svn commit: r455056 - head/devel/py-naiveBayesClassifier

[Dan Langille]

> On Nov 28, 2017, at 11:40 AM, Alexey Dokuchaev <danfe at FreeBSD.org> wrote:
> 
> On Tue, Nov 28, 2017 at 09:26:55AM -0700, Adam Weinberger wrote:
>>> On 28 Nov, 2017, at 9:19, Alexey Dokuchaev <danfe at FreeBSD.org> wrote:
>>> On Tue, Nov 28, 2017 at 04:12:32PM +0000, Dan Langille wrote:
>>>> [...]
>>>> +COMMENT=	Yet another general purpose naive bayesian classifier
>>>> [...]
>>>> +++ head/devel/py-naiveBayesClassifier/pkg-descr
>>>> @@ -0,0 +1,4 @@
>>>> +yet another general purpose Naive Bayesian classifier.
>>>> +(under heavy development)
>>> 
>>> Inconsistent capitalization in COMMENT vs. pkg-descr; also, typically
>>> port description text should contain a little more information about
>>> the port (see section 3.2.1 of the PHB for some hints).
>> 
>> This is something that we have really regressed on lately. Our pkg-descrs
>> are becoming very terse, often simply repeating the COMMENT. Someone, I
>> forget who, recently added a bunch of ports without even listing WWW. We
>> have many ports broken into sections or variants where they all share the
>> same descr, so now I have no idea which one I actually want to install.
> 
> Big +1, I appreciate the support Adam.
> 
>> Folks, the descr is our catalogue; it's the first place that end-users
>> learn about ports and decide which one they're looking for. Take the 2
>> minutes to write a meaningful description, [...]
> 
> Exactly.  I can give an advice: if upstream lacks elaborate description,
> and you're having difficulty coming up with one of your own, take a look
> at some GNU/Linux distros that provide the same (or similar) package.
> I recall adding `sysutils/radeontop' and not being happy with pkg-descr
> (in r436717) until I found better description in Linux Mint package, cf.
> r436868.  HTH,

Nice idea.  I went looking for this port elsewhere but failed to find it elsewhere.

I tried searching for links to the sourcecode, failed on that search too.

Instead, I found some descriptions Naive Bayes Classifier and used those.

-- 
Dan Langille - BSDCan / PGCon
dan at langille.org