royale-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Carlos Rovira <carlosrov...@apache.org>
Subject Re: Royale and Websites (was Re: About website actual work in progress)
Date Mon, 01 Jan 2018 15:44:35 GMT
Hi Alex and all, And first of all Happy new year to all of you! Hope 2018
be a great year for all of you! :)



2017-12-30 8:50 GMT+01:00 Alex Harui <aharui@adobe.com.invalid>:

>
> I agree with what you say.  What I am suggesting is making enough changes
> to the site to attract people who can help us create a "most usable
> product".  I don't think we have enough people to do that as fast as I'd
> like, so I'd rather tweak what you have to do what I see in many shopping
> centers:  a sign goes up saying that Store XX is coming soon and they are
> hiring.  Only later does that sign come down and they have a grand
> opening.  Similarly, I would like to tweak enough of what you have to say
> that we need "pioneers" and folks who like the bleeding-edge.  Even one or
> two more folks who can work well with others would be great.  I haven't
> gone through every page you've done, but on the main page, I think we only
> need to change the NPM section in some way.
>

IMHO, and talking about my experience with my own business and other stuff
I tried to build and market in my professional life:
making noise of a product that is not enough mature, use to result in a
very bad strategy that make the opposite effect and make
people run away from it instead of buy or adopting it.

So, my plan would be to focus in make Royale do what people coming would
expect to do, and maybe here is where we have some
discrepancies on what we are trying to achieve.

So don't worry now about website. It only needs some tweaks, and little
work to make it ready to publish.
We need to talk about things in Royale plan itself.



> One reason for my mockup of the website in MXML was to make sure folks
> truly understood the power of the extensible component model for Royale.
> You are still focused on traditional interactive Applications, and I agree
> that migrating Flex apps is a key market for us, but you also seemed to be
> interested in folks creating new projects, and I wanted to illustrate that
> Royale, with MXML and an extensible component model can be a benefit there
> too.  I may be crazy, but I believe that Royale can be useful in many more
> places than what folks think of as traditional Applications.  I still
> think we want to use Royale to produce our web site some day (which
> doesn't have to be now).
>

I agree. I believe Royale could have a target far more open that Flex has,
but I think we have here a warning:
people coming to Royale will not search for a tool to make website on the
first term. They will come to make
things in HTML like they produced with Flex (or Angular, or something
similar). But if they succeed, they will
want to use Royale for all other things. So focusing on something that is
still not the min focus is a mistake for me.
We should focus in make Applications as easy as we did with Flex, and that
is not happening right now.
I tried to start building an App with Royale for the theme feature and get
stuck as I started with Slider (I still couldn't
get back to it and see Peter's changes). So that's for me where we are
failing. We need to have components that work.
And right now as we start working, things are still not working. That's my
main point. If I want to start working in styling and
Theming but the basic components are not working, think in people coming to
us for the first time...Hope you could
figure their opinion about Royale although we have a great and good looking
website, or even if we can build some
basic website. They simply will gone since what they expect to get out of
the box is not working.



>
> In mocking up the site, I realized there might be content in your proposed
> site that is not ALv2 compatible, so I am suggesting changing that now,
> otherwise, it may be much more noticeable the day we want to cut over to
> the Royale version of the site, since the Royale version will probably
> have to release-able as an Apache release.


> I'm in no hurry to switch over to the Royale version of the site.  I just
> want us to take a few minutes here and there to continuously improve it.
> I still don't know where the line is between theme and content, but if the
> colors and fonts aren't part of the theme and you can tell us what you
> used, we can make sure the fonts will be ALv2 compatible and the Royale
> mockup will look a little better.
>

That's what I want to said that we have no problem at all with licenses on
the site.
If you read the ticket we created with legal they express full support of
our website in the current
state, so we can publish it and have it as long as we need. And when we
reach the point
where we can change to a Royale website we can do that (tomorrow or 5 years
in the future).
For that reason, don't worry about this and go with the things that our
users really wants for us


>
> Meanwhile, you've been saying for a few months now that you want a better
> UI set, but really, I have no idea what work needs to be done when you say
> that.  I don't doubt that there are bugs and missing features in our UI
> set, but IMO, we are not staffed with a QA team nor are we staffed to do
> things for "completeness".  We pretty much just try to do what potential
> users ask for on the mailing list.  Priority is given for folks migrating
> Flex apps.  Someone asked for modules so I put together basic modules.
> Someone asked for a TreeGrid, Peter put one together.  Harbs put something
> in about Validation.  Someone asked about I18N, I showed a way to do that.
>  If you want to migrate an app (even Tour de Flex) and can show us
> specifically what is broken, we'll try to fix it (or better yet, help you
> fix it).
>

To make theming and style working I think we need to team to make this work
you or Peter have a huge skill in architecting but don't have UX skills. In
the other
hand, I can make UX thing work, but If I find stuck with something that
doesn't work (i.e: Slider)
I can lose lots of time figuring it and fixing it. So, my best bet is: we
should join. We can
choose a component (i.e: Slider) and try to build an App Example that use
our framework changes
to introduce styling and themeing. And try to make it good as better as we
can to match
most of what Flex did in the Slider case, then go with other and so on...

In the end, a new user coming to us, should start building with Slider
directly without find anything that
make stop their work. That should be our goal, at least with 15-20
components.


>
> Of course it would be better if we could match Flex 4.6, but I'll be happy
> if we can approximate Flex 1.0.  That's where it all started.  And Flex
> 1.0 didn't have modules or I18N or a TreeGrid.  IMO, we have to be clever
> and smart about where we spend our energy and time and continuously
> improve and continuously try to recruit new people.  We can't be like a
> corporation with a fixed staff that builds something with limited public
> input and then launches it.
>

I like the approach of the strand-beads, or PAYG, but in the end that
should be something that will make
people choose as they build. The problem right now is that we want people
come to us and we have lots of pieces
but many of them doesn't work well, or directly doesn't work at all. PAYG
and compositions vs inheritance will be
a key factor in we reach something that works like Flex (or near it), We
can target Flex 1.0 since in that version
most of components worked pretty well, Button, TextInput, Slider, List,
ComboBox, CheckBox, RadioButton.
I was trapped by Flex since I started to build, and things worked out of
the box!! That's not happening with Royale right now,
and we thing that people will understand that and expect they use all that
is done so they can build on their own, but
that will never happen since people expect to adopt a framework that make
his life easier, and Royale in the current state
will make the opposite.

>
> I hope we do have different things in mind.  I know my mind does not have
> all of the answers.  We need good ideas from different people, including
> yourself.  We have this mailing list to try to get some level of mutual
> understanding, but folks are still free to scratch their own itch.  We
> don't even need to all agree, we just have to try to not get in each
> other's way, and try to help folks succeed in scratching their itch if it
> makes sense.
>

Right, we only need to reach a state where we can make enough noise to make
people come (publish website, spreading the word to the four winds,...)
but we should ensure we have something that really works or the result will
be people saying very bad things about Royale, and that's not what we want.
We want the opposite. For that reason this is very important.



>
> So, I don't know how much time you have, but in summary my requests of you
> would be (in order):
>

going through this list

>
> 1) Make some tweaks to the site to try to recruit more committers instead
> of users
>
This is difficult since what we are supporting Royale is currently what we
are and
all know what we want to get. getting more people on board will happen when
people
try us, see Royale work and want to improve it.

> 2) Tell us the colors and fonts used on the site and whether you chose
> those colors and fonts or whether they are part of the theme
>
As I said, colors and fonts are chosen by me. But that's not a problem,
even nothing about licenses in website

> 3) Replace ET-Line font in your version of the site or show us that it is
> ALv2 compatible
>
Againg, we don't have license problems as Apache legal give us green light
in the current state.

> 4) Try to build something with Royale so you can be more specific about
> what is missing/broken.  Maybe you could try to make our ASDoc example
> look and work better.  That might expose some things the UI set needs that
> is more tangible.
>
> I can go again over Royale Theme example app and try Peter's changes.
Hopefully the Slider could be working now,
but if that's ok, we should try to team to revise each component and try to
make it working out-of-the-box, in look and feel,
layout, events and so on.

If we don't get that, don't expect Royale to convince people out there, or
don't expect that our component model and features
do that, since they are great if are features build over a good working
technology. That's the main point I want to try to express here.





> I pretty much agree that we don't want to give users the impression that
> Royale is a UI set that is the equivalent of React before it is, but I
> think we need to get a release out soon and try to establish a workflow
> where we can release more often, so we can continuously improve the
> releases.  And the UI set probably won't be as ready as you would like it
> in these first releases.  I just think we need to do that in order to try
> to attract one or two more committers to help us make a better UI set, and
> attract one or two more users who are willing to ride a bumpy road with us
> so they can be a testimonial to attract other users and slowly build up
> momentum. I don't think we can make big leaps.
>

I'm with you that we need a first release as soon as we can to build over
the rest. But only to put the wagon on rails
For me that release only would be something like the actual website. Tools
to prepare Royale to get what it really needs:
A good working UI set that looks good and behaves ok so I can build a first
app without much hassle. At that point all
the tools could be in the end be use to show and spread the word to people.
And that people will find at that time that Royale
works and would eventually join to us as user or as devs to improve it.

Making things in a incorrect order, would make us have a horrible result, I
guess

Thanks

Carlos

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message