Client Update v0.9.19

Published at August 18, 2020 ·  3 min read

Share on:

This release includes a big update!

For a while I’ve gone back and forth about adding an HTTP inspector. I wasn’t sure how difficult it would be to write a decent simple inspector and how muc time and effort would be required. So I delayed this feature and thought it might not be necessary.

However, I’ve found it difficult to lure developers to Packetriot and I’ve had several customers and users ask for it since our launch.

Version v0.9.18 didn’t introduce any new functionality but it was a redesign and rewrite of the clients’ backend HTTP muxer. After adding features over a year it became a patchwork and began introducing bugs…

The redesign shrank the code base and made it more flexible for additions. After thinking about some more it was clear that I could implement an inspector with much less effort and code than I had estimated in the past.

Luckily it was true and I’m glad to introduce our HTTP Inspector!

HTTP Inspector

HTTP Inspector

HTTP Inspector

The inspector is currently invoked using the newly added inspect command. This command works identically to the http command. You can specify a destination host, port and also a path to serve static assets from. The inspect command will enable the local inspector web application and record all requests and response for your inspection :)

You can visit the inspector using the URL http://127.0.0.1:4000. If that port is taken it will increment from 4000.

Right now the only actions is replaying requests, which is as simple as selecting a request, which will highlight it, and then clicking the replay button.

More details and example usage scenarios are available in our updated docs.

Roadmap

We’re not done implementing the inspector but wanted to share a version of it that is quite useable and could help invite some feedback and requests from you. I’m planning to add a feature to export a request to a cURL command next.

In addition, I thought it would be helpful to enable saving requests to disk so that the inspector can be used offline and not require a connected tunnel. This could help developers and may let users on the Free tier stretch their bandwidth allotment.

Other Updates

We’ve updated our HTTP Basic Auth to respond back with the correct status code 401 and a the realm in the response header which will signal most browsers to prompt the user for a username and password.

This corrects the behavior and make it more useful.

Thanks!

I’m excited that this redesign and rewrite of our HTTP muxer made it easy (and quick) to write a useable HTTP inspector. I hope you enjoy this new feature, it was fun writing it and seeing it take shape.

It’s stable but I’m thinking of it as beta until more users begin to use it. Please send any feedback, issues or discovered bugs via email or let us know on Twitter.

Cheers!