Hi Phil,
While Google may seem like it has infinite resources to coordinate and
build bridges between all of it's projects, it isn't remotely true. Google
is a collection of teams and individuals that have their own set of goals
and priorities and of course, time limitations, that also fit together
within the greater direction of their larger organizations and company.
It's simply not feasible for every project to even share the same goals,
much less be seamlessly integrated even if they did.
Ultimately, Blockly and Polymer don't necessarily share the same goals or
audiences.
Blockly is indeed a cool project, especially as a teaching tool. It's been
around even longer than 3 years as part of various project Neil has worked
on, and at least I am keenly aware of it and have talked to Neil several
times before. Blocky attempts to make syntax errors impossible, and to
guide novice programmers towards the discovery of constructs that are
compatible with each other. Blockly is still an imperative programming
language, just not a text-based one, and while it limits syntax errors it
does much less to prevent logic errors. Blockly is also not very well
known, not exactly fast to program in for experienced programmers, and not
likely to be a popular choice for experienced engineers, nor designers and
more general web developers that we target with Polymer, it's mostly
targeted at education.
Polymer is not trying to completely replace JavaScript, but to offer an
easier way to build custom elements. Our philosophy and methodology include
some aspects, like declarative templating and data-binding, that do
replace imperative programming (this is language-agnostic even) for some
focused tasks, but for other tasks programmers still use JavaScript. If you
wish you can use a compile-to-JS language like Dart, TypeScript,
CoffeeScript, etc., but we by default use the lingua-franca of the web, and
feel strongly that this is the productive choice to reach and be usable by
the broadest audience.
And so, we have limited resources, we should spend them on things that will
be the most useful to the most users in our audience. Right now that
includes things like more elements, testing infrastructure, performance
tuning, addressing high-demand features in data-binding, styling and
layout, modules, etc., and my personal favorite: Polymer Designer. We know
there's more to do than that (and it's not an exhaustive list of what we're
working on), and if you're frustrated that we haven't done the thing you
care about, we're more frustrated that we don't have infinite time.
We are very sympathetic to the idea that programming and building web apps
should be easier. It's why we want elements to be usable by
non-programmers. It's why some elements simply wrap other APIs to make them
usable via markup and data-binding. And it's why we work on Polymer
Designer.
If Blockly can be shown to fit well in this world of custom elements, then
I'm sure it's something that interested community members can drive, and we
would help out as much as we can. I certainly can see how it might be used
to build the logic portion of elements.
I'm personally interested in non-imperative logic building for use in
Polymer, but my current experiments focus on a non-imperative dataflow
programming approach that fits nicely with our data-binding system. If
you're interested in that, I'll be sure to announce any interesting
developments in the usual places.
Cheers,
Justin
On Mon, Jun 1, 2015 at 1:51 PM, phil cleaver gigglestheth...@gmail.com
wrote:
I'm the biggest fan of Google and everything they have done. Google
Blockly and Polymer are my absolute two favourite things that's going on
online right now. Google tells us that with Polymer they have completely
reinvented the way the web works, and they have. What they have done with
Polymer is mind-blowingly awesome. It does away with most of the
JavaScript needed to write apps and now you do it all in HTML.
Throw in Firebase, another Google company and they have utterly and
completely made all our lives (as long as you're a Web Developer) far far
easier. It's nothing short of genius.
Polymer has been around for 2 years now, Blockly has been around for at
least 3 years, so why oh why am I struggling to make these two things work
together? It is completely unfathomable to me that a company with as many
resources as it has, a company so ahead of it's time has made 2 completely
awesome products, yet has apparently spent no time getting the left hand to
talk to the right one, and linking them up.
There is a Blockly Web Component at
https://github.com/gmp26/blockly-element. It kind of works, but there's
no data-binding on the workspace, or on the generated code, making it
(almost) totally useless for anything serious. It works when the
blockly-element is in regular DOM, but put 2 of them in shadow DOM and see
if you can find the workspace xml? I've spent all in all probably a
