* Re: [Edbrowse-dev] a new strategy
2014-04-16 9:58 [Edbrowse-dev] a new strategy Karl Dahlke
@ 2014-04-16 14:47 ` Adam Thompson
2014-04-16 14:50 ` Cleverson Casarin Uliana
1 sibling, 0 replies; 3+ messages in thread
From: Adam Thompson @ 2014-04-16 14:47 UTC (permalink / raw)
To: Karl Dahlke; +Cc: Edbrowse-dev
[-- Attachment #1: Type: text/plain, Size: 3452 bytes --]
On Wed, Apr 16, 2014 at 09:58:37AM +0000, Karl Dahlke wrote:
> A couple months ago I talked about creating a full featured dom,
> and comprehensive scans pre and post javascript,
> and all sorts of connections between the two worlds,
> and we may still need to do that some day,
> but I think I'm postponing all that for the indefinite future.
> Even unemployed, it may be more work than I have time for,
> with all the other projects I have going on.
> And I don't think it's a lot of bang for the buck.
> I really don't.
> Damn few websites rebuild themselves completely using js.
Yeah, and doing it properly's going to mean an almost complete rewrite of much
of the browsing internals.
> My new approach is plodding and incremental.
> What sites do we want to work,
> and what will it take to make them work?
>
> Informational is pretty good.
> You can search google and read wikipedia,
> and even search and pull audio down from youtube.
I didn't know you could get youtube to work, I gave up and started using youtube-dl for that.
> I think e-commerce is the next big step.
> Blind folks can't easily go to the store;
> how much more important then it is to order online.
> I pulled down the home page for amazon.com,
> set debug level 3, and browsed.
For me it's not so much a case of not being able to get to the shops (where I live's
fairly easy from a mobility perspective and has decent bus services etc),
more that much of what (certainly I)
need or want to get is only available online (at least at sane prices).
> Good news, it doesn't pull in any other javascript files.
This surprises me.
> Some pages pull in a dozen other js files and you have to paste it all together
> to understand what is happening.
Yeah, and it's usually a bunch of minified things (almost always involving
jquery or prototype.js or the like).
> But it's all here.
>
> And good news, it is many little js functions,
> not one big monster function.
>
> Bad news, each function has been turned into dog vomit,
> but if I put newlines in after semicolons and braces I can almost
> read it again.
Yep, it was actually in doing this that I noticed the inability to substitute
on extremely long lines which I mentioned (and you fixed) a while back.
> Good news, the browse has very few errors.
> Most of these js snippets run. A few do not.
> I've been looking at the errors one by one.
> See recent pushes.
> I fixed a tag we didn't recognize,
> I fixed <frame> to understand ID= attribute,
> which it wasn't doing, so those errors fixed,
Shouldn't almost every tag recognise the id attribute?
> and now I'm working on the next error, object.parentNode.
What's that supposed to do?
> This is just the home page of course,
> I haven't tried to place an order yet.
Yeah, that may be a bit more "fun".
> I dread the day I have to look at facebook.
> I think facebook is absolutely stupid,
> but obviously I'm in the minority.
In that case so am I.
> Well anyways this is what I am thinking.
> Get the popular websites to run,
> and hopefully many others will run as well.
I kind of agree, though I think if we can get the things working for jquery
(basically I think this is some form of ajax set up,
probably with json support as well) then we'll make many sites run much better.
I'm really not sure how to do ajax in edbrowse though.
Cheers,
Adam.
[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 473 bytes --]
^ permalink raw reply [flat|nested] 3+ messages in thread