Sunday, 28 February 2010

Bandwidth theft

It seems I committed the heinous crime of linking to someone's site's images in a previous posting on here. Apparently this is considered "bandwidth theft" which is a rather odd idea.

The original site had a good item on various types of BT socket and joint boxes, but only a couple of pictures were relevant to illustrate my blog posting. I was careful not to copy the images though.

I have since changed the post so that it has a link to the whole page and not just a couple of smaller images. To me this seems to mean a lot more bandwidth used as people load the whole page and all of the images on following the link, and not just the ones I originally referenced. I have not heard the term "bandwidth theft" before in this context but insisting on a link to the page not just the relevant images seems to counter the suggestion of "bandwidth theft" somehow. The web server in question was happy to serve the images even though it knew that they were referenced from another page (referrer header in request) and so one has to assume that the author was happy about serving such images to my readers. Apparently not.

The logic of the complaint seems to elude me slightly, but as a breach of an arbitrary protocol, I do offer my apologies.

16 comments:

  1. It's an odd idea to me.. the web is made up of links between different sites. That's the point..

    The only two people who really have any say are

    (a) the site owner (and there are measures they can take to stop linking if they want), and
    (b) the copyright holder of the image - and I'm not sure even they have much say on mere linking as you're not copying.

    It's considered polite for clicking on a linked image to go back to the original site.. but even that isn't very widespread.

    As far as the "bandwidth theft" thing goes I believe that's down to advertising revenue - by linking directly you deprive the owner of page views that they get paid for. I don't agree with the argument personally, but at least it makes some sense.

    ReplyDelete
  2. The clicking on image to original makes sense.
    I had not thought of that. The advertising revenue is a possible explanation of the origin of that phrase I guess.

    I am glad it is not quite as clear cut as I was starting to think I had missed something.

    Doubly odd in this instance was that the copyright in the images was the Zen customer not the site owner that was complaining, anyway.

    ReplyDelete
  3. I wonder if they will pursue Google for 'bandwidth theft' too
    http://images.google.com/images?hl=en&q=bt+external+nte

    ReplyDelete
  4. Yeah, the objection really is that you're serving the image from their server (and thus, using *their* bandwidth), but embedded in your own article instead of theirs. In other words, you're getting all the "benefit" of the image, without the associated costs.

    I'm not sure about Google Images, but I'm pretty sure they host the thumbs they use themselves - so no "bandwidth theft" going on there.

    ReplyDelete
  5. But the server served the images knowing the referrer was not the original page but someone else's, so it was asked "I'm another web page will you give me that image" and it said "yes". If you ask for something and are given it, that is not theft is it!

    As for google, if they make copies then that is possibly far worse, copyright violation!

    ReplyDelete
  6. True, you can configure your server not to serve hotlinked images. But most people either don't know how to, or just don't bother - I'm not sure a failure to block something constitutes consent.

    I think it's pretty much a non-issue on something as small as this blog; the principle, however, is that you're deriving benefit from a service someone else is paying for. Imagine if you'd got slashdotted; the site you'd hotlinked the images from would suddenly get hit for a ton of bandwidth, despite having received no actual readership for it.

    ReplyDelete
  7. On the Google side of things, I think that under US law they'd be able to claim 'reasonable use'. I think they'd also react positively if you asked them to remove any material you had copyright on.

    ReplyDelete
  8. Err, its not a simple failure to block it. The browser asks and the server sends. It does not take it. And it even says where it got the link from.

    ReplyDelete
  9. Pfft. You're trying to split hairs there - just because most web servers by default allow hotlinking doesn't imply consent. If I forget to lock my car, that doesn't imply consent for anyone else to get in.

    Like I said, in this instance I think it's a non-issue, and if the original site is going to get all antsy about it then they damn well *should* have preventing hotlinking - I'm just puzzled that you're having a hard time getting the issue.

    ReplyDelete
  10. Not at all the same. trying to bend your analogy to reality - imaging you let your wife drive your car - she says "can I drive the car" and you say "yes"...

    But one say a stranger comes along and says "can I drive the car" and you say "yes".

    That is *not* stealing.

    The browser *asked* the server for the file, and the server chose to give it. The browser even says why it wants the file - to show in a different web page - when it asked.

    It asks and is given - no way that can be theft, sorry.

    ReplyDelete
  11. Except the browser isn't asking the owner of the site - it's asking the server direct. To push this silly analogy even further beyond sanity, you're saying that as long as the stranger asks the *car* and the car doesn't object, that's permission :)

    Like I said, the basic problem is that the hotlinked site is bearing the bandwidth cost of images which, *to the user*, appear to be part of your article. While you and I both know what happens under the covers, the average user doesn't care what server the browser is talking to.

    ReplyDelete
  12. The fact the average user does not know how it works does not change how it works.

    The *server* is set up by or at the request of the site and copyright owner as they require. If they set it up to give images to just anybody then that is what they have set up and that is clearly not theft.

    It is like leaving brochures in a shop saying "Please take one". You may have intended it only for your customers buying other things to take one, but if someone just walks in and takes one that is *not* theft.

    ReplyDelete
  13. Failing to prevent something is not the same as granting permission. The brochures example is close, but I would suggest a closer one would be sitting in your local pub surfing on the free WiFi provided by the next-door Starbucks. The technology may be allowing you, but you're still using a resource intended for customers of *the other* establishment.

    I think we may have to agree to disagree :)

    ReplyDelete
  14. Not really no. It is more like asking starbucks if you can use their wifi and them saying yes.

    The protocol is clear - you *ask* for the image and you are then given it.

    ReplyDelete
  15. Think yoursef lucky.

    Some sites configure their servers to substitute the Goatse image for any hotlinks.

    I once had a guy rip off a site I made for a customer, calling it "THE OFFICIAL [site] WEBSITE". He hotlinked all the photos I'd gone to financial costs (digital cameras weren't always cheap!) and time to obtain. Lo-and-behold, he was reported (and taken permanently offline) for violating his webhost t&c's when his whole site sprouted a million penises. Imagine that ;).

    ReplyDelete