Semper Fi

[unixronin]

(I swear the music was not front-loaded.  The true irony of this musical serendipity didn't become apparent until I previewed my first draft.)

Sailor Jim Johnston linked to this Snopes article about USMC Gunnery Sgt Michael Burghardt, nicknamed "Iron Mike" by his buddies.  Read the story, and you'll understand why they call him that.  I read it, and figured more people needed to see the photo (and to read the story, several months old though it may be).

Snopes.com, however, which got the photo from the Omaha World-Herald, doesn't want anyone else getting it from Snopes in turn -- accredited or otherwise.  So they put a monkeywrencher on it that replaces it, if YOU try to link it in turn from THEM, with a rude face, almost ... gee, almost like it was their copyrighted photo in the first place, and not something they copied from the Omaha World-Herald.

So, yeah, here's the Stars & Stripes article on Iron Mike.  And here's a NON-monkeywrenched copy of the photo.  To Gunny Michael "Iron Mike" Burghardt, a hearty "Semper Fi."  And to Snopes.com ... well, Iron Mike says it all, really.

Hey, Snopes guys?  Iron Mike looks a lot better with his ass hanging out in the breeze than you do.

Comments

[midiamin.livejournal.com]

Hey, Snopes guys? Iron Mike looks a lot better with his ass hanging out in the breeze than you do.

C'mon. This has nothing to do with image rights. It's about resources. It's common practice to check referrers to make sure that if you're downloading an image from our server, you're doing it from the context of one of our pages (in most cases so it will be displayed complete with our advertising). You asked their server to serve up an image for your post. They said no. End of story.

[unixronin.livejournal.com]

While your point in principle is taken, I somehow find myself having difficulty believing that Snopes, as a mostly-text site, has resource problems stemming from people linking to occasional images on their site -- much less problems that outweigh the benefits of additional visitors those links bring to them. And, frankly, y'know, it's increasingly a hyperlinked world -- you post things for people to see them, and in this day and age, if you post something for people to see, and you think it's worth anyone seeing in the first place, you should have a prior expectation that some small proportion of the people who view it are going to link to it.

Now, objecting to people who steal your own, copyrighted images and present them as their own work .... that, I can 100% see. But that's not what Snopes is doing.

[unixronin.livejournal.com]

Further to that: I'm probably a little hair-triggered on the subject at present, after having recently gone looking for an image for another post, and finding that the first site I found it on, which had the article I was referring to, would neither allow it to be linked to, NOR downloaded -- either way, they substituted a GIF handslapping you for "stealing their copyrighted content" -- EVEN THOUGH IT WASN'T EVEN THEIR IMAGE IN THE FIRST PLACE.

So I went off and found another site with a similar write-up on the same subject and the same set of photos distributed by the same Very Large Company's PR department, and linked to their article instead.

[midiamin.livejournal.com]

I somehow find myself having difficulty believing that Snopes, as a mostly-text site, has resource problems stemming from people linking to occasional images on their site

While Snopes regularly posts articles with no attached media, a significant percentage of their articles are centered around images and video clips that people find unbelievable. I would be willing to bet that their image/video content outweighs their text content.

Consider also that the media served by Snopes is, by its nature, unbelievable, not to mention viral in its propagation. URLs propagate even more freely.

Suppose that the URL for a Snopes image propagates out to, oh, 5,000 average people who then put it in their blogs. As their friends read and refresh said blog pages, suppose each image is downloaded 100 times. That's 500,000 requests for that one image served with no recognition of the server. That number is almost certainly higher for the more amusing pics. The breakdown is likely different for video clips, but I would guess the bandwidth impact is even more severe.

much less problems that outweigh the benefits of additional visitors those links bring to them.

Bunkum. If you img-src a pic into your journal, J. Random Webuser is going to look at the image, respond appropriately and never give a single thought to where the image came from unless maybe the person writing the entry is kind enough to explicitly provide a link to the source. A number of people who see a particularly interesting image are going to do a "copy image location" and plonk it into their own journals or whatever forums they happen to be playing on. A very small percentage of readers out there are going to edit the link and check out the rest of the site.

if you post something for people to see, ... you should have a prior expectation that some small proportion of the people who view it are going to link to it.

Yup. And by checking referrers, Snopes can exert pressure on people like you to link to their web page (which I'm sure they encourage) instead of just their copy of the image. In the specific case of Snopes, making sure that all of the images served by their site (many of which are faked) are accompanied by an explanation of the image is also in line with the service that they are trying to provide to the community.

Now, objecting to people who steal your own, copyrighted images and present them as their own work .... that, I can 100% see. But that's not what Snopes is doing.

Correct. They are objecting to you and the potentially thousands of people like you using their bandwidth to display someone else's images on your page with your commentary. It's their bandwidth. Cope.


As to your followup, when I've run into the download problem you describe, it's usually been due to interaction between referrer checking and caching, but admittedly not always.

[unixronin.livejournal.com]

Suppose that the URL for a Snopes image propagates out to, oh, 5,000 average people who then put it in their blogs. As their friends read and refresh said blog pages, suppose each image is downloaded 100 times. That's 500,000 requests for that one image served with no recognition of the server. That number is almost certainly higher for the more amusing pics. The breakdown is likely different for video clips, but I would guess the bandwidth impact is even more severe.

OK, point. Even though the majority of Snopes articles are text-only (or were the last time I was poking around there to any great extent), the kiloword factor does tip the balance disproportionately for a linked image.

And by checking referrers, Snopes can exert pressure on people like you to link to their web page (which I'm sure they encourage) instead of just their copy of the image.

...And which I did. (Usually in such cases I'll make the image a link back to the original article too; in this case, I forgot to do it in the end, being distracted by being pissed off.)

Really, though, if they'd just replaced it with a banner asking people to either link the entire article or make a local copy of the image, I wouldn't have had a problem with that. I just found their particular choice of replacement image, sans any kind of explanation, unnecessarily rude. It's one of those cases where one moment's ill-considered decision can create an unfavorable impression forever.

[rain-manpnw.livejournal.com]

Um, Alaric? I'm afraid you're dead wrong on this one, lad. And if you think about it calmly, you'll understand why. It has nothing to do with copyright; it has to do with bandwidth. It's considered bad netiquette to just link to elements on someone else's website instead of linking to the website itself.

If all the page elements are hosted on your server, then when someone looks at your page, you are the one paying for the bandwidth. If, however, you link to elements on someone else's server, suddenly you're using their bandwidth to serve your customers.

Consider if I decided to create a gallery including several pictures from Seatoria 2000. I could do it one of three ways.

1. I could contact you and work out a deal where I either mirrored your galleries, paying for my bandwidth for people who looked at them, or I could provide thumbnails on my site with direct links to Babcom.com. The second would impact your bandwidth, but do it by redirecting visitors to your site, which you might want for advertising or other reasons.

2. I could simply steal the pics off babcom.com and save them on my server, then set up the web pages that way. You'd find out about it after the pics had spread far and wide across the web, and you'd go after me for stealing your work. That's copyright.

3. I could link directly to the images on Babcom.com from the page I created, then, oh, post a link to my page on Slashdot. Suddenly, your bandwidth is hosed, and you haven't the slightest idea why until you can check your logs. Even if all I did was link a few pics from my LJ, you would still eat the cost of providing the data, without any recompense.

How do you stop it? Simple. You set your server to check who is asking for the data. If it's a page from your site, go ahead. If it's someone else, you can ignore them, or redirect them to as polite or impolite a "No" answer as you like. It's your bandwidth, why should you do the work for them?

It's Snope's bandwidth, why should they do your work for you?

PS - Congrats on the news about the nerve block ;)

[unixronin.livejournal.com]

See the discussion above.