Hi Luca, some inspiration you can take from integration of TX to Solaris. It brought few new databases to nss and it happend during time of public development in ON gate, so you can find the history in Mercurial and see what developers of TX needed to do.
Best regards, Milan tamb...@studenti.unina.it píše v st 10. 02. 2010 v 12:45 +0100: > Hi Bayard, > in my previous request I was very brief and unspecific. I try > to provide you more informations. > > The goal of my project is adding two new databases in the nss. > > First database will be called x509pkc and it will store X509 PKCs. > In the files backend its syntax will be > username X509_PKC_PEM_FORMAT > > Second database will be called x509ac and it will store X509 ACs. > In the files backend its syntax will be > username X509_AC_PEM_FORMAT > > The principal backend to two databases is not files, but ldap. > > The frontend for this two database will be provided by functions: > getx509pkcnam(const char *name) > getx509acnam(const char *name) > > At the moment OpenLDAP supports x509 PKC and AC syntaxes, > PKC syntax is a work of Kurt Zeilinga, AC syntax is my work > delevoped for the thesis of first level university deegree. > I submitted this work at OpenLDAP comunity and it was included > in OpenLDAP suite. > > Can you suggest me some documents and relevant materials to study, > helpful for this project ? > > Thank you. > > Luca Tamburo > > Bayard Bell <buffer.g.overf...@googlemail.com> ha scritto: > > Luca, > > > > As this is a student assignment, the proper spirit seems to require > > answering mostly with pointers to further relevant materials or to > > further considerations that appear minimally necessary to arrive at > > adequate answers. > > > > For item 1, I expect the comments in /usr/include/nss_common.h > > should be adequate to evaluate your understanding. > > > > For item 2, I don't see why you'd expect to modify a system header > > in adding a new name service to an architecture that's supposed to > > be extensible by third parties. The header in question is supposed > > to be code for nss provider implementations from Sun, so you > > probably don't want to give it anything other than exemplary value. > > A few other points: > > > > it's not correct to refer to getXXXbyYYY as a system call > > I don't really follow your understanding of the source organisation > > (why separate source for each of the pw lookup functions?) > > you're not proposing to create multiple backends, just a single new > > backend for a new database to query pw data (again, reading > > nss_common.h indicates that the term backend is either a synonym for > > database or specifies the instantiation of a backend source pair) > > you've talked about source organisation but not specified the file > > object(s) or installation requirements > > > > Cheers, > > Bayard > > > _______________________________________________ > opensolaris-code mailing list > opensolaris-code@opensolaris.org > http://mail.opensolaris.org/mailman/listinfo/opensolaris-code _______________________________________________ opensolaris-code mailing list opensolaris-code@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/opensolaris-code
