Open stoft opened 6 years ago
Hi @stoft
Thanks for your feedback. I have a couple of points to mention here.
I'm marking this as a feature request.
@kamalaknn thank you!
"Denoting timeouts in ms is an industry-wide adopted practice. In most of the cases, the latency of APIs is less than a second. That forces users to enter timeouts in decimal notation for most of the time."
This is under general settings. Do developers really set it to below 1s since it will affect all of their requests? :) Also ref. this article where it's explicitly mentioned in the troubleshooting as being a possible issue: https://www.getpostman.com/docs/v6/postman/sending_api_requests/troubleshooting_api_requests
But I'm not so well versed in the Runner e.g. so there may be other implications of that setting that I'm unaware of.
It is super confusing that postman just gives a whitelabel error page when its own internal timeout (often set too low by user) is the only thing that caused the request to fail.
At least for this situation, have postman report that it canceled the call as it exceeded the user-specified timeout.
"Denoting timeouts in ms is an industry-wide adopted practice. In most of the cases, the latency of APIs is less than a second. That forces users to enter timeouts in decimal notation for most of the time."
It is true that API-designers often strive for sub-second response times. However, it is rare that API testers are unwilling to wait more than 1 second for a response, and extremely unlikely that they would appreciate receiving a generic whitelabel error page from postman whenever their timeout setting is exceeded.
I would be interested in hearing some use cases where for this timeout maximum, people prefer to specify in ms. As stoft suggested - is it more important for bulk / automated running?
After losing some time and banging my head in the wall, trying to reinstall etc. to solve what I thought was a bug, possibly corrupted collection issue, network issue etc. I have the following usability feedback:
When Postman times out due to hitting the timeout limit it would be nice if it could be more explicit that it timed out. Admittedly I could be better at reading the fine print under "Why this might have happened", but considering it's Postman itself enforcing the limit it should know this and communicate it to the user.
Also the timeout setting default is 'ms' which is easy to miss. Possibly change it to 's' with optional decimal values? Or type it out to 'milliseconds'.
Thank you for an excellent, invaluable tool!