I tested before stable. But when there was the first possible Beta the design was basically already defined (Navigation, elements, etc.). The work for creating workflows, checking the needs of users, create personas, all the ux design phase would have been needed to be done long before. When I saw the interface for the first time, most strategic decisions were already made.
Then with the first or second update ComfortUpdate broke my test installation and it took me a while to get around and make a new clean install to be able to test again. And the end of the year is probably the busiest time of the year for us in market research, which means we need to finalize projects before the end of the year. And then Christmas down here falls together with the summer holidays and the 2 weeks around Christmas and New Year are the only ones where we can actually take off some time without any problems, as most clients are also off, so no request for quotes, etc. So the release candidates came actually at a very bad time for me to make any tests.
But, as said before. The actual problem is that once we got to see the new interface for the first time (demo installations from Louis) it was actually too late already to do some good research and creating a sustainable concept with different scenarios and persona (different profile of users that might use Limesurvey) in mind. I admit that at this time I could have made some suggestions for strategic changes. But for this i would have to do some research, speak with other users how they use the interface, see what user profiles are more common, define important workflows, see how we can best incorporate this recurring workflows into a good navigation, make wireframes, make sketches, as people about those suggestions, get back to create new wireframes based on suggestions, etc. And I didn't think that there would be any changes at that point, that this would be incorporated into the design, which was basically already defined somewhere by someone. I don't have insights into this process, but I from experience with clients I can imagine how this happened.
I think 2.5 would have been a good chance to really create a very good GUI for Limesurvey that looks good and works even better. Not that it doesn't work, but there are so many clitches that show that the GUI was designed by someone who usually doesn't create surveys or works with Limesurvey. And hey, we can't really blame the developer for it. He/she is usually not a designer, or a UX expert, or a everyday user of LS. What probably went wrong is that no one was asked. We have quite a few users here in the forum that use LS on a daily basis. Why not trying to get them on board? They might not be developers, but they are users.
And I don't take myself out of this critique. Seeing the outcome, I think I should have been more pro-active. But this is hard if you don't really know what is going on and is planned behind the scenes. And after the last experience, where I had offered help quite actively, I wasn't too sure if this help was evenwanted (went down the same path last time when a new GUI was planned). And of course, all this research and UX design work would have taken some time and delayed the release of LS 2.5.
Well, now we have to see what we can still fix with this GUI. I found a few flaws when I was working with it. Can't remember from the top of my head now, but mainly the very inconsitent navigation concept is annoying. Many options are missing where you would need them, it is not clear how to really get to certain places, etc. But those problems are very structural. I have to admit that I was very enthusiastic when I saw the new interface, because visually it is a good step ahead (even if not as modern and up to date as I would have liked). But then the everyday usage comes in and the flaws of the new design show.
Topic Author
.
