On Sun, 31 Jul 2016, 1:46 PM Scott Cotton <w...@iri-labs.com> wrote: > Neat, maybe a bit like sat4j does in eclipse dependency management? >
Probably similar I am sure. For standard distro package management systems I think they have to be concerned with the idea of installing new or upgrading existing or removing unneeded. Since ours is just an environment management system, we have a slightly less complicated goal. We just need to find the total working resolve of the package set so that we can build a shell environment from the combination > For generating the cause of a failure, do you mean "failed assumptions" in > picosat speak or > unsat core or some other feature (proofs, etc)? > I would have to look at it again, but I think it was the dimac proof that you get from a failed solve when picosat is compiled with trace support (which it is in the pigosat bindings) . It's the actual progression of clauses and assumptions that lead to a fail. > Scott > > 2016-07-30 0:01 GMT+02:00 Justin Israel <justinisr...@gmail.com>: > >> >> >> On Sat, 30 Jul 2016, 2:22 AM Scott Cotton <w...@iri-labs.com> wrote: >> >>> Thanks. >>> >>> Out of curiosity, are you using sat solvers in visual effects pipelines? >>> >> >> I was using it for a new version of the package dependency resolver >> component of our environment management system. >> I was able to model the entire "repository" of software package >> descriptions and their dependencies. Then add the specific package >> selections for an environment as constraints and resolve that back into the >> full list of needed packages. >> The useful bit in pigosat is that I can have it generate the cause of the >> failure to solve and convert that back into human readable reporting. >> >> >>> Scott >>> >>> >>> Le vendredi 29 juillet 2016 00:18:14 UTC+2, Justin Israel a écrit : >>>> >>>> Neat. I had been using pigosat <https://github.com/wkschwartz/pigosat> for >>>> a project. It vendors the picosat C dependency right in. >>>> Cool to see a pure Go solution. >>>> >>>> Justin >>>> >>>> >>>> On Sunday, July 24, 2016 at 9:33:31 AM UTC+12, Scott Cotton wrote: >>>>> >>>>> I'm happy to announce the first public beta release of mini, available >>>>> at >>>>> github <http://github.com/IRIFrance/gini>. >>>>> >>>>> Gini is a SAT solver with some related tools built for solving the >>>>> canonical NP-complete SAT problem. SAT solvers have many applications in >>>>> formal verification and discrete optimisation, >>>>> often acting as an indispensable component in these domains. >>>>> >>>>> Gini is written in 100% pure go and thus far, our core CDCL solver >>>>> either outperforms or is competitive with analogs in C/C++ like picosat >>>>> and >>>>> minisat. Additionally, internal measures of raw speed such as >>>>> mega-props/second are good and independent of variations arising from >>>>> heuristics. >>>>> >>>>> By bringing a high quality SAT solver to go, we hope to enable >>>>> competitive innovations in the go community which tackle combinatorial >>>>> explosion symbolically. >>>>> >>>>> Gini is in first beta public release, following the recent SAT >>>>> competition. To maintain performance in the long term, we plan to have >>>>> gini compete in sat races and sat competitions annually. To this end, we >>>>> are happy to collaborate with gophers, the curious, raw speed junkies, >>>>> algorithm officianados, and logicians alike. >>>>> >>>>> Cheers, >>>>> >>>>> >>>>> -- >>>>> Scott Cotton >>>>> President, IRI France SAS >>>>> http://www.iri-labs.com >>>>> >>>>> >>>>> -- >>> You received this message because you are subscribed to the Google >>> Groups "golang-nuts" group. >>> To unsubscribe from this group and stop receiving emails from it, send >>> an email to golang-nuts+unsubscr...@googlegroups.com. >>> For more options, visit https://groups.google.com/d/optout. >>> >> > > > -- > Scott Cotton > President, IRI France SAS > http://www.iri-labs.com > > > > > > -- > Scott Cotton > President, IRI France SAS > http://www.iri-labs.com > > > -- > You received this message because you are subscribed to the Google Groups > "golang-nuts" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to golang-nuts+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/d/optout. > -- You received this message because you are subscribed to the Google Groups "golang-nuts" group. To unsubscribe from this group and stop receiving emails from it, send an email to golang-nuts+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
