Ditto here. I actually use Net10 (which uses either AT&T or Verizon's towers/data/tech, depending on which phone you buy or bring)... $35/mo for unlimited talk/text, and 2GB data (then throttled after that, but with no further charges). So really, why the frig would I pay Verizon or AT&T $100/mo for something that I wouldn't use? In the 5 years I've been with Net10, I think I've gone over the 2GB cap once, a year ago (when I was farting around with rooting).
I would say the economy peaked last summer, when in June and early July, there were hundreds of DevOps job postings for the local area I am in (Austin, TX). In six weeks, the number of those was reduced by over 90%.
Minor point of order: The term is overused, and passe'. Christ sakes, every sysadmin I know in Portland quickly ran to LinkedIn and tacked the word "DevOps" to his resume last year, even if they never professionally touched any of the common tools or processes involved with the job. Doesn't matter though: Everyone and his dog knows how to run/use Puppet** these days, and thanks to the Forge and hiera, you don't even have to really know how to write a module these days (unless you need to do something unique/niche/etc). Toss in Docker, Consul, Ansible, etc, and you really don't need near as many people who get in there and do grunt-work for your company. Now getting people who can actually put together a usable automation process and who knows how to make it all work at a production level of reliability... different story. That said, you don't need nearly as many of those to make that happen either.
Anyrate, that's not exactly a usable metric (and neither is any other buzzword, really. I mean, c'mon, that's like gauging the economy in 1996 by how many MCSE job postings were open...)
** Let's be honest here: to 90% of PHBs, "DevOps" == "Guy who knows how to use Puppet". Never meant anything else, common sense be damned.
If it weren't for doing little side projects that involved the new skills/goodies, I'd still be an AIX sysadmin*. Ugh.
Go find something that needs done (but no one is really doing) in the company using those fancy new/cool/marketable skills, then go do it in your spare cycles, learning as you go.
* Funny thing, I became the sysadmin because I had a known habit of side-projects that required using the computer systems. That's how I got into this biz 22+ years ago (the previous sysadmin had flunked his drug test, and my boss at the time thought that his fresh-outta-school EE who tinkered on the work computers would be a perfect replacement).
Not completely related, but to answer your question... why not just publish paper-only stuff? Lots of self-publishing resources out there. For example, CreateSpace can publish his books on paper and sell it for him on Amazon for a nominal fee per copy, and for an additional fee, pimp his book to bookstores.
I guess what I'm getting at is, your bother is limiting himself if he only does eBooks.
No real sysadmin is going to use a $20 a year account just to (maybe) rely on onedrive. You're either rolling your own exchange server or renting email from google or office 365.
TBH, in a small enough operation you can skip even that... a simple Postfix/Dovecot/Spamassassin rig with IMAP enabled will do the job just as quickly, and for far less money (the entire thing can be parked on an old cast-off *nix server with a decent amount of disk space, or on a small AWS instance if that's how your small business rolls.)
In that case, I suspect this service is for the rubes, then... or does the ad injection happen in the emails themselves, out of reach of AdBlock/uBlock/etc?
This is gonna be a real nasty turnoff if they expect me to stay with Android (that, or drive me towards rooting the thing from now on...)
First off, I do not need or want *any* app digging deeper into the phone OS (where it can siphon off even more of my personal meta-info to sell). Oh, and imagine, if you will, your little built-in alarm clock app going off in the morning, then showing you a stupid advertisement as you reach for the thing. No frickin' thanks.
Second, More nagging... yay? Bad enough that some apps flip their shit when you don't rate them on Google Play.
In all seriousness, maybe they *should* get a team together and 'rip the bandage off' now, before another decade elapses and the thing gets even hairier...
Dynamics NAV (it was originally called Navision) used to be rather nice to work with... until, err, Microsoft starting making changes to it (around 2008-ish?). You could actually watch it turning to shit with each new version put out post-acquisition. I think I stopped bothering when I left the company I was working with, and went out of my way to tell subsequent employers to avoid the hell out of it.
On critical stuff, you want to make it a habit to mv stuff you're not familiar with somewhere (/tmp works most cases), test the system, test affected applications, double-check once more, and *then* rm.
On rm itself, I make it a habit to type the rm, double-check the command forwards and (literally) backwards, and only when satisfied hit enter. Ain't perfect, but I've caught potential disaster more times than I can count by bad regex, misplaces spacing, and other dumb tricks by reading it forwards and (literally!) backwards.
PS: The very first time I screwed up on rm, I learned the hard way to never, ever, ever type rm -rf.* to blanket-remove hidden files. Tends to nuke your entire server, including NFS mounted disks.
Originally, he was talking about Prohibition, but yeah, same-same.
Tax it high enough, and smuggling/home-growing/stolen black-market tobacco becomes profitable and enticing to folks who have enough creativity and the means to do it.
Believe it or not, this was a big thing in the 1970s-1990s in New York state, where smokes would be smuggled in from Kentucky (or other places where they were cheap), then sold at a price far lower than the New York smokes - yet was still highly profitable for the smuggler to do it.
Now Australia is rather isolated, but a large fishing boat or two and a determined bunch of black marketeers could still make out fairly well...
For the console players? Yeah, no. They most likely have pre-baked meshes to work with (to account for different-but-common facial shapes), and slap a texture on one made from a photograph. That's even more useless as a biometric base...
A 3-D scan usually results in a very messy pile of vertices/polygons that resemble the human face. The modeler (for the game) then takes that slop and dumps approximately a zillion polys and vertices from it, creating a mesh that has at least some sort of symmetry (for ease of rendering) and a *lot* less polygons (a typical game engine used today would explode if it had to simultaneously render, on-the-fly, two full basketball teams of figures at various distances, each with 50-75k-poly meshes just for their heads, let alone the additional burden of bodies, fabric dynamics, scenery, oh, and the meshes for the refs, etc.)
Usually when the modeler is done, the head might have up to 1k polygons on it (and that's really pushing things). Add Subdividing Surfaces, and you can squeeze the polycount way the hell down, even from that. The likeness is made-up for by texture (skinning), vertex placement (which will by necessity involve a lot of movement from original), displacement/bump texturing, and perhaps (if you have the GPU cycles to spare) a bit of animation that resembles the personality of the dude being scanned (say, a trademark smile).
Meanwhile, biometric data on a human face only calculates a relatively smaller number of points (that are not as easily affected by small variations in emotion, muscular movement, etc), and has fuck-all to do with what a human would expect to see. Neither one will resemble the other in any way, shape, or form, because they're made for two totally different purposes. For example: biometrics will have, say, vertices marking the center of each pupil, whereas a game mesh doesn't give a flying frig about the eyes beyond telling subdivision make them spheres and by the way here's the UV Map image to set the texture on them.
TL;DR: You cannot (as a practical matter) make a usable mesh of the guy from biometric data, nor can you make usable biometric data from the mesh used in-game.
Sheeit... you're kidding, right? Any EE that can't figure down basic wiring, look up Code, know how to keep the neutral line balanced, and get it done to ensure it's safe and compliant? Yeah, that person isn't worth a damn as an EE, seriously. When you compare it to the regs, boundaries, and environment adaptation required to rig-up industrial control systems? House wiring is a do-it-in-your-sleep piece of cake.
I get what you were getting at, but really... that bit wasn't the best example.
(Okay... *recent* EE grads who don't know shit outside of Verilog might have a hard time with it, but c'mon...)
"Companies", as in normal-sized critters, do this all the time. When Megacorps do it, it warrants attention.
In this case, straightforward restructuring of debt makes sense... Microsoft isn't growing like it used to, which only reinforces the need to sell bonds (as opposed to increasing shares of growth stock to cover it, or relying on future market income to wipe out the debt in short order.)
I see it as confirmation that Microsoft's growth is sputtering out, and they know it. Not saying they're dying by any stretch, but more along the lines of Microsoft becoming what IBM has been for a decade now... a maintenance-mode growth curve.
Sure, you will have rockstars there as well (I know quite a few living in Pune - they're trying to move here), but for every rockstar, you have something like 10,000 total incompetents whose code will require a massive overhaul just to get built without fatal errors.
Usually ends up costing more than its worth once you add it all up.
"Critical Mass" indicates that there are more facilities coming online, or at least publicly planning to. No indication of that in TFA... in fact, the closest they got is this:
For now, gas peaker plants still win out on price for projects that aren’t constrained by space, emissions, or urgency, said Ron Nichols, President of SCE, the California utility responsible for most of the biggest battery storage contracts. 3 But that may change in the next five years, he said.
"...may change in the next five years..." is nowhere near actual activity that would indicate a "critical mass" in industry.
How about they call us when it actually gets in motion - regionally, if not nationally or globally.
Oracle is good for only one thing these days: RAC.
For everything else, there's PostgreSQL, MSSQL (if you're into that), MySQL... and funny enough, a simple heartbeat script with STONITH capability, some decent replication tweaks, and a load balancer in front of the cluster can do most of what RAC can do on a practical level - at least enough to not really justify the $$$$$$$$ spent on RAC (again, in most cases.)
Put this way: Unless you're, say, running a brokerage and use a RTOS for sub-millisecond trading in front of that DB, or have a multi-zillion-dollar legacy system that was built around Oracle proprietary SQL/code? Hate to tell you, but Oracle is not really a standout these days.
(well, if you're an Oracle-certified DBA or analyst, I guess you can contrive an excuse or two to justify spending all that certification money...)
Slashdot Mirror
Ditto here. I actually use Net10 (which uses either AT&T or Verizon's towers/data/tech, depending on which phone you buy or bring)... $35/mo for unlimited talk/text, and 2GB data (then throttled after that, but with no further charges). So really, why the frig would I pay Verizon or AT&T $100/mo for something that I wouldn't use? In the 5 years I've been with Net10, I think I've gone over the 2GB cap once, a year ago (when I was farting around with rooting).
I would say the economy peaked last summer, when in June and early July, there were hundreds of DevOps job postings for the local area I am in (Austin, TX). In six weeks, the number of those was reduced by over 90%.
Minor point of order: The term is overused, and passe'. Christ sakes, every sysadmin I know in Portland quickly ran to LinkedIn and tacked the word "DevOps" to his resume last year, even if they never professionally touched any of the common tools or processes involved with the job. Doesn't matter though: Everyone and his dog knows how to run/use Puppet** these days, and thanks to the Forge and hiera, you don't even have to really know how to write a module these days (unless you need to do something unique/niche/etc). Toss in Docker, Consul, Ansible, etc, and you really don't need near as many people who get in there and do grunt-work for your company. Now getting people who can actually put together a usable automation process and who knows how to make it all work at a production level of reliability... different story. That said, you don't need nearly as many of those to make that happen either.
Anyrate, that's not exactly a usable metric (and neither is any other buzzword, really. I mean, c'mon, that's like gauging the economy in 1996 by how many MCSE job postings were open...)
** Let's be honest here: to 90% of PHBs, "DevOps" == "Guy who knows how to use Puppet". Never meant anything else, common sense be damned.
...that's where side projects come in.
If it weren't for doing little side projects that involved the new skills/goodies, I'd still be an AIX sysadmin*. Ugh.
Go find something that needs done (but no one is really doing) in the company using those fancy new/cool/marketable skills, then go do it in your spare cycles, learning as you go.
* Funny thing, I became the sysadmin because I had a known habit of side-projects that required using the computer systems. That's how I got into this biz 22+ years ago (the previous sysadmin had flunked his drug test, and my boss at the time thought that his fresh-outta-school EE who tinkered on the work computers would be a perfect replacement).
Wait... what did you mean by "suddenly"?
Not completely related, but to answer your question... why not just publish paper-only stuff? Lots of self-publishing resources out there. For example, CreateSpace can publish his books on paper and sell it for him on Amazon for a nominal fee per copy, and for an additional fee, pimp his book to bookstores.
I guess what I'm getting at is, your bother is limiting himself if he only does eBooks.
Seems like a solution using a problem.
No real sysadmin is going to use a $20 a year account just to (maybe) rely on onedrive. You're either rolling your own exchange server or renting email from google or office 365.
TBH, in a small enough operation you can skip even that... a simple Postfix/Dovecot/Spamassassin rig with IMAP enabled will do the job just as quickly, and for far less money (the entire thing can be parked on an old cast-off *nix server with a decent amount of disk space, or on a small AWS instance if that's how your small business rolls.)
In that case, I suspect this service is for the rubes, then... or does the ad injection happen in the emails themselves, out of reach of AdBlock/uBlock/etc?
...does this mean the overall XBox ROI has finally been reached/surpassed (what, 18-19 years later?)
Dude, AT&T ain't much better.
This is gonna be a real nasty turnoff if they expect me to stay with Android (that, or drive me towards rooting the thing from now on...)
First off, I do not need or want *any* app digging deeper into the phone OS (where it can siphon off even more of my personal meta-info to sell). Oh, and imagine, if you will, your little built-in alarm clock app going off in the morning, then showing you a stupid advertisement as you reach for the thing. No frickin' thanks.
Second, More nagging... yay? Bad enough that some apps flip their shit when you don't rate them on Google Play.
...and don't try to cross-upgrade (to a competing product).
...or replace a busted GFX card.
Yeah, thinking this to be a very bad idea for the consumer.
In all seriousness, maybe they *should* get a team together and 'rip the bandage off' now, before another decade elapses and the thing gets even hairier...
They likely store their comments as separate files - one per comment.
(no, really... has no one in Redmond ever heard of making their shit modular?)
Dynamics NAV (it was originally called Navision) used to be rather nice to work with... until, err, Microsoft starting making changes to it (around 2008-ish?). You could actually watch it turning to shit with each new version put out post-acquisition. I think I stopped bothering when I left the company I was working with, and went out of my way to tell subsequent employers to avoid the hell out of it.
This, right here... holy shit this!
On critical stuff, you want to make it a habit to mv stuff you're not familiar with somewhere (/tmp works most cases), test the system, test affected applications, double-check once more, and *then* rm.
On rm itself, I make it a habit to type the rm, double-check the command forwards and (literally) backwards, and only when satisfied hit enter. Ain't perfect, but I've caught potential disaster more times than I can count by bad regex, misplaces spacing, and other dumb tricks by reading it forwards and (literally!) backwards.
PS: The very first time I screwed up on rm, I learned the hard way to never, ever, ever type rm -rf .* to blanket-remove hidden files. Tends to nuke your entire server, including NFS mounted disks.
But... but... The Cloud! The Cloud is our DR solution!
(*chuckle*)
Originally, he was talking about Prohibition, but yeah, same-same.
Tax it high enough, and smuggling/home-growing/stolen black-market tobacco becomes profitable and enticing to folks who have enough creativity and the means to do it.
Believe it or not, this was a big thing in the 1970s-1990s in New York state, where smokes would be smuggled in from Kentucky (or other places where they were cheap), then sold at a price far lower than the New York smokes - yet was still highly profitable for the smuggler to do it.
Now Australia is rather isolated, but a large fishing boat or two and a determined bunch of black marketeers could still make out fairly well...
Oh - forgot...
For the console players? Yeah, no. They most likely have pre-baked meshes to work with (to account for different-but-common facial shapes), and slap a texture on one made from a photograph. That's even more useless as a biometric base...
Worse than that, really...
A 3-D scan usually results in a very messy pile of vertices/polygons that resemble the human face. The modeler (for the game) then takes that slop and dumps approximately a zillion polys and vertices from it, creating a mesh that has at least some sort of symmetry (for ease of rendering) and a *lot* less polygons (a typical game engine used today would explode if it had to simultaneously render, on-the-fly, two full basketball teams of figures at various distances, each with 50-75k-poly meshes just for their heads, let alone the additional burden of bodies, fabric dynamics, scenery, oh, and the meshes for the refs, etc.)
Usually when the modeler is done, the head might have up to 1k polygons on it (and that's really pushing things). Add Subdividing Surfaces, and you can squeeze the polycount way the hell down, even from that. The likeness is made-up for by texture (skinning), vertex placement (which will by necessity involve a lot of movement from original), displacement/bump texturing, and perhaps (if you have the GPU cycles to spare) a bit of animation that resembles the personality of the dude being scanned (say, a trademark smile).
Meanwhile, biometric data on a human face only calculates a relatively smaller number of points (that are not as easily affected by small variations in emotion, muscular movement, etc), and has fuck-all to do with what a human would expect to see. Neither one will resemble the other in any way, shape, or form, because they're made for two totally different purposes. For example: biometrics will have, say, vertices marking the center of each pupil, whereas a game mesh doesn't give a flying frig about the eyes beyond telling subdivision make them spheres and by the way here's the UV Map image to set the texture on them.
TL;DR: You cannot (as a practical matter) make a usable mesh of the guy from biometric data, nor can you make usable biometric data from the mesh used in-game.
Most electrical engineers can't re-wire a house.
Sheeit... you're kidding, right? Any EE that can't figure down basic wiring, look up Code, know how to keep the neutral line balanced, and get it done to ensure it's safe and compliant? Yeah, that person isn't worth a damn as an EE, seriously. When you compare it to the regs, boundaries, and environment adaptation required to rig-up industrial control systems? House wiring is a do-it-in-your-sleep piece of cake.
I get what you were getting at, but really... that bit wasn't the best example.
(Okay... *recent* EE grads who don't know shit outside of Verilog might have a hard time with it, but c'mon...)
"Companies", as in normal-sized critters, do this all the time. When Megacorps do it, it warrants attention.
In this case, straightforward restructuring of debt makes sense... Microsoft isn't growing like it used to, which only reinforces the need to sell bonds (as opposed to increasing shares of growth stock to cover it, or relying on future market income to wipe out the debt in short order.)
I see it as confirmation that Microsoft's growth is sputtering out, and they know it. Not saying they're dying by any stretch, but more along the lines of Microsoft becoming what IBM has been for a decade now... a maintenance-mode growth curve.
Umm, have you *seen* the code from such locales?
Sure, you will have rockstars there as well (I know quite a few living in Pune - they're trying to move here), but for every rockstar, you have something like 10,000 total incompetents whose code will require a massive overhaul just to get built without fatal errors.
Usually ends up costing more than its worth once you add it all up.
Not to mention safety. I'd hate to be the nearest Fire Department to that place...
Good article, but...
"Critical Mass" indicates that there are more facilities coming online, or at least publicly planning to. No indication of that in TFA... in fact, the closest they got is this:
"...may change in the next five years..." is nowhere near actual activity that would indicate a "critical mass" in industry.
How about they call us when it actually gets in motion - regionally, if not nationally or globally.
Well there's RAC, but a decent bit of scripting, replication tweaks, and a competent load balancer can pretty much obviate that as well.
*chuckle*...
Oracle is good for only one thing these days: RAC.
For everything else, there's PostgreSQL, MSSQL (if you're into that), MySQL... and funny enough, a simple heartbeat script with STONITH capability, some decent replication tweaks, and a load balancer in front of the cluster can do most of what RAC can do on a practical level - at least enough to not really justify the $$$$$$$$ spent on RAC (again, in most cases.)
Put this way: Unless you're, say, running a brokerage and use a RTOS for sub-millisecond trading in front of that DB, or have a multi-zillion-dollar legacy system that was built around Oracle proprietary SQL/code? Hate to tell you, but Oracle is not really a standout these days.
(well, if you're an Oracle-certified DBA or analyst, I guess you can contrive an excuse or two to justify spending all that certification money...)