This. It isn't like backup software is anything new. The best backup software I've seen would be Retrospect, because it did compression, deduplication, encryption, and it allowed one to move backed up data to new media, verify the data, expire old backups, move selected sets to WORM media, and so on. It was a great way to archive files because you could just put the files into it, hit "archive", and pretty much forget about it.
Then, there are programs like NetBackup and TSM/Spectrum Protect, which have had deduplication for decades. Why isn't this common? The only program I know that allows multiple hosts to back up to a single repository is obnam, and after heavy testing, I had so much data lost, it wasn't worth continuing.
Downside is that it stopped supporting virtually all Blu-Ray optical drives, and it is so expensive. So, the best we have on the consumer end is Veeam. Stuff like Macrium Reflect, Acronis TrueImage are OK... but I have had issues with both silently failing and not bothering to notify the user that backup jobs are not going on. However, none of these allow one to take older backed up files and move those to other media.
Which leaves us with the tools available:
Veeam -- best of breed, offers deduplication, encryption, compression, and bare metal recovery. For free. The more advanced features cost more. However, you do need some place to back up your data to... which is likely an external HDD or NAS, both of which are vulnerable to ransomware.
Windows Server backup -- decent, but vulnerable to ransomware.
CrashPlan/Mozy/Carbonite/BackBlaze/etc. Cloud computing is OK... but has not stood the test of time, and one's network connection may make a restore not feasible, or expensive (if one has to ask for a removable HDD with the data.) Plus, what is to stop a cloud provider from telling everyone to go pike off and shut down, or charge them at extreme rates if they want their data back. Cloud providers are like banks, before the 1929 crash... great and fun, but should something seriously happen, one is basically hosed.
Sync/backup tools to a NAS -- Both Synology and QNAP have backup and folder/file sync tools to their products.
For Macs, there is Time Machine... but that only backs up the root partition and OS drive.
Linux, there is no way to image and restore the OS unless you do offline backups. However, there are a lot of good file/directory backup tools, be it using btrfs and snapshots, borg backup, plain old tar, or many other utilities.
Yes, one can get protection with a mishmash of utilities, but an ideal would be a tape drive that you plug in, or one attached to a NAS so backups get dumped to disk (or SSD), then migrated to tape.
Can we get a consumer-level tape drive, something that works with USB 3.x, costs around $1000, with media costing $10-25 a cartridge, with an actual archival life, and some basic features like AES-256 encryption and compression in hardware? Having WORM media would be nice as well.
This would solve a ton of storage/backup issues. A lot of people don't have the network connection to make cloud backups (much less complete disaster recoveries) feasible. People may not trust the cloud. Regulations may prohibit use of a remote backup provider. Or, someone just likes having the peace of mind of physical control of their data on media which might last past the end of this decade.
There are tons of high capacity optical offerings available (Sony has some)... but other than specialty markets, they are not worth it. What would be nice would be a tape or even an optical drive for backups.
Optical wouldn't be bad either. 400 CD jukeboxes are a solved problem, so having a BD-like format with 2-3 terabytes per disk would solve a lot of backup problems, especially with WORM capability to fend off ransomware.
This. I'd like IoT devices to communicate to a hardened hub (or perhaps hubs for redundancy), and the hubs do the work. Each device would have a manifest of what servers it talks to as well, so any communication outside of the pre-arranged files gets blocked.
This is so simple, it is just crazy that this hasn't been made into a standard.
My question about "securing". Is this more to secure the device against the user wanting to do stuff with it (anti-jailbreak), or secure it against remote bad guys? I worry every time I see anything government based going into security, because I expect more DMCA type stuff, and not stuff that actually keeps the bad guys out.
What I don't get is why the round keys are even still used. The basic Abloy style disk detainer locks have fallen out of patent for almost a century, and even a version of that like what is in the Sargent & Greenleaf Environmental padlock which uses fewer disks, would be more than good enough for a basic security container. Heck, a lever lock, which is 1700s technology, would be useful and decently secure (hell, AT&T/Ma Bell used a variant of lever locks (29B/30C) for decades which were extremely pick resistant in pay phones. (Matt Blaze did an excellent writeup of this.)
In fact, with the ease of modern machining, the fact that Ace locks are still used is just befuddling. One can make a lock with fewer moving parts that has far more security. One doesn't need Abloy PROTEC2 level either. A STRATTEC lock, which is what Ford and GM use for automotive tasks would be more than enough for a basic security container, providing not just pick resistance, but decent resistance to tough usage conditions (dusty environments, etc.)
I just think it is befuddling that in this era, we are still using ACE locks and CH751-style disk tumbler locks. If a cheap pin tumbler lock has to be used, even the cheap Chinese makers are turning out dimple locks that have some pick resistance. Realistically, we should be using things like EVVA MKS, Abloy PROTEC2 CLIQ, or other locks which have no real way to make a tool to pick or bypass nondestructively.
I have seen that before. I once dealt with a "DBA" who didn't understand that "DROP TABLE" autocommits. Thnk $DEITY I had backups of the MySQL database, which were sent to a NAS that I had my manager purchase for "general admin tasks". Said DBA was supposedly a Linux "admin"... but when he got the "deer in headlights" look when I asked him to fetch updates from the internal mirrored repos first, since the machines were not a VLAN not accessible to or from the Internet.
This. IT is stable enough [1] that it doesn't need even a B. S. in CS to work well. It should be a trade, and vendor independant. Certs are pointless, because if one uses plumbing, why would you need to know ProPex's specific pipes in order to know plumbing in general?
Plus, it sets a standard. Someone can be a chatter-monkey, but it would be like an A/C repairman without their TACL license (here in Texas).
We need a licensing and trade body. We already have junior, mid-level, and senior IT, might as well make that apprentice, journeyman, and master.
[1]: Stable as in in 10 years, we will still have the same issues as now. Mail will still have spam, databases will still have bad table designs, we will have ransomware, and so on.
There have been processes for behavioral tracking for years now. The trick is to root the device, yank the Chinese certificates out of your root CA store [1], add outgoing blocks on the iptables level to ensure that it doesn't phone home, add some ad blocking, and you will have a decent phone for a cheap price. Ideally, install an OS like LineageOS (if available.)
[1]: It is interesting to see what both Apple and Android device makers stick in the root CA store. It is wise to reduce that number.
If fusion were to be able to be done, it would fundamentally change every aspect of our society.
I will propose an assertion: Energy = wealth.
If fusion becomes inexpensive and commercially available, perhaps along the "too cheap to meter" line, there would be a lot of things that are doable, which we could do, which we couldn't before:
1: Desalination plants on a large scale, combined with water pipelines. Once the warlords are out of the way, African droughts and famines would be over, and there would be a lot more arable, fertile land available.
2: Thermal depolymerization can be used as a very effective way to recycle plastics. Combine that with a ship, and it can actually harvest the plastic in the Pacific Gyre and turn it back into fuel.
3: Direct CO2 extraction out of the atmosphere, perhaps reusing it as fuels.
4: The ability to create stuff that would be prohibitivily expensive. Same thing happened with aluminum. Before electricity was available, getting aluminum from bauxite was extremely expensive. With energy cheaply available, titanium would be able to be used more.
5: The ability to do transportation networks which are wasteful on fuel right now. Cheap fuel + electric vehicles mean a bus service that can even handle rural areas with 1-2 hours on a street.
So, all, and all, if fusion is available, it will fundamentally change life as we know it, just as electricity changed things. So, it is worth keeping at it.
We already have Google, Amazon, and Apple in this space. Why do we need another smart speaker? More exactly, why do we need a smart speaker in the first place?
The ironic thing is that secure system design isn't anything new. There have been books on this since the 1970s. It is a solved problem. The issue is that actually bothering to implement defense in depth is something companies don't want devs to spend time on. Again, the "security has no ROI" mantra.
Were things designed from the ground right using proven security techniques, this wouldn't be an issue.
Android has similar, with long-pressing on an app's notifications. The ability to disable notifications was in result due to a "service", AirPush that got installed with various apps, which would spam the notification bar with crap.
Sticking the head in the sand isn't going to help things. Other countries are ahead with AI development, and if a country is to remain a superpower, having AI research (as well as supercomputers to back it up) is a must. AI is useful to figure out scenarios on how a country is going to attack and figure out the best defense for it, and ultimately, AI may replace generals as the best way of pushing forward in a theater of combat, just as chessmasters and Go veterans have been set aside.
If AI research is banned in the US, China and Russia will happy continue carrying the torch.
Even then, a phone isn't exactly a conspicuous consumption device. From what I see, Louis Vuitton is making money hand over fit with their "Eye Trunk" case which goes for quad digits. A good iPhone case would have gotten more income than a phone. Especially if the case had a Bluetooth item with the concierge button or other niceties.
The problem is that Vertu wound up in a market that was "neither fish nor fowl". If they sold expensive, bespoke, phones designed from the ground up for the individual, similar to how Prevost and Newell make buses, and sold them for six digits, they would still be around. If they sold phones for 50-100% above the normal rate mass-produced (perhaps by some OEM/ODM out of Taiwan or Japan), and sold them in volume, they likely would have a decent business model. However, for a phone (which is a mass produced, disposable, gadget with an expiration date), it doesn't make sense to charge high prices.
Vertu might have been better off designing a "timeless" smart watch, where it would be self-winding, use a capacitor instead of a battery, using a very low power screen, minimizing the amount of apps (or complications) to save battery life.
Or, they could have done something unique: Vertu could have made a PDA which connects to the user's phone via Bluetooth. This device would not need to have state of the art stuff on it, but designed from the ground up for being timeless and elegant. That way, the "common" smartphone can remain in a pocket or purse, while the PDA, a status symbol, could be used. Then, Vertu could look at early PalmOS, which was designed to be simple, yet effective, and design a device around it, perhaps with the capacity for voice transactions (via Bluetooth like the HTC Mini). That way, they would have something unique to sell that would last a long time (when the smartphone gets tossed, re-pair with the new one, continue as before.)
Here in Austin, there are some roach coaches that won't accept cash, and are usually running Square, PayPal, or some other item. I carry cash, but that is for emergencies (card gets declined [1].)
[1]: I've had my bank be overly cautious and freeze my card, especially when I head out for a trip to rural Texas and buy gas at a small town, as well as get something to eat there.
This is how countries stop being world leaders. When businesses and scientists leave for greener pastures, you know things are hosed. I'd assert that the startup visa rule will do worse for the US economy than the past decisions to give businesses breaks by letting China have the manufacturing jobs.
Notes at least gave everyone their own private key that was used for everything.
My question is, can S/MIME, or even Symantec's PGP fill in this gap for secure E-mail? Symantec's PGP had the ability to use ADK (additional decryption keys) for recovery, and work pretty well. It would be another add-on to Outlook, but done right, a compromise of a mail server would be mitigated by doing this.
It makes sense that they are pitching it for energy-time units. A typical RV battery would cost about $100 for a kilowatt-hour worth of charge on average. However, lead-acid batteries get permanently damaged if drawn below 50% SoC, so $250 per kWh is a pretty good deal, factoring in standard battery prices.
It looks like the Aussies are getting their money's worth, just on the battery capacity side. The LiFePO4 batteries used tend to be long lasting, so this is something that would require little in upkeep (no watering systems needed.)
The only place those seem to be found here in the US are decking screws.
I wonder if it is because the US was first in a lot of fields, so chose the suboptimal solutions (120 VAC, imperial system, no red/yellow or green/yellow traffic combinations to warn about a change in the light, etc.) Phillips screws were deliberately chosen by Henry Ford because they cam out before snapping... but with the fact that every mechanic uses a precisely calibrated torque wrench, that style of fastener needs to hit the ashbin of history and be replaced by the Robertson type, Torx, or something modern.
Hybrids bring one advantage to the table, especially as congestion gets worse: At idle, the IC engine can be turned off so the vehicle takes zero power, other than to handle the HVAC and electronics. This means less pollution overall.
Of course, a hybrid with an inverter is useful if one has a house or condo and power goes out. It is a lot quieter and likely more fuel efficient than a generator.
Torx screws actually have a purpose than being a new fastener. They don't cam out, unlike Phillips screws. Even the Philllips Screw Company offers Hex Stix, which are similar to Torx/hexilobe screws.
It also can put you in the hospital. Doing 100+ hours a week with a shitty sleep schedule likely means a craptastic diet and no exercise. This is a great recipe for major health problems, and there is a good chance that insurance won't cover any of it.
DevOps is good, but you better know the fashion program of the week. If you know GitHub Enterprise and BitBucket, but not GitLab, the interviewer will stop and show you the door. A few months ago, it was kubernates.
Traditional system admin stuff is shrinking. When I interviewed last year, companies were going whole-hog on services like Amazon Lambda, because it allowed them to fire their server and ops staff, and keep only one "IT" person -- the guy who handed out the IAM AWS accounts. Everything else was done by offshored dev houses.
Slashdot Mirror
This. It isn't like backup software is anything new. The best backup software I've seen would be Retrospect, because it did compression, deduplication, encryption, and it allowed one to move backed up data to new media, verify the data, expire old backups, move selected sets to WORM media, and so on. It was a great way to archive files because you could just put the files into it, hit "archive", and pretty much forget about it.
Then, there are programs like NetBackup and TSM/Spectrum Protect, which have had deduplication for decades. Why isn't this common? The only program I know that allows multiple hosts to back up to a single repository is obnam, and after heavy testing, I had so much data lost, it wasn't worth continuing.
Downside is that it stopped supporting virtually all Blu-Ray optical drives, and it is so expensive. So, the best we have on the consumer end is Veeam. Stuff like Macrium Reflect, Acronis TrueImage are OK... but I have had issues with both silently failing and not bothering to notify the user that backup jobs are not going on. However, none of these allow one to take older backed up files and move those to other media.
Which leaves us with the tools available:
Veeam -- best of breed, offers deduplication, encryption, compression, and bare metal recovery. For free. The more advanced features cost more. However, you do need some place to back up your data to... which is likely an external HDD or NAS, both of which are vulnerable to ransomware.
Windows Server backup -- decent, but vulnerable to ransomware.
CrashPlan/Mozy/Carbonite/BackBlaze/etc. Cloud computing is OK... but has not stood the test of time, and one's network connection may make a restore not feasible, or expensive (if one has to ask for a removable HDD with the data.) Plus, what is to stop a cloud provider from telling everyone to go pike off and shut down, or charge them at extreme rates if they want their data back. Cloud providers are like banks, before the 1929 crash... great and fun, but should something seriously happen, one is basically hosed.
Sync/backup tools to a NAS -- Both Synology and QNAP have backup and folder/file sync tools to their products.
For Macs, there is Time Machine... but that only backs up the root partition and OS drive.
Linux, there is no way to image and restore the OS unless you do offline backups. However, there are a lot of good file/directory backup tools, be it using btrfs and snapshots, borg backup, plain old tar, or many other utilities.
Yes, one can get protection with a mishmash of utilities, but an ideal would be a tape drive that you plug in, or one attached to a NAS so backups get dumped to disk (or SSD), then migrated to tape.
Can we get a consumer-level tape drive, something that works with USB 3.x, costs around $1000, with media costing $10-25 a cartridge, with an actual archival life, and some basic features like AES-256 encryption and compression in hardware? Having WORM media would be nice as well.
This would solve a ton of storage/backup issues. A lot of people don't have the network connection to make cloud backups (much less complete disaster recoveries) feasible. People may not trust the cloud. Regulations may prohibit use of a remote backup provider. Or, someone just likes having the peace of mind of physical control of their data on media which might last past the end of this decade.
There are tons of high capacity optical offerings available (Sony has some)... but other than specialty markets, they are not worth it. What would be nice would be a tape or even an optical drive for backups.
Optical wouldn't be bad either. 400 CD jukeboxes are a solved problem, so having a BD-like format with 2-3 terabytes per disk would solve a lot of backup problems, especially with WORM capability to fend off ransomware.
This. I'd like IoT devices to communicate to a hardened hub (or perhaps hubs for redundancy), and the hubs do the work. Each device would have a manifest of what servers it talks to as well, so any communication outside of the pre-arranged files gets blocked.
This is so simple, it is just crazy that this hasn't been made into a standard.
My question about "securing". Is this more to secure the device against the user wanting to do stuff with it (anti-jailbreak), or secure it against remote bad guys? I worry every time I see anything government based going into security, because I expect more DMCA type stuff, and not stuff that actually keeps the bad guys out.
What I don't get is why the round keys are even still used. The basic Abloy style disk detainer locks have fallen out of patent for almost a century, and even a version of that like what is in the Sargent & Greenleaf Environmental padlock which uses fewer disks, would be more than good enough for a basic security container. Heck, a lever lock, which is 1700s technology, would be useful and decently secure (hell, AT&T/Ma Bell used a variant of lever locks (29B/30C) for decades which were extremely pick resistant in pay phones. (Matt Blaze did an excellent writeup of this.)
In fact, with the ease of modern machining, the fact that Ace locks are still used is just befuddling. One can make a lock with fewer moving parts that has far more security. One doesn't need Abloy PROTEC2 level either. A STRATTEC lock, which is what Ford and GM use for automotive tasks would be more than enough for a basic security container, providing not just pick resistance, but decent resistance to tough usage conditions (dusty environments, etc.)
I just think it is befuddling that in this era, we are still using ACE locks and CH751-style disk tumbler locks. If a cheap pin tumbler lock has to be used, even the cheap Chinese makers are turning out dimple locks that have some pick resistance. Realistically, we should be using things like EVVA MKS, Abloy PROTEC2 CLIQ, or other locks which have no real way to make a tool to pick or bypass nondestructively.
I have seen that before. I once dealt with a "DBA" who didn't understand that "DROP TABLE" autocommits. Thnk $DEITY I had backups of the MySQL database, which were sent to a NAS that I had my manager purchase for "general admin tasks". Said DBA was supposedly a Linux "admin"... but when he got the "deer in headlights" look when I asked him to fetch updates from the internal mirrored repos first, since the machines were not a VLAN not accessible to or from the Internet.
This. IT is stable enough [1] that it doesn't need even a B. S. in CS to work well. It should be a trade, and vendor independant. Certs are pointless, because if one uses plumbing, why would you need to know ProPex's specific pipes in order to know plumbing in general?
Plus, it sets a standard. Someone can be a chatter-monkey, but it would be like an A/C repairman without their TACL license (here in Texas).
We need a licensing and trade body. We already have junior, mid-level, and senior IT, might as well make that apprentice, journeyman, and master.
[1]: Stable as in in 10 years, we will still have the same issues as now. Mail will still have spam, databases will still have bad table designs, we will have ransomware, and so on.
There have been processes for behavioral tracking for years now. The trick is to root the device, yank the Chinese certificates out of your root CA store [1], add outgoing blocks on the iptables level to ensure that it doesn't phone home, add some ad blocking, and you will have a decent phone for a cheap price. Ideally, install an OS like LineageOS (if available.)
[1]: It is interesting to see what both Apple and Android device makers stick in the root CA store. It is wise to reduce that number.
You sacrificed goats? I was lucky enough to just get away with chickens, especially when using differential SCSI.
If fusion were to be able to be done, it would fundamentally change every aspect of our society.
I will propose an assertion: Energy = wealth.
If fusion becomes inexpensive and commercially available, perhaps along the "too cheap to meter" line, there would be a lot of things that are doable, which we could do, which we couldn't before:
1: Desalination plants on a large scale, combined with water pipelines. Once the warlords are out of the way, African droughts and famines would be over, and there would be a lot more arable, fertile land available.
2: Thermal depolymerization can be used as a very effective way to recycle plastics. Combine that with a ship, and it can actually harvest the plastic in the Pacific Gyre and turn it back into fuel.
3: Direct CO2 extraction out of the atmosphere, perhaps reusing it as fuels.
4: The ability to create stuff that would be prohibitivily expensive. Same thing happened with aluminum. Before electricity was available, getting aluminum from bauxite was extremely expensive. With energy cheaply available, titanium would be able to be used more.
5: The ability to do transportation networks which are wasteful on fuel right now. Cheap fuel + electric vehicles mean a bus service that can even handle rural areas with 1-2 hours on a street.
So, all, and all, if fusion is available, it will fundamentally change life as we know it, just as electricity changed things. So, it is worth keeping at it.
We already have Google, Amazon, and Apple in this space. Why do we need another smart speaker? More exactly, why do we need a smart speaker in the first place?
The ironic thing is that secure system design isn't anything new. There have been books on this since the 1970s. It is a solved problem. The issue is that actually bothering to implement defense in depth is something companies don't want devs to spend time on. Again, the "security has no ROI" mantra.
Were things designed from the ground right using proven security techniques, this wouldn't be an issue.
Android has similar, with long-pressing on an app's notifications. The ability to disable notifications was in result due to a "service", AirPush that got installed with various apps, which would spam the notification bar with crap.
Sticking the head in the sand isn't going to help things. Other countries are ahead with AI development, and if a country is to remain a superpower, having AI research (as well as supercomputers to back it up) is a must. AI is useful to figure out scenarios on how a country is going to attack and figure out the best defense for it, and ultimately, AI may replace generals as the best way of pushing forward in a theater of combat, just as chessmasters and Go veterans have been set aside.
If AI research is banned in the US, China and Russia will happy continue carrying the torch.
Even then, a phone isn't exactly a conspicuous consumption device. From what I see, Louis Vuitton is making money hand over fit with their "Eye Trunk" case which goes for quad digits. A good iPhone case would have gotten more income than a phone. Especially if the case had a Bluetooth item with the concierge button or other niceties.
The problem is that Vertu wound up in a market that was "neither fish nor fowl". If they sold expensive, bespoke, phones designed from the ground up for the individual, similar to how Prevost and Newell make buses, and sold them for six digits, they would still be around. If they sold phones for 50-100% above the normal rate mass-produced (perhaps by some OEM/ODM out of Taiwan or Japan), and sold them in volume, they likely would have a decent business model. However, for a phone (which is a mass produced, disposable, gadget with an expiration date), it doesn't make sense to charge high prices.
Vertu might have been better off designing a "timeless" smart watch, where it would be self-winding, use a capacitor instead of a battery, using a very low power screen, minimizing the amount of apps (or complications) to save battery life.
Or, they could have done something unique: Vertu could have made a PDA which connects to the user's phone via Bluetooth. This device would not need to have state of the art stuff on it, but designed from the ground up for being timeless and elegant. That way, the "common" smartphone can remain in a pocket or purse, while the PDA, a status symbol, could be used. Then, Vertu could look at early PalmOS, which was designed to be simple, yet effective, and design a device around it, perhaps with the capacity for voice transactions (via Bluetooth like the HTC Mini). That way, they would have something unique to sell that would last a long time (when the smartphone gets tossed, re-pair with the new one, continue as before.)
Here in Austin, there are some roach coaches that won't accept cash, and are usually running Square, PayPal, or some other item. I carry cash, but that is for emergencies (card gets declined [1].)
[1]: I've had my bank be overly cautious and freeze my card, especially when I head out for a trip to rural Texas and buy gas at a small town, as well as get something to eat there.
This is how countries stop being world leaders. When businesses and scientists leave for greener pastures, you know things are hosed. I'd assert that the startup visa rule will do worse for the US economy than the past decisions to give businesses breaks by letting China have the manufacturing jobs.
Notes at least gave everyone their own private key that was used for everything.
My question is, can S/MIME, or even Symantec's PGP fill in this gap for secure E-mail? Symantec's PGP had the ability to use ADK (additional decryption keys) for recovery, and work pretty well. It would be another add-on to Outlook, but done right, a compromise of a mail server would be mitigated by doing this.
I wonder if a battery system like this can replace a baseline power plant for a small town. If it could, it would be very useful.
It makes sense that they are pitching it for energy-time units. A typical RV battery would cost about $100 for a kilowatt-hour worth of charge on average. However, lead-acid batteries get permanently damaged if drawn below 50% SoC, so $250 per kWh is a pretty good deal, factoring in standard battery prices.
It looks like the Aussies are getting their money's worth, just on the battery capacity side. The LiFePO4 batteries used tend to be long lasting, so this is something that would require little in upkeep (no watering systems needed.)
The only place those seem to be found here in the US are decking screws.
I wonder if it is because the US was first in a lot of fields, so chose the suboptimal solutions (120 VAC, imperial system, no red/yellow or green/yellow traffic combinations to warn about a change in the light, etc.) Phillips screws were deliberately chosen by Henry Ford because they cam out before snapping... but with the fact that every mechanic uses a precisely calibrated torque wrench, that style of fastener needs to hit the ashbin of history and be replaced by the Robertson type, Torx, or something modern.
Hybrids bring one advantage to the table, especially as congestion gets worse: At idle, the IC engine can be turned off so the vehicle takes zero power, other than to handle the HVAC and electronics. This means less pollution overall.
Of course, a hybrid with an inverter is useful if one has a house or condo and power goes out. It is a lot quieter and likely more fuel efficient than a generator.
Torx screws actually have a purpose than being a new fastener. They don't cam out, unlike Phillips screws. Even the Philllips Screw Company offers Hex Stix, which are similar to Torx/hexilobe screws.
It also can put you in the hospital. Doing 100+ hours a week with a shitty sleep schedule likely means a craptastic diet and no exercise. This is a great recipe for major health problems, and there is a good chance that insurance won't cover any of it.
DevOps is good, but you better know the fashion program of the week. If you know GitHub Enterprise and BitBucket, but not GitLab, the interviewer will stop and show you the door. A few months ago, it was kubernates.
Traditional system admin stuff is shrinking. When I interviewed last year, companies were going whole-hog on services like Amazon Lambda, because it allowed them to fire their server and ops staff, and keep only one "IT" person -- the guy who handed out the IAM AWS accounts. Everything else was done by offshored dev houses.