On Aug 30, 2013, at 5:13 PM, markus <[email protected]> wrote: > >> But I also don't think there's fundamentally anything *wrong* with >> using whits, since they accurately model the notion of containment, >> with respect to order: classes have a "start", and a "finish", and >> there's stuff in between. It doesn't really feel like a hack to me. >> Ironically, they *do* feel like a hack for solving the problem of >> many-to-many relationships, which was what they were intended to fix. >> But as far as representing containment goes, I think they function >> well. > > Actually, they were intended to address the general problem of which > many to many relationships, ordering, containment, etc. are examples. > >> It's easy to traverse the spliced graph with only a single type of >> edge,...The biggest failing is really the loss of semantic >> information, since it's not as easy in the transformed graph to >> identify containment, which is a meaningful relationship. > > That's a common feature of compilation in any language though. You > transform into a form that's more regular/easier to traverse by > factoring out information that, while meaningful, isn't necessary for > the task at hand. :)
That totally works as long as you don't expose it to the user, but the dot graphs are (until recently with PuppetDB) the only real way for a normal user to see relationships, which resulted in whits being exposed. Along with other issues. So it was not just an implementation issue, it was a user experience issue. Which is why I care. -- Luke Kanies | http://about.me/lak | http://puppetlabs.com/ | +1-615-594-8199 -- You received this message because you are subscribed to the Google Groups "Puppet Developers" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at http://groups.google.com/group/puppet-dev. For more options, visit https://groups.google.com/groups/opt_out.
