[Architecture] code conventions
christophestrobbe at yahoo.co.uk
Wed Aug 29 11:24:17 EDT 2012
Hi Colin, Antranig,
I was mainly interested in a tool that can be embedded in an Ant script, so the link task can be run automatically as part of every build etc, instead of pasting my code in a web-based form. I have created a JSHint task as part of my preparatory work for the smart house simulation. Can your fork of JSLint be downloaded somewhere so I can create an Ant task for it? (I tried the Fluid wiki to find a link, but I couldn't reach it.)
> From: Colin Clark <colinbdclark at gmail.com>
>To: Antranig Basman <Antranig.Basman at colorado.edu>
>Cc: architecture at lists.gpii.net
>Sent: Friday, 3 August 2012, 23:15
>Subject: Re: [Architecture] code conventions
>Digging up an old thread; responses inline...
>On 2012-07-21, at 12:39 PM, Antranig Basman wrote:
>> Personally I have found the Crockford whitespacing rules rather annoying in their full detail but they had been something we had mostly agreed on as a community (Fluid) - and certainly without this feature at all, there is nothing to prevent very badly formatted code which doesn't even operate consistent indentation. The only whitespace rules which are left when "white" is removed are the ones which prohibit mixtures of tabs and spaces on the same line - this is an incredibly lax level of support which I think if it became the only supported configuration on jshint we would become unhappy with.
>> Prosecuted to the ultimate detail, this is an issue which could consume a huge amount of our time :) Personally I would be happier with something that supplied *some* level of whitespace validation, although preferably less than Crockford's - but it is an issue of who has the time to fiddle with and maintain a suitable version of jslint. Probably our "infrastructure time" in the short run would be better spent on our testing infrastructure which I am turning to shortly...
>I think from experience we know how easy it is to fall into bike shed arguments  over our personal code and whitespace tastes.
>The approach we took in the Fluid community, which I think makes a lot of sense, is to just pick someone else's standards and stick to them. That way we're all a little unhappy, and no one wins. :) Sticking with Crockford's traditional whitespace standards gives us a chance to move on without having to have the same old arguments or waste someone's time hacking on his code base.
>That said, I would love it if we eventually find the time to update our fork of JSLint and evaluate whether or not we can eventually join the JSHint community. Some day, but I imagine it's not on anyone's list of priorities at the moment, and for good reason. I think your efforts on testing infrastructure will prove far more fruitful for now.
>Technical Lead, Fluid Project
>Architecture mailing list
>Architecture at lists.gpii.net
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Architecture