User: Password:
|
|
Subscribe / Log in / New account

Gräßlin: FLOSS after Prism: Privacy by Default

Gräßlin: FLOSS after Prism: Privacy by Default

Posted Aug 23, 2013 2:13 UTC (Fri) by elanthis (guest, #6227)
In reply to: Gräßlin: FLOSS after Prism: Privacy by Default by maxiaojun
Parent article: Gräßlin: FLOSS after Prism: Privacy by Default

> Why bug reporting requires registration and E-mail address?

Actually, this is a severe usability problem. Almost every time a user wants to file a bug they have to (a) find the bug reporting site, (b) find and click "sign up", (c) fill in three pages of info, (d) wait for an email that does not always show up instantly, (e) go to a confirmation or login page, (f) finally start filling in a rigid form designed by people who think every user is an experienced QA lead, and then (g) potentially get stuck receiving email updates for the next 5 years on a bug that is apparently too contentious or difficult to just fix.

The vast majority of people never even get to step (a). A very large number of people getting past that stop at (b)-(e). (g) and the pain of going through the prior steps trains more people to not even bother with (a). The system is broken.

I've been very happy with an open bug report form using Akismet and some other strategies to eliminate spam. The form has an _optional_ sign up field offering Google or Facebook or OpenID login. The bug report entry is just a big text field.

If I get a low-quality report with no contact email, I just hit the big easy red "Ignore" button in the admin-side. The workflow is incredibly simple. I get a number of bug reports I'm quite certain I never would have gotten otherwise. For larger projects, we have trainers QA teams who filter the bugs from users before passing them to devs, so low-quality reports never waste a minute of engineering time.

The best systems collect as much information as possible (I realize the irony given the nature of this thread) and then filter things down to relevant kernels. Even if 90% of the bugs received are too low-quality to spend time trying to fix, trends of what kinds of bugs received, categories (selected by the triage team) and keywords, and so on all help build up very real and useful data that doesn't exist when only the most diehard and desparate users are filing bugs.


(Log in to post comments)


Copyright © 2017, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds