Re: [Haskell-cafe] Package descriptions on hackage

2011-09-12 Thread Ivan Lazar Miljenovic
On 12 September 2011 23:29, Ryan Newton  wrote:
> On Sun, Sep 11, 2011 at 1:14 PM, wren ng thornton  wrote:
>>
>> On 9/11/11 6:37 AM, Neil Mitchell wrote:
>>>
>>> Why not email the maintainers of packages you think need a better
>>> description - ideally giving suggestions? I'd welcome that for any of
>>> my packages.
>>
>> +1.
>>
>
> +1
> Actually this thread itself is helpful.  It made me cringe and realize I was
> guilty.  In my/our case I think we just overlooked it.

I just realised from you mentioning this that I only added to the
description of graphviz that it's for graph visualisation in the
releases I made last month... in my defence, this had been added in
the darcs repo a while ago, but I hadn't made a release in so long...

-- 
Ivan Lazar Miljenovic
ivan.miljeno...@gmail.com
IvanMiljenovic.wordpress.com

___
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe


Re: [Haskell-cafe] Package descriptions on hackage

2011-09-12 Thread Ryan Newton
On Sun, Sep 11, 2011 at 1:14 PM, wren ng thornton  wrote:

> On 9/11/11 6:37 AM, Neil Mitchell wrote:
>
>> Why not email the maintainers of packages you think need a better
>> description - ideally giving suggestions? I'd welcome that for any of
>> my packages.
>>
>
> +1.
>
>
+1

Actually this thread itself is helpful.  It made me cringe and realize I was
guilty.  In my/our case I think we just overlooked it.
___
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe


Re: [Haskell-cafe] Package descriptions on hackage

2011-09-11 Thread Brandon Allbery
On Sun, Sep 11, 2011 at 13:14, wren ng thornton  wrote:

> On 9/11/11 6:37 AM, Neil Mitchell wrote:
>
>> Why not email the maintainers of packages you think need a better
>> description - ideally giving suggestions? I'd welcome that for any of
>> my packages.
>>
>
> +1.
>
> Of course, requiring feedback in order to improve places an undue burden on
> the users; the maintainers should be trying to save users from that labor.
> But everything has to start somewhere.
>

Not to mention that not all users can actually help there; I'm still
smarting a bit after hitting the functional lenses package and bouncing hard
(even after digging out and reading the paper, which made more sense than
the haddock...).

-- 
brandon s allbery  allber...@gmail.com
wandering unix systems administrator (available) (412) 475-9364 vm/sms
___
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe


Re: [Haskell-cafe] Package descriptions on hackage

2011-09-11 Thread wren ng thornton

On 9/11/11 6:37 AM, Neil Mitchell wrote:

Why not email the maintainers of packages you think need a better
description - ideally giving suggestions? I'd welcome that for any of
my packages.


+1.

I always love to hear that my packages are useful to someone. And if 
someone says "hey this is useful, but here's something to make it even 
more useful" that's awesome.


Of course, requiring feedback in order to improve places an undue burden 
on the users; the maintainers should be trying to save users from that 
labor. But everything has to start somewhere.


--
Live well,
~wren

___
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe


Re: [Haskell-cafe] Package descriptions on hackage

2011-09-11 Thread Herbert Valerio Riedel
On Sun, 2011-09-11 at 11:37 +0100, Neil Mitchell wrote:
> There are several problems here:
[..]
> Why not email the maintainers of packages you think need a better
> description - ideally giving suggestions? I'd welcome that for any of
> my packages.

Maybe something similiar to http://kernelnewbies.org/KernelJanitors
could be done, e.g. "HackageJanitors".

Those HackageJanitors could also help making new GHC releases smoother
by providing efficiently the usual minor fixes to hackage libraries
breaking by new GHC releases (updating library version constraints,
fixes related to GHC becoming less tolerant, adapting to Cabal changes).

Maybe they could also be given the privilege to upload trivial fixes to
hackage directly (possibly with a version indicating such a
maintainer-override upload) if the library maintainer takes too long to
respond or has generally opt-ed in for to such a "auto"-maintaince
service (maybe by a new flag in .cabal files).

cheers,
hvr


___
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe


Re: [Haskell-cafe] Package descriptions on hackage

2011-09-11 Thread Neil Mitchell
There are several problems here:

1) Not everyone can write beautiful clear English descriptions, it
takes a certain skill.
2) The person writing the description is the author, who knows all the
details, but the person reading the description doesn't  - writing for
a different audience is an even harder skill to master.
3) It's easy to miss something when updating a package.
4) Quality documentation places an ongoing maintenance burden on the
package, and while test suites etc. make code maintenance easy, I
don't know any way to automatically check documentation!

Why not email the maintainers of packages you think need a better
description - ideally giving suggestions? I'd welcome that for any of
my packages.

Thanks, Neil

On Mon, Sep 5, 2011 at 11:22 PM, Ivan Lazar Miljenovic
 wrote:
> On 5 September 2011 23:59, Joachim Breitner  wrote:
>> Dear hackage package authors,
>>
>> this is a short message from your distribution package creators: Please,
>> if possible, write good, not too short descriptions, and also keep them
>> up to date. Of course, users browsing hackage will benefit as well.
>
> Also for potential users trying to work out what your library does!
>
> Something that I find particularly frustrating is all the libraries of
> the form "hFoo: Haskell bindings to the Foo library"; well, what _is_
> the Foo library?
>
> --
> Ivan Lazar Miljenovic
> ivan.miljeno...@gmail.com
> IvanMiljenovic.wordpress.com
>
> ___
> Haskell-Cafe mailing list
> Haskell-Cafe@haskell.org
> http://www.haskell.org/mailman/listinfo/haskell-cafe
>

___
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe


Re: [Haskell-cafe] Package descriptions on hackage

2011-09-05 Thread Ivan Lazar Miljenovic
On 5 September 2011 23:59, Joachim Breitner  wrote:
> Dear hackage package authors,
>
> this is a short message from your distribution package creators: Please,
> if possible, write good, not too short descriptions, and also keep them
> up to date. Of course, users browsing hackage will benefit as well.

Also for potential users trying to work out what your library does!

Something that I find particularly frustrating is all the libraries of
the form "hFoo: Haskell bindings to the Foo library"; well, what _is_
the Foo library?

-- 
Ivan Lazar Miljenovic
ivan.miljeno...@gmail.com
IvanMiljenovic.wordpress.com

___
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe


[Haskell-cafe] Package descriptions on hackage

2011-09-05 Thread Joachim Breitner
Dear hackage package authors,

this is a short message from your distribution package creators: Please,
if possible, write good, not too short descriptions, and also keep them
up to date. Of course, users browsing hackage will benefit as well.

Keep in mind that the users do not know what other packages this relates
to. So if the package is part of a larger set of corresponding packages,
include a general blob and then explain this particular packages. This
is especially true for packages that contain the base types for a suite
of packages.

Also think about all the possible keywords that someone might enter in a
search engine when looking for something like your package, and ideally
explain how your library relates to these keywords so that the user can
estimate whether it useful for him even before reading the API.

Here a few negative examples, not to expose particular authors, but
selected because they were the most resent that the Debian Haskell Team
had packaged.

=

The http-enumerator package

This package uses attoparsec for parsing the actual contents of the HTTP
connection. The only gotcha is the withHttpEnumerator function,
otherwise should do exactly what you expect. 

(A list of supported features would be great, maybe a short note what
enumerator means. Also, the second sentence is incomprehensible to me
and the latest version does not contain a withHttpEnumerator function
any more)

=

The representable-functors package

Representable functors 

(Clearly not helpful. Here is what Iulian came up for for the Debian
package, from the individual module documentation, but I still think the
author could do better

This package provides a generalized Store comonad, parameterized by a
Representable functor (the representation of that functor serves as the
index of the store) and a generalized State monad, parameterized by a
Representable functor (the representation of that functor serves as the
state).

Representable functors on Hask all monads, being isomorphic to a reader
monad.

Representable contravariant endofunctors over the category of Haskell
types are isomorphic to (_ -> r) and resemble mappings to a fixed range.

Representable endofunctors over the category of Haskell types are
isomorphic to the reader monad and so inherit a very large number of
properties for free.)

=

The data-lens package

Haskell 98 Lenses 

(Also not helpful. And probably a very useful package! Iulian wrote „A
lense is a composable notion of a purely functional field accessor.“
which is still quite short, but explains the use of the package much
better)

=

The algebra package

Constructive abstract algebra 

(Seems to be the “main” package of a series of packages. Yet another
reason for having a good description. Here a suggestion, again by
Iulian:

“This package provides algebraic structures, such as groups, fields,
rings, modules. It also provides bands, also known as idempotent
semigroups (band is a semigroup where every element is equal to its own
square), coalgebras, semirings (rigs), idempotent semirings, also known
as dioids.“)


=


I know that most people have better things to do than to write API
documentation, so it is not unexpected that, after having written the
API documentation after all, they don’t spend too much time on the
description. But think about it: It is the first thing possible users
are going to see from your package. So if you want to attract users, do
take your time to write a comprehensive and comprehensible description.

Thanks,
Joachim

-- 
Joachim "nomeata" Breitner
  m...@joachim-breitner.de  |  nome...@debian.org  |  GPG: 0x4743206C
  xmpp: nome...@joachim-breitner.de | http://www.joachim-breitner.de/



signature.asc
Description: This is a digitally signed message part
___
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe