Web Page Import

I am using LIttle Snitch along with Scrivener 2.2 (15858) on OS X 10.7.3.

In my current project, when I choose Import… Web Page…, Scrivener tries to connect to Wikipedia in some apparently random language (first Croation, then Hindi, then some language I can’t recognize that uses the Arabic script) before the sheet for the URL to be imported slides down. Denying the connection does not seem to affect functionality, but why should it happen at all?

In a new project created using the Short Story template, the same behavior occurs, except that the attempted connection is consistently to en.wikipedia.org.

The same happens with a new project created with the Blank template: an attempted connection to en.wikipedia.org on port 80.

Is this a bug or intended behavior? If the latter, then why? Searching the forums hasn’t led to an answers, so any light someone could shed on the issue will be welcome.


I’ve no idea, I’m afraid - certainly Scrivener doesn’t try to connect to Wikipedia in Croation or Hindi or any other language. Import Web Page has absolutely zero code doing anything like this, so the only thing I can think of is that something else (not directly related to Scrivener) is afoot. Very strange!

All the best,

Are you accessing the Internet through an anonymous proxy or something? That sounds a bit like the result you might get if you are using VPN or something. What happens if you temporarily switch off LS? Do you have any other software that can import web pages, like DEVONthink, EagleFiler or Yojimbo? So long as it is a fairly normal Cocoa citizen (I’m not sure if Evernote would count, for example) chances are high it uses the same Apple framework Scrivener does.

The weird thing is that, when using Import Web Page, Scrivener doesn’t connect to the net at all until after you have chosen which web page you want to import and hit “OK” - just calling up the import web page panel doesn’t connect to the internet at all. So this has to be something external. But even when you hit “OK”, the only URL called upon is the one specified for import. (Of course, if that URL has links to other sites, I suppose it is possible that Apple’s WebArchive code might call out to them, but that’s all.)

That’s where a proxy would sneak in though. The proxy sits on your computer and intercepts all Internet traffic, altering it according to how it is programmed to work. An anonymizer would route traffic through multiple machines, which would eventually result in the final request coming from another part of the planet. Since sites like Wikipedia and Google are designed to detect where the request is coming from and offer up a local version of the site, that could explain it. Same could go for a VPN, though that would probably be a static result. If the VPN server is running in India then all of your Internet traffic would be piped through India, making it look like that’s where you are located.

Try using a specific language link for the Wikipedia article: en.wikipedia.com/XXX instead of www. That should force the site to serve up English content no matter where the request is coming from.