[petervandijck]

I don't see why my loved ones need to be on the internet.

If they have to be, the main value is long-term-ness. Whatever info/pictures I put out there, I'd likely want that to still be there in 10, 20, 50 years. Nobody has done that type of long-termness well, so far, that I know off.

What I might want to put online is their digital archive: the stuff they chose to share online while alive, archived for a long time.

[chwahoo]

I really like the idea of making it easy for people to share pictures and stories of loved ones who have passed one. I particularly think people who couldn't make it to the funeral would enjoy the chance to read the stories that friends have posted.

I'm not sold on permanent archiving as a selling point, given that companies (particularly startups) aren't permanent. I do hope they explore ways to support themselves by selling a bound paper form of the contributions to a site or allow users to download a nicely designed PDF containing the pictures and stories.

[pg]

They don't need to be, but people often wish to do something to memorialize people who've died, and this approach is in many ways better than e.g. putting up a slab of stone with their name on it, or publishing a notice in a newspaper.

[IgorPartola]

I don't think this is limited to people that have passed on. I want my photos out there for a long time, but I don't want to hand everything over to flickr/picasa, etc.

[moserware]

What about archive.org?

[IgorPartola]

That's not quite it. For one, I have no control over what gets archived. My site with the photos would likely not get much traffic since it's only interesting to my family and friends. Also, it's a lot of content: currently around 4GB and growing. Instead, I'd like to be able to pre-pay for a managed server to run with a guarantee for the next say 100 years. If in that time, any of my descendants decide that what I have there is worthwhile they can renew it for another 100.

[ciupicri]

Regarding long-term-ness, http://www.kensteadman.com/ might be a 14 years old website or at least 11 years old judging by the guest book entries.

[Be_Silly]

No in the internet domain or electronic, but Wilhelm Research have epic amounts of data on film longevity, and you can download freely their 800page illustrated history / report book which is awesome. You find out things like Kodak knowingly sold crappy film stock that would fade to uselessness within a few years, which explains why there's so many 70's and 80's vintage movies no-onecan get on DVD, because the originals are spoiled short of painstaking and hugely expensive restoration. Can't remember if this applied to consumer film, but Kodak sure lost their reputation bad, letting Fujifilm grab the market.

Side note: the true losses i've experienced of family records were precipitated by ignorant relatives disposing of things they had no right to. My only ever data loss was also caused by carelesness or malice by a third party.

As a direct result of this i realy got into learning about everything from backup tape wear to filw system formats, data integrity. I suppose that might be a silver lining, certainly i am able to use what i learned most weeks for work.

But my real concern is how little people understand bit-rot and data integrity in general for storage. My view is nothing is yet truly archival quality. One of the things i learned was the way careful movie studios (used to?) separate their color originals into monochrome R,G,B reels, because the mono stock lasts so much longer, doesn't fade. Also i learned about Technicolor, which uses organic dyes which don't fade or hardly fade. But the original technicolor process is no longer used, apparently the last few lab setups were sold to China. I think China maybe got a good deal, and i wonder if we won't see it making a revival? That thought came courtesy of Andy Grove's recent interview where he observes that if we throw the "low tech" jobs away to other ambitious countries, eventually they innovate with it, like RCA / Westinghouse televisions and then Sony who then outsoruced to Korea who are the present hib of plasmas / lcds and so on.

What i mean is, don't outsource your family records and memories. Write stuff on paper.Print digital picstures on the best current Epson or Canon ink and put them in a dry dark place. Put a lock on it or put it under a attorney managed storage contract so your kids can get access but not throw it away in a fit of upset or stupidity.

As for this service, i admit i've not read much on their site, but where is the front and center discussion of how they've addressed the myriad ways data gets lost / corrupted?

I looked hard at trying to start a service similar to this, but not with the same angle. No way could i make the numbers work on a "freemium" model. Some website outfits do diligent archive style backup for pro photogs, and they charge very serious amounts of money, and i couldn't get any of those on record as to storage policies. It's not just having ZFS, but every app which touches your data better be doing checksums and more, and when i modelled my own idea, i had a 4 - tier multi redundant system that extensively used WORM tape for anything flagged "archive". The numbers got hard at the point the equipment his EOL or cumulative MTBF required preventative reinvestment. I was looking hard at trying to find insurance cover so that, should three data centers all fail, i could at least give some financial condolence to customers / cover lawsuits, but never got right at the kind of person who could quote me. I suppose that would be "cat risk" and i'd need a big "RE" not a front line insurer, but that was only part of my problem. If data is truly private - it might be of a marriage which went wrong and you only want those records to be available after a decent amount of time has passed - you need to build a fairly sensible, but necessarily complex, security model. Then there's the simpler fun of where to store keys, and how do you make any of this easy to use . .

I fear that projects like this risk more tears than they solve problems. The obituary thing is just a gimmick. What print paper runs an obit (for real,not as a paid insertion) without someoe caring to contact them and explain why this person was significant. Not only that but the few obits of friends of mine had to be written on a volunteer basis, even for pretty well known people, because the paper (a major national) allocated no budget for that beyond very high profile people.

I don't want to simply knock this project, because it's an interesting idea, but i think it's flawed and fragile. I don't want to place trust with a faddish organisation when, heck, i've been burned by trusted family and weak courts and straight up accidents and worst of all, couldn't care less attitudes. Only you care enough to do this. If we all did, maybe family life would get a lot nicer for some of us?

Just a thought. Best to all.

[PhotoGuy]

"Organic dyes" have nothing to do with Technicolor print stability. All films and most colorants (in ink jet, dye transfer, etc.) use organic dyes or organic pigments. What made old Technicolor so stable was that it could produce silver separations of the three primary color records, and it was the silver negatives, which are extremely stable, that can be used to strike new color prints (with their organic dyes). Trouble is, this is a very expensive process. Disney used it for its animated features which it intended to release again and again over many decades. Most studios went for the much less expensive dye originals which are less stable, especially if poorly processed or stored. (New viewing prints can be struck from properly stored master prints.) Modern color movie films are quite stable (science marches on), which is largely why the costly Technicolor separation process gradually disappeared.

Digital systems pose their own problems, of course. There's media obsolescene (8-inch floppies, anyone?), digital integrity (bit drift, etc.), and media integrity (oops, that tape broke again). Much of what keeps archivists busy is deciding protocols for the continuous migration of images from one generation of media to another.

For really good up to date information on image permanence, I recommend the independent, non-profit (unlike Wilhelm) Image Permanence Institute at the Rochester Institute of Technology. They've got some excellent, free downloads on both the basics and the advanced stuff. Wilhelm's book may be free, but it's very old and out of date. http://www.imagepermanenceinstitute.org/

[PhotoGuy]

Back then Wilhelm knew that to get his operation off the ground in Iowa, he had to make a splash, so he attacked the big guy, Kodak. Actually Kodak's films were largely state of the art when they came out, which of course is not nearly as good as everyone's films and papers are today. Many imaging researchers question Wilhelm's methods (he won't let anybody in his laboratory) and he made totally wrong predictions about Epson ink jet prints when they first came out (he overlooked ozone effects). He also overweights the effects of light on image lifetime when over 95% of prints are stored most of the time in the dark, where thermal effects are the actors. None the less, his ratings can be useful if read cautiously. He currently has given a "Best in Class" permanence rating to Kodak's ESP ink jet prints, which when you consider the low cost of Kodak's ink jet catridges (compared to HP, Lexmark, Epson and the rest) makes this a good choice.

[PhotoGuy]

Oh, I forgot. Another good site is http://www.savemymemories.org. It's run by an international imaging association to which most major imaging companies belong. Anything on the site has had to pass muster with a board of imaging scientists.