Hey Team, Kim brought this to my attention and we discussed it at length, so she opted to have this conversation with the entire community (which is brilliant and the right thing to do). What follows are my $0.02:
As a veteran participant in this codebase (back in the Peter Lean days) the current organization makes sense to me and I think it is a vast improvement over the previous layout. That restructure was needed to clearly extract the core functionality of RCMES and make it into the OCW we have today. IF we want to have OCW act as a library/module which higher level tools and apps are built on, then I think it should be treated as such. While this codebase isn't as popular as say Lucene/SOLR we are kind of talking about a similar situation where the low level library (OCW) should be powering the other apps like mccsearch. On the extreme end we could spin the "apps" off into separate codebases and they just use OCW as a dependency. I thought about this for a while and I see the benefit of helping folks that are new to OCW be able to incubate their app within the OCW project (argument against spinning off the apps). They will get a lot more exposure being released and maintained within OCW than just another project on Github. Using the mccsearch 'app' as an example here, because the code is now part of the OCW project that code will automatically have more 'owners and maintainers'. The one minor potential downside is mccsearch must release on the OCW schedule, and it will have to share the attention with OCW (still not a bad thing). After typing all this out, I think having an "apps" directory (or something like it) would clearly define what is part of the core library and what has been built on top of OCW. In the future if/when mccsearch, ocw-cli or ocw-ui were to outgrow OCW, then we will have to deal with spinning it off. Until then I think a simple directory to house all the apps that are "Powered By OCW" is a solid move. Thanks for your time and attention, Cameron On Mon, Oct 27, 2014 at 1:06 PM, Michael Joyce <[email protected]> wrote: > I'm mostly +0 on this. I don't mind a restructuring but I don't think it's > terribly critical either. However, if someone wants to take a stab at > rearranging some folders I'm fine with it. > > > -- Joyce > > On Mon, Oct 27, 2014 at 9:29 AM, Loikith, Paul C (329C-Caltech) < > [email protected]> wrote: > > > Kim, this sounds like a good suggestion. We should talk more and flush > > out details when we meet next week. > > > > Paul > > > > Sent from my Verizon Wireless 4G LTE DROID > > > > > > "Whitehall, Kim D (398J)" <[email protected]> wrote: > > > > Hey guys, > > as we prep for our next release, I was thinking about the file structure > > of the distribution. It is my feeling that as it stands the file > structure > > doesn’t clearly delineate that OCW is a library that can perform various > > manipulations and analysis of earth science datasets. In the same way > there > > is an examples folder, I am thinking that it may be useful to have an > > applications folder (or something similar) for ocw-ui, ocw-cli, ocw-ui > and > > mccsearch folders can be moved as they are all applications utilizing the > > OCW libs, as opposed to keeping them as separate folders in the main > > distribution. > > any thoughts? > > > > > ************************************************************************** > > Kim Whitehall, Ph.D. > > Scientific Applications Software Engineer II > > Computer Science for Data Intensive Applications Group (398M) > > NASA Jet Propulsion Laboratory, > > 4800 Oak Grove Drive, > > Pasadena, CA 91109, USA > > Mailstop: 158 - 252D > > Phone: (+1) (818) 354-5162 > > Email: [email protected]<mailto:[email protected]> > > > > > > > > > > > > > > > -- Sent from a Tin Can attached to a String
