Dear Tore Unfortunately no dump will help you with this process. The "changed:" attribute is almost useless. For what you want to do it IS useless. The syntax requires all objects to have at least one mandatory "changed:" attribute. The user can set the date to almost anything. It can be tomorrow or ten years ago. When the object is changed, there is no requirement to make any changes to this date or add any additional "changed:" attributes. It is user generated and only has meaning to the user. What you want to do would only work if these were auto generated timestamps by the update software. But that is not the case. Regards Denis Walker Business Analyst RIPE NCC Database Group

(Apologies for breaking threading, I just subscribed to the list.)

In general, the proposal seems fine to me and I agree with its goals.

In preparing for my APWG presentation next week, I've been wanting to find the answer to the question «how many IPv4 PA Assignments have been registered between dates A and B?». I thought I could do this by examining the inetnum database export - but that didn't work out, due to the fact that all the "changed:" attributes gets rewritten to "unread‌‌@ripe.net 20000101". So while I don't care about the e-mail addresses, I'd need the dates to be left intact. Also, the current algorithm folds all changed attributes into a single one. It's not clear from the Labs article whether or not this practise will stop - I hope it will.

Would it by any chance be possible to get a sneak preview of the new inetnum dump, so that I could find the numbers I want before heading to Dublin? If so that would be greatly appreciated!

Best regards, Tore Anderson