On Jun 13, 8:57 am, "Diez B. Roggisch" <[EMAIL PROTECTED]> wrote: > > a field should be marked alternativeid only if it should be unique. > > not the other way ;-) . > > Erm - no. The way it is it's the right way - if something is supposed to > be working as uniquely identifying something, it must be a unique column. > we said the same thing. no problem :) . > > valid policy may be to require the forgetter to enter username AND > > email. this way, someone able to intercept the mail of an user should > > provide minimal proof of identity. > > > conversely, if we'd aggree to send a password to any email address > > someone requests a remainder for, there would also be no problem to > > send a list of username and password pairs. > > What's the usecase for someone to have several accounts - with one email? > > About 100% of all online applications I'm registered require me to > provide a unique mail, or even use my mailaddress as login. > here is the difference: the login name should be unique for our system, since this is the token to identify the user on our system, but the email address should be treated in the same spirit as a street address, and we don't usually care to register 2 members of the same family. this is not the case in my home, but you can still find lots of families that are using the same email address they received from their isp, especially in windows. in computer terms, enforcing unique email addresses for users un our systems, is somehow maintaining an one to one relationship with the "table" of global email addresses. > So I consider this the sensible default in the TG identity model. > > But you can change that for your apps in about 20 secs. So - where is > the reason for this discussion? > i have only proposed to change this default, since choosing it is obstructing freedom, thus i think there should be a reason for doing this. sorry to bother, but enforcing this cleanly is more than a 20 seconds to do (look for example at the tg-registration project). for the same project, to make it work both ways based on the identity submodule, is it a way to query the model as of the uniquety of a field? > Diez thanks for your patience, alex --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "TurboGears" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/turbogears?hl=en -~----------~----~----~----~------~----~------~--~---
