Thanks Emre and Marlon for the critical feedback. I have updated my proposal accordingly, which can be viewed at: http://www.google-melange.com/gsoc/proposal/public/google/gsoc2014/nadeemanjum/5632763709358080
Any other suggestions are welcome. Thanks, Nadeem On Fri, Mar 21, 2014 at 12:54 AM, Marlon Pierce <[email protected]> wrote: > Thanks for your interest, Nadeem. Some comments: > > * You should clarify in the beginning the advantages for both projects. > Airavata would get a tool for quickly generating user interfaces in > multiple languages. Benefits to GenApp are ability to run long running, > non-interactive jobs on XSEDE and similar resources and the ability to > compose and execute workflows using Airavata's workflow engine. Correct > and expand my previous statements for your proposal. Section 1.3 says > "what" but not "why". > > * Section 1.2: The component list you have here is old and incomplete, > so please update. We have put a lot of work into the Thrift-based > version of the API, which is the preferred way to build Airavata > clients. This supersedes the registry API client you have listed in this > section. > > * Section 1.3: XBaya is currently not working with the Thrift APIs and > will need a lot of updating. This would be an entire GSOC by itself. If > workflow composition of GenApp modules is a critical deliverable, it may > be better to have a new "XBaya-lite" composition tool (based on Airavata > API) for composing workflows in GenApp than to use XBaya. See also > comment below. > > * Section 2.2: Airavata's workflow interpreter is also not exposed in > the current API. I think this would be the better way to handle GenApp > workflows instead of XBaya, and there may be some work to complete and > fully test the workflow parts of the API. Also, if you were to make an > XBaya-lite, this would need to be done first. > > * Section 2.3: What is the current GenApp messaging system and its > requirements? Note also the Airavata messaging system is old and is on > our long-term to-do list to replace it. For this GSOC, we would need to > define an implementation-independent Component Programming Interface > (which we are doing with other components) and also decide best how this > should be exposed through the API to GenApp. This would shield GenApp > from any changes. > > > Marlon > > On 3/20/14 8:15 AM, Nadeem Anjum wrote: > > Hi All, > > > > Here is my proposal for GSoC 2014: > > Integration of GenApp with Apache Airavata: > > > http://www.google-melange.com/gsoc/proposal/public/google/gsoc2014/nadeemanjum/5632763709358080 > > > > Please provide your feedback and suggestions. > > > > Thanks, > > Nadeem. > > > >
