Hi Gimhana.
On Tue, 20 Mar 2018 14:36:10 +0530, Gimhana Nadeeshan wrote:
Hello devs,
I have updated my draft proposal with @Gilles's suggestions[Draft
Proposal
V1.1
<https://docs.google.com/document/d/1sqSa0hrYc2AD75RZyJRkeqCOBOqTOeMnPaBsE9U5YhU/edit?usp=sharing>]
and I think I need some more clarifications on below suggestions
I haven't read the new version yet; I'll try to answer
some of the below questions.
== "Background" section ==
number of listed/active contributors
histogram of component's sizes (lines of code)
How to recognize "active contributors" ?
Good question!
Perhaps organize a survey? :-)
In the ML
A possible source, but probably not very efficient. I certainly
do not suggest to perform a "manual" counting of who talks about
what. ;-)
Actually I don't know how to make automated queries (if possible).
You might want to have a look at "kibble":
http://kibble.apache.org/
I asked that "Commons" projects be added to their "live demo".
But never got much time to explore the information it extracts
from various data sources.
I'd guess that it would be quite interesting to get more acquainted
with that tool. Don't hesitate to subscribe to their ML (I'm not);
you might then report here what you found useful and which a lot of
us may not be aware of.
or GitHub?
AFAICT, it would be completely biased. Indeed if one looks at
that page, for example:
https://github.com/apache/commons-rng/graphs/contributors
there is absolutely no trace that someone not mentioned there
performed 88% of all commits. [And this is much lower than the
actual number of deletions/additions...]
What do you
mean by "histogram of component's sizes (lines of code)" ?
There is a (command-line) tool called "cloc".
You could "clone" the repositories (of a selection of the
active or popular components) and run it on the "src/main"
directory to get some indication of the size of the projects.
== "Deliverables" section ==
* less dependencies" (an example?)
* "Advanced mathematical functionalities": other than what
exists now? Or do you mean new interfaces (e.g. in
accordance with the APIs provided by JDK8)?
Most of the classes in "math4.stat" contains "math4.exception"
classes. And
some classes in "correlations module" dependent to "RealMatix"
interfaces.
Can those be considered as dependencies ?
Yes. And we don't want that any of the new component to depend
on "Commons Math" code. [With perhaps an exception for the "test"
scope.]
Can't exceptions be substituted
with inbuilt java exceptions ?
Certainly. You should take a look at what is done in "Commons
Numbers".
@Gilles would you please explain this matrix
issue because I didn't get it much....
Design issues were identified a long time ago.
You should be able to find them by doing a "Search issues"
in JIRA.
Bottom-line is we don't want to depend on an API that must
be changed (at some point).
Although not ideal, a workaround is to copy the necessary
functionality over to the new component but ensure that it
is *not* part of the API.
Better would be to tackle the issues themselves but it has
proven difficult and was postponed several times...
== "Implementation" section ==
* "Design goals": give concrete examples.
I noted some examples for Design Goals in my proposal. But I'm not
sure I
that I wrote it correctly. (And don't know those are the examples you
expect me to mention.) Please clarify those too.
IIRC, you mention streams. So for example, you could show how
the contribution would enhance usage (comparing "before"/"after").
== "Results" section ==
Hope to get comment from PMC...
[Wish list, design requirements, mentor(s), etc.]
mentor(s)?? @Gilles,@Eric wont't you guys be the mentors of this
project ??
I'm asking this because I'm new to ASF and GSoC !! And I'm appreciate
to
know how this is working !!
I don't know whether there is an official "mentor" role for GSoC,
and if so, what that implies. This is also new for me; so I hope
that people can give advice about the "administrative" side.
For the contents, Eric indeed proposed to participate, but I
don't know how available he is (and how this will fit with
the GSoC timetable, which I also don't know).
It seems that Eric's contributions are currently extremely
asynchronous... :-}
In the meantime, you could also review the open issues for "Commons
Numbers":
https://issues.apache.org/jira/projects/NUMBERS/
This is quite important as almost all other "Commons Math"
spin-offs will have some dependency on this new component;
hence a release of "Commons Numbers" must precede a release
of either "Commons Statistics" or "Commons Geometry".
Yep Wow sure...I'm on my way right now CM Numbers!!
Whenever you find something you can handle tackle, please
submit a PR.
Thanks,
Gilles
[...]
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org
For additional commands, e-mail: dev-h...@commons.apache.org
