Project Silica | Hacker News Reader

Project Silica

microsoft.com

78 pointshliyan14 comments

Microsoft Project Silica, long term storage - https://news.ycombinator.com/item?id=38003883 - Oct 2023 (1 comment)

Microsoft's project Silica: Is glass the future of long term storage? - https://news.ycombinator.com/item?id=37895643 - Oct 2023 (13 comments)

Project Silica 2022 - https://news.ycombinator.com/item?id=36089007 - May 2023 (1 comment)

Project Silica proof-of-concept stores movie on quartz glass - https://news.ycombinator.com/item?id=21482904 - Nov 2019 (102 comments)

Love that this scifi trope is coming to life, I really do. Also pretty funny is the inevitability of dumb memes and porn being etched into glass for essentially eternity. We're gonna produce a lot of 'cave paintings'.

I wonder if/when this will scale down to become commercially available to consumers.

I think these new formats should be just a complementary ways of preserving what civilization produces and we shouldn't abandon old physical media. There's a risk of relying only on digital media - we can screw ourselves pretty badly and we can be affected by things that we can't even control.

I was asked the question about long term storage of news back in 2005. The answer hasn't really changed, fundamentally there is nothing you can buy that will give you rapid access other than a working computer and todays storage technology and it will need replacing and redundency. You need to budget to constantly move to new machines and storage technology and often you only have a 5 years to move off some technologies after they go obsolete.

The file formats are a bit stickier. The further you stray from basic text the less time the format tends to last and then you have the potential additional cost of developing a format conversion tool as well.

I see the appeal of this if its got sufficient access speed there are a bunch of organisations that will see value in keeping data archived permanently safe. But the problem of binary formats over the long term will persist and I am not sure there is much we can do about that. "Project Silica" will also become an obsolete storage medium as well, I don't think this changes the fundamentals.

I would be happier if they were talking about a zip drive like thing so we could all have this rather than a cloud service.

> But the problem of binary formats over the long term will persist and I am not sure there is much we can do about that.

It doesn’t solve the problem, but you could perhaps mitigate it somewhat by having part of each unit of storage contain more me or more parsers for other formats. Eventually a standard “archival format” could evolve that way, but even if not you would be able to cross reference across multiple parsers in multiple units to to have a sort of Rosetta Stone for translations.

I would imagine the laser recorder of Project silica would be prohibitively expensive for consumer. But I wouldn't mind to pay $100 if they could just write everything I had in the cloud and mail me the "disk" backup.

That is assuming the reader of this disc to be extremely affordable.

I only need a few of these to have my life's digital asset somewhat safely stored and protected.

> I would imagine the laser recorder of Project silica would be prohibitively expensive for consumer.

the laser alone is going to be on the order of $100k.

I'm a bit turned off by the "humanity's needs" for storage in this article, where humanity would probably be more respected in an environment where longer term storage would be scarcer and curated. Hovering up every byte ever produced and storing it forever undoubtedly serves someone's [n|gr]eed, but it ain't "humanity's".

What on earth do you mean? Yes, MS is of course a for-profit company, but the long-term storage situation in academia alone is dire. I know from experience that most research data at my university exist as a single copy on a single external hard drive in a drawer (with a partial disorganised copy currently in processing on a laptop drive somewhere). This may be anything from the only records of a now dead language, or medical research data. Add to this that many seem to "forget" their data when it has been published on. Archiving is a contract with future generations and we are horribly prepared. The storage medium of the rosetta stone, handily beats the flimsy, digital solutions the vast majority use. Or did I completely misunderstand your point?

This is really cool. Hopefully the shuttle robots don't accidentally crack the glass!

Is this different than the quartz long-term data storage developed by Hitachi?

There's not much known on the Hitachi project and there's not been any news since it was first mentioned in 2012.

However, project silica is also using Quartz glass.

I would assume that the silence around Hitachi's product reflected a lack of market interest, so I'm curious about why Microsoft wants to walk the same path.

Maybe Microsoft can find a market for it, they do have enormous datacenters and lasers are always getting cheaper.

Especially due to femtosecond lasers becoming very common for laser eye surgery, driving down cost a lot.

Reply on news.ycombinator.com