some pages just won't load
Michael Butash
michael at butash.net
Wed Jul 4 11:12:08 MST 2012
Honestly it sounds like infrastructure/code problems on their side more
than browser scripting, but different versions of browsers interpret
things differently. I see big differences between milestone releases of
chrome/chromium how certain webapps work differently, and a complicated
"social networking" site is no different.
Your issue sounds like a hung session that eventually timed out,
reconnected, and got its resource finally. Bad/broken scripting (or
server) causes odd timing events too with menus and other mouse-over
events that sometimes I see not fully complete, even flicker. Load
balancers cause this kind of errant display when dealing with
non-stateful code, but largely depends on what framework and language
they use for their content menus. I see lots of craziness like that
with asp or any windows-y code in anything but ie.
You don't "need" notscript, only your symptoms sound like when js *is*
broken from it disallowing scripting by default. It won't help you
here, but still good to have.
Most sites want js actually, only im selective about what gets allowed.
Keeps crazy scripting and tracking to a minimum when they truly add no
value (to me).
You'll see things like double-click, intellitxt, and various other
parasitic sites that try to run scripts to track, advertise, and in
other ways exploit local scripting for their business necessity. I find
most times I only need to enable scripting on one site, the parent site,
and leave the other 9 blocked to function. Sites like gawkers are
terrible, requiring 4-5 domains just to function for content delivery.
I avoid them as poorly designed and now annoying.
This reduces the overall possibility someone will infect you with a
drive-by script attack (rogue ad in facebook seems most common). Kept
me virus free for duration of windoze use with noscript+firefox, but it
reduces marketing nausea under linux as well using notscript+chrome.
I use this as kind of a gauge how much a site is out to screw me. Sadly
more do than don't. RSS is a good way to bypass it as well to get
content off a site without direct scripting.
-mb
On 07/04/2012 08:34 AM, Michael Havens wrote:
> Hmmmmm..... this is interesting. I git an email stating someone left me
> a message and I followed the link and everything loaded correctly. It
> must have taken a while to take effect (I guess). Anyways.... how will I
> be able to tell if a site need JS?
>
>
> On Wed, Jul 4, 2012 at 8:06 AM, Michael Havens <bmike1 at gmail.com
> <mailto:bmike1 at gmail.com>> wrote:
>
> But it works with my ubuntu box but not on the mint laptop. (both
> running chromium) Thanks for looking at the site.
> Okay, so I installed the notscripts extension, set the password, and
> restarted, and added hi5 to the white list.... but none of those
> steps helped any.
>
>
> On Wed, Jul 4, 2012 at 12:59 AM, Michael Butash <michael at butash.net
> <mailto:michael at butash.net>> wrote:
>
> Noscript (firefox) or notscript (chrome equivalent) are
> extensions, security, default denies scripts and breaks anything
> remotely web2.0ish with good reason. Necessary evil, especially
> if you use windoze. I use it mostly to deny advertisements or
> other ill attempts at getting more script access than i wish to
> give questionable vendors. Sites using them are questionable
> enough to allow as it is.
>
> I'm thinking it's more crap scripting that doesn't work entirely
> compatible with chrom(e|ium), ie errata/bug. I've seen some odd
> scripting differences using chrome under windows or chromium
> under linux on enterprise-y necessary crapware (ahem, cisco acs
> and others) that I can't explain other than scripting
> fixes/changes between versions trying to make sense of ambiguous
> code.
>
> That *social* site looks as though it will test your scripting
> to see what it can extract from your computer for user
> information, expect compatibility issues outside of IE that it
> would just otherwise use to mirror your hard disk to their
> server. :)
>
> -mb
More information about the PLUG-discuss
mailing list