Hi. Le mer. 27 mars 2019 à 13:43, Abhishek Dhadwal <dhadwal1...@gmail.com> a écrit : > > > > Hello Sir, > > >I don't know what are the hard requirements for the schedule, but > >I would not make such delineation between coding, reviewing and > >testing. > The Elements Of A Quality Proposal in Google’s Guide > (https://google.github.io/gsocguides/student/writing-a-proposal) state the > following about deliverables and timelines: > Deliverables > Include a brief, clear work breakdown structure with milestones and > deadlines. Make sure to label deliverables as optional or required. You may > want your plan to start by producing some kind of white paper, or planning > the project in traditional Software Engineering style. It’s OK to include > thinking time (“investigation”) in your work schedule. Deliverables should > include investigation, coding and documentation.
So, indeed, what I described (quoted below) is referred to here as "investigation". Then "coding" could be assumed to include "testing" and "review", and even "documentation" (as it is a prerequisite to commit that all codes come with complete Javadoc). > I attempted to follow it in my implementation of the milestone timeline. > What be your advice on improving/altering the timeline provided in the > documentation ? I don't know whether we follow the "traditional Software Engineering style"... According to what happens in practice, top-level would be * investigation (with sub-tasks such as collecting reference implementations) * coding (with subtasks such as producing numbers from reference implementations, porting those implementations to Java, write unit tests and Javadoc) These two "steps" can be repeated for each RNG which you are going to contribute. But as I said there is no strict order: if you are stuck with one (e.g. missing a reference implementation, or cannot compile it for some reason), you could start working on another that may prove more straightforward. HTH, Gilles > >You could perhaps also mention that you'll need to compile the > >"reference" implementations (in C probably), and most importantly > >devote some time to looking for those (a.o. also compare several > >implementations of the same algorithm (and with the algorithm > >description), if there is no "authoritative" origin for the code (i.e. > >when the person who wrote the code is not the person who > >designed the algorithm). > >Depending on when you actually find all the needed resources > >for a given task, the schedule will be quite different... > > I’ll add those tasks and work on the schedule as required. > > Thanking You, > Yours Faithfully, > Abhishek --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org For additional commands, e-mail: dev-h...@commons.apache.org
