Hi

[bobymicroby]

Hello @ianopolous .

I'm writing you because I have forked your repo and wanted to let you know about it.
The main reason for forking was using Gradle as a build tool. It will make
artifact publishing to maven central ,cross platform testing and integration with http://openjdk.java.net/projects/code-tools/jmh/ an easy task.

The other reason for forking was to document the methods ( the most tedious one as well...) .
And the third, but not last reason was to apply a set of checkstyle rules and formatting to the code.

P.S I have made some micro benchmarks , and here are the results

TweetNaClBenchmark.cryptoBox         thrpt   10     701.672 ±   38.688  ops/s
TweetNaClBenchmark.cryptoBoxAfternm  thrpt   10  126582.308 ± 4855.731  ops/s 

P.P.S I guess that you will be reserved against me making a PR because of the code formatting and the replacement of three for loops with System.arrayCopy , but I just wanted to let you know about my work :)

P.P.P.S all tests are passing correctly.

Have a nice one.

Bobby.

[ianopolous]

Hi Borislav,

Thanks for your email. I'm quite happy to merge in most of what you've
done, so long as it supports building with either make or gradle, all is
good. So feel free to make a pull request along those lines.

Many thanks,
Ian

"Those who say it cannot be done should not interrupt those doing it."

• Michael Johnson's coach

On Wed, Nov 25, 2015 at 8:19 PM, Borislav Ivanov [email protected]
wrote:

Hello @ianopolous https://github.com/ianopolous .

I'm writing you because I have forked
https://github.com/bobymicroby/tweetnacl-java/tree/bobymicroby/gradleBuild
your repo and wanted to let you know about it.
The main reason for forking was using Gradle as a build tool. It will make
artifact publishing to maven central ,cross platform testing and
integration with http://openjdk.java.net/projects/code-tools/jmh/ an easy
task.

The other reason for forking was to document the methods ( the most
tedious one as well...) .
And the third, but not last reason was to apply a set of checkstyle rules
and formatting to the code.

P.S I have made some micro benchmarks , and here are the results

TweetNaClBenchmark.cryptoBox thrpt 10 701.672 ± 38.688 ops/s
TweetNaClBenchmark.cryptoBoxAfternm thrpt 10 126582.308 ± 4855.731 ops/s

P.P.S I guess that you will be reserved against me making a PR because of
the code formatting and the replacement of three for loops with
System.arrayCopy , but I just wanted to let you know about my work :)

P.P.P.S all tests are passing correctly.

Have a nice one.

Bobby.

—
Reply to this email directly or view it on GitHub
#3.

[bobymicroby]

Hi Ian,

My comments are in #4

[ianopolous]

The resulting Test.jar from "make test" should be runnable with "java -jar build/test.jar", but currently this fails because of the wrong main class, and the jar mainfest's classpath not finding the junit jars.

[bobymicroby]

I will check it out .

On Fri, 27 Nov 2015 21:05 Ian Preston [email protected] wrote:

The resulting Test.jar from "make test" should be runnable with "java -jar
build/test.jar", but currently this fails because of the wrong main class,
and the jar mainfest's classpath not finding the junit jars.

—
Reply to this email directly or view it on GitHub
#3 (comment)
.