to Rollbar not just exceptions enabling you to go back and check for missed exceptions.

You can easily send "deploy" events with curl, so that you know to which version the errors apply. Errors are grouped automatically by name/type, but if the re-occur after a deploy, they'll be shown again above the deploy in the timeline.

Asana - Create and resolve Asana tasks

Bitbucket Issues - Create and manage Bitbucket Issues

Campfire - Get alerts in your Campfire room

Flowdock - Get messages in your flow

GitHub Issues - Create and manage GitHub Issues

HipChat - Get alerts in your HipChat room

JIRA - Create, resolve, and reopen JIRA issues

PagerDuty - Trigger and resolve incidents in PagerDuty

Pivotal Tracker - Create and manage Pivotal Tracker stories

Slack - get alerts in a chatroom or group

Sprintly - Create and manage Sprintly items

Trello - Create and manage cards in Trello

VictorOps - Trigger and recover alerts in VictorOps

Webhook - Post JSON payloads to arbitrary URLs

Email - get notifications and daily summaries

Source map support is used to reference original source code in cases where the code is minified.

Rollbar allows ignoring specific exceptions simply by adding the error message to the ignoredMessages array.

Understanding the console, network, and user events that led up to an error makes it possible to recreate it in development.

Reports show what are the important errors to focus on fixing.

All messages sent into console are included with error messages for additional context and logging.

See when new errors are happening in Slack or Hipchat rooms.

TrackJS supports custom user IDs, which can be random strings that we, the customer can assign to a certain user. This helps a lot when handling a support request or a bug report from customers.

You can see how your site is doing every day with a summary of all the interesting activity and errors that should be checked into.

You will always get an engineer who can understand and answer your question right away whenever you need help.

Hooks in the code that let you customize just about anything, you can get the error object before it's sent to the server and do whatever you like with it.

Error messages and stacktraces are shown with snippets of the raw source at the location. This gives a ton of context to quickly find the error.

To ignore spurious errors.

TrackJs is very useful to catch console errors.

A snapshot of how you are doing, comparing page view volume to error volume.

No surprise bills caused by changes in usage.

Some general information like browser, operating system and viewport used by the client may help to determine source of trouble more precisely.