-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

All in all, I'd have to agree with the sentiments of this. I don't think
anything should be taken personally although I know some personalities
might be inclined to react emotionally at the words used, but if you
look at the sentiment I believe it's right on. Your users may be
developers, but not necessarily for your project. Treat them to a 'user'
site.

Brian

Tim O'Brien wrote:
> On 3/8/06, John Casey <[EMAIL PROTECTED]> wrote:
>> <snip>
> 
> 
> 
>> Something to think about is maybe not having the initial Maven page be a
>>> Maven site.  ASF sites in general seems to be more Developer focused
>> than
>>> user focused.  What if the initial Maven site were more like the front
>> pages
>>> of Mozilla or Rails.  An attractive logo, links to resources and
>> materials,
>>> an introduction.  The home page should be user focused, Maven developers
>> are
>>> a minority of the audience here.
>> Are you saying that the developer-centric tendency is appropriate for
>> ASF project web sites, then? I'd tend to say that for a product like
>> Maven (not an API, like commons-cli, for instance) it's worth striving
>> to help the user. Since Maven is an Apache product/project, I'd say that
>> having a developer site on a different physical location would be
>> bad...they're different aspects of the same product/project. That said,
>> I think we need to section the developer docs off and put them a click
>> or so in from the main landing page...probably with their own landing
>> page.
> 
> 
> I think I agree with you.
> 
> When I said "developer-centric" I meant
> "apache-developer-centric". IMO, more projects need to think about the user
> who has "30 seconds to figure out the best way to download/use Derby".  The
> current Maven site has too many links, too much distracting people from what
> should be a simple message - "Download Maven, you won't believe how you
> developed without it.  We promise.".    I'm not saying that we should all
> become marketing people, but it's something to consider.
> 
> 
>> It's not a simple hierarchy, but then, we have a great deal of variation
>> among our audience members. As these audience members [possibly]
>> transition from users to contributors and so on, we don't want a
>> separation like this to get in their way...there should be *some*
>> cohesiveness, I would think...
> 
> 
>  I'm not saying this to be contrary, bear with me:  most people that use
> Maven don't care to know much about the internals or how it is being
> developed.  They simply want to download a product that works, is
> well documented, and use it.  A small minority of those users will get an
> itch that needs to be scratched or decide that the gift of free software
> should be repaid by involvement in a community.  So, if you wrote up a
> survey, and quizzed people who use Maven every day, I think you'd probably
> find that most of them think Jason van Zyl is a famous magician and the
> Meritocracy is a spell in Dungeons and Dragons.   I'm not saying this as a
> cynical jibe, but to make the point that regardless of the Maven website, we
> will always about an equal mix - out of 100 - 95 people download Maven, 5
> subscribe to the users list for a week, maybe 4 ask questions on the user
> list, and, if you are lucky, 1 submits a patch.  And even that's
> overestimating, apply that forumla to the httpd project and it would have
> 10^5 patches submitted per year.
> 
> Turn your statement around.  "....audience members [possibly] transition
> form users to contributors".   Assume that a simpler user focused launch
> page made it easier for people to adopt Maven.  If this "separation" of
> user-focused and developer-focused documentation increases the population of
> users, we've increase the pool of potential contributors.   I'm betting on
> the fact that as users "root around" the documentation, they'll catch on to
> the fact that this is the ASF they will pick up the community.
> 
>  I think a good strategy is to make the landing page for Maven as simple as
> possible, with a good short sell, as little text as possible, link to
> download, and some news and announcements.  The Maven launch page should be
> very similar to the Mozilla launch page - there are still links to the
> developer pages.
> 
> Instead we've got a link named "Netbeans Module", a link on the top of the
> page to "Maven" followed by a link to "Maven 2.x" and "Maven 1.x".  All the
> links have a 20% change of having a completely different skin.  There are
> links to a Wiki and external sites that are not labeled as such.  There is a
> "Project Team" report (http://maven.apache.org/team-list.html) that doesn't
> list have of the people involved in Maven.  Did you know there are no
> contributors to the Maven project and that there are only 8 people working
> on Maven"?   For some reason, I know what the "Actual Time" is for all the
> people on the project team down to the second (I've never figured the need
> for that field out), but good luck customizing the team-list support. :-)
> Here's another good one, click on a Guide, then click on the banner logo.
> That links back to /guide, click on the same banner logo, that links back to
> the home page.   The Maven site is full of these inconsistencies, and it's
> not something that people can just fix with patches.  IMO, the site plugin
> needs serious rework.  It doesn't create good web sites, especially
> for multi-projects.
> 
> I guess I've just pissed off the Maven development team by telling all of
> you that Maven doesn't do a good job of creating web sites for
> users.  There, I said it.  Maven creates so-so web sites for developers, but
> doesn't do a good job creating web sites for end-users.  Maybe that was
> never the goal.  Maybe I should just shut up and write the
> maven-end-user-site plugin?  But, it's one reason why I didn't get too
> exctied about contributing to the site last year.  I don't think another 100
> APT documents are the right direction, I think you need to think long and
> hard about the audience, and I think someone other than Hani Sulemani needs
> to say "Maven sites suck"   :-)  The fact that one of the core committers
> for Maven, had to send out an email entitled "Making the current web site
> suck less", is not a good sell for generating sites with Maven.
> 
> 
>>>
>>> On 3/8/06, John Casey <[EMAIL PROTECTED]> wrote:
>>>> +1
>>>>
>>>> Maybe we can put a banner at the top of each page that marks the
>> version
>>>> it refers to or something. If we styled the reference doco as a manual,
>>>> it could be part of the page layout that ties it together. I'd be
>>>> willing to trade a bit of the look&feel for that sort of information,
>> as
>>>> it seems to me that it would reduce confusion.
>>>>
>>>> -john
>>>>
>>>> Tim O'Brien wrote:
>>>>> Having to choose between publishing the latest and greatest docs and
>>>> only
>>>>> the released version is a problem that Maven seems to have created for
>>>>> itself.  Same issue comes up in other projects frequently - Commons
>> has
>>>> a
>>>>> problem because some of the sites only publish on a release.  Latest
>> and
>>>>> greatest are almost never there.
>>>>>
>>>>> What about publishing the latest and greatest docs to another
>> directory?
>>>>> The Maven site gets pushed to a directory that has a version of a
>>>>> label.  http://maven.apache.org/version/1.0
>>>>> , http://maven.apache.org/version/2.0.2, and
>>>>> http://maven.apache.org/version/trunk.  This way the Maven site can
>> have
>>>> a
>>>>> nightly publish of the most current Maven site to Trunk every single
>>>> day,
>>>>> but still keep legacy docs around intact for people using older
>> versions
>>>> of
>>>>> the product.  The "consumer" site can point to the latest release, and
>>>> the
>>>>> "developer" site can point to "trunk".  The Maven site plugin would
>> need
>>>>> some mechanism for adding a skin to a site to clearly identify it as
>>>>> "Development".
>>>>>
>>>>>
>>>>>
>>>>> On 3/7/06, Wendy Smoak <[EMAIL PROTECTED]> wrote:
>>>>>> On 3/7/06, Brett Porter <[EMAIL PROTECTED]> wrote:
>>>>>>
>>>>>>> * I'm still a little torn on where plugin docs go. No hurry on this,
>>>> but
>>>>>>> something to ponder. We definitely need to make the references for
>>>> those
>>>>>>> integrate better. Site/skin inheritance will help
>>>>>> No matter where they go, I think they need to be updated more often.
>>>>>> Random example... the assembly plugin docs are wrong, and have been
>>>>>> that way for months. (it's descriptorId, not
>>>>>> maven.assembly.descriptorId.)
>>>>>> * http://maven.apache.org/plugins/maven-assembly-plugin/howto.html
>>>>>>
>>>>>> I would like to see the "latest and greatest" docs on the main site.
>>>>>> Yes, they'll be ahead of the released version, but not by much, and
>>>>>> (hopefully) not for long.When the answer to a lot of "X doesn't work"
>>>>>> questions is "It's fixed in the trunk, use a snapshot," it would be
>>>>>> nice to have the snapshot docs available in a centralized place.
>>>>>>
>>>>>> This also makes it more fun to contribute to the documentation,
>>>>>> because you get to see your work "in print" right away.
>>>>>>
>>>>>> Thanks for updating the main site. :)
>>>>>>
>>>>>> --
>>>>>> Wendy
>>>>>>
>>>>>> ---------------------------------------------------------------------
>>>>>> To unsubscribe, e-mail: [EMAIL PROTECTED]
>>>>>> For additional commands, e-mail: [EMAIL PROTECTED]
>>>>>>
>>>>>>
>>>> ---------------------------------------------------------------------
>>>> To unsubscribe, e-mail: [EMAIL PROTECTED]
>>>> For additional commands, e-mail: [EMAIL PROTECTED]
>>>>
>>>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: [EMAIL PROTECTED]
>> For additional commands, e-mail: [EMAIL PROTECTED]
>>
>>
> 

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (MingW32)

iD8DBQFED8RGaCoPKRow/gARAtP4AKCJQwdMKZ9zLx3TtznXjNspY4L2PQCfSqDt
nRLUZCjKfHinJS/Tog+xJec=
=xRRx
-----END PGP SIGNATURE-----

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to