Really good suggestions, @Alex Harui <[email protected]> . I could add text
and links to the "migrating" page based on what you outlined. I can get
that done sometime this week, but am not going to jump on it today in case
the conversation continues and an enhanced idea bubbles up.

Andrew

On Tue, Oct 8, 2019 at 12:48 PM Alex Harui <[email protected]> wrote:

> I'm going to try to address all of Chris Velevitch's several threads from
> today in this response.
>
> We have a page on Migrating Flex Apps here:
> https://apache.github.io/royale-docs/create-an-application/migrate-an-existing-app/migrate-from-flex.html
> But it is hard to find (took me several clicks to find it).  And it
> doesn't address the questions that I think Chris is asking, which are:
>
> -Which component sets should I use?
> -After I choose a component set, how do I figure out how something that
> worked in Flex works in that component set?
>
> We've offered Chris the opportunity to help with the doc, which would be
> great, but I'd settle for having Chris and others simply add the kinds of
> questions they have so we can add them to our FAQ and/or the docs.
>
> We may need to make it easier for Chris and others to decide on component
> sets first.  The way I think of it now, the trade offs are:
> A) I need the result to be as small and fast as possible and have time to
> rewrite the UI -> Use Basic and Strands and Beads
> B) I want to rewrite the UI to get a more modern look-and-feel -> Use Jewel
> C) I want to touch as little code as possible -> Use MXRoyale and
> SparkRoyale.
>
> If you choose C, you should not have to know anything about Strands and
> Beads.  MXRoyale and SparkRoyale should hide that from you.  Skinning
> should work like it used to.  Your app will not be as small and fast as if
> you spend more time rewriting it to Basic or Jewel, but TourDeFlex seems to
> be running ok.  As Alina and Pashmina finish up, we'll see how a really big
> Flex migration performs in Royale.
>
> It has taken them longer than expected, but for every bug or missing
> feature that is emulated correctly, the time to migrate for the next person
> goes down.  And as 2019 becomes 2020, there will be less time to migrate
> for some folks, so having more people on the emulation helps everyone else.
>
> So, we should figure out where to place information like this so Chris and
> others find it more easily.  Once a component set/migration strategy is
> chosen, it focuses a lot of other questions, such as how to do skinning, or
> how much about Strands and Beads you have to know.  In fact, it would be
> great to find a place to recommend to folks that when they write to us with
> questions that they specify which component set(s) they are using because
> the answer can be quite different based on the component sets involved.
>
> Suggestions from Chris and others as to how to make this information more
> accessible are encouraged.
>
> My 2 cents,
> -Alex
>
>
> On 10/8/19, 8:18 AM, "Carlos Rovira" <[email protected]> wrote:
>
>     Hi,
>
>     I mean that like RO, that's a non visual component used to perform
>     communications, we have other "components" or "code" that can be
>     implemented as beads or strands.
>     The concept behind is that Royale have the minimum code needed for
>     functionality, and compose other parts (whatever the purpose of the
> code
>     could be). So while UI Components use to refer to code that are visual
> and
>     use to be a Button, a List or a Checkbox, strands are more agonistic
> about
>     that particular use case, so seems, IMHO a better way to describe it
> from
>     that point of view.
>
>     El mar., 8 oct. 2019 a las 10:14, Chris Velevitch (<
>     [email protected]>) escribió:
>
>     > So you are suggesting a RemoteObject is a headless component who's
>     > base class only supports the model and controller plugins and that UI
>     > classes extend the headless component to add in the view?
>     >
>     > On Tue, 8 Oct 2019 at 15:52, Carlos Rovira <[email protected]>
>     > wrote:
>     > >
>     > > Hi Chris,
>     > >
>     > > thanks for sharing your thoughts.
>     > >
>     > > IMHO, not always a "strand" is a "visual component". This relation
> is not
>     > > always true. a strand could be a non visual component (for example
> the
>     > > implementation of RemoteObject in the Network library). And a bead
> could
>     > be
>     > > a strand itself.
>     > >
>     > > I think the term component is right in most cases and accomplish a
>     > meaning
>     > > purpose, but strand/beads concept comes to give another subset of
> meaning
>     > >
>     > > just my opinion about this.
>     > >
>     > > Carlos
>     > >
>     > >
>     > >
>     > > El mar., 8 oct. 2019 a las 9:23, Chris Velevitch (<
>     > [email protected]>)
>     > > escribió:
>     > >
>     > > > The use of the terms "strands" and "beads" still doesn't make
> sense to
>     > > > me because they are concepts I have never heard before and it is
>     > > > creating a barrier to acceptance and deepens the learning curve.
> As
>     > > > far as I can tell, it's something to do with visual/UI
> components.
>     > > >
>     > > > The section "Strands and Beads" should ideally be titled "Visual
>     > > > Components" because that section is about visual components and
> is
>     > > > alluding to the fact visual components are standalone microcosms
> of
>     > > > the MVC pattern and the ability to treat the model, view and
>     > > > controller as plugins to the component. The statement that
> components
>     > > > are "strands" is, in my mind, misleading because it doesn't make
> sense
>     > > > to interchange terms components and strands if a strand is a
>     > > > component. In fact, diving into the code, the "addBead" function
> gives
>     > > > the "bead" a reference to the component the "bead" is being
> added to.
>     > > >
>     > > > It is clear from the documentation that "beads" are "plugins"
> and the
>     > > > documentation should be talking about extending components with
>     > > > plugins. All references to "bead" should be replaced with
> "plugin" and
>     > > > all references to "strand" be replaced with either
> "hostComponent", or
>     > > > "parent" or appropriately something else.
>     > > >
>     > > > We seem to be neglecting the rich heritage that we have gotten
> from
>     > > > Adobe Flex and I don't see the point giving existing concepts new
>     > > > names.
>     > > >
>     > >
>     > >
>     > > --
>     > > Carlos Rovira
>     > >
> https://nam04.safelinks.protection.outlook.com/?url=http%3A%2F%2Fabout.me%2Fcarlosrovira&amp;data=02%7C01%7Caharui%40adobe.com%7C56331bd7a799497db41308d74c02c990%7Cfa7b1b5a7b34438794aed2c178decee1%7C0%7C0%7C637061447162721530&amp;sdata=zhUI1U2%2F0jU9Od5Pm2pE%2B6RKrJxm97BafPqJ%2FVxrB%2Bg%3D&amp;reserved=0
>     >
>     >
>     >
>     > --
>     >
>     >
>     > Chris
>     > --
>     > Chris Velevitch
>     > m: 0415 469 095
>     >
>
>
>     --
>     Carlos Rovira
>
> https://nam04.safelinks.protection.outlook.com/?url=http%3A%2F%2Fabout.me%2Fcarlosrovira&amp;data=02%7C01%7Caharui%40adobe.com%7C56331bd7a799497db41308d74c02c990%7Cfa7b1b5a7b34438794aed2c178decee1%7C0%7C0%7C637061447162721530&amp;sdata=zhUI1U2%2F0jU9Od5Pm2pE%2B6RKrJxm97BafPqJ%2FVxrB%2Bg%3D&amp;reserved=0
>
>
>

-- 
Andrew Wetmore

http://cottage14.blogspot.com/

Reply via email to