Forum OpenACS Q&A: Re: Call for Volunteers: Redesign inside and out

Posted by Joe Cooper on
After reading this thread, I immediately had the urge to scream loudly, "why does everyone care what color the bike shed is?"

The OpenACS nuclear power plant is melting down and people are spending vast amounts of time proposing new (very bad) project names and logos. At least half of the packages are broken in 5.2.0 in small or large ways, the bug-tracker at has a seriously annoying bug that prevents users from closing their own bugs, bugs remaining unanswered for months or years is par for the course, documentation is wrong, following the tutorial results in non-working code.

The OpenACS logo is just fine. While it may be fun to imagine that a slick new image will make OpenACS vastly more popular, it's not the logo or project name that is keeping people away from OpenACS in droves.

Fixing the real problems does not require a new logo, a new website design, or even a project manager (despite it being a cool package, that I find very useful), and in fact, arguing about them merely distracts folks into thinking that they're doing something positive for the project when they're just arguing about the color of the shed.

What needs to be done for OpenACS boils down to (this is not new...I'm not the first to notice these things):

  • Documentation. Correct documentation, more specifically, since currently there's lots of documentation that hasn't been correct in years. A user that is introduced to OpenACS development with a tutorial that results in non-working code is not going to come away with a good impression of the project.
  • Bugs need to be brought under control. There are hundreds of bugs that have been open without response for months or years. Even patches to fix problems don't always result in closing of a bug. This is also very frustrating for new developers--enthusiasm for helping out dies very quickly when not only bug reports but bug fixes are ignored.
  • New development information needs to trickle up out of the forums. I learned of a slew of function deprecations in 5.2.0 only when Torben pointed me to a forum post from February of 2005 that mentioned them. A wiki documentation project might be the way to make this happen. Clearly the DocBook-based process is not working.