I don't shoot people images and I don't buy images either, so maybe it is just me: I was looking at the image for a while before reading any text/caption. "Couple having a disagreement" was definitely not what I saw.

Does this image sell? And what is it used for, I wonder?

I don't think it will be possible to provide limitless search results at least for common keywords. So when searching for e.g. "flower" the program must decide. Now way around it.

How about a combined approach? Like
pick X of the Y sites networked partly based on sites that have the most matches (for "flower" in this case) and partly based on sites with heaviest concentration (which probably means they specialize in "flower" images). Then, if that's not enough, let user ask for more.

Interesting suggestions, Jareso. However, I wonder how this could work in reality.

Say I have 1000 images with an average of 30 keywords each and, now, after loooong hours of revising titles and descriptions, they have an additional average of 300 title/description words each. So that's a total of 330.000 words.

Now I want to have those translated into 50 languages you suggest. That's 16.5 million words to be translated.

A huge task. I know because I work as a translator (English > German) for a living. I usually charge US$0.16 per word. OK, I am expensive. So I might be able getting a decent German translation for US$0.05 per word and many other languages might be even cheaper than that. However, even if I manage to cut cost down to US$0.02 per word on average I will be facing a total bill of US$330.000.

So how do you do it?   

Quote from: Pilens on May 07, 2013, 13:02

That was quick, Chico, Thank you.

I noticed there are a few links in the list that point to (photo-)sites that are technically no Symbiostock sites.

let me know which sites those are & i'  move them to the friends of symbio list -- all the sites on the network list should be functioning symbio sites, even if they're not fully up to speed yet

OK, I PM you a list

That was quick, Chico, Thank you.

I noticed there are a few links in the list that point to (photo-)sites that are technically no Symbiostock sites.

Wow. I notice the list of SY sites grows longer every day. Great. I need to update my list, too 

Could you please change "Pietzko Outdoor Photography" to "Imago Borealis"? - Thanks.

Great! The IPTC feature seems to be coming along nicely. Most of my wishes are already covered. There is, however, one extra thing I'd love to see. Let me explain:

You suggested that we change image titles so they are not the same like they are with the micros. I think that was great advice. However, I'd like to keep all my image data in one place (that is my LR catalog). So I was hoping I could use e.g. the IPTC headline field for entering my modified SY titles, then export from LR and upload to my SY site. Now, if your IPTC feature would let me choose to use the headline as title (or copy the headline to the title field) I'd be a really happy camper. Do you think that could be coded?

Dummy me! I thought "pingback" is a funny name for spam site. Turns out it is actually a wordpress function 

This explains it: http://www.mcbuzz.com/2011/what-is-a-pingback-search-marketing-101/

I found pingbacks turned off generally but some pages still had it explicitly turned on, which overrules the general setting. I will keep it turned off because I don't want to spam-police my site daily.

I've developed a deep rooted allergy for categories from dealing with the useless info some micro sites need to have for every single image. 

Here is how I understand this: It is one categorization per SY site  - Really, I think I can deal with that: Once, I'd go through the whole category list and tick every category item that I've photos that fit in. Maybe even with options like "well covered", "so-so"", "just a few"... Some advanced future system might be even be able to do that by query the "image"-keywords.  When I have done my shooting and uploaded a brand new killer batch  I'd come back and tick a new "site"-category. Done! Not too bad, I think.

And here's my understanding on how my "not too-hard" categorization work would be used: A detailed category taxonomy on a per site NOT per image level would open doors to an extremely powerful search and networking tool. This list contains now for a certain (sub-)category all sites that cover it with images. It could be available, regularly updated and synchronized over the SY network on every single SY site. Buyers can search within those categories and get useful results because the SY system knows now which sites in the network to look up and display results from. Buyers can just "browse" the network (hop from site to site) guided by this list that always shows the top x SY sites that cover a chosen category featuring a number of y images automatically found with the "site"-category term used as a "image"-keyword search. - This could rock!!!

The category taxonomy I imagine should be detailed enough so that at the end of its many tree-like branches an average number of z links to Symbiostock sites can be found that have images falling under the corresponding specific category. It should be still coarse enough that we could display a meaningful "site"-category cloud on our SY sites that inspires our buyers to look for something they weren't even dreaming about.

The parameters of x, y, and z need to be determined by site performance, network (search) performance, manageability of (search) results to display, ...

If done right this will be a blast!

Can't wait for the search and networking update and seeing the next step evolving. Great work, Leo.

I am very cautious about PE, too. Their perpetual beta state combined with big mouth announcements that didn't hold too much water when discussed here on MSG left a sour taste. I might give PE a try, though, if and when they come across with something much more convincing than what they have to show for now, after a year or more in beta.

I am definitely not holding my breath for that to happen. Hope we can succeed with SY without any help from PE or from a Nigerian prince, for that matter. 

To add an end to this thread.

There are two networking ideas developing.... an open distributed model, and a slightly more controlled central model. ( I want to be in both)

Both of these networks will rely on peer review ( who you link to) and the market to control the content and quality level.

I am excited for the future of where this is going!

Glenn

Yes, I guess that summarizes this thread well. I also want to be in both network models. And, yes, of course, I am excited about this, too.

Inevitably your going to get the sort of audience like at www.deviantart.com who capitalizes on the easy access to mess up the business as we know it. So the "dark side" goes a little beyond just the illegal stuff...into the bad quality, questionable, and just downright stupid  .

And the more I hear, the more I'd like people to be able to police their own issues. Instead of centralizing so many things on main site, I wanted to create a plugin (inspired by Cascoly) which manages network needs on a larger level. So you can take your SY site and make it into a sort of host, thereby not only creating an index for networks (either the whole SY network, or just your hub of 10-30 sites who agree to similar philosophies), and running things that way. It would be able to collect and manage data, formulate extended search engine abilities, and the list goes on. Like an SY site greatly modded.

I cannot tell enough how much I must keep myself separate from the politics. By way of illustration, I enjoy building awesome cars, not making rules for the road.

The BIGGEST priority for SY is to keep artists in control, and a central site completely un-needed. Its actually a fun web science experiment too.

+10

Quote from: Pilens on April 22, 2013, 23:50

....As long as we stay with 5 hand-selected network buddies this won't be a problem at all. If we somehow expand the network/change the type of networking, I agree, this can easily become a problem.

I think what we should look for is some de-centralized networking solution that on one hand would greatly expand the reach of the network search and on the other hand would still be "self-"controlled (in a sense that each SY member's decisions with whom to network with will leave the bad apples stranded outside the network [of serious and responsible members]). Maybe up to 20 links (or more if you are still comfortable all of them are good partners) instead of just 5 would do the trick? Maybe every one of these many links could provide daily(?) a keyword list (with statistics) of images available. The search then would first look up these lists and decides then based on number/quota/whatever of matching keywords in each list which of the many links to use to actually show search results from. I am just brainstorming here and have no idea if this could work.

that's along the lines of what I've been thinking, too.  anyone participating in this expanded network would agree to have data collected and made available for searches, etc.  the keywords lists is a good idea, since with just 5 sites, some will match my site, but especially early, many wont have any results, but with 20 or more, the odds improve. again, those who don't want to link to similar sites can choose not to participate.

If this would really work, I mean keeping those lists updated and searching them in a timely manner, it would definitely improve image buyers' experience with the SY network search. Again, Opt-in/out makes all the difference. (Funny that some big corporations just don't get that  )