I'm very interested in following suggestions for improvement, so to clarify my "contribution" on this, I'm just thinking out loud as an Ubuntu server n00b about real world use of eBox 1.0 in production environments. Not large scale stuff like server farms, etc. but SOHO and SMB environments where resources (money and brains) are limited -- the very things which make eBox, even at 1.0, attractive enough to put to the test (along with ourselves). Else I'd just run Ubuntu server like the purists maintain I should: sans eBox. I know we're not altogether on different pages here from earlier comments.
Unless I'm mistaken, eBox will draw more n00bs like me. We'll be a little edgy and out of our comfort zone (there's no comparison between being responsible for, e.g., a Linksys BEFVP41, and an eBox doing all and more of what the blue box did). We'll need to know how to assure management that we have it "all under control." When we suggest taking eGroupWare for a drive as a replacement to a cobbled and unreliable implementation of Outlook, we need to look and sound more confident than we feel. To get there, we have two choices: We can wait until eBox is more robust on the update and recovery fronts, or we can think out loud among ourselves about functional implementation strategies for today, here and now.
That's all I'm doing, really. I've already gotten eBox in the production door by keeping a low profile and by touting the potential. Now it's time either to make good with what we've got or to admit being overly ambitious and move on. A major screwup now will jeopardize the eBox experiment in my little world for at least a couple of revisions into the future. The people using "my" network aren't interested in the back end in itself. They just want to work, and want it to work. Time and money are at stake.
Anybody out of the SOHO and SMB markets that is interested in eBox is interested in either doing things already being done measurably and reliably better or in finding solutions to existing problems. In my case, it's both. And in my case, I have to add Ubuntu server n00bishness to the scenario (clearly there are several here who know what they're doing outside eBox; I'm not among them). I'm hoping just to be among the first of many Ubuntu server n00bs to end up here, since I find eBox an exciting and useful tool. Thinking out loud about the challenges people like me might face in real world implementation today is intended to be my little contribution to the greater discussion about the long term and the possibilities, and to the success of eBox.
The End.