Can You Answer These Twelve Internet Security Questions?

[Aardvark]

On Sat, 26 May 2012 09:26:06 -0400, FromTheRafters wrote:

> Dustin wrote:
>> FromTheRafters<erratic@nomail.afraid.org> wrote in
>> news:jpp7lu$pgk$1@dont-email.me:
>>
>>> Aardvark wrote:
>>>> On Fri, 25 May 2012 21:26:44 +0000, Aardvark wrote:
>>>>
>>>>>> Perhaps SM does pre-rendering in newsgroups?
>>>>>>
>>>>>>
>>>>> Possibly. I might try it in Konqueror and whatever other browsers I
>>>>> have installed here.
>>>>
>>>> Just tried it in both Konq and SM. Nothing.
>>>>
>>>> If you can still open the page, could you check the source and see if
>>>> you can divine what might be amiss?
>>>>
>>> Maybe it's something in:
>>>
>>> <script type="text/javascript" src="setcookie.js"></script>
>>>
>>> or
>>>
>>> <script type="text/javascript"
>>> src="http://analytics.aweber.com/js/awt_analytics.js?id=6HSL"></script
>>>>
>>>>
>>> It uses style sheets too, but that's pretty common practice.
>>>
>>>
>>>
>>>
>> firefox v12 opened it with no hassle. Noscript blocking by default.
>
> I'm not all that familiar with FF's capabilities. Do you know if, even
> with Noscript blocking by default, any pre-fetching is still attempted?
> After all - it still appears as a link in the HTML even though the
> script itself won't be allowed to execute.
>

Wouldn't the simplest method of operation for something like NS be
running the source through NS first, then every time it encounters a pair
of <script> </script> tags either ignoring them or rewriting the source
without them, then sending the edited or reformed HTML to the rendering
engine? Either way, by the time the code actually gets to FF itself,
there are no script tags. Might that affect the source code subsequent to
the script introduction,causing some kind of lock-up in the loading?

> Most of the links (mostly graphics) are relative links, except the
> analytics one.

See above. If NS runs as I have described above, relativity or otherwise
of the script links is purely academic, isn't it? Unless, of course, some
of the script subsequent to one of the tags is dependent in some way on
the script?



--
"Tell me, truthfully, do *you* download music from The Pirate Bay for
your personal use and listening pleasure?"
Clever guy BD in post <gLidnTjP2tgGPy7SnZ2dnUVZ7qKdnZ2d@bt.com>

[FromTheRafters]

Aardvark wrote:
> On Sat, 26 May 2012 09:26:06 -0400, FromTheRafters wrote:
>
>> Dustin wrote:
>>> FromTheRafters<erratic@nomail.afraid.org> wrote in
>>> news:jpp7lu$pgk$1@dont-email.me:
>>>
>>>> Aardvark wrote:
>>>>> On Fri, 25 May 2012 21:26:44 +0000, Aardvark wrote:
>>>>>
>>>>>>> Perhaps SM does pre-rendering in newsgroups?
>>>>>>>
>>>>>>>
>>>>>> Possibly. I might try it in Konqueror and whatever other browsers I
>>>>>> have installed here.
>>>>>
>>>>> Just tried it in both Konq and SM. Nothing.
>>>>>
>>>>> If you can still open the page, could you check the source and see if
>>>>> you can divine what might be amiss?
>>>>>
>>>> Maybe it's something in:
>>>>
>>>> <script type="text/javascript" src="setcookie.js"></script>
>>>>
>>>> or
>>>>
>>>> <script type="text/javascript"
>>>> src="http://analytics.aweber.com/js/awt_analytics.js?id=6HSL"></script
>>>>>
>>>>>
>>>> It uses style sheets too, but that's pretty common practice.
>>>>
>>>>
>>>>
>>>>
>>> firefox v12 opened it with no hassle. Noscript blocking by default.
>>
>> I'm not all that familiar with FF's capabilities. Do you know if, even
>> with Noscript blocking by default, any pre-fetching is still attempted?
>> After all - it still appears as a link in the HTML even though the
>> script itself won't be allowed to execute.
>>
>
> Wouldn't the simplest method of operation for something like NS be
> running the source through NS first, then every time it encounters a pair
> of<script> </script> tags either ignoring them or rewriting the source
> without them, then sending the edited or reformed HTML to the rendering
> engine?

I suppose, but the simplest way isn't always the way things are actually
done. )

Browsers want to do stuff in the background while the user is reading
what is already displayed so that it can appear to be faster at loading
the next page the user might navigate to. If NS runs as you say, like a
proxy (the old proxomitron comes to mind) then FF would never know that
those links existed - but as a plug-in I'm not sure what parses the page
first.

BTW, that page loads in a snap now in IE - even a refresh. It might be
cached by my ISP though.

> Either way, by the time the code actually gets to FF itself,
> there are no script tags. Might that affect the source code subsequent to
> the script introduction,causing some kind of lock-up in the loading?

I don't think so, it's a mystery.

>> Most of the links (mostly graphics) are relative links, except the
>> analytics one.
>
> See above. If NS runs as I have described above, relativity or otherwise
> of the script links is purely academic, isn't it?

Yes, it would never be fetched if all of the script containers and their
contents were removed before the browser got the page.

> Unless, of course, some
> of the script subsequent to one of the tags is dependent in some way on
> the script?

Then we're back to lousy web design, IMO there should be any reliance on
JS being enabled. I didn't see anything like that on that page, and I
didn't download and look at the scripts themselves when I read the HTML
page, just noticed that they were there.