The things you list aren't requirements for a BOINC project at all. It would be nice if WCG was a little more like a standard BOINC project, and there are a few parts of BOINC that WCG have changed or omitted - but not the ones you complain of.

Your download complaint in particular is totally unfounded. Downloading is done by BOINC, and by BOINC alone. The only possible difference is that WCG uses a BOINC feature that many other BOINC projects ignore - namely SSL encryption of downloads.

Duplicate usernames are generally a bad idea under any circumstances. No idea why Berkeley chose that route. And the forum? Well, you're here, aren't you?

Thanks uplinger for putting zoomable feature to the list. I'm sure more users will appreciate it.

I'm familiar with Work Unit, Results, Replitation, Quarum Daily Quota and other BOINC related stuff (that is mostly invisible on WCG as it is limited BOINC project).
My point was - wow, only couple results for testing for such a large project? They will be done within couple of hours. I guess I have been wrong in my estimation there.
Usually - on BOINC project - there is a quick feedback among testers and the team...or at least I'm used to it on BOINC. If there was BOINC forum, it would be different...WCG would have been able to attract more testers.
I hope more feedback from users is comming...or there are other ways?

Didactylos,
things I required are not neccesary for those install&omit users.
A standard navigation menu (there are a lot of feature omitted), standard view of BOINC database (users, hosts, workunits, results etc) plus the missing BOINC forum is something that BOINC users miss there.

I know that BOINC handle downloads but the lagging of computer is only on WCG. It's not that computer is slow (it is a fast Core 2 Duo computer with pelnty of RAM, BOINC running from ultra-fast i-RAM etc.). Perhaps the SSL solution.

Duplicate users names might not be a good idea. Definitely not on single project (e.i. non- or preBOINC). But once you enable duplicate names, you can let the user having same name on every project. I'm not much into statistics (where a different name on each project makes a real mess) but having same login/password, one personality etc.
And when you come to a project, where your name was taken by a non-BOINC user...try to take look at it from participants view.

Why Berkeley chose that route? I think because unique 32-byte semi-random identification was used before (erlier versions of BOINC).
During the transition from web resitration to a more automatic procedure done by BOINC manager upon installation, there was no need for it any longer.
Now, users needs a table with evrey project and each needs a column of user and password in order to log-in to projects when some disallow unique names.

The question can be put the other way round - why WCG doesn't allow duplicate names? Are user names used in internal databases and it doens't used some non-personal identification internally?

Forum. Yeah - I'm here. It doesn't asnwer my question - why WCG doesn't support Opera while every other BOINC project does? It's yet another aspect that makes WCG less BOINCish.

Hi Hon_za, welcome to the World COMMUNITY Grid......which is not a real BOINC website, but a DC where 80% of the members crunch on UD Agent and 20% on BOINC, where WCG is in gradual transition.

There seems to be a new wind blowing, lately some, absolutely only some new members coming in and lifting there leg and spraying the first tree in sight.....does not work......any knowledge u have to add is welcome but bring it to us members in a COMMUNITY tone and acquaint yourself with that mixed UD / BOINC forums COMMUNITY.

thank u.

I think, while it mostly works in Opera, there may be some issues with the Javascript. Part of my proposal to make it all compliant included the removal of all the unnecessary Javascript. It doesn't do anything that can't be done with pure html.

And on the username issue: actually, BOINC doesn't care much about usernames. The way it identifies a particular user is using the CPID (cross-project identifier) which is based, not on your username, but on the email you use to register with. Email addresses should be unique, and this method allows stats sites to group together all the statistics of a particular user while ignoring the username. WCG did have a few problems early on with their CPID imlementation, but I understand they have all been fixed ages ago.

Didactylos,

I'm glad to hear the both Opera and usernames issues were already addressed.
Any outcome from past discussion?