Reply to post: Re: Dubious

Re: Dubious

Agree.

I really like the idea of figuring out what is actually being used. Hopefully with an eye to deprecating some of the stuff that isn't.

However, when you flag Canvas and SVG as no-gooders, that leaves open what the suggested replacement is. I have little nostalgia for mid-90s websites and even less for sites using Flash. For better or worse, browser-based approaches have become a primary delivery mechanism for software services. Not least because much of our software now expects to have connectivity and would not operate well in pure standalone mode.

Leaving aside considerations about the cloud being a good or bad thing, if you need a network aware app that also provides rich interactivity, that leaves you with a few options. Locally-installed software, apps a la iOS/Android. Or a browser-based app with JS and SVG/Canvas support.

Personally, for all its weaknesses, I find browser-delivered apps with a suitably hardened browser and a sparse NoScript JS whitelist rather less dangerous than installing a lot of programs or apps locally. I actually trust open source repositories for Linux/OSX utilities/programming tools a fair bit, but not much else. Not app stores. And certainly not stuff floating around on the web waiting to be downloaded and installed.

The less stuff I install locally, the happier I am.

i.e. OK to jettison the unused cruft, but suggest rather than just criticize for the more useful bits. Doesn't make this study wrong per se, but it's only part of the picture. And website bloat/slowness is an coding issue, mostly not the fault of the tools themselves (Flash excepted).

p.s. maybe the SVG spec itself requires winnowing. I recall MS self-serving justification for not having SVG support in IE (prior to IE9) being that the spec was bloated with stuff like file uploading facilities, best left to HTML. Maybe they were correct in that instance.