Re: website draft 4, help wanted

[Graham Percival]

On Wed, Jul 08, 2009 at 11:54:05AM +0200, Jan Nieuwenhuizen wrote:
> On di, 2009-07-07 at 23:26 -0700, Graham Percival wrote:
> 
> Hi Graham,
> 
> > I disagree; I think we've spent *too much* time on the look/feel
> 
> Yes, I understand that's how you feel. It's exactly why I sent my mail,
> and why I reacted so strongly on this.

Trying to prompt me to work harder, I see... it's not necessary.
I was visiting my old university last weekend; my lower output
wasn't related to a lack of frustration.


> > Yes, I know that you want stuff that's attractive.
> 
> What is the main public you have in mind when you set your priorities
> for the website?

In order,
- existing users who might be pursuaded to become contributors.
- new users who can't figure out how lilypond works.
- other people (such as my family) who don't know what lilypond
  is.

> Looking at our future user base, I feel that 1 is *by far* the largest
> and most interesting group

Ok, that's nice.  What, precisely, is your problem with the
extensive Introduction materials we have?

> I mean, if we somehow manage to scare off all people in category 1,
> our community will stop growing?

No.  Our community will die IF WE DON'T HAVE MORE CONTRIBUTORS.
That's why I put them at the top of *my* list.

> In an slightly indirect way, the best thing we can do for our current
> users, be it a category 2 or 3 user, is to grow our community?

As long as potential contributors can easily find info about
contributing, then I'll agree that adding more users is good.

> In our current culture, things seem to be speeding up, esp. the
> internet experience often enjoys very short attention moments.
> When you are looking for something, the internet can be very big,
> you do not have the time to explore everything in depth.  Sometimes,
> you only take a few seconds to decide: this thing is crap, this
> this may be it.

Then it's a good thing that our Introduction is so nicely laid
out.  Potential users can see at a glance whether they want to
read about Features, see Examples, etc.

> So, what I would like, is to A) have a front page where at
> first glance, as many category 1 persons as possible will
> hold in their tracks long enough to find  I) just the right
> next page to read, and D) get more interested.  Be it
> the essay, the crash course, tiny or complex examples.

I) and D) are already done!!!

> Do I make any sense?  I do not want to force anything on
> you, I really like what you already did, it is very important
> and it is probably impossible to make something that looks
> good without sensible content.  However, please reread this 
> rant until you can freely decide to take over my priority list? :-)

No, that's not my job.  (see below)

> Possibly we can get Valentin to step over his 'there
> are cool things I cannot do with css, I cannot do anything'
> and hack a CSS + header + margins that are as crisp as
> http://news.lilynet.net in just a day or two.

I am accepting patches to lilypod-general.css.  Or rather, Patrick
will be accepting them.

> And I mean, do we have big mentioning, pictures/screeshots from
> his opera?

We already have a section for Productions, or something like that.
If users contribute material (that stuff goes to Jonathan), then
we can have a beautiful list, with pngs of notation, photos from
the stage, etc etc.

IF USERS CONTRIUTE MATERIAL, that is.

> Valentin's site attracts my initial attention and gives a professional
> feel.  Much more than both the lilypond.org sites.  Possibly he/we
> could/should  merge this site even with lilypond.org.  I don't know.

Content is separate from presentation.  If Valentin, or you, or
Flozzy the Fairy Funky Flutist, sends patches to the CSS
(including images to load from the css), that's great!


Now for *my* rant.

As I've said before, I stopped using lilypond 4-5 years ago.  Last
Fall, I briefly got back into improving the engraving of my old
compositions, but that stopped when I went to Singapore.

So why am I still around?  I'm really bugged by inefficiency.
Users ask questions, Mats gives the same answers over and over
again: inefficient.  So I improved the docs.

I decide to leave, but remember that it took me two weeks to get
my first patch, it was horribly frustrating, and it was horribly
inefficient.  So I started GDP to train my replacements.

Potential contributors can't figure out git, ask questions, we
have the same confused discussion over and over again:
inefficient.  So I started the CG.

Potential programmers can't figure out how to get started, and the
existing programmers have learned that most well-intentioned
offers of programing never pan out when they have to actually do
work, so they ignore all those requests about getting started:
inefficient.  Another reason to start the CG, and to continually
nag people to dump instructions in the CG.
(we direct new contributors to the CG, so that they can
demonstrate their willingness and ability to do work.  Then we
know that it's worth spending hours helping and training them,
rather than hoping that an unknown contributor will be in the
25-40% of new contributors who will actually stick around)


You know what kind of website I think we should have?  I think we
should have whatever website the community is willing to make.  So
far, "the community that's willing to make" a website consists of
me, Patrick, Jonathan, and Francisco.  Plus many others who give
suggestions... but when it comes to actually sitting down and
writing the content, hacking the perl, and whatnot, it's just us.

Which, in itself, is a horrible waste.  Every hour that I spend
working on the website is an hour that I'm not working on writing
instructions about how to make releases.  Whenever I retire from
building releases, I don't want my replacement to spend *another*
5 months trying to figure out how to build the regtests correctly
or run the upload script.

Every hour I spend on the website is an hour that I'm not
rearranging LM 1 and the AU.  You know, we had a sincere offer to
help write docs for alternate editors, 2 or 3 weeks ago.  But
*that* can't be done until I've made the space for that in LM 1.

Every hour I spend on the website is an hour that I'm not
investigating+discussing Reinhold's ajax-search stuff, which I
*promised* that I'd do "in 1 week" on June 23.

Every hour I spend on the website is an hour that I'm not checking
over the CG, which needs to be done before we can start GOP, which
is a (the?) prime time to gather more contributors.

Every hour I spend on the website is an hour that I'm not writing
another column for the LilyPond Report, which may or may not be
holding up the next issue.

Every hour I spend on the website is an hour that I'm not dealing
with the 58 emails containing material which should be added or
fixed in the docs.  (and yes, all your GDP graduates reading this,
I _am_ slightly miffed at you.  I spent a lot of time training you
21 people so that I wouldn't have to deal with this stuff any
more)

Every hour I spend on the website is an hour that I'm not starting
my own Tadpole stuff, which is particularly annoying since I was
really hoping to start doing programming bugfixes myself, this
summer.


Given all those tasks... and all the potential contributor effort
that's waiting for me to finish those tasks... it really doesn't
make sense for me to be gathering a list of productions and
nagging famous lilypond users to send pictures.  Which,
thankfully, I've offloaded to Jonathan.  But it also doesn't make
sense for me to be rewriting the Old News from the current website
into texinfo to add to the new website, but I'll probably end up
doing that.

In some ways, I actually cringe more when I see Patrick working on
the website.  Not because he doesn't do a great job of the CSS and
perl hacking (he *does* do a great job of this), but because he
could be working on so many other things.  CSS isn't hard; there's
dozens of users that _could_ be doing it.  But improving the SVG
backend of lilypond _is_ hard.  Nobody else is doing _that_.
Fixing build bugs in GUB is hard; you and he are the only people
currently working on that.  Frankly, if the really annoying
lilypond bugs like #34 (grace notes) ever get fixed, there's a
good chance it'll be done by Patrick.  So why the bloody mao is he
the main person working on the CSS?  We should get a relatively
inexperienced contributor / web-savvy user to do that stuff.


*That* is why I'm not going to spend (significant) time screwing
around with the appearance.  If somebody is seriously interested
in this task, I'll gladly mentor them.

You know the phrase "in a democracy, the population receives the
government they deserve"?  That's how I feel about lilypond.  The
community receives the program/documentation/website they deserve.

- Graham