Java in the web client is dead (so is Silverlight / Flash)... go Javascript / HTML5 if you have to do things on the client. Java on the server side... isn't going away for many decades.
The only downside of Java is that it's rather heavy for "one-of-a-kind" web pages. There's a lot of setup that you have to learn (Maven archtypes help) before you get HTML on the web browser. But as soon as you need something that can scale, talk to disparate systems, support unit testing, etc., it's far better then PHP. PHP just falls apart once you get past a handful of PHP pages.
If by SSD support you mean TRIM, that came in around the era of Vista or Win7. WinXP definitely didn't have it and SSDs definitely got slower after a year or two of use (until you redid the system).
The HTC One (m8), released this year, also has a battery stretch feature.
Overall, very happy with the HTC One (m8) other then I wish it was about 1/2" to 3/4" smaller. HTC did a good job with the UI and it's very snappy, makes my 18 month old Asus TF700T Transformer tablet feel slow (both are quad-core units).
You really don't know what you're missing. For business laptops, we've made the switch to 100% SSDs for 2-3 years now (ever since they dropped down to $1.50-$1.75 per GB). Granted, these are all uses who can function with only a 128GB SSD. Which holds true for probably 90% of office workers who have access to a file server (instead of storing business critical data on their HD).
Now, instead of waiting on their HD to seek around and find information (a boot process measured in minutes, program loading times measured in 10s of seconds), boot-up takes under 20-30s and program loading times are near instant. What you *will* notice is that your CPU is now the bottleneck (oops). For development work or any thing where you need to do two or three things at once, or run something disk-intensive like a scan or search of files, SSDs are a must-have. I will regularly kick off compiles / version control updates / searches, and still be able to use the machine for other things while it thinks.
Just makes sure you have a good backup system in place. On the Windows-side, I recommend Acronis True Image writing to a 2nd old-style HD inside the case. Or an external 1TB USB3 drive that you leave connected during the backup window. That is not because SSDs are unreliable (unless you buy crap like OCZ), it's because their failure modes are such (if the controller goes crazy) that data recovery is highly unlikely.
No, becasue the only food they can afford is salt laden fatty food.
If you are willing to spend a minimum amount of time cooking, things like rice, lentils, beans that you soaked overnight in the fridge, potatoes, budget cuts of meat, frozen veggies, quick-oats are all easily affordable and don't come laden with salt unless you add it. None of it requires special expertise to cook (most of that consists of "put in pot of boiling water for 10-20 minutes"). It's not going to be high cuisine, but it will be nutritious and filling.
Once you learn how to boil water and cook things in the boiling water, then you can graduate to "make a stew on Sunday, serve it as leftovers on top of rice / potatoes the rest of the week". You know, like your grandparents did back during the 1920s and 1930s.
Even if it's not free the price is not particularly high; basically $50/dose for immunization against Diptheria, Tetanus, Pertussis, and Polio; add another $60/dose for MMR.
That might not seem high to you, but for many lower-income families, that's still a lot of money.
$100 for MS Windows license
$150 for a good CPU
$50 for RAM
$80 for a good motherboard
$60 for a good PSU
$80 for a good case
$100 for a SSD
$80 for a magnetic drive (to backup that SSD to)
$150 for a gfx card
Adjust to taste. You could spend up to $250 for a powerful video card, or as little as $100 for a more bargain-side card. CPUs can also be scaled down to an $80 model or up to a $250 model without getting too much out of the sweet spot. The sweet spot for both the CPU and video card is around $150 +/- $25. You get very good bang for the buck at that price point.
If money is truly an issue, go with a 10k RPM SATA drive instead of the SSD. Performance will be good enough for a budget machine.
So you could probably whittle the above down to $650 and still be able to play the latest games at modest frame rates (24-36fps).
That 50k iops were measured in what? 4k operations? 16k? What? I could claim I can pull 1 million amps out of my house socket, which would be true... as long as the voltage is 0.0018v. IOPs are just as meaningless.
In general, real-world usage, a good rule-of-thumb is 100:1 speed up for random seeks when comparing SSDs to 7200 RPM drives. Maybe only 50:1 for 15k SAS drives.
Since enterprise SSDs are only about 2x-3x the cost of the equivalent sized 15k SAS drive, you have to ask whether that 50x-100x improvement in seek speed is worth the 2x-3x drive cost.
(Rough cost of enterprise SSD is $1.50-$2.50 per GB right now.)
For a lot of use cases, where your drive spindles are 100% busy frequently, SSDs are a good solution. They're cost-efficient if you were having to short-stroke a bunch of 15k SAS drives in order to get enough performance. If you were short-stroking your SAS drives and only using 1/3 of the drive space, why not use 3x fewer SSD drives of the same size and save space / money / power in the rack?
Who is going to match Apple for top-of-the-line laptops, which a professional can use for 5-6 years before replacement?
Probably the Lenovo Thinkpad T-series is still up to snuff. Build quality on a recent T440 purchase is pretty good.
Personally, I'm still using a T61p from mid-2007. Purchased it with a 4-year warranty and made sure to use that warranty during the 4th year to get worn out bits replaced.
It doesn't need to be perfect. If cracking it still takes some time, it lowers their resources. And it can still be unbreakable for attackers with fewer resources at their disposal.
Encryption is the easy part. Managing those encryption keys is the really really hard part. And if you screw up managing those encryption keys, the attackers don't need to spend those resources to crack your encryption.
Plus, encryption is no silver-bullet. There's still traffic analysis that can give the game away.
Simple. The hover bike chase scenes in RotJ mattered, because lives were at stake, specifically those of the main characters that we cared about, not some nebulous planetary population. They also made sense within the movie plot.
The pod racer setup was so horribly contrived... just an excuse to show a hutt and show off some special effects as a way to make some quick cash and setup Anakin as the golden child.
Itanium 64-bit failed for one reason. It could not execute 32-bit workloads at the same speed as an equivalently priced 32-bit Xeon chip.
With an AMD Opteron chip, you got the best of both worlds. A chip that could do 64bit, and it could run your existing 32bit software as fast as your old CPU.
That made moving to Opteron a no-brainer decision. You got better performance from having a newer chip, even if you weren't ready to jump to a 64bit OpSys. PLUS, when you finally did move to 64bit operating systems, your CPU chip was ready and waiting.
The other reason that AMD ate Intel's lunch for a while was that they were the first ones to drive the cost of dual-core chips below $150. Intel was still charging $300+ for a dual-core CPU while you could pickup AMD Athon x2s for under $150. And dual-core makes a huge difference in how responsive the machine feels compared to a single-core CPU.
A) The built-in SQL engine in Base is crap and uses weird syntax for SQL statements. Stupidity like having to double-quote all table names and column names. When you are doing mock-up work, you want to be using local tables, not have to constantly create and destroy tables on some database server.
B) Base is very poor at import/export. In MSAccess, if you have a CSV file, you can import that and MSAccess will offer to create a brand new table for you, with proper field names and you can muck with the column types during the import process. In Base, you have to create the database table, by hand, ahead of time. This means for a simple CSV import of some random data, it takes you much longer to do in the Base world.
Base also suffers horribly when you want to pull data from source A, B and C, and output it to source D. That's something that MSAccess handles easily with linked tables (either linking to other MSAccess databases, or ODBC data sources, or whatever).
C) There's no visual design tool for update / delete / append queries. You have to write the SQL by hand. In MSAccess, you can put together an append query that pulls in 100+ fields very easily. In Base, you have to write out that query yourself, by hand.
It can be handy if I'm dragging data from multiple sources like an Excel spreadsheet, CSV file and MySQL database, via ODBC connections and be able to build queries on all these sources (even if it can be as slow as a dog).
That's something that ooBase (or LibreOffice Base) has yet to get right - a good database tool lets you pull from *anywhere* and put to *anywhere* with minimal effort. MSAccess has very good import/export capability, which makes it easy to pull in a CSV, massage it, and then output something else.
The other issue with ooBase/LibreBase is that you cannot visually design insert / update / delete queries using their QBE interface. Instead you have to write out all of the SQL. Add to that the stupid idea to use a non-standard SQL engine that requires weird syntax not supported by the mainstream databases (like pgsql). In ooBase you have to put double quotes around every table and field name.
As much as I want to use ooBase/Libre at the office, MSAccess still beats it hands-down for data manipulation.
If you watch "Return to Midway", during one of the first dives, they have an implosion of some part on their ROV. The quote said it was similar in force to a stick of TNT going off at close range.
A larger catastrophic implosion would naturally result in an even larger amount of damage.
Plus, the ROV that was lost was only trailing a slim fiber optic cable for control signals, not for power, and definitely not strong enough to haul up a ton or two of metal off the ocean floor.
Depends how often the user downloads and installs something like a new program. There are still plenty of sites out there with shady "add-ons" bundled into the program installer. They'll take a legitimate program, which has no adware/malware attached, re-bundle it, and then SEO their way to the top of the search results.
We also block a few hundred executable scripts attached to spam at the mail gateway each week. So that vector is alive and well.
For everything else web-related (infected ads being most common, followed by hijacked servers) there is NoScript / FlashBlock. They are probably the most prevalent, because there are so many opportunities (if you browse a few hundred web pages per day, that's a lot of chances).
Until they fail in less than a month because you exceeded theeeeeeeeeeeeeir write endurance. Thereee's a ressaon no one uses SSDs in seserrvvvvvvvveeeeeeers.
That's never been true. Enterprise SSDs are hitting the point where they are only 2x-3x more expensive then 15k RPM 2.5" SAS drives. For workloads where you don't need terabytes of storage space, but you *do* need the IOPS, a small array of 8 SSDs outperforms the 24 or 48 disk 15k SAS solutions.
An Intel DC S3700 400GB drive advertises that it will last 10x400GB of writes per day for 5 years. That's pretty good write endurance, even for a database workload.
Over the next 2 years, you are either going to see 15k SAS RPM drives drop drastically in $/GB or vanish from the market as SSDs encroach. The price difference of 2x-3x is low enough and the performance gain is high enough that you can do the same things with fewer drives.
That and the abandonment of density increases on spinning media.
They haven't abandoned trying to make spinning hard drives have higher capacity, they just hit the technical limits on the current technology cycle.
Seagate already announced a 6TB 3.5" for next year. With possibly room for a 10x improvement in density over 5-10 years. So a 3.5" drive with 20-30TB of storage is now likely within reach.
Depends on how you are doing the backups. Say you only have 1TB of data, you could back it up once per day appending and only need to swap out tapes every 9 months or so.
Unless they make the tape material out of unobtainium, odds are high that you would break the tape well before the 9 months is up. LTO-5 end-to-end read/write durability is only about 200 passes. Half that number if you do a full tape verify after write (which counts as an additional R/W pass).
If you use a different tape for each day of the week, with a 2-week cycle (10 total tapes), you need to consider replacing tapes after 2-3 years (52-78 uses).
Initially I thought that might not be true because tape will still be linear and you would still need to move it past the read heads.
Let's assume a 100m long piece of tape that holds 1TB. That means each meter of tape is about 10GB/m of storage density. If tape flows past the head at a rate of 1m per minute, you are looking at read/write rates of 167 MB/s.
Now, let's change the bit density of the tape to be 50 GB/meter. The tape still flows past the read/write head at 1m/min. Which gives us read/write rates of 5x greater (~833MB/s).
The only variable here is whether the read/write head can react fast enough to the data rate (which is a function of how high a frequency the hardware can operate at).
Most small businesses, unless they deal in hundreds of thousands of high-resolution images or hundreds/thousands of audio/video files are not storing 100TB of data, or backing up that much.
Some will have backup needs as small as 50-100GB and most can probably fit all their data onto a single 1TB drive / tape.
Any mechanism, really. I wouldn't count on a tape drive to be working in 5 years, but some tape drive will be available to read your tapes.
The problem with that is that at 10pm on a Friday night, after a server failure, when your only tape drive won't read the tape back properly... who do you call? Any tape drive that you order online isn't going to arrive until Tuesday at the earliest and very few local stores carry a $3500 tape drive.
That's why I dislike tape for smaller businesses if you don't have special retention requirements or multi-terabyte backups each week. If you're going to go the tape route, you really need to have a recovery plan where you can get your hands on a replacement tape drive in under an hour. Which means you should be buying at least two or three tape drives when you setup the system so that you have a spare.
And those extra $3500 tape drives sitting on the shelf tend to wipe out any cost-effectiveness that tape has over hard drives.
You can buy an awful lot of 2.5" 1TB USB3 drives for the cost of a single tape drive, let alone 2-3 tape drives.
Is that "long lasting" or is it sad that 9-10 years out of a laptop is considered long?
I think it's more a function of "your machine is no longer out-of-date 18 months after it shipped". Used to be that laptops had to be replaced every 2-3 years, because the newer models were sporting larger screens, faster CPUs, more RAM, bigger HDs. A 3 year old laptop was *really* slow compared to a brand new model, or it had a sub-par screen, or was just extremely limited in other ways.
Ever since multi-core CPUs hit the market, combined with the plateau in per-core performance, a 3-5yr old machine is still a very viable machine. Newer systems are maybe only 20-30% faster per core. Gone are the days where CPU performance doubled every 12 months (now only true if you keep adding cores and your workload is easy to run in parallel).
For instance, I'm still using a Thinkpad T61p from mid-2007. it has 8GB RAM, Win7, pair of SSDs and still performs well enough that I'm not ready to replace it yet. It lives in a docking station, so the screen/keyboard don't see much wear and tear. I made sure to take it in for service (new keyboard / mouse / system board) before the 4yr warranty ran out.
The only downside is that it doesn't have enough CPU power (2.2GHz Core2 Duo) to meet all of my needs, but I have an octo-core AMD chip on the desktop for those needs. But the Thinkpad is still the machine I use 95% of the time for work.
Access is still causing businesses to slowly go bankrupt.
For small ad-hoc databases or situations where you need to pull data together from 3 different sources, massage it a few times, then output it to a 4th format... MSAccess is a wonderful tool.
LibreOffice still can't easily import from a CSV. In MSAccess, we can pull in a CSV file quickly, and MSAccess will either create a table for you or let you append to an existing table. There's even the "make table" query type which will take your SELECT statement and create a new table with the proper field names and data types.
Linking to multiple data sources in LibreOffice is difficult, at best.
There is no way to have LibreOffice help you write an update / insert / append query. Instead you have to write it in the bastardized version of SQL called HSQL, where every field and table name has to be enclosed in double-quotes.
As a front-end to a database server like PostgreSQL, LibreOffice Base is passable, but not great. While being less of a toy then it was 3-4 years ago, there's still things that it does not do well, that are needed on a daily basis if it is going to supplant MSAccess.
The $/GB ratio between 15k SAS and enterprise SSD is now under 2.0. Which is really squeezing the profit margins on 15k RPM SAS. I think that is where old-style HDs are going to die first in the market, once enterprise SSD can reach near price-parity with the 15k RPM SAS drives.
Slashdot Mirror
Java in the web client is dead (so is Silverlight / Flash)... go Javascript / HTML5 if you have to do things on the client. Java on the server side... isn't going away for many decades.
The only downside of Java is that it's rather heavy for "one-of-a-kind" web pages. There's a lot of setup that you have to learn (Maven archtypes help) before you get HTML on the web browser. But as soon as you need something that can scale, talk to disparate systems, support unit testing, etc., it's far better then PHP. PHP just falls apart once you get past a handful of PHP pages.
If by SSD support you mean TRIM, that came in around the era of Vista or Win7. WinXP definitely didn't have it and SSDs definitely got slower after a year or two of use (until you redid the system).
The HTC One (m8), released this year, also has a battery stretch feature.
Overall, very happy with the HTC One (m8) other then I wish it was about 1/2" to 3/4" smaller. HTC did a good job with the UI and it's very snappy, makes my 18 month old Asus TF700T Transformer tablet feel slow (both are quad-core units).
You really don't know what you're missing. For business laptops, we've made the switch to 100% SSDs for 2-3 years now (ever since they dropped down to $1.50-$1.75 per GB). Granted, these are all uses who can function with only a 128GB SSD. Which holds true for probably 90% of office workers who have access to a file server (instead of storing business critical data on their HD).
Now, instead of waiting on their HD to seek around and find information (a boot process measured in minutes, program loading times measured in 10s of seconds), boot-up takes under 20-30s and program loading times are near instant. What you *will* notice is that your CPU is now the bottleneck (oops). For development work or any thing where you need to do two or three things at once, or run something disk-intensive like a scan or search of files, SSDs are a must-have. I will regularly kick off compiles / version control updates / searches, and still be able to use the machine for other things while it thinks.
Just makes sure you have a good backup system in place. On the Windows-side, I recommend Acronis True Image writing to a 2nd old-style HD inside the case. Or an external 1TB USB3 drive that you leave connected during the backup window. That is not because SSDs are unreliable (unless you buy crap like OCZ), it's because their failure modes are such (if the controller goes crazy) that data recovery is highly unlikely.
No, becasue the only food they can afford is salt laden fatty food.
If you are willing to spend a minimum amount of time cooking, things like rice, lentils, beans that you soaked overnight in the fridge, potatoes, budget cuts of meat, frozen veggies, quick-oats are all easily affordable and don't come laden with salt unless you add it. None of it requires special expertise to cook (most of that consists of "put in pot of boiling water for 10-20 minutes"). It's not going to be high cuisine, but it will be nutritious and filling.
Once you learn how to boil water and cook things in the boiling water, then you can graduate to "make a stew on Sunday, serve it as leftovers on top of rice / potatoes the rest of the week". You know, like your grandparents did back during the 1920s and 1930s.
Even if it's not free the price is not particularly high; basically $50/dose for immunization against Diptheria, Tetanus, Pertussis, and Polio; add another $60/dose for MMR.
That might not seem high to you, but for many lower-income families, that's still a lot of money.
A good gaming PC is about $850.
$100 for MS Windows license
$150 for a good CPU
$50 for RAM
$80 for a good motherboard
$60 for a good PSU
$80 for a good case
$100 for a SSD
$80 for a magnetic drive (to backup that SSD to)
$150 for a gfx card
Adjust to taste. You could spend up to $250 for a powerful video card, or as little as $100 for a more bargain-side card. CPUs can also be scaled down to an $80 model or up to a $250 model without getting too much out of the sweet spot. The sweet spot for both the CPU and video card is around $150 +/- $25. You get very good bang for the buck at that price point.
If money is truly an issue, go with a 10k RPM SATA drive instead of the SSD. Performance will be good enough for a budget machine.
So you could probably whittle the above down to $650 and still be able to play the latest games at modest frame rates (24-36fps).
That 50k iops were measured in what? 4k operations? 16k? What? I could claim I can pull 1 million amps out of my house socket, which would be true... as long as the voltage is 0.0018v. IOPs are just as meaningless.
In general, real-world usage, a good rule-of-thumb is 100:1 speed up for random seeks when comparing SSDs to 7200 RPM drives. Maybe only 50:1 for 15k SAS drives.
Since enterprise SSDs are only about 2x-3x the cost of the equivalent sized 15k SAS drive, you have to ask whether that 50x-100x improvement in seek speed is worth the 2x-3x drive cost.
(Rough cost of enterprise SSD is $1.50-$2.50 per GB right now.)
For a lot of use cases, where your drive spindles are 100% busy frequently, SSDs are a good solution. They're cost-efficient if you were having to short-stroke a bunch of 15k SAS drives in order to get enough performance. If you were short-stroking your SAS drives and only using 1/3 of the drive space, why not use 3x fewer SSD drives of the same size and save space / money / power in the rack?
Who is going to match Apple for top-of-the-line laptops, which a professional can use for 5-6 years before replacement?
Probably the Lenovo Thinkpad T-series is still up to snuff. Build quality on a recent T440 purchase is pretty good.
Personally, I'm still using a T61p from mid-2007. Purchased it with a 4-year warranty and made sure to use that warranty during the 4th year to get worn out bits replaced.
It doesn't need to be perfect. If cracking it still takes some time, it lowers their resources. And it can still be unbreakable for attackers with fewer resources at their disposal.
Encryption is the easy part. Managing those encryption keys is the really really hard part. And if you screw up managing those encryption keys, the attackers don't need to spend those resources to crack your encryption.
Plus, encryption is no silver-bullet. There's still traffic analysis that can give the game away.
Simple. The hover bike chase scenes in RotJ mattered, because lives were at stake, specifically those of the main characters that we cared about, not some nebulous planetary population. They also made sense within the movie plot.
The pod racer setup was so horribly contrived... just an excuse to show a hutt and show off some special effects as a way to make some quick cash and setup Anakin as the golden child.
Itanium 64-bit failed for one reason. It could not execute 32-bit workloads at the same speed as an equivalently priced 32-bit Xeon chip.
With an AMD Opteron chip, you got the best of both worlds. A chip that could do 64bit, and it could run your existing 32bit software as fast as your old CPU.
That made moving to Opteron a no-brainer decision. You got better performance from having a newer chip, even if you weren't ready to jump to a 64bit OpSys. PLUS, when you finally did move to 64bit operating systems, your CPU chip was ready and waiting.
The other reason that AMD ate Intel's lunch for a while was that they were the first ones to drive the cost of dual-core chips below $150. Intel was still charging $300+ for a dual-core CPU while you could pickup AMD Athon x2s for under $150. And dual-core makes a huge difference in how responsive the machine feels compared to a single-core CPU.
OpenOffice or LibreOffice's Base.
A) The built-in SQL engine in Base is crap and uses weird syntax for SQL statements. Stupidity like having to double-quote all table names and column names. When you are doing mock-up work, you want to be using local tables, not have to constantly create and destroy tables on some database server.
B) Base is very poor at import/export. In MSAccess, if you have a CSV file, you can import that and MSAccess will offer to create a brand new table for you, with proper field names and you can muck with the column types during the import process. In Base, you have to create the database table, by hand, ahead of time. This means for a simple CSV import of some random data, it takes you much longer to do in the Base world.
Base also suffers horribly when you want to pull data from source A, B and C, and output it to source D. That's something that MSAccess handles easily with linked tables (either linking to other MSAccess databases, or ODBC data sources, or whatever).
C) There's no visual design tool for update / delete / append queries. You have to write the SQL by hand. In MSAccess, you can put together an append query that pulls in 100+ fields very easily. In Base, you have to write out that query yourself, by hand.
It can be handy if I'm dragging data from multiple sources like an Excel spreadsheet, CSV file and MySQL database, via ODBC connections and be able to build queries on all these sources (even if it can be as slow as a dog).
That's something that ooBase (or LibreOffice Base) has yet to get right - a good database tool lets you pull from *anywhere* and put to *anywhere* with minimal effort. MSAccess has very good import/export capability, which makes it easy to pull in a CSV, massage it, and then output something else.
The other issue with ooBase/LibreBase is that you cannot visually design insert / update / delete queries using their QBE interface. Instead you have to write out all of the SQL. Add to that the stupid idea to use a non-standard SQL engine that requires weird syntax not supported by the mainstream databases (like pgsql). In ooBase you have to put double quotes around every table and field name.
As much as I want to use ooBase/Libre at the office, MSAccess still beats it hands-down for data manipulation.
If you watch "Return to Midway", during one of the first dives, they have an implosion of some part on their ROV. The quote said it was similar in force to a stick of TNT going off at close range.
A larger catastrophic implosion would naturally result in an even larger amount of damage.
Plus, the ROV that was lost was only trailing a slim fiber optic cable for control signals, not for power, and definitely not strong enough to haul up a ton or two of metal off the ocean floor.
Depends how often the user downloads and installs something like a new program. There are still plenty of sites out there with shady "add-ons" bundled into the program installer. They'll take a legitimate program, which has no adware/malware attached, re-bundle it, and then SEO their way to the top of the search results.
We also block a few hundred executable scripts attached to spam at the mail gateway each week. So that vector is alive and well.
For everything else web-related (infected ads being most common, followed by hijacked servers) there is NoScript / FlashBlock. They are probably the most prevalent, because there are so many opportunities (if you browse a few hundred web pages per day, that's a lot of chances).
Until they fail in less than a month because you exceeded theeeeeeeeeeeeeir write endurance. Thereee's a ressaon no one uses SSDs in seserrvvvvvvvveeeeeeers.
That's never been true. Enterprise SSDs are hitting the point where they are only 2x-3x more expensive then 15k RPM 2.5" SAS drives. For workloads where you don't need terabytes of storage space, but you *do* need the IOPS, a small array of 8 SSDs outperforms the 24 or 48 disk 15k SAS solutions.
An Intel DC S3700 400GB drive advertises that it will last 10x400GB of writes per day for 5 years. That's pretty good write endurance, even for a database workload.
Over the next 2 years, you are either going to see 15k SAS RPM drives drop drastically in $/GB or vanish from the market as SSDs encroach. The price difference of 2x-3x is low enough and the performance gain is high enough that you can do the same things with fewer drives.
That and the abandonment of density increases on spinning media.
They haven't abandoned trying to make spinning hard drives have higher capacity, they just hit the technical limits on the current technology cycle.
Seagate already announced a 6TB 3.5" for next year. With possibly room for a 10x improvement in density over 5-10 years. So a 3.5" drive with 20-30TB of storage is now likely within reach.
Depends on how you are doing the backups. Say you only have 1TB of data, you could back it up once per day appending and only need to swap out tapes every 9 months or so.
Unless they make the tape material out of unobtainium, odds are high that you would break the tape well before the 9 months is up. LTO-5 end-to-end read/write durability is only about 200 passes. Half that number if you do a full tape verify after write (which counts as an additional R/W pass).
If you use a different tape for each day of the week, with a 2-week cycle (10 total tapes), you need to consider replacing tapes after 2-3 years (52-78 uses).
Initially I thought that might not be true because tape will still be linear and you would still need to move it past the read heads.
Let's assume a 100m long piece of tape that holds 1TB. That means each meter of tape is about 10GB/m of storage density. If tape flows past the head at a rate of 1m per minute, you are looking at read/write rates of 167 MB/s.
Now, let's change the bit density of the tape to be 50 GB/meter. The tape still flows past the read/write head at 1m/min. Which gives us read/write rates of 5x greater (~833MB/s).
The only variable here is whether the read/write head can react fast enough to the data rate (which is a function of how high a frequency the hardware can operate at).
Most small businesses, unless they deal in hundreds of thousands of high-resolution images or hundreds/thousands of audio/video files are not storing 100TB of data, or backing up that much.
Some will have backup needs as small as 50-100GB and most can probably fit all their data onto a single 1TB drive / tape.
Any mechanism, really. I wouldn't count on a tape drive to be working in 5 years, but some tape drive will be available to read your tapes.
The problem with that is that at 10pm on a Friday night, after a server failure, when your only tape drive won't read the tape back properly... who do you call? Any tape drive that you order online isn't going to arrive until Tuesday at the earliest and very few local stores carry a $3500 tape drive.
That's why I dislike tape for smaller businesses if you don't have special retention requirements or multi-terabyte backups each week. If you're going to go the tape route, you really need to have a recovery plan where you can get your hands on a replacement tape drive in under an hour. Which means you should be buying at least two or three tape drives when you setup the system so that you have a spare.
And those extra $3500 tape drives sitting on the shelf tend to wipe out any cost-effectiveness that tape has over hard drives.
You can buy an awful lot of 2.5" 1TB USB3 drives for the cost of a single tape drive, let alone 2-3 tape drives.
Is that "long lasting" or is it sad that 9-10 years out of a laptop is considered long?
I think it's more a function of "your machine is no longer out-of-date 18 months after it shipped". Used to be that laptops had to be replaced every 2-3 years, because the newer models were sporting larger screens, faster CPUs, more RAM, bigger HDs. A 3 year old laptop was *really* slow compared to a brand new model, or it had a sub-par screen, or was just extremely limited in other ways.
Ever since multi-core CPUs hit the market, combined with the plateau in per-core performance, a 3-5yr old machine is still a very viable machine. Newer systems are maybe only 20-30% faster per core. Gone are the days where CPU performance doubled every 12 months (now only true if you keep adding cores and your workload is easy to run in parallel).
For instance, I'm still using a Thinkpad T61p from mid-2007. it has 8GB RAM, Win7, pair of SSDs and still performs well enough that I'm not ready to replace it yet. It lives in a docking station, so the screen/keyboard don't see much wear and tear. I made sure to take it in for service (new keyboard / mouse / system board) before the 4yr warranty ran out.
The only downside is that it doesn't have enough CPU power (2.2GHz Core2 Duo) to meet all of my needs, but I have an octo-core AMD chip on the desktop for those needs. But the Thinkpad is still the machine I use 95% of the time for work.
Access is still causing businesses to slowly go bankrupt.
For small ad-hoc databases or situations where you need to pull data together from 3 different sources, massage it a few times, then output it to a 4th format... MSAccess is a wonderful tool.
LibreOffice still can't easily import from a CSV. In MSAccess, we can pull in a CSV file quickly, and MSAccess will either create a table for you or let you append to an existing table. There's even the "make table" query type which will take your SELECT statement and create a new table with the proper field names and data types.
Linking to multiple data sources in LibreOffice is difficult, at best.
There is no way to have LibreOffice help you write an update / insert / append query. Instead you have to write it in the bastardized version of SQL called HSQL, where every field and table name has to be enclosed in double-quotes.
As a front-end to a database server like PostgreSQL, LibreOffice Base is passable, but not great. While being less of a toy then it was 3-4 years ago, there's still things that it does not do well, that are needed on a daily basis if it is going to supplant MSAccess.
The $/GB ratio between 15k SAS and enterprise SSD is now under 2.0. Which is really squeezing the profit margins on 15k RPM SAS. I think that is where old-style HDs are going to die first in the market, once enterprise SSD can reach near price-parity with the 15k RPM SAS drives.