Re: [Cdk-jchempaint] [Cdk-devel] Javascript JCP

[gilleain t. <gil...@gm...>]

Ok, forgot to mention the standalone app use-case. Sure - that won't go
away. Indeed, much of the drawing code is in the core CDK anyway, to allow
Bioclipse to use it. On top of that is 'just' a swing app with some
buttons, I suppose. I know there's more to it - but don't worry, that
should remain as-is, for sure.

The only use-case that wouldn't be handled would be a web page with just a
sketcher, with no back-end. That would be a shame - especially considering
the 2D-to-3D examples that Jmol/JSmol has been experimenting with.

At the moment, I'm just trying out various technologies to see which is
nicest to use. I'll let people know when there is something too look at.

thanks,

Gilleain

On Mon, May 22, 2017 at 1:07 PM, Nina Jeliazkova <jel...@gm...>
wrote:

>
>
> On 22 May 2017 at 15:00, John Mayfield <joh...@gm...>
> wrote:
>
>> For a web front end -  there are other JS chemical editors out there
>>> indeed.
>>
>>
>> Ah... but a lot of them are pretty rubbish :-).
>>
>
> Very much agree ;)   [1]
>
> My point is different - JCP as a java dependency for standalone
> application is a valid use case - and it is better to be kept as such,
> without introducing web service dependency.
>
> (also agree that single object model is not necessary the right choice for
> all use cases)
>
>
> Regards,
> Nina
>
>
> [1] https://twitter.com/10705013/status/448160308042162177
>
>
>> We were trying to find once recently at NextMove to handle reactions and
>> basically with the exception of MarvinJS they were all terrible move agents
>> to reactants etc add superfluous grouping.
>>
>> John
>>
>> On 22 May 2017 at 12:55, John Mayfield <joh...@gm...>
>> wrote:
>>
>>> 2) Port (cross-compile? transpile?) JCP to Javascript
>>>
>>>
>>> This would be my preference, but my understanding is with GWT etc would
>>> essentially be a re-write. IIRC MarvinJS (https://www.chemaxon.com/prod
>>> ucts/marvin/marvin-js/) is done using GWT but you'll notice it's very
>>> different from their MarvinSketch applet.
>>>
>>> I think it's best to do a lighter weight model for the sketcher with the
>>> "chemical intelligence" handled by WS as you suggest. I actually don't
>>> think a sketcher should use the same object model (OM) at all. There are
>>> lots of things useful to capture in a sketch which just doesn't make sense
>>> to represent in data.
>>>
>>> After working with it a fair bit I feel ChemDraw have basically nailed
>>> the OM required for sketching, it's not perfect but very close and captures
>>> many intricacies.
>>>
>>> e.g. Character attach information allows you do the following:
>>>
>>>       OH
>>>>       |
>>>> -C(CH2CHO)-
>>>
>>>
>>> e.g. Which side of a bond the double bond line goes:
>>> http://www.cambridgesoft.com/services/documentation/sdk/chem
>>> draw/cdx/properties/Bond_DoublePosition.htm
>>>
>>> John
>>> ​
>>>
>>
>>
>> ------------------------------------------------------------
>> ------------------
>> Check out the vibrant tech community on one of the world's most
>> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
>> _______________________________________________
>> Cdk-devel mailing list
>> Cdk...@li...
>> https://lists.sourceforge.net/lists/listinfo/cdk-devel
>>
>>
>
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> Cdk-jchempaint mailing list
> Cdk...@li...
> https://lists.sourceforge.net/lists/listinfo/cdk-jchempaint
>
>