The drawing board

Well, I figured out by myself that the blogging process is not certified to run under crunch time. I secretly hoped that I’d be the one to pull off the magnificent feat of being able to keep this blog updated no matter what would happen, but alas no such thing.

Still, this evening’s post is about the time I worked on “sexying up” a WYSIWYG editor. It was a noble cause – one I referred to previously when I talked about the importance of starting. I should have known better in the first place, but no – I dared to question and propose a change to the way this editor had always looked and felt like.

For weeks all went well, team morale was great and the development was being done at a steady rate. Ideas came, prototypes were made, tests were done, you’re familiar with the routine. When the project started to get close to the deadline, things suddenly had to take an unexpected turn. Due to technical limitations, I was facing a dillema. On one side we had a slick looking interface, which was hampered by a few quirks: the result of countless weeks of hard-work, powered by coffee, cookies and Dilbert strips… On the other  one there was a good enough interface, which had some nuisances that date so far back that are probably documented in the Old Testament – and as such, people grew up with them and have adjusted themselves to them. It was like using windows 95: after experiencing 10 BSODs, people started taking it for a feature.

Still, some users had already been using this slick version, and they were quite fond of the change. The problem was that they were not immediately aware that some of the quirks they were experience were not present in the previous version. They had been introduced by this change – and were not mendable easily (believe me, we’ve tried). They still didn’t care, they didn’t want to see! Only after some sense-slapping were they able to admit that those problems were real.

I can’t blame them, honest: it was like we had presented them to supermodels, had put them to date them for a couple of weeks, and now we were telling them “sorry, you’ll be better off making up with the ex-girlfriends/boyfriends you ditched for these, ’cause these are history”. It was the ogre’s choice, and I wasn’t proud of it. At first, the feeling that pervade my spirit was that countless weeks of work had lead to nothing and we were back to the drawing board – Wyle E. Coyote style.

The truth was that these sexier changes hampered the operations in the editor. Though it was more immediate to the user in many cases, in some scenarios it would take the user a bit more of effort to do normal things such as placing things on screen. This would ultimately lead to frustration, since this didn’t happen in the previous version of the editor, and these few occurences actually accounted for 40% of the regular user’s use cases.

The ultimate decision was to tag the code as it was right now – for future reference, -  revert the look and feel to the original one, keep the fixes made along the way and now start fixing things that could improve user satisfaction with this codebase. It was not what we wanted, but it was probably the best solution at that stage.

What I learned first hand from this is that no matter how similar the end result is to the starting point, the whole trip makes up for it. There will no longer be careless remarks such as “Ha, how come no one has thought of this first?” and “Why didn’t they do it like this then?”. Now I’ll be the living incarnation of the “Why”s of this whole project - especially the “Why”s that justify the “Don’t”s. And that is saying a lot.

Also, I was hugely surprised to see users sticking up for the sexier – but ultimately hampered – solution, and sad to go back to the old not-so-hampered-yet-sexy-as-a-goat solution. From a developer’s point of view – and myself as an user (albeit an over-informed one) – that still doesn’t make sense, and I sincerely hope it was just an initial reaction. I’m planning on keeping an eye on this particular topic, and getting feedback from users after this reverted-with-slight-improvements version is released.

Cheers.

Advertisement

Explore posts in the same categories: Usability