Re: POCs, Prototypes, experiments ...

2018-10-26 Thread Daniel Takabayashi
+1 Em sex, 26 de out de 2018 às 16:47, Lucas Bonatto Miguel < lucasb...@apache.org> escreveu: > Makes perfect sense to me. I believe that if the majority of people agrees > with this architecture we could create issues and start implementing it. > > +1 > > > On Fri, Oct 26, 2018 at 4:09 PM Daniel

Re: POCs, Prototypes, experiments ...

2018-10-26 Thread Lucas Bonatto Miguel
Makes perfect sense to me. I believe that if the majority of people agrees with this architecture we could create issues and start implementing it. +1 On Fri, Oct 26, 2018 at 4:09 PM Daniel Takabayashi < daniel.takabaya...@gmail.com> wrote: > I see... what about this new version? > > > https://

Re: POCs, Prototypes, experiments ...

2018-10-26 Thread Daniel Takabayashi
I see... what about this new version? https://docs.google.com/drawings/d/shajxIpLJHxxMbFgDXiPuhg/image?w=602&h=468&rev=1630&ac=1&parent=1ySERHGBXbHeyCMRookq5UfTuFkzzU0ugtjvR3rF3deY I changed the toolbox ...from 1 component to 3... docker client, cli and repl. Both (cli and repl) now are using the

Re: POCs, Prototypes, experiments ...

2018-10-26 Thread Lucas Bonatto Miguel
For sure, both suggestions are to make sure we're not killing the CLI and that we have a component (the SDK client) that is shared by CLI and REPL. Thanks for clarifying where the system calls will live. - Lucas On Fri, Oct 26, 2018 at 2:39 PM Daniel Takabayashi < daniel.takabaya...@gmail.com> w

Re: POCs, Prototypes, experiments ...

2018-10-26 Thread Daniel Takabayashi
Hi Lucas, could you please justified your two suggestions? The general ideia is to use the Docker DSK to execute local commands, to consume outputs and also to execute Marvin "specific language" commands. Today almost all CLIs commands in Marvin is a mixed of systems commands and library method ca