Linus' latest investment is DOOOOOMED to be a giant catastrophe!!!

[Thomas A. Fine]

[clickbait headline intended]

 

[Editing to add: this is about a new proposed Linus investment, in a brand new network storage company, that was discussed on the latest WAN show, starting at 2:06:03.]

 

First, I have no issues with disclosed investments like this, generally.  Or even this in particular, in terms of ethics or whatever.

 

I just think this sounds like a very risky proposition.  As much as I am not at all a pro-corporate person, and as much as I want the "little guy/gal" to have access to "big guy/gal" technologies, I do, from experience, think that certain technologies really need to have a fairly substantial enterprise behind them.

And I think network storage solutions are one of these technologies.

 

I'm a sysadmin, and I work at a major astrophysics laboratory, one of the biggest in the world.  Astrophysics is a field which uses a lot of data. A LOT.  People outside of the science community probably don't know much about science labs, but the primary driver of funding is grants.  Nearly all projects are paid for by grants (from wherever), meaning each particular project has its own specific budget.  Because of this, or because of the traditions of the institution, each project group works very independently.  As a sysadmin, we provide a standard environment to scientists, and make whatever modifications we can reasonably accommodate, up to a point.  A point, after which, some groups decide to do their own system administration because of how our "packaged" product doesn't fit their needs.

 

One of the most common ways this happens is with network storage.  We use a major NETwork APPliance vendor who's name you might guess, to provide storage for all basic user space as well as project space.  Project space must be paid out of grant money per project, and we charge a very fair price to projects for their network attached storage (which includes online snapshots taken hourly, daily and weekly up to three weeks, as well as a regular backup of all data to tape, and offsite tape storage.  And high availability, and high bandwidth connections to the servers, etc.  And drive failures are no big thing to us.  Often drives show up by delivery before anyone on staff even knew an old drive had failed. The price also includes staff support to provide all these services.

 

The point of this meandering story is that when some smaller projects see our quote for what we charger for storage, they balk.  They come back and say "yeah but I can buy this software raid for a tenth (or whatever small fraction) of that price.  We always try to talk them out of it..  We explain everything they'll get for the price, and, more importantly, what they're giving  up if they do it themselves.  We do not back up privately created and managed storage.  We make this VERY clear.  We give them a long talk about the importance of setting up a backup system.  In particular we emphasize the importance of TESTING the backup system they've set up, regularly, to make sure it serves their purpose.

 

Despite this warning, groups sometimes go the DIY route.  And every year or two, we end up with a scientist at our door, with a broken raid that won't rebuild, and a backup system that they thought worked, but didn't.   Years of data of multiple scientists have been lost this way.  There have been cases where they never bothered to set up a backup system, because despite our warnings, it's a RAID, right?  There's been cases where they've set up a backup system, and even tested it.  But, their backup system was a mirroring system, and they never tested their cobbled together scripts for what happens when a disk actually fails (the mirroring software dutifully mirrors the now empty failed disk mount point to the mirror... erasing it.  There's been all sorts of different failure modes.  The data loss has been devastating and unrecoverable.

 

Now, on the one hand, this is a GREAT argument that what Linus is talking about is absolutely needed.  And it SEEMS LIKE it should be totally doable.  But when it comes to a reliable, turnkey storage solution, it's incredibly hard to beat a solution that has big resources behind it.  Because of the seriousness of data loss (in certain applications), and because of the liability to the vendor, I think this is a situation where the cost of a really reliable storage solution winds up being much higher than whatever thumbnail solution you can come up with.  I also think that Linus' experience in particular, with network storage that's used personally for video libraries (not the worst thing in the world to lose all the data), or used professional as a video library, where there's a cloud storage solution (YouTube) inherently built in to your system without having to think about it, that Linus perhaps hasn't had to deal with what truly devastating data loss can look like firsthand. (I've only seen it as a sysadmin who couldn't be blamed for it; I'd hate to have to see this if such loss came as a result of a product I produced and supported.)

 

Having said all that, I understand market progression.  Eventually, someone will probably crack this nut and there will be a cheap (well, cheaper, but still not as cheap as you think), easy-to-use, yet reliable off-the-shelf thing that can be used to store any kind of data, no matter how critical it is.  I just don't think we're close to that yet.

 

Maybe I'm wrong.

 

With Frameworks, Linus was investing in something that was radical only from a philosophical viewpoint.  Technically, it's just a laptop.  And when he invested in it, it already existed.  This product is vaporware.  And it's a space where many others have tread, and yet failed to come up with solution that actually both succeeded, and remained with accessible pricing.  Unlike the Frameworks investment, this investment is a much bigger gamble — it's a bet that this is both the right time, and the right entity, to open up an entirely new market niche.

Again, I have no ethical complaint about Linus' investment choice with respect to LTT, LMG, or whatever.  I just think this one is a much riskier proposition than Frameworks.

[JoHeJo06]

20 minutes ago, Thomas A. Fine said:

[clickbait headline intended]

 

First, I have no issues with disclosed investments like this, generally.  Or even this in particular, in terms of ethics or whatever.

 

I just think this sounds like a very risky proposition.  As much as I am not at all a pro-corporate person, and as much as I want the "little guy/gal" to have access to "big guy/gal" technologies, I do, from experience, think that certain technologies really need to have a fairly substantial enterprise behind them.

And I think network storage solutions are one of these technologies.

 

I'm a sysadmin, and I work at a major astrophysics laboratory, one of the biggest in the world.  Astrophysics is a field which uses a lot of data. A LOT.  People outside of the science community probably don't know much about science labs, but the primary driver of funding is grants.  Nearly all projects are paid for by grants (from wherever), meaning each particular project has its own specific budget.  Because of this, or because of the traditions of the institution, each project group works very independently.  As a sysadmin, we provide a standard environment to scientists, and make whatever modifications we can reasonably accommodate, up to a point.  A point, after which, some groups decide to do their own system administration because of how our "packaged" product doesn't fit their needs.

 

One of the most common ways this happens is with network storage.  We use a major NETwork APPliance vendor who's name you might guess, to provide storage for all basic user space as well as project space.  Project space must be paid out of grant money per project, and we charge a very fair price to projects for their network attached storage (which includes online snapshots taken hourly, daily and weekly up to three weeks, as well as a regular backup of all data to tape, and offsite tape storage.  And high availability, and high bandwidth connections to the servers, etc.  And drive failures are no big thing to us.  Often drives show up by delivery before anyone on staff even knew an old drive had failed. The price also includes staff support to provide all these services.

 

The point of this meandering story is that when some smaller projects see our quote for what we charger for storage, they balk.  They come back and say "yeah but I can buy this software raid for a tenth (or whatever small fraction) of that price.  We always try to talk them out of it..  We explain everything they'll get for the price, and, more importantly, what they're giving  up if they do it themselves.  We do not back up privately created and managed storage.  We make this VERY clear.  We give them a long talk about the importance of setting up a backup system.  In particular we emphasize the importance of TESTING the backup system they've set up, regularly, to make sure it serves their purpose.

 

Despite this warning, groups sometimes go the DIY route.  And every year or two, we end up with a scientist at our door, with a broken raid that won't rebuild, and a backup system that they thought worked, but didn't.   Years of data of multiple scientists have been lost this way.  There have been cases where they never bothered to set up a backup system, because despite our warnings, it's a RAID, right?  There's been cases where they've set up a backup system, and even tested it.  But, their backup system was a mirroring system, and they never tested their cobbled together scripts for what happens when a disk actually fails (the mirroring software dutifully mirrors the now empty failed disk mount point to the mirror... erasing it.  There's been all sorts of different failure modes.  The data loss has been devastating and unrecoverable.

 

Now, on the one hand, this is a GREAT argument that what Linus is talking about is absolutely needed.  And it SEEMS LIKE it should be totally doable.  But when it comes to a reliable, turnkey storage solution, it's incredibly hard to beat a solution that has big resources behind it.  Because of the seriousness of data loss (in certain applications), and because of the liability to the vendor, I think this is a situation where the cost of a really reliable storage solution winds up being much higher than whatever thumbnail solution you can come up with.  I also think that Linus' experience in particular, with network storage that's used personally for video libraries (not the worst thing in the world to lose all the data), or used professional as a video library, where there's a cloud storage solution (YouTube) inherently built in to your system without having to think about it, that Linus perhaps hasn't had to deal with what truly devastating data loss can look like firsthand. (I've only seen it as a sysadmin who couldn't be blamed for it; I'd hate to have to see this if such loss came as a result of a product I produced and supported.)

 

Having said all that, I understand market progression.  Eventually, someone will probably crack this nut and there will be a cheap (well, cheaper, but still not as cheap as you think), easy-to-use, yet reliable off-the-shelf thing that can be used to store any kind of data, no matter how critical it is.  I just don't think we're close to that yet.

 

Maybe I'm wrong.

 

With Frameworks, Linus was investing in something that was radical only from a philosophical viewpoint.  Technically, it's just a laptop.  And when he invested in it, it already existed.  This product is vaporware.  And it's a space where many others have tread, and yet failed to come up with solution that actually both succeeded, and remained with accessible pricing.  Unlike the Frameworks investment, this investment is a much bigger gamble — it's a bet that this is both the right time, and the right entity, to open up an entirely new market niche.

Again, I have no ethical complaint about Linus' investment choice with respect to LTT, LMG, or whatever.  I just think this one is a much riskier proposition than Frameworks.

what reason do you have to say all this?

[mr moose]

2 minutes ago, JoHeJo06 said:

what reason do you have to say all this?

If I read it right, the reason is that new products have to be both astronomically better than current offerings and have sufficient resources to be able to deliver the goods to large networks of buyers in order to become anything.  And that a lot of "new' products no matter how good they seem on paper just never cut the mustard for a multitude of reasons.

 

It actually put me in mind of music, there are literally millions of very talented artists who would be great in front of thousands of fans,  but in order to succeed you need to be way more than even very talented. You need to be able to supply what the industry demands without a reputation or income long before you start to sniff at success.

 

 

[DigitalGoat]

22 minutes ago, Thomas A. Fine said:

[clickbait headline intended]

 

First, I have no issues with disclosed investments like this, generally.  Or even this in particular, in terms of ethics or whatever.

 

I just think this sounds like a very risky proposition.  As much as I am not at all a pro-corporate person, and as much as I want the "little guy/gal" to have access to "big guy/gal" technologies, I do, from experience, think that certain technologies really need to have a fairly substantial enterprise behind them.

And I think network storage solutions are one of these technologies.

 

I'm a sysadmin, and I work at a major astrophysics laboratory, one of the biggest in the world.  Astrophysics is a field which uses a lot of data. A LOT.  People outside of the science community probably don't know much about science labs, but the primary driver of funding is grants.  Nearly all projects are paid for by grants (from wherever), meaning each particular project has its own specific budget.  Because of this, or because of the traditions of the institution, each project group works very independently.  As a sysadmin, we provide a standard environment to scientists, and make whatever modifications we can reasonably accommodate, up to a point.  A point, after which, some groups decide to do their own system administration because of how our "packaged" product doesn't fit their needs.

 

One of the most common ways this happens is with network storage.  We use a major NETwork APPliance vendor who's name you might guess, to provide storage for all basic user space as well as project space.  Project space must be paid out of grant money per project, and we charge a very fair price to projects for their network attached storage (which includes online snapshots taken hourly, daily and weekly up to three weeks, as well as a regular backup of all data to tape, and offsite tape storage.  And high availability, and high bandwidth connections to the servers, etc.  And drive failures are no big thing to us.  Often drives show up by delivery before anyone on staff even knew an old drive had failed. The price also includes staff support to provide all these services.

 

The point of this meandering story is that when some smaller projects see our quote for what we charger for storage, they balk.  They come back and say "yeah but I can buy this software raid for a tenth (or whatever small fraction) of that price.  We always try to talk them out of it..  We explain everything they'll get for the price, and, more importantly, what they're giving  up if they do it themselves.  We do not back up privately created and managed storage.  We make this VERY clear.  We give them a long talk about the importance of setting up a backup system.  In particular we emphasize the importance of TESTING the backup system they've set up, regularly, to make sure it serves their purpose.

 

Despite this warning, groups sometimes go the DIY route.  And every year or two, we end up with a scientist at our door, with a broken raid that won't rebuild, and a backup system that they thought worked, but didn't.   Years of data of multiple scientists have been lost this way.  There have been cases where they never bothered to set up a backup system, because despite our warnings, it's a RAID, right?  There's been cases where they've set up a backup system, and even tested it.  But, their backup system was a mirroring system, and they never tested their cobbled together scripts for what happens when a disk actually fails (the mirroring software dutifully mirrors the now empty failed disk mount point to the mirror... erasing it.  There's been all sorts of different failure modes.  The data loss has been devastating and unrecoverable.

 

Now, on the one hand, this is a GREAT argument that what Linus is talking about is absolutely needed.  And it SEEMS LIKE it should be totally doable.  But when it comes to a reliable, turnkey storage solution, it's incredibly hard to beat a solution that has big resources behind it.  Because of the seriousness of data loss (in certain applications), and because of the liability to the vendor, I think this is a situation where the cost of a really reliable storage solution winds up being much higher than whatever thumbnail solution you can come up with.  I also think that Linus' experience in particular, with network storage that's used personally for video libraries (not the worst thing in the world to lose all the data), or used professional as a video library, where there's a cloud storage solution (YouTube) inherently built in to your system without having to think about it, that Linus perhaps hasn't had to deal with what truly devastating data loss can look like firsthand. (I've only seen it as a sysadmin who couldn't be blamed for it; I'd hate to have to see this if such loss came as a result of a product I produced and supported.)

 

Having said all that, I understand market progression.  Eventually, someone will probably crack this nut and there will be a cheap (well, cheaper, but still not as cheap as you think), easy-to-use, yet reliable off-the-shelf thing that can be used to store any kind of data, no matter how critical it is.  I just don't think we're close to that yet.

 

Maybe I'm wrong.

 

With Frameworks, Linus was investing in something that was radical only from a philosophical viewpoint.  Technically, it's just a laptop.  And when he invested in it, it already existed.  This product is vaporware.  And it's a space where many others have tread, and yet failed to come up with solution that actually both succeeded, and remained with accessible pricing.  Unlike the Frameworks investment, this investment is a much bigger gamble — it's a bet that this is both the right time, and the right entity, to open up an entirely new market niche.

Again, I have no ethical complaint about Linus' investment choice with respect to LTT, LMG, or whatever.  I just think this one is a much riskier proposition than Frameworks.

So because it's risky he shouldn't attempt it? With that attitude LMG would not have existed.

Personal videos not the worse thing in the world to lose? In whos opinion? Data value is subjective.

Linus is aware of the impact of data loss and needed to rely Wendel's expertise to write custom scripts in order to recover most of the lost data.

I am certain that if Linus is going to invest in anything then he will employ the correct people for the job, I am sure he is aware of his own limitations.

Several YouTubers have lost their video collections regardless of them being backed up by a big tech company, those videos are their source of income and probably are as important to them as any scientists data sets.

[Thomas A. Fine]

3 minutes ago, DigitalGoat said:

So because it's risky he shouldn't attempt it? With that attitude LMG would not have existed.

Personal videos not the worse thing in the world to lose? In whos opinion? Data value is subjective.

Linus is aware of the impact of data loss and needed to rely Wendel's expertise to write custom scripts in order to recover most of the lost data.

Obviously Linus can and will do what he wants.  I guess the TL;DR of my posting is, this investment is a VERY different animal than his Framework investment, with far more risk (and Framework involved plenty of risk).

 

I get that Linus says it isn't about making money, also, but... vanity projects have a poor history.  And again to differentiate between this and Framework, that investment seemed more like some form of industry leadership (right to repair), while this one... doesn't.

 

It's more like just random frustration with a product, than some important industry issue.  Linus' pet peeve. (And on top of that it does not match my current vendor pet peeve, which is mesh wifi interoperability - now that would be worth investing in ).

[Blue4130]

55 minutes ago, Thomas A. Fine said:

[clickbait headline intended]

 

First, I have no issues with disclosed investments like this, generally.  Or even this in particular, in terms of ethics or whatever.

 

I just think this sounds like a very risky proposition.  As much as I am not at all a pro-corporate person, and as much as I want the "little guy/gal" to have access to "big guy/gal" technologies, I do, from experience, think that certain technologies really need to have a fairly substantial enterprise behind them.

And I think network storage solutions are one of these technologies.

 

I'm a sysadmin, and I work at a major astrophysics laboratory, one of the biggest in the world.  Astrophysics is a field which uses a lot of data. A LOT.  People outside of the science community probably don't know much about science labs, but the primary driver of funding is grants.  Nearly all projects are paid for by grants (from wherever), meaning each particular project has its own specific budget.  Because of this, or because of the traditions of the institution, each project group works very independently.  As a sysadmin, we provide a standard environment to scientists, and make whatever modifications we can reasonably accommodate, up to a point.  A point, after which, some groups decide to do their own system administration because of how our "packaged" product doesn't fit their needs.

 

One of the most common ways this happens is with network storage.  We use a major NETwork APPliance vendor who's name you might guess, to provide storage for all basic user space as well as project space.  Project space must be paid out of grant money per project, and we charge a very fair price to projects for their network attached storage (which includes online snapshots taken hourly, daily and weekly up to three weeks, as well as a regular backup of all data to tape, and offsite tape storage.  And high availability, and high bandwidth connections to the servers, etc.  And drive failures are no big thing to us.  Often drives show up by delivery before anyone on staff even knew an old drive had failed. The price also includes staff support to provide all these services.

 

The point of this meandering story is that when some smaller projects see our quote for what we charger for storage, they balk.  They come back and say "yeah but I can buy this software raid for a tenth (or whatever small fraction) of that price.  We always try to talk them out of it..  We explain everything they'll get for the price, and, more importantly, what they're giving  up if they do it themselves.  We do not back up privately created and managed storage.  We make this VERY clear.  We give them a long talk about the importance of setting up a backup system.  In particular we emphasize the importance of TESTING the backup system they've set up, regularly, to make sure it serves their purpose.

 

Despite this warning, groups sometimes go the DIY route.  And every year or two, we end up with a scientist at our door, with a broken raid that won't rebuild, and a backup system that they thought worked, but didn't.   Years of data of multiple scientists have been lost this way.  There have been cases where they never bothered to set up a backup system, because despite our warnings, it's a RAID, right?  There's been cases where they've set up a backup system, and even tested it.  But, their backup system was a mirroring system, and they never tested their cobbled together scripts for what happens when a disk actually fails (the mirroring software dutifully mirrors the now empty failed disk mount point to the mirror... erasing it.  There's been all sorts of different failure modes.  The data loss has been devastating and unrecoverable.

 

Now, on the one hand, this is a GREAT argument that what Linus is talking about is absolutely needed.  And it SEEMS LIKE it should be totally doable.  But when it comes to a reliable, turnkey storage solution, it's incredibly hard to beat a solution that has big resources behind it.  Because of the seriousness of data loss (in certain applications), and because of the liability to the vendor, I think this is a situation where the cost of a really reliable storage solution winds up being much higher than whatever thumbnail solution you can come up with.  I also think that Linus' experience in particular, with network storage that's used personally for video libraries (not the worst thing in the world to lose all the data), or used professional as a video library, where there's a cloud storage solution (YouTube) inherently built in to your system without having to think about it, that Linus perhaps hasn't had to deal with what truly devastating data loss can look like firsthand. (I've only seen it as a sysadmin who couldn't be blamed for it; I'd hate to have to see this if such loss came as a result of a product I produced and supported.)

 

Having said all that, I understand market progression.  Eventually, someone will probably crack this nut and there will be a cheap (well, cheaper, but still not as cheap as you think), easy-to-use, yet reliable off-the-shelf thing that can be used to store any kind of data, no matter how critical it is.  I just don't think we're close to that yet.

 

Maybe I'm wrong.

 

With Frameworks, Linus was investing in something that was radical only from a philosophical viewpoint.  Technically, it's just a laptop.  And when he invested in it, it already existed.  This product is vaporware.  And it's a space where many others have tread, and yet failed to come up with solution that actually both succeeded, and remained with accessible pricing.  Unlike the Frameworks investment, this investment is a much bigger gamble — it's a bet that this is both the right time, and the right entity, to open up an entirely new market niche.

Again, I have no ethical complaint about Linus' investment choice with respect to LTT, LMG, or whatever.  I just think this one is a much riskier proposition than Frameworks.

Have a link or a TLDR? What did he invest in? I gather it's some sort of data service...

[Poinkachu]

5 minutes ago, Thomas A. Fine said:

Obviously Linus can and will do what he wants.  I guess the TL;DR of my posting is, this investment is a VERY different animal than his Framework investment, with far more risk (and Framework involved plenty of risk).

 

I get that Linus says it isn't about making money, also, but... vanity projects have a poor history.  And again to differentiate between this and Framework, that investment seemed more like some form of industry leadership (right to repair), while this one... doesn't.

 

It's more like just random frustration with a product, than some important industry issue.  Linus' pet peeve. (And on top of that it does not match my current vendor pet peeve, which is mesh wifi interoperability - now that would be worth investing in ).

Well, I can say this for sure.
I know some rich peoples in real life that made a way more questionable investment, and some of it were made just because they were pissed at something.
I came to call it : "Rich people's behavior", it's sounds generalizing i know. But yeah well, it's not really something someone who isn't rich can/would do afterall.

When I asked them about it a few years later, most of them simply answered that they knew it's very risky and bound to fail.

[Thomas A. Fine]

20 minutes ago, Blue4130 said:

Have a link or a TLDR? What did he invest in? I gather it's some sort of data service...

Sorry, I left this out.  On the latest WAN show he described his high likelihood of investing in a brand new business that will develop a new network attached storage solution that's more user-friendly.  Not sure of the time stamp, but maybe roughly 2/3 of the way trhough.  At 2:06:03 in the video.

 

It sounds like a startup from someone that does know what they are doing (from within the industry), but still, I see it as very high risk.

[Radium_Angel]

1 hour ago, Thomas A. Fine said:

network attached storage solution that's more user-friendly.

From professional experience, when I hear the words "user friendly" I think "lacking features".

Cisco switches, by way of example, are anything but user friendly, but they are the Gold Standard when it comes to switches for a reason.

 

 

[ToboRobot]

3 hours ago, Thomas A. Fine said:

[clickbait headline intended]

 

[Editing to add: this is about a new proposed Linus investment, in a brand new network storage company, that was discussed on the latest WAN show, starting at 2:06:03.]

 

First, I have no issues with disclosed investments like this, generally.  Or even this in particular, in terms of ethics or whatever.

 

I just think this sounds like a very risky proposition.  As much as I am not at all a pro-corporate person, and as much as I want the "little guy/gal" to have access to "big guy/gal" technologies, I do, from experience, think that certain technologies really need to have a fairly substantial enterprise behind them.

And I think network storage solutions are one of these technologies.

 

I'm a sysadmin, and I work at a major astrophysics laboratory, one of the biggest in the world.  Astrophysics is a field which uses a lot of data. A LOT.  People outside of the science community probably don't know much about science labs, but the primary driver of funding is grants.  Nearly all projects are paid for by grants (from wherever), meaning each particular project has its own specific budget.  Because of this, or because of the traditions of the institution, each project group works very independently.  As a sysadmin, we provide a standard environment to scientists, and make whatever modifications we can reasonably accommodate, up to a point.  A point, after which, some groups decide to do their own system administration because of how our "packaged" product doesn't fit their needs.

 

One of the most common ways this happens is with network storage.  We use a major NETwork APPliance vendor who's name you might guess, to provide storage for all basic user space as well as project space.  Project space must be paid out of grant money per project, and we charge a very fair price to projects for their network attached storage (which includes online snapshots taken hourly, daily and weekly up to three weeks, as well as a regular backup of all data to tape, and offsite tape storage.  And high availability, and high bandwidth connections to the servers, etc.  And drive failures are no big thing to us.  Often drives show up by delivery before anyone on staff even knew an old drive had failed. The price also includes staff support to provide all these services.

 

The point of this meandering story is that when some smaller projects see our quote for what we charger for storage, they balk.  They come back and say "yeah but I can buy this software raid for a tenth (or whatever small fraction) of that price.  We always try to talk them out of it..  We explain everything they'll get for the price, and, more importantly, what they're giving  up if they do it themselves.  We do not back up privately created and managed storage.  We make this VERY clear.  We give them a long talk about the importance of setting up a backup system.  In particular we emphasize the importance of TESTING the backup system they've set up, regularly, to make sure it serves their purpose.

 

Despite this warning, groups sometimes go the DIY route.  And every year or two, we end up with a scientist at our door, with a broken raid that won't rebuild, and a backup system that they thought worked, but didn't.   Years of data of multiple scientists have been lost this way.  There have been cases where they never bothered to set up a backup system, because despite our warnings, it's a RAID, right?  There's been cases where they've set up a backup system, and even tested it.  But, their backup system was a mirroring system, and they never tested their cobbled together scripts for what happens when a disk actually fails (the mirroring software dutifully mirrors the now empty failed disk mount point to the mirror... erasing it.  There's been all sorts of different failure modes.  The data loss has been devastating and unrecoverable.

 

Now, on the one hand, this is a GREAT argument that what Linus is talking about is absolutely needed.  And it SEEMS LIKE it should be totally doable.  But when it comes to a reliable, turnkey storage solution, it's incredibly hard to beat a solution that has big resources behind it.  Because of the seriousness of data loss (in certain applications), and because of the liability to the vendor, I think this is a situation where the cost of a really reliable storage solution winds up being much higher than whatever thumbnail solution you can come up with.  I also think that Linus' experience in particular, with network storage that's used personally for video libraries (not the worst thing in the world to lose all the data), or used professional as a video library, where there's a cloud storage solution (YouTube) inherently built in to your system without having to think about it, that Linus perhaps hasn't had to deal with what truly devastating data loss can look like firsthand. (I've only seen it as a sysadmin who couldn't be blamed for it; I'd hate to have to see this if such loss came as a result of a product I produced and supported.)

 

Having said all that, I understand market progression.  Eventually, someone will probably crack this nut and there will be a cheap (well, cheaper, but still not as cheap as you think), easy-to-use, yet reliable off-the-shelf thing that can be used to store any kind of data, no matter how critical it is.  I just don't think we're close to that yet.

 

Maybe I'm wrong.

 

With Frameworks, Linus was investing in something that was radical only from a philosophical viewpoint.  Technically, it's just a laptop.  And when he invested in it, it already existed.  This product is vaporware.  And it's a space where many others have tread, and yet failed to come up with solution that actually both succeeded, and remained with accessible pricing.  Unlike the Frameworks investment, this investment is a much bigger gamble — it's a bet that this is both the right time, and the right entity, to open up an entirely new market niche.

Again, I have no ethical complaint about Linus' investment choice with respect to LTT, LMG, or whatever.  I just think this one is a much riskier proposition than Frameworks.

Do you have all the details on his investment?  Do you have any details future that what he said on WAN show?

If the answers are no, I don't think you have enough information to know if he is right or wrong, and this is all unfounded speculation.

I would expect better from a professional sysadmin.

[Thomas A. Fine]

5 minutes ago, ToboRobot said:

I would expect better from a professional sysadmin.

Sir, this is an Internet Forum...

[Blue4130]

2 hours ago, Thomas A. Fine said:

Sorry, I left this out.  On the latest WAN show he described his high likelihood of investing in a brand new business that will develop a new network attached storage solution that's more user-friendly.  Not sure of the time stamp, but maybe roughly 2/3 of the way trhough.  At 2:06:03 in the video.

 

It sounds like a startup from someone that does know what they are doing (from within the industry), but still, I see it as very high risk.

I just watched it. To me it seems like you are expecting them to be going for the enterprise market, at least that is what your initial post seemed to imply. To me it looks like they are going more for the home lab crowd, which is more tolerant to unpolished over $. 

[Thomas A. Fine]

2 minutes ago, Blue4130 said:

I just watched it. To me it seems like you are expecting them to be going for the enterprise market, at least that is what your initial post seemed to imply. To me it looks like they are going more for the home lab crowd, which is more tolerant to unpolished over $. 

One major point I was trying to make is that there is no such thing.  A cheaper, easier-to-use storage product will be used in enterprises.  There's no way to stop that, and it would be weird to try.  Therefore, the product has to function with high reliability in that environment.

[XWAUForceflow]

4 hours ago, Thomas A. Fine said:

Sorry, I left this out.  On the latest WAN show he described his high likelihood of investing in a brand new business that will develop a new network attached storage solution that's more user-friendly.  Not sure of the time stamp, but maybe roughly 2/3 of the way trhough.  At 2:06:03 in the video.

 

It sounds like a startup from someone that does know what they are doing (from within the industry), but still, I see it as very high risk.

Is it a risky investment? Absolutely, that's why Linus himself called it an Angel investment. Those are always risky, but he also said it would be around a quarter or a million CAD. Honestly even if it doesn't pan out it's not going to be a big deal for Linus. I mean look at his house, the amount of money he is pouring into that is insane. Linus is rich, he can afford to loose the money. It would certainly sting, but it wouldn't be a big issue.

I mean he basically agreed to throw away 50k on that meme investment. I think he can afford to throw 250k at something he does believe in. And hey, with him being an investor and potentially using his name as a reference there is a chance that they will get that thing up and running.

[Middcore]

So what? It's his money and he's done dumber stuff with it, lest we forget the GameStop thing.

[divito]

4 hours ago, Thomas A. Fine said:

A cheaper, easier-to-use storage product will be used in enterprises.  There's no way to stop that, and it would be weird to try.  Therefore, the product has to function with high reliability in that environment.

So because enterprise users will try to use prosumer products instead of the pricier enterprise products, it "has to function with high reliability in that environment?" 

That's not a thing, and never will be. Companies will make products for their desired market and consumers, prosumers and enterprise will have to purchase with their needs and budgets in mind, as it's always been.

[Finkle Garbled Crap]

On 12/11/2022 at 4:31 AM, Thomas A. Fine said:

[clickbait headline intended]

 

[Editing to add: this is about a new proposed Linus investment, in a brand new network storage company, that was discussed on the latest WAN show, starting at 2:06:03.]

 

First, I have no issues with disclosed investments like this, generally.  Or even this in particular, in terms of ethics or whatever.

 

I just think this sounds like a very risky proposition.  As much as I am not at all a pro-corporate person, and as much as I want the "little guy/gal" to have access to "big guy/gal" technologies, I do, from experience, think that certain technologies really need to have a fairly substantial enterprise behind them.

And I think network storage solutions are one of these technologies.

 

I'm a sysadmin, and I work at a major astrophysics laboratory, one of the biggest in the world.  Astrophysics is a field which uses a lot of data. A LOT.  People outside of the science community probably don't know much about science labs, but the primary driver of funding is grants.  Nearly all projects are paid for by grants (from wherever), meaning each particular project has its own specific budget.  Because of this, or because of the traditions of the institution, each project group works very independently.  As a sysadmin, we provide a standard environment to scientists, and make whatever modifications we can reasonably accommodate, up to a point.  A point, after which, some groups decide to do their own system administration because of how our "packaged" product doesn't fit their needs.

 

One of the most common ways this happens is with network storage.  We use a major NETwork APPliance vendor who's name you might guess, to provide storage for all basic user space as well as project space.  Project space must be paid out of grant money per project, and we charge a very fair price to projects for their network attached storage (which includes online snapshots taken hourly, daily and weekly up to three weeks, as well as a regular backup of all data to tape, and offsite tape storage.  And high availability, and high bandwidth connections to the servers, etc.  And drive failures are no big thing to us.  Often drives show up by delivery before anyone on staff even knew an old drive had failed. The price also includes staff support to provide all these services.

 

The point of this meandering story is that when some smaller projects see our quote for what we charger for storage, they balk.  They come back and say "yeah but I can buy this software raid for a tenth (or whatever small fraction) of that price.  We always try to talk them out of it..  We explain everything they'll get for the price, and, more importantly, what they're giving  up if they do it themselves.  We do not back up privately created and managed storage.  We make this VERY clear.  We give them a long talk about the importance of setting up a backup system.  In particular we emphasize the importance of TESTING the backup system they've set up, regularly, to make sure it serves their purpose.

 

Despite this warning, groups sometimes go the DIY route.  And every year or two, we end up with a scientist at our door, with a broken raid that won't rebuild, and a backup system that they thought worked, but didn't.   Years of data of multiple scientists have been lost this way.  There have been cases where they never bothered to set up a backup system, because despite our warnings, it's a RAID, right?  There's been cases where they've set up a backup system, and even tested it.  But, their backup system was a mirroring system, and they never tested their cobbled together scripts for what happens when a disk actually fails (the mirroring software dutifully mirrors the now empty failed disk mount point to the mirror... erasing it.  There's been all sorts of different failure modes.  The data loss has been devastating and unrecoverable.

 

Now, on the one hand, this is a GREAT argument that what Linus is talking about is absolutely needed.  And it SEEMS LIKE it should be totally doable.  But when it comes to a reliable, turnkey storage solution, it's incredibly hard to beat a solution that has big resources behind it.  Because of the seriousness of data loss (in certain applications), and because of the liability to the vendor, I think this is a situation where the cost of a really reliable storage solution winds up being much higher than whatever thumbnail solution you can come up with.  I also think that Linus' experience in particular, with network storage that's used personally for video libraries (not the worst thing in the world to lose all the data), or used professional as a video library, where there's a cloud storage solution (YouTube) inherently built in to your system without having to think about it, that Linus perhaps hasn't had to deal with what truly devastating data loss can look like firsthand. (I've only seen it as a sysadmin who couldn't be blamed for it; I'd hate to have to see this if such loss came as a result of a product I produced and supported.)

 

Having said all that, I understand market progression.  Eventually, someone will probably crack this nut and there will be a cheap (well, cheaper, but still not as cheap as you think), easy-to-use, yet reliable off-the-shelf thing that can be used to store any kind of data, no matter how critical it is.  I just don't think we're close to that yet.

 

Maybe I'm wrong.

 

With Frameworks, Linus was investing in something that was radical only from a philosophical viewpoint.  Technically, it's just a laptop.  And when he invested in it, it already existed.  This product is vaporware.  And it's a space where many others have tread, and yet failed to come up with solution that actually both succeeded, and remained with accessible pricing.  Unlike the Frameworks investment, this investment is a much bigger gamble — it's a bet that this is both the right time, and the right entity, to open up an entirely new market niche.

Again, I have no ethical complaint about Linus' investment choice with respect to LTT, LMG, or whatever.  I just think this one is a much riskier proposition than Frameworks.

 

 

 

 

Honestly, from my point of View I see where your coming from, but I also disagree to a stant... 

 

Alright, 

 

1.) I don't really care 1 way or Another if there are People Rich out there Selling me stuff, I never had issues with it before... 

 

2.) Never really cared who even sold that stuff, either, nor even ran, or maintained it... 

 

3.) I never gave a Single crap, simply because People just left me alone, & let me live my life how I wanted, & if my life suck? It was only my fault, no one elses, but my fault... 

 

Oky? 

 

Today, it's all about forcing BS crap on the good People, crapping all over one's like me, & expecting to just straight up forcefully helping me get richer, just simply, because People think that's what I want? While they shove a Bogus fake life Style in my face, & attack me at the same time while they help me with a better life... 

 

What I really want, is People to just leave me alone, get off my back, & let me just live... 

 

-----------------------------------

 

To me, Money is becoming the New Addiction of America, & Hatefully Crapping on People, & breaking the Laws are also becoming an Addiction, too 

 

And to me, as much as i'm laughing at the Stupidity of it all, I just don't find it funny, nor do I find the humor in how People are just oky with thinking this is better Today at all... 

 

-----------------------------------

 

The truth is, if the little guy could not make it in a Market they had Total Control over? How the hell is the little guy going anywhere by forcefully crapping on everyone, & forcefully trying to get Rich, as the Rich also craps on them, & Quits the Market at the same time, & just leaves them in ta Massive mess anyways? 

 

I still say, things were better in 2016, I straight up do not care about the problems, & crap that was going on... I would give up being Rich, just for 1 more Day of 2016 again... 

 

Today had done nothing, but crap on People, & left People in a complete Crappy situation, & I just hate those who thought this was better, then what we had before... 

 

I'd rather choose a warm bed, food on table, roof over me, & peaceful Neighbors, & such... Then to call "whatever the hell you even remotely think Today" is better, -_- 

 

That's just my View... 

 

My View is? 

 

I don't need a Cold Mansion with 1,000's of Empty Rooms, loads of Enemies, & a Yatch / Plane in my life... 

 

I need a warm bed, some friends,

some food, & a life... 

--- That's it, & that's all I need, -_- 

 

--------------------------------------------

 

You don't Agree? 

 

As for Linus Tech Tips? He's probably not 

understanding, or getting that Message from 

his Viewing Audience, or has sorta lost touch 

with his Fan Base... I dunno, I haven't watched his Videos in a while, cause of Political BS issues kinda intervined, & since I only recently got Internet back... Soooo? Maybe i'm wrong?

[denishay]

On 12/11/2022 at 10:31 AM, Thomas A. Fine said:

[...]

Maybe I'm wrong.

[...]

Again, I have no ethical complaint about Linus' investment choice with respect to LTT, LMG, or whatever.  I just think this one is a much riskier proposition than Frameworks.

[...]

Yes, I think you have "tunnel vision" because of how you are daily immersed in tech.

 

There is a huge need for a turnkey solution better than "external USB drive" without paying the "Synology" tax. And I'm pretty sure that's the rationale behind it, nothing more, nothing less.

[NobleGamer]

On 12/11/2022 at 9:22 AM, Thomas A. Fine said:

One major point I was trying to make is that there is no such thing.  A cheaper, easier-to-use storage product will be used in enterprises.  There's no way to stop that, and it would be weird to try.  Therefore, the product has to function with high reliability in that environment.

Well more companies are buying high end gaming/consumer laptops (e.g. ASUS) so that they don't have to pay a bunch more for traditional business laptops (e.g. ThinkPad) that are more expensive and/or are less performant (especially in AI, 3D, or CPU intensive tasks).

 

Yet I don't get any impression that the gaming brands are shifting to more reliable than usual parts & such as a result of enterprise.

 

That's why I can't make the leap that every or at least a startup network storage company has to cater to enterprise expectations from the get go. If somehow a consumer startup gets most of their sales from Enterprise, then that's a different story where they have to confront supply & demand as opposed to your supposition that some enterprise demand exists, and so the company's fundamental supply needs to meet enterprise needs.

 

2 hours ago, denishay said:

There is a huge need for a turnkey solution better than "external USB drive" without paying the "Synology" tax. And I'm pretty sure that's the rationale behind it, nothing more, nothing less.

I respect that as someone who has done backups to an external drive, a WD consumer NAS, and a Synology NAS.

 

Practically speaking, the challenge I see is that the WD NAS was turnkey enough for me and meets the "huge need", as I believe that type of product is turnkey enough for the general consumer who wants a NAS.

 

Are there various downsides to WD NAS? Absolutely, but I don't know that most consumers are so fed up with it that they want to ditch the brand, as they could just get another NAS of the same brand, start backups on the new drive, and trash the other one.

 

So if you are right about the rationale, and we consider what's currently in that gap, then that means the target is a WD type NAS simplicity without the drawbacks but cheaper than Synology.

 

I wish this were as enticing as Framework from a market standpoint, but it's hard for me to see.

[Sol_NB]

I mean pretty sure Linus wants to invest money in it because he believes in the company and product and thinks it’s cool. Not sure there is more to it then that 

[Soundbored]

questionable copy pasta?