After reading the thread I don't really have a whole lot to add. Here are a few thoughts anyway:
DOING IT RIGHT:
*Hard scientific results and transparency in sharing these results.
*Sharing F@H resources with researchers around the world.
*Very few technical issues in all the years and platforms I've used. 6.23 was literally set and forget on the boxes at my law office.
*Love the project details in v7. They should be standardized and expanded: A concise paragraph for the layman, followed with details for bio majors.
DOING IT:
*The heat from all these boxes is great in the winter, but killers in the summer. I take older boxes offline for good when summer starts.
*Oh lawdy do I want me a $upermicro quad G34 build...
*'lectricity costs be gettin' me down yo.
*It's weird how difficult it is to explain F@H to someone.
DOING IT ASKEW:
*v7 is a great improvement, but still a bit cryptic. Needs more explanations of acronyms, more checkboxes to toggle, and should eliminate the nonsense with flags. This software should be made such that my grandmother can look at it and understand what is going on. (I'd still have to install it for her though
)
*GPU lag issues in Windows. Not sure how much latitude PG has over this though, but I was surprised v7 was released to the front page without this issue resolved.
*Related to lag is the default option with 7.1.52 for SMP & GPU usage. If a noob experiences lag, they may simply delete the program instead of spending time to figure out it's a GPU issue and address it.
*Further related to noobs: These install options should all be clearly explained. Not everyone knows what SMP, GPU, or uniprocessor means. There are too many assumptions here on the part of the software.
*Real time visuals of progress on projects. Not just a progress bar, but perhaps a (simulated) 3D model of a structure we're helping to fold/build/understand. There should be indicators in the graphic of exactly where your particular WU is. There should be color coding so that we can see what has been processed, what is being worked on, and what is left to be done.
*GPU support in Linux. This may affect only a small number of folders, but these tend to be power users who contribute disportionately more than an average user.
*GPU support in OSX. The gains may be outweighed by the development effort on this one however.
*Would like more stats about the project in general.
*Everyone in this forum grasps the overarching issue of misfolding proteins. However, to the general public, protein is something you get in meat and a folded protein is a burrito. I feel the project would capture the imagination of the lay-user much better if it concentrated on scary-sounding disease names rather than the more nebulous-sounding 'protein folding'. A few years ago this was my tact when commandeering the computers at my law office. I told everyone the project was looking for a cure for Alzheimer's.
*what campbbri said a few pages back about trying to get hardware review sites such as Anandtech and Tom's Hardware to utilize a F@H benchmark utility sounds brilliant. I would really have liked something like this a few months ago when I was on the lookout for a new GPU.
PS
*What's up with the viewer anyway? I've been folding for about 8 years and have never seen anything other than the demo. I'm not much into pretty graphics, but it really engages a lot of people. Believe it or not I showed the demo to a lawyer at my office. He didn't care about anything else I showed him, but when he saw a 3D molecule moving around on the screen he said "Wow, they're really serious about this." I suspect this is not an uncommon reaction.
I consider the initial configuration to be a part of the installation.