Re: javascript transpilers

Mike Dewhirst Wed, 31 Jan 2018 16:35:57 -0800

On 1/02/2018 9:50 AM, Matthew Pava wrote:

I just want to point out something called WebAssembly.  It's supported across 
all modern browsers.  Presumably, you can write a program in Python, convert it 
to WebAssembly, and it will run in your browser.  I haven't worked with it 
myself, but I am keeping my eye on it because I would like to think it will 
become more widely adopted in the not-so-distant future.
http://webassembly.org/


Cute demo!

If all the browser makers have agreed to make it work that does make itinteresting. But they all agreed on Javascript as well. I guess theysaid they have one scripting language and don't need Python as another.Better to introduce a compiler. What could possibly go wrong? All weneed now is a Python to WebAssembly transpiler ;)


Cheers

M



-----Original Message-----
From: django-users@googlegroups.com [mailto:django-users@googlegroups.com] On 
Behalf Of Mike Dewhirst
Sent: Wednesday, January 31, 2018 4:29 PM
To: django-users@googlegroups.com
Subject: Re: javascript transpilers

On 30/01/2018 6:33 PM, Andréas Kühne wrote:

2018-01-30 5:18 GMT+01:00 Mike Dewhirst <mi...@dewhirst.com.au
<mailto:mi...@dewhirst.com.au>>:

     On 29/01/2018 9:03 PM, Nick Sarbicki wrote:

         Going SPA when you don't know JS is very much a baptism of fire.

         Angular is personally my favourite too and essentially forces
         the SPA model. But it has a much bigger learning curve
         compared to Vue and React (the other two common names thrown
         about nowadays). The others are also easier to not force into
         a SPA model.

         Figure what's the minimum requirement first ;-).


     Single page is pretty close to the way the application currently
     works so that is promising. It is a lightweight training
     framework. Simple architecture of courses, instruction(s) and
     questions with answers aggregating a total score for the student.

     I won't publicise the URL because the server isn't heavy-duty. But
     if there is any individual interest I'm happy to give it off-list.
     Today I released the beta to the first client who (generously) is
     using it for his troops to do online training on how to use some
     Windows software.

     So back to the minimum requirement ...

     The single (question) page has previous-question and next-question
     links which morph into previous-instruction and next-instruction
     links when at the first and last question respectively. So that is
     the SPA similarity.

     When you click the [Save] button after answering a question there
     is no indication the answer has been saved without adding a line
     of text (via the view) after answer.save()

     Refreshing the page resends the [Save] and reprints the line of
     text saying the [Save] was successful. Too clunky.


If you are using django in the "normal" way - you should just be able
to add the message to the messages contrib package. Then the message
should show when saving? Because when you click on the save button you
should post your form and that method should (when successful) add the
message?

Hadn't thought of that! Must investigate the messages package.

     It was that behaviour which made me start this thread.

     I checked out typescript and it looks great but it is javascript.
     It seems to let you code in ECMAScript 6 which is looking more
     like Python and as Nick mentioned, has a compilation step (which I
     suppose is like a minification step anyway) to get it into the
     version currently supported by all browsers.


Typescript is more of a object oriented language than javascript is -
at least compared to ECMAScript 5. It also has some support for typing
- and great unittest support - which is criticle when writing
javascript....

Agreed

     That prompted me to look more closely at a few of the transpilers
     in the link I posted in the thread-starter. One stood out for me
     ... http://www.transcrypt.org/

     It seems the compilation to javascript takes place in the CI
     process and it can be debugged in Python or Javascript. You can
     use both javascript and/or python libraries. Written by a Dutchman
     which is possibly interesting ;)


What you have to remember is that you still need to transpile it when
running in development mode. So it really doesn't matter where you do
your transpilation. You need to setup the transpilation even in
development - a webb browser can't run python natively.

True. However, Transcript produces Javascript. Interestingly you debug in 
python or Javascript.

I'll have to allocate some time to investigate but in any case I think DRF 
comes first.

     The Angular thing is interesting.
     
https://medium.com/@chriscordle/why-angular-2-4-is-too-little-too-late-ea86d7fa0bae

<https://medium.com/@chriscordle/why-angular-2-4-is-too-little-too-lat

e-ea86d7fa0bae>

     
https://www.infoworld.com/article/3213244/javascript/whats-new-in-angular-5-easier-progressive-web-apps.html

<https://www.infoworld.com/article/3213244/javascript/whats-new-in-ang

ular-5-easier-progressive-web-apps.html>

     >From those articles, it seems React is lighter-weight than
     Angular and for me that probably means easier to learn. It is
     still daunting though. Vue also assumes javascript know-how.


Yes react is more lightweight. However it is like comparing the flask
framework with django. The difference is very similar. React is more a
library and angular is a framework.

I suspect React is super efficient. Facebook use it and they shave very close 
to the bone.

     I think my best plan is to postpone a decision and make a start on
     DRF anyway. No matter what happens that is a prerequisite.

     I hope it isn't wishful thinking but Transcrypt looks more
     attractive than learning javascript plus a javascript framework.


The main problem here is that you will still need to learn javascript
- because you need to understand what will happen when your
client-side code fails (and it will :-)). And my other point is still
valid. Look at the user base. When you have problems with transcrypt,
how many users can you rely on to get help? If you have a pure
javascript problem - how many users can you rely on for help?

Excellent points. But it is difficult to dislodge wishful thinking.
There will be two sets of code. Python which I hope I can debug and 
automatically generated Javascript which definitely will (I agree) have bugs. I 
would have had to debug Javascript anyway and I reckon forced debugging 
Javascript will let me skip the awful step of deciding to learn Javascript. I 
mean that's a depressing decision to have to make.

Maybe by the time I get DRF working Transcrypt will have a few more users? I 
really want to avoid Javascript. And I know Transcrypt would not have been 
written unless the authors felt the same way. If it works there will be a flood 
of users. It comes with a lot of promise(s).

I do hope that you find a solution that works for you though :-)

Haven't made any decisions yet - apart from DRF and the messages investigation.

Many thanks for helping :)

Mike

     Thanks to all for the excellent and very professional feedback.

     Cheers

     Mike


Regards,

Andréas



         On Mon, Jan 29, 2018 at 9:15 AM Andréas Kühne
         <andreas.ku...@hypercode.se
         <mailto:andreas.ku...@hypercode.se>
         <mailto:andreas.ku...@hypercode.se
         <mailto:andreas.ku...@hypercode.se>>> wrote:

             If that is the case - I would recommend looking into
         creating a
             single page app - with Django restframework on the backend
         if you
             want to continue with Django. For the frontend you will
         however be
             required to learn some sort of JS framework - however this
         doesn't
             have to be THAT bad. I personally like angular 5 - which is
             written in typescript - a javascript based language - that
             actually removes most of the pain. HOWEVER JS knowledge is
         good to
             have anyway....

             There are other frameworks that don't require that much JS
             knowledge and Jason suggested GWT which sounds like it
         could be
             worth a shot, as long as it doesn't add to much
         complexity. The
             main reason for using that would be a large user base, so
         you can
             get help. The python transpiled JS versions don't really
         have the
             greatest user base :-)

             Regards,

             Andréas

             2018-01-29 0:00 GMT+01:00 Mike Dewhirst
         <mi...@dewhirst.com.au <mailto:mi...@dewhirst.com.au>
             <mailto:mi...@dewhirst.com.au
         <mailto:mi...@dewhirst.com.au>>>:

                 That seems unanimous :(

                 Jani asked what problem I need to solve. Without going
         into
                 analytical detail, all the code is serverside and now the
                 project needs to move towards AJAX for user experience
         reasons.

                 M

                 On 29/01/2018 5:44 AM, Nick Sarbicki wrote:


                     Agree with everything already said.

                     Only thing I want to add is that while js has
         historically
                     been a horrible language, the modern standards
         have helped
                     rectify this. Modern js can be almost pleasant to work
                     with nowadays.

                     I wouldn't spend time transpiling python to js.
         Just try
                     to learn js and maybe transpile from a modern
         standard to
                     the browser standard. It's almost as easy as python to
                     pick up.

                     Django compressor helps greatly in the last case.


                     On Sun, 28 Jan 2018, 16:49 Avraham Serour,
                     <tovm...@gmail.com <mailto:tovm...@gmail.com>
         <mailto:tovm...@gmail.com <mailto:tovm...@gmail.com>>
                     <mailto:tovm...@gmail.com
         <mailto:tovm...@gmail.com> <mailto:tovm...@gmail.com
         <mailto:tovm...@gmail.com>>>> wrote:

                         If you don't want and don't have the time to
         learn JS
                     you may as
                         well consider outsourcing the frontend work, I've
                     worked in places
                         that did just that and everybody was happy, the
                     internal team
                         liked python and the outsourced company people
         liked
                     UX and JS.

                         On Sun, Jan 28, 2018 at 6:09 PM, Jani Tiainen
                     <rede...@gmail.com <mailto:rede...@gmail.com>
         <mailto:rede...@gmail.com <mailto:rede...@gmail.com>>
                         <mailto:rede...@gmail.com
         <mailto:rede...@gmail.com> <mailto:rede...@gmail.com
         <mailto:rede...@gmail.com>>>>
                     wrote:

                             Hi,

                             You don't describe to what extent you need
         to use
                     JavaScript -
                             IOW, what problem you need to solve with
         JavaScript?

                             Depending you needs you might get away
         with basic
                     knowledge
                             and usage of some helpful JavaScript
                     library/framework.

                             On Sun, Jan 28, 2018 at 12:44 AM, Mike
         Dewhirst
                             <mi...@dewhirst.com.au
         <mailto:mi...@dewhirst.com.au>
                     <mailto:mi...@dewhirst.com.au
         <mailto:mi...@dewhirst.com.au>>
                     <mailto:mi...@dewhirst.com.au
         <mailto:mi...@dewhirst.com.au>

                     <mailto:mi...@dewhirst.com.au
         <mailto:mi...@dewhirst.com.au>>>> wrote:

                                 I have avoided javascript like the plague.
                     However it
                                 seems I have to bite the bullet.
                     Interestingly, there are
                                 now a number of Python -> Javascript
         transpilers.

         http://stromberg.dnsalias.org/~strombrg/pybrowser/python-browser.html

<http://stromberg.dnsalias.org/%7Estrombrg/pybrowser/python-browser.ht

ml>

<http://stromberg.dnsalias.org/%7Estrombrg/pybrowser/python-browser.html<http://stromberg.dnsalias.org/%7Estrombrg/pybrowser/python-browser.ht

ml>>

<http://stromberg.dnsalias.org/%7Estrombrg/pybrowser/python-browser.html<http://stromberg.dnsalias.org/%7Estrombrg/pybrowser/python-browser.ht

ml>>

                                 Where would be a good place to start a
                     discussion on the
                                 topic?

                                 I am self-employed  with multiple
         demands on
                     my time so I
                                 cannot afford to allocate sufficient
         time to
                     learn it let
                                 alone undertake formal Javascript
         training. My
                     hope is to
                                 pick it up by osmosis and wonder if a
                     transpiler would help.

                                 Thanks for any pointers

                                 Mike

                                 --             You received this message
                     because you are subscribed to
                                 the Google Groups "Django users" group.
                                 To unsubscribe from this group and stop
                     receiving emails
                                 from it, send an email to
         django-users+unsubscr...@googlegroups.com
         <mailto:django-users%2bunsubscr...@googlegroups.com>