Sebastien Roy wrote:
> On Wed, 2009-09-30 at 20:11 +0200, Milan Jurik wrote:
>
>> Yes, getprinterbyname() is private interface and lp is the only
>> consumer. The question is how lp on other platforms is consuming NIS
>> maps for printing because websearch reveals that "printers in NIS maps"
>> is not Solaris only thing (but not on Linux probably).
>>
>> Still it is unclear how CUPS will take care or not (for now "not") about
>> "printers" in nsswitch. I am not saying it is proper solution to use
>> that interface but it should be evaluated before disbanding lp in the
>> next phases.
>>
>
> "Not this case" seems appropriate to me. CUPS uses a perfectly modern
> desktop printer configuration tool that has a discovery feature that
> displays local printers (i.e. those that I'd actually want to print to)
> that use modern protocols. This is in contrast to NIS which has an
> incompatible notion of scope (NIS domain) that results in applications
> having access to potentially hundreds of printers I'd never want to
> print to. I'd rather see the printers NIS map Obsoleted rather than
> extended (and focus more on service discovery rather than
> "directory"-based solutions), but that's just my personal opinion.
> Anyway, not this case.
>
>
[ I've removed the case number. Because it's *not* this case as already
noted.]
I think this question, directory vs. discovery, is a fairly important
one. Certainly discovery based services are often easier for end-users,
but the question is, what other discovery services do we use for things
*besides* printers. Sites that administer all of their configuration in
a directory (NIS or LDAP or whatever) probably want to keep doing that.
Changing the administrative *paradigm* might be upsetting for some sites.
Additionally, sites might have very good reasons enforce a directory
based approach. For example, you don't want someone adding a queue that
appears to be the private printer for the HR department, but really just
duplicates a copy of everything printed, and then forwards the job to
the "real" HR printer... While directory based approaches might not
completely prevent this vector, they are probably a very good start at
ensuring that non-tech-savvy users don't misprint to "unauthorized" devices.
- Garrett
