But I think the real issue is that UX is more work than most developers expect, and especially to iterate over time.
UX mistakes (or imperfections) are visible for all to see, and correctness is sometimes subjective. Imagine if all of our users were able, qualified, and effectively required to evaluate source code! We do some lazy/misguided/opinionated stuff too, sometimes! :)
I didn't mean to dis Taiga above. It's thoroughly functional, if not quite right for the users I was testing with. I tried Taiga because I know the other options mostly suck, and my next-step-test is not definitively better!
I'm not sure, I still see this dismissiveness. Hell, open any HN thread about design and see how many will ask what designers even do all day, or talk about them as if they're "less than" developers.
I can proudly say that our team not only cares about design, but has put Design and User Experience at the center of the product development. We are designing an improved interface, with special accessible approach, and also having special care about the way space is used (information density like you said). Likewise, we know this is a critical point in terms of a project management tool.
I'm really looking forward to sharing more advances! For now, we are sharing some of them on Taiga community (https://community.taiga.io/), please feel free to post your thoughts, ideas, and feedback.
I'm a strong proponent of open source and highly usable tools for critical project infrastructure.
I really wanted Taiga to work for us, and I look forward to revisiting it in the new version.
- The "Save" icon is tiny and hard to find in some layouts (right hand gutter, near the top and far away from your linear text entry flow). It's a floppy disk! Navigating away does not warn you that you have unsaved edits.
- Issues can be converted to Stories after review. Great. But there is no indication that the Issue has been converted, and reporters can continue to edit Issues, and the changes are not synced to the new Story item. This was very confusing for some people, and created way more friction than could be reasoned away.
- Information density was OK with a small quantity of items/comments/etc, but too sparse for a confident view with large quantities.
We moved this project over to test GitHub Projects, which is an improvement but also lacking in some important (different) areas.
...
All of these tools have always sucked on some or several dimensions. It's probably impossible to not suck.
JIRA can be made to work, though I will never do so again.
Pivotal Tracker is solid but overwhelming for reporters/casual users, and requires custom code to integrate with GitHub (but API is decent).
Sometimes I miss Trac and Redmine, if only because anything that bothered me enough could be fixed in our self-hosted instance.