More on Drupal and usability

Over at Drupal's "Development" mailing list there is a discussion titled : "How many modules is too many?"

The question hinges around the issue of performance : The more modules the system has, the slower it works because it has to load all those bits and pieces of code from all over the place.

The issue seems like only affecting coders, that is, the people who develop modules and/or work on the core of Drupal. Yet this is not the case at all.

At Drupal's "Support" mailing list Edward Peter's points us to the work around he had to implement in order to solve a simple usability issue he had with SimpleNews.

What follows is the email I was going to send out to Edward but have opted to publish on my blog instead (because it breaks the rule of being more than 3 paragraphs long). I definitely want to hear your thoughts on this :

Hi Edward,

I am forwarding this to the development list. There is a thread about module bloat and how this affects performance and I think your post here reflects the problem that a lot of people face when using contributed modules ( http://lists.drupal.org/archives/development/2007-11/msg00465.html ).

I believe your post is really timely to that conversation. What follows is just a "out-loud" musing on both discussions, so here it goes ...

As an expert user of Drupal yet not a coder, I often times find myself frustrated because there seems to be no "Golden rules of Usability" for module developers. I wish there were a way for developers to tap into the usability needs and insights of expert users like you in order to look ahead to potential implementations of their modules and take into consideration the kind of "module bloat" that could impact a Drupal site (as per Khalid's assessment here : http://2bits.com/articles/server-indigestion-the-drupal-contributed-modu... )

So, to take your example with SimpleNews, all you needed was to have the ability to add 2 new fields to the block for collecting 2 more pieces of data. Yet, in order to go around that limitation, you have now 4 extra modules : Profile, Captcha, Logintoboggan and Poormanscron. (NOTE TO READERS : Original post is at http://drupalproject.iofc.org/node/172 ).

Let's just say that well, since profile is in the default distro, it doesn't count. OK, that's 3 more modules. Then again, you'd probably would have anyways used poormanscron. That leaves us with 2 extra modules CAPTCHA and LOGINTOBOGGAN. And for one more justification, let's just say CAPTCHA is there for anti-spamming purposes. We are left with the one extra module.

Yet, from just wanting two extra fields on a block, you've now gone to having added 5 extra instances to the END-USER EXPERIENCE : They have to go to a registration page (1) with the added fields (2) and captcha (3) that once filled out will take them to a landing page (4) with a link back to their original place (5).

That's not counting and addressing what all this running around and tinkering did to YOUR tech/expert/webdev user experience How many hours did it take you before you found the solution? I'm going to take a gander of, give or take, 20 hours (you did try out CiviCRM after all!)

You found a great and workable yet not-so-simple solution to SimpleNews' limitations. Yet it is the kind of "module bloat" that Khalid says is bad for Drupal sites. I don't see though how you could have avoided the solution you found. You're workaround is a great example of the way expert/webdev users of Drupal will use the modules when there is no simple 1-2 way ahead.

There's a lot of us who use the system and added modules precisely just like you, lending to performance issues that do not reflect the benchmarks tested on the basic distribution. Yet, I don't know if there is an easy answer to this other than maybe Drupal.org may want to think of a way to engage more expert users in the software development process exactly to take into consideration potential extensibility issues like the one you encountered and resolved.

Maybe I am wrong, maybe I am not. Yet, given that usability is the #1 complaint I hear from expert-users and potential end-users who would like to use Drupal but opt not to, I feel like yours is precisely the kind of example we need to discuss.

Cheers,
liza