* Jay Berkenbilt <[EMAIL PROTECTED]> [2004-12-25 23:28]: > I guess updating this field for existing applications isn't really > that important, but maybe it would be worth doing in these cases > where it can pretty much be done in bulk?
I thought of this, but it won't necessarily be accurate. Anyway, I updated the fields for applicants now where it was easily verifiable that the date is accurate. > Also, now that there's a field for date of FD approval FWIW, I added "Awaiting FD Approval" and "On hold at FD stage" stats to the page now. Not that I believe that these statistics tell very much (*), but I'm sure people will want them anyway. > does this make it possible to update "Mode days" and "Median days" > for "Awaiting DAM approval" in the aggregate statistics section of > the NM status page? Okay, I just figured out what's going on there. (I didn't write those stats.) The statistics only look at things which happened in the last 2 months (which I guess is a good idea because the statistics reflect the current situation; note that his is actually explained at http://nm.debian.org/stats.php). Since nobody got approved in the last 2 months, it shows ??, but this will fix itself once people get approved again. (*) Let's give an example. An AM approves an applicant on day X, then the FD checks the report on day Y and finds that it is not complete. FD tells AM to do more stuff and in the meantime sets the application to "not complete". Later, the AM submits the information (X2) and FD updates the date Y2. Now, the statistics saying how fast FD is will take Y2 into account rather than Y but in reality it took FD only Y-X days to read the report. X2-Y was because of the AM, and Y2-X2 because of FD. To handle this properly, we'd need multiple dates in the database and add up days, but I'm not going this way. (Obviously, we could also set the "AM approves" date to X2 but that is not done for various reasons. e.g. FD approves (Y) would then be earlier than AM approves (X2) until FD checks the new report (Y2) and I'm pretty sure some scripts won't like this; there are some other problems afaik.) As an example of how meaningless the stats are, it currently shows that AM assignment is max 552 days. This is because two applicants ([EMAIL PROTECTED], [EMAIL PROTECTED]) who originally applied in 2003 and had an AM got reassigned to a new AM recently and so obviously AM assignment shows 2004. In reality, they didn't wait 552 days though... -- Martin Michlmayr http://www.cyrius.com/
