Virtual Images and SEO?

Hey,

From past experience, Google should be okay with virtual images rather than a bunch of images all crammed onto one site.

When using images from Zazzle, have you thought of adding a "rel=nofollow" to the end of each link? Just so that Google doesn't get the sites SEO mixed up. . . plus I have found that to actually improve rankings.

If you were to have the rel=nofollow at the end of the links, I don't think that Google will not index those specific thumbnails from your site because you are telling Google not to follow those links. . . so they will show up from their original source.

This is just my suggestion, but yes, I agree with your programmer on using virtual images rather than lots and lots of thumbnails.

Hope this helps!

-Josh

I really like your site by the way!

I don't know what your doing with the images (redirects?) but it looks like they're indexed right now.

As per Paul's confusion, what do you mean by a virtual image?

There's no such term related to web sites, only Virtual image - Wikipedia, the free encyclopedia

Your making this complicated & setting yourself up for a massive failure If your entire site is hotlinking images from an external domain. All zazzle has to do is change the image root folder URL or any other part of the image URL & your entire site will be toast as far as images.

Trust me on this...

Years ago I hosted hundreds of files on mediafire for free, the URLs expired exactly 1 year later. It took me weeks to fix that mess. Hotlinking all the images on your site is a very similar situation.

Live & learn...

I meant to say "My goal here is to have Google think I'm not showing 10 Zazzle images."

Bottom line: I wanted to download all 200,000 product thumbnails from Zazzle and upload them to my server. My programmer says instead of doing that, he can just rewrite the Zazzle thumbnail URL to make it look like the product thumbnail is being hosted on familycrestdn.com when in fact it is not.

I am the programmer y0kkles is referring to. Thanks to everyone for your responses. I thought it was time for me to weigh in here and clarify a few things. Sorry for what I can already see will be a long post.

The way y0kkles's site currently works is that a visitor looks up his surname and finds information about its history, a link to purchase products, one image for the design itself (a "real" jpeg file which lives on the client's site) and 10 product images (which come from Zazzle). y0kkles's question is entirely about the product images, not the design image. As yukon pointed out, the design images are already being indexed.

Right now, the 10 product images are of course img tags, with a src at zazzle.com. y0kkles's thinking is that by changing those src URLs to be at familycrestdb.com instead of zazzle.com, the images themselves will be indexed at familycrestdb.com, and Google is less likely to flag familycrestdb.com as an affiliate. I'm no SEO expert, but that seems reasonable to me. (And as an added bonus as you'll see below, the image URLs at familycrestdb.com are much friendlier than those at zazzle.com.)

His original idea was to save those images as independent files on his site. I have suggested that it doesn't matter whether there are actual images files at the URL location or not, and therefore that it's a waste of time and effort to save all of those images as files (20,000+ names * 10 images per = 200,000+ image files). And, as I'll get to below, it turns out there is a technical issue preventing the images from being saved in the way I'd intended, which is what precipitated this whole question.

Now we get into the semi-technical stuff. I realize that my terminology (actual images vs. virtual images) may not be commonplace, so I'll explain what I mean. An "actual image" is a real .jpg file (or whatever image format) that lives as a stand-alone file on the server. A "virtual image" is one that is created on the fly by a program. There is no corresponding .jpg (or equiv) file on the server. The image is generated in memory on the server and delivered to the browser, without any real corresponding file existing.

The question boils down to this: y0kkles is already changing the img src URLs from zazzle to familycrestdb. The question is, does it matter if there are "actual" .jpg files in those locations? Or can that URL be served instead by a php script which generates the images on the fly ("virtual image")?

Let's think about Zazzle for a minute. y0kkles has 20,000+ designs in this particular Zazzle store, and although the familycrestdb site is only showing 10 products, he has up to 50 or so for each design. And he has several other similar Zazzle stores. So it's probably safe to say that this one Zazzle designer is alone responsible for at least a million product images, and probably way more than that. And he's one of I don't know how many designers. So Zazzle has millions if not tens or hundreds of millions of product images to deal with.

Each product image from Zazzle consists of a design superimposed on a product with an added watermark. It is highly unlikely that the best Zazzle's engineers could come up with was to actually create these hundreds of millions of .jpg files. What's going on behind the scenes, I'm sure, is that the image is being created on the fly when requested. I'm sure they have a script which combines the three images (blank product, design, and watermark) into one virtual image and sends that virtual image to the browser.

In other words, the Zazzle URLs are pointing to virtual images, not actual images. Maybe they cache them, but I suspect not, or at most selectively.

y0kkles request of me consists of essentially two parts: 1) create URLs (for the same images) that live on familycrestdb.com instead of zazzle.com; and 2) change the links on the surname page to use the local URLs instead of the Zazzle URLs.

1 is done and functioning, but not quite as originally intended. Here's how it was designed to work.
- The page at www.familycrestdb.com will include img tags with src at images.familycrestdb.com. (That's 2 above, and not implemented yet. More correctly, not deployed yet.) Here's a real example of such an image URL:
http://images.familycrestdb.com/english/jones-shirt.jpg
- When each of those img src links is requested by the browser, the web server will look for a real .jpg file matching the location. If it exists, it will be served up and no php script is called (no magic here).
- If the .jpg file does not exist, then a php script at images.familycrestdb.com will be called. The php script will know the URL that was requested.
- The php script identifies the corresponding zazzle URL for the product image. Here's a real example:
http://rlv.zcache.com/abbot_family_c..._8naxt_210.jpg
- The php script fetches the image from that URL, but doesn't send it to the browser. It creates an image object in memory.
* The php script then saves the image object as a file, so that next time this image is requested, the image file will be found and this whole process will stop where I wrote "no magic here" above.
- But this time, the browser that's trying to display the page with 10 product links on it needs to receive an image in response to its request. The php script still has an image object in memory, so it returns that image to the browser.

Short recap: the first time an image URL at images.familycrestdb.com is requested, the image file will be created and a virtual image will be served. From that point on, when that image URL is requested, there will be a real image file, so the script will never be called again for that URL.
I sure hope I've taken something that seemed confusing and clarified it. It really isn't all that complicated.

The line above marked with an * instead of a - is the part that doesn't work as expected. It works fine on the server where I developed it, but the client's host does not allow php scripts to create files or make directories. That makes saving the files (via the script as described above) impossible, without changing the host.

My recommendation to y0kkles has been to simply skip the step that doesn't work. I can comment out two lines of code and the virtual images that come from Zazzle are delivered as the result of a request to images.familycrestdb.com. The URL http://images.familycrestdb.com/english/jones-shirt.jpg is requested, the .jpg file doesn't exist so the php script is called, and the php script delivers the image found at http://rlv.zcache.com/abbot_family_c..._8naxt_210.jpg.

With all that background info, the question y0kkles is asking is whether the fact that there is no actual .jpg file at http://images.familycrestdb.com/english/jones-shirt.jpg have a negative effect on his SEO.

Thanks in advance for any insights any of you may have.