Microsoft have always been cowards when it comes to taking even the tiniest risks, and moving things forward, and dropping legacy. Its why the newest surface laptops still come with USB-A ports. They're just so fucking terrified of losing customers if they drop them, and thus they're slowing down the full adoption of USB-C and extending the transition indefinitely. Thats how little confidence they have in the merits of their products. They lean on ancient tech as some kind of competitive advantage.
It's not cowardice, it's commitment to backwards compatibility. There are files and software created decades ago that still work on modern Windows machines. That's admirable in a sense, though it does limit their ability to deliver something new that rocks the boat. But it means that business and customers are assured that things won't break if they stick with Microsoft, which is a large part of why they're so successful, they've willingly tied themselves to legacy.
Having different priorities to Apple is not a weakness, it's just different.
Also, I would like a USB-A port on my Mac. It'd be very useful.
It's also a commitment to security breaches. That legacy software isn't always being updated with security patches and even if it is, not at the same frequency as the more modern versions of said software.
As an IT professional, I will not recommend that as soon as a service pack or update comes along we immediately apply it but within a period of 6-12 months we have to start looking at staying with the times.
They announce that they’re going to break with the past, release something that points to a legacy-free future.
Then a chap from marketing reminds them that if they do that, then they’re going to have to build market share from scratch.
And then the whole idea gets binned … for another five years.
Microsoft is trapped in a legacy hellhole they cannot escape. As time marches on Windows gets ever more bloated because of this captivity. Apple has the luxury of just telling the legacy luddites to fuck off. Every Apple blog and every Apple discussion forum is loaded with outraged users livid because their legacy hardware and software no longer perform. The dropping of 32bit compatibility is just one example. Microsoft could never do that. Apple simply tells the legacy crowd too bad, so sad. The kicker is all those outraged legacy types stay with the platform anyway.
"Legacy types", like companies and people actually making a living out of their computers? Like it or not, but there are old, but working and essential, SW and HW out there, keeping industries and people alive. I hate the inconsistencies of the Win UX, but our company never has to hesitate when there are updates. Strangely enough, it just works (which is not something one can say about macOS any longer).
They announce that they’re going to break with the past, release something that points to a legacy-free future.
Then a chap from marketing reminds them that if they do that, then they’re going to have to build market share from scratch.
And then the whole idea gets binned … for another five years.
Microsoft is trapped in a legacy hellhole they cannot escape. As time marches on Windows gets ever more bloated because of this captivity. Apple has the luxury of just telling the legacy luddites to fuck off. Every Apple blog and every Apple discussion forum is loaded with outraged users livid because their legacy hardware and software no longer perform. The dropping of 32bit compatibility is just one example. Microsoft could never do that. Apple simply tells the legacy crowd too bad, so sad. The kicker is all those outraged legacy types stay with the platform anyway.
That may be one of the reasons why MacOS is an "also ran".
Support for older hardware is one of the strengths Windows has. We saw that with WIndows 7 as well as with IE: Microsoft wanted to move on but its users, particularly corporate users had too much invested in those so called "legacy systems". Currently the 14 year old Thinkpad I use for financial work is running Windows 8.1. But, when I get a break from yard work and tutoring my grandson, I plan to stick in an SSD and upgrade it Windows 10. Why not? It runs fine, the upgrade will cost almost nothing and will stop me from having to sink money into a new machine for a few more years.
Another example is COBOL, the business language the proliferated in 30-40 years ago. Today many businesses still run on business critical systems developed with it.
The truth is: while hardware can continue to move forward, it is the software that businesses rely on and where the investment lies. They aren't going to walk away from that investment quickly or easily. And, in some cases, like COBOL, they can't. The resources to replace it aren't available.
See my reply to "crowley" about legacy software and security holes.
While many businesses may still run COBOL programs, I'm pretty certain that IBM has upgraded their mainframe hardware and OS, is it still called 360? I'm also pretty certain that the COBOL compiler of old (70's, 80's) isn't the same as the COBOL compiler of today which runs on much more modern architecture.
Think of SQL Server 7 meant to run on Windows NT systems and SQL Server 2019 meant to run on todays's modern systems. At a most basic level they're just DBMS's but I can assure you, they're quite different.
As an IT person I can tell you it is irresponsible of IT departments hanging on to older versions of their software. But often we do get trumped by accountants and budgets and we find ourselves scrambling to make massive upgrades from SQL Server 2008 --> 2019 or Oracle 10g --> 19c. And the more bean counters wait, the more it costs.
They announce that they’re going to break with the past, release something that points to a legacy-free future.
Then a chap from marketing reminds them that if they do that, then they’re going to have to build market share from scratch.
And then the whole idea gets binned … for another five years.
Microsoft is trapped in a legacy hellhole they cannot escape. As time marches on Windows gets ever more bloated because of this captivity. Apple has the luxury of just telling the legacy luddites to fuck off. Every Apple blog and every Apple discussion forum is loaded with outraged users livid because their legacy hardware and software no longer perform. The dropping of 32bit compatibility is just one example. Microsoft could never do that. Apple simply tells the legacy crowd too bad, so sad. The kicker is all those outraged legacy types stay with the platform anyway.
That may be one of the reasons why MacOS is an "also ran".
Support for older hardware is one of the strengths Windows has. We saw that with WIndows 7 as well as with IE: Microsoft wanted to move on but its users, particularly corporate users had too much invested in those so called "legacy systems". Currently the 14 year old Thinkpad I use for financial work is running Windows 8.1. But, when I get a break from yard work and tutoring my grandson, I plan to stick in an SSD and upgrade it Windows 10. Why not? It runs fine, the upgrade will cost almost nothing and will stop me from having to sink money into a new machine for a few more years.
Another example is COBOL, the business language the proliferated in 30-40 years ago. Today many businesses still run on business critical systems developed with it.
The truth is: while hardware can continue to move forward, it is the software that businesses rely on and where the investment lies. They aren't going to walk away from that investment quickly or easily. And, in some cases, like COBOL, they can't. The resources to replace it aren't available.
See my reply to "crowley" about legacy software and security holes.
While many businesses may still run COBOL programs, I'm pretty certain that IBM has upgraded their mainframe hardware and OS, is it still called 360? I'm also pretty certain that the COBOL compiler of old (70's, 80's) isn't the same as the COBOL compiler of today which runs on much more modern architecture.
Think of SQL Server 7 meant to run on Windows NT systems and SQL Server 2019 meant to run on todays's modern systems. At a most basic level they're just DBMS's but I can assure you, they're quite different.
As an IT person I can tell you it is irresponsible of IT departments hanging on to older versions of their software. But often we do get trumped by accountants and budgets and we find ourselves scrambling to make massive upgrades from SQL Server 2008 --> 2019 or Oracle 10g --> 19c. And the more bean counters wait, the more it costs.
Yeh, I get the "newer is better" stuff.
But, those old systems were built like battleships to be solid, reliable and to dependably get the the job done. And many of them still do. It's not about "accountants". It's about executives who want systems that "just work" -- year, after year, after year.... And, in a mission critical system, that trumps "new" -- even if it is "better".
And, your example of going from a crappy DBS to another crappy DBS shows you likely have likely never seen a quality application system.
They announce that they’re going to break with the past, release something that points to a legacy-free future.
Then a chap from marketing reminds them that if they do that, then they’re going to have to build market share from scratch.
And then the whole idea gets binned … for another five years.
Microsoft is trapped in a legacy hellhole they cannot escape. As time marches on Windows gets ever more bloated because of this captivity. Apple has the luxury of just telling the legacy luddites to fuck off. Every Apple blog and every Apple discussion forum is loaded with outraged users livid because their legacy hardware and software no longer perform. The dropping of 32bit compatibility is just one example. Microsoft could never do that. Apple simply tells the legacy crowd too bad, so sad. The kicker is all those outraged legacy types stay with the platform anyway.
That may be one of the reasons why MacOS is an "also ran".
Support for older hardware is one of the strengths Windows has. We saw that with WIndows 7 as well as with IE: Microsoft wanted to move on but its users, particularly corporate users had too much invested in those so called "legacy systems". Currently the 14 year old Thinkpad I use for financial work is running Windows 8.1. But, when I get a break from yard work and tutoring my grandson, I plan to stick in an SSD and upgrade it Windows 10. Why not? It runs fine, the upgrade will cost almost nothing and will stop me from having to sink money into a new machine for a few more years.
Another example is COBOL, the business language the proliferated in 30-40 years ago. Today many businesses still run on business critical systems developed with it.
The truth is: while hardware can continue to move forward, it is the software that businesses rely on and where the investment lies. They aren't going to walk away from that investment quickly or easily. And, in some cases, like COBOL, they can't. The resources to replace it aren't available.
Cobol from 30 or 40 years ago? That only takes us back to 1990 or 1980. Cobol has been around since the 1960's, defined in 1959 as part of the CODASYL process.
Cobol came into being at about the same time as Fortran, Lisp, and Algol.
The truth about Cobol is not the language, but the legacy managers. The Cobol language has incorporated many newer features in addition to having DB2 relational database the backend. But the old Cobol programs were written against card readers and mag tapes, and they still today treat DB2 as a dumb sequential record system. And managers won't allow, much less require, the use of the newer more robust Cobol language features.
It's legacy mindsets, not legacy programs, hardware, etc. Businesses don't have to walk away from Cobol -- they can take the relatively trivial step to refactor their Cobol programs to use the new Cobol features, and SQL against the DB2 database. The result would be more reliability and half the existing Cobol programs could be each be replaced by a single SQL query.
They announce that they’re going to break with the past, release something that points to a legacy-free future.
Then a chap from marketing reminds them that if they do that, then they’re going to have to build market share from scratch.
And then the whole idea gets binned … for another five years.
Microsoft is trapped in a legacy hellhole they cannot escape. As time marches on Windows gets ever more bloated because of this captivity. Apple has the luxury of just telling the legacy luddites to fuck off. Every Apple blog and every Apple discussion forum is loaded with outraged users livid because their legacy hardware and software no longer perform. The dropping of 32bit compatibility is just one example. Microsoft could never do that. Apple simply tells the legacy crowd too bad, so sad. The kicker is all those outraged legacy types stay with the platform anyway.
That may be one of the reasons why MacOS is an "also ran".
Support for older hardware is one of the strengths Windows has. We saw that with WIndows 7 as well as with IE: Microsoft wanted to move on but its users, particularly corporate users had too much invested in those so called "legacy systems". Currently the 14 year old Thinkpad I use for financial work is running Windows 8.1. But, when I get a break from yard work and tutoring my grandson, I plan to stick in an SSD and upgrade it Windows 10. Why not? It runs fine, the upgrade will cost almost nothing and will stop me from having to sink money into a new machine for a few more years.
Another example is COBOL, the business language the proliferated in 30-40 years ago. Today many businesses still run on business critical systems developed with it.
The truth is: while hardware can continue to move forward, it is the software that businesses rely on and where the investment lies. They aren't going to walk away from that investment quickly or easily. And, in some cases, like COBOL, they can't. The resources to replace it aren't available.
See my reply to "crowley" about legacy software and security holes.
While many businesses may still run COBOL programs, I'm pretty certain that IBM has upgraded their mainframe hardware and OS, is it still called 360? I'm also pretty certain that the COBOL compiler of old (70's, 80's) isn't the same as the COBOL compiler of today which runs on much more modern architecture.
Think of SQL Server 7 meant to run on Windows NT systems and SQL Server 2019 meant to run on todays's modern systems. At a most basic level they're just DBMS's but I can assure you, they're quite different.
As an IT person I can tell you it is irresponsible of IT departments hanging on to older versions of their software. But often we do get trumped by accountants and budgets and we find ourselves scrambling to make massive upgrades from SQL Server 2008 --> 2019 or Oracle 10g --> 19c. And the more bean counters wait, the more it costs.
Yeh, I get the "newer is better" stuff.
But, those old systems were built like battleships to be solid, reliable and to dependably get the the job done. And many of them still do. It's not about "accountants". It's about executives who want systems that "just work" -- year, after year, after year.... And, in a mission critical system, that trumps "new" -- even if it is "better".
And, your example of going from a crappy DBS to another crappy DBS shows you likely have likely never seen a quality application system.
What is crappy about SQL Server or Oracle DBMS’s? Or are you one of those mainframe is king DB2 guys? Also keeping up with modern versions of OS’s and software isn’t just change for the sake of change.
It’s also about security.
Would you recommend to your users to stick to with the old 32 bit versions of their software because they still work? Would you recommend that they stick with Windows NT? Lotus Notes?
They announce that they’re going to break with the past, release something that points to a legacy-free future.
Then a chap from marketing reminds them that if they do that, then they’re going to have to build market share from scratch.
And then the whole idea gets binned … for another five years.
Microsoft is trapped in a legacy hellhole they cannot escape. As time marches on Windows gets ever more bloated because of this captivity. Apple has the luxury of just telling the legacy luddites to fuck off. Every Apple blog and every Apple discussion forum is loaded with outraged users livid because their legacy hardware and software no longer perform. The dropping of 32bit compatibility is just one example. Microsoft could never do that. Apple simply tells the legacy crowd too bad, so sad. The kicker is all those outraged legacy types stay with the platform anyway.
That may be one of the reasons why MacOS is an "also ran".
Support for older hardware is one of the strengths Windows has. We saw that with WIndows 7 as well as with IE: Microsoft wanted to move on but its users, particularly corporate users had too much invested in those so called "legacy systems". Currently the 14 year old Thinkpad I use for financial work is running Windows 8.1. But, when I get a break from yard work and tutoring my grandson, I plan to stick in an SSD and upgrade it Windows 10. Why not? It runs fine, the upgrade will cost almost nothing and will stop me from having to sink money into a new machine for a few more years.
Another example is COBOL, the business language the proliferated in 30-40 years ago. Today many businesses still run on business critical systems developed with it.
The truth is: while hardware can continue to move forward, it is the software that businesses rely on and where the investment lies. They aren't going to walk away from that investment quickly or easily. And, in some cases, like COBOL, they can't. The resources to replace it aren't available.
See my reply to "crowley" about legacy software and security holes.
While many businesses may still run COBOL programs, I'm pretty certain that IBM has upgraded their mainframe hardware and OS, is it still called 360? I'm also pretty certain that the COBOL compiler of old (70's, 80's) isn't the same as the COBOL compiler of today which runs on much more modern architecture.
Think of SQL Server 7 meant to run on Windows NT systems and SQL Server 2019 meant to run on todays's modern systems. At a most basic level they're just DBMS's but I can assure you, they're quite different.
As an IT person I can tell you it is irresponsible of IT departments hanging on to older versions of their software. But often we do get trumped by accountants and budgets and we find ourselves scrambling to make massive upgrades from SQL Server 2008 --> 2019 or Oracle 10g --> 19c. And the more bean counters wait, the more it costs.
Yeh, I get the "newer is better" stuff.
But, those old systems were built like battleships to be solid, reliable and to dependably get the the job done. And many of them still do. It's not about "accountants". It's about executives who want systems that "just work" -- year, after year, after year.... And, in a mission critical system, that trumps "new" -- even if it is "better".
And, your example of going from a crappy DBS to another crappy DBS shows you likely have likely never seen a quality application system.
What is crappy about SQL Server or Oracle DBMS’s? Or are you one of those mainframe is king DB2 guys? Also keeping up with modern versions of OS’s and software isn’t just change for the sake of change.
It’s also about security.
Would you recommend to your users to stick to with the old 32 bit versions of their software because they still work? Would you recommend that they stick with Windows NT? Lotus Notes?
What would be the problem with running 32 bit software that work? There's nothing "secure" with 64 bits. A lot of mission critical systems are 32 (or even 16) bit. Ever worked in automotive or aviation?
They announce that they’re going to break with the past, release something that points to a legacy-free future.
Then a chap from marketing reminds them that if they do that, then they’re going to have to build market share from scratch.
And then the whole idea gets binned … for another five years.
Microsoft is trapped in a legacy hellhole they cannot escape. As time marches on Windows gets ever more bloated because of this captivity. Apple has the luxury of just telling the legacy luddites to fuck off. Every Apple blog and every Apple discussion forum is loaded with outraged users livid because their legacy hardware and software no longer perform. The dropping of 32bit compatibility is just one example. Microsoft could never do that. Apple simply tells the legacy crowd too bad, so sad. The kicker is all those outraged legacy types stay with the platform anyway.
That may be one of the reasons why MacOS is an "also ran".
Support for older hardware is one of the strengths Windows has. We saw that with WIndows 7 as well as with IE: Microsoft wanted to move on but its users, particularly corporate users had too much invested in those so called "legacy systems". Currently the 14 year old Thinkpad I use for financial work is running Windows 8.1. But, when I get a break from yard work and tutoring my grandson, I plan to stick in an SSD and upgrade it Windows 10. Why not? It runs fine, the upgrade will cost almost nothing and will stop me from having to sink money into a new machine for a few more years.
Another example is COBOL, the business language the proliferated in 30-40 years ago. Today many businesses still run on business critical systems developed with it.
The truth is: while hardware can continue to move forward, it is the software that businesses rely on and where the investment lies. They aren't going to walk away from that investment quickly or easily. And, in some cases, like COBOL, they can't. The resources to replace it aren't available.
See my reply to "crowley" about legacy software and security holes.
While many businesses may still run COBOL programs, I'm pretty certain that IBM has upgraded their mainframe hardware and OS, is it still called 360? I'm also pretty certain that the COBOL compiler of old (70's, 80's) isn't the same as the COBOL compiler of today which runs on much more modern architecture.
Think of SQL Server 7 meant to run on Windows NT systems and SQL Server 2019 meant to run on todays's modern systems. At a most basic level they're just DBMS's but I can assure you, they're quite different.
As an IT person I can tell you it is irresponsible of IT departments hanging on to older versions of their software. But often we do get trumped by accountants and budgets and we find ourselves scrambling to make massive upgrades from SQL Server 2008 --> 2019 or Oracle 10g --> 19c. And the more bean counters wait, the more it costs.
Yeh, I get the "newer is better" stuff.
But, those old systems were built like battleships to be solid, reliable and to dependably get the the job done. And many of them still do. It's not about "accountants". It's about executives who want systems that "just work" -- year, after year, after year.... And, in a mission critical system, that trumps "new" -- even if it is "better".
And, your example of going from a crappy DBS to another crappy DBS shows you likely have likely never seen a quality application system.
What is crappy about SQL Server or Oracle DBMS’s? Or are you one of those mainframe is king DB2 guys? Also keeping up with modern versions of OS’s and software isn’t just change for the sake of change.
It’s also about security.
Would you recommend to your users to stick to with the old 32 bit versions of their software because they still work? Would you recommend that they stick with Windows NT? Lotus Notes?
What would be the problem with running 32 bit software that work? There's nothing "secure" with 64 bits. A lot of mission critical systems are 32 (or even 16) bit. Ever worked in automotive or aviation?
No one ever said that 64 bit is more secure.
Legacy software is just that. Many older versions are just not supported any longer and even if they are, many aren’t getting the critical security patches that more modern versions are, at least not at the same frequency.
Never worked in automotive industry but that sounds like hyper specialized equipment.
The topic of discussion here isn’t electronic fuel injectors. It’s your standard Windows/ Linux/ MacOS desktop or server OS and whether said OS should support legacy software (often 32 bit).
They announce that they’re going to break with the past, release something that points to a legacy-free future.
Then a chap from marketing reminds them that if they do that, then they’re going to have to build market share from scratch.
And then the whole idea gets binned … for another five years.
Microsoft is trapped in a legacy hellhole they cannot escape. As time marches on Windows gets ever more bloated because of this captivity. Apple has the luxury of just telling the legacy luddites to fuck off. Every Apple blog and every Apple discussion forum is loaded with outraged users livid because their legacy hardware and software no longer perform. The dropping of 32bit compatibility is just one example. Microsoft could never do that. Apple simply tells the legacy crowd too bad, so sad. The kicker is all those outraged legacy types stay with the platform anyway.
That may be one of the reasons why MacOS is an "also ran".
Support for older hardware is one of the strengths Windows has. We saw that with WIndows 7 as well as with IE: Microsoft wanted to move on but its users, particularly corporate users had too much invested in those so called "legacy systems". Currently the 14 year old Thinkpad I use for financial work is running Windows 8.1. But, when I get a break from yard work and tutoring my grandson, I plan to stick in an SSD and upgrade it Windows 10. Why not? It runs fine, the upgrade will cost almost nothing and will stop me from having to sink money into a new machine for a few more years.
Another example is COBOL, the business language the proliferated in 30-40 years ago. Today many businesses still run on business critical systems developed with it.
The truth is: while hardware can continue to move forward, it is the software that businesses rely on and where the investment lies. They aren't going to walk away from that investment quickly or easily. And, in some cases, like COBOL, they can't. The resources to replace it aren't available.
Cobol from 30 or 40 years ago? That only takes us back to 1990 or 1980. Cobol has been around since the 1960's, defined in 1959 as part of the CODASYL process.
Cobol came into being at about the same time as Fortran, Lisp, and Algol.
The truth about Cobol is not the language, but the legacy managers. The Cobol language has incorporated many newer features in addition to having DB2 relational database the backend. But the old Cobol programs were written against card readers and mag tapes, and they still today treat DB2 as a dumb sequential record system. And managers won't allow, much less require, the use of the newer more robust Cobol language features.
It's legacy mindsets, not legacy programs, hardware, etc. Businesses don't have to walk away from Cobol -- they can take the relatively trivial step to refactor their Cobol programs to use the new Cobol features, and SQL against the DB2 database. The result would be more reliability and half the existing Cobol programs could be each be replaced by a single SQL query.
COBOl has always accessed and manipulated many file systems: sequential, ISAM, VSAM, IMS, etc. DB2 was just one more and it was added back in the 80's -- mostly as a replacement for IMS which was a hierarchical database as opposed to DB2's relational structure. And COBOL programs tend to be far more complex than simple data retrieval and update. If that's all its doing - regardless of the file structure its accessing - it is an extremely short and simple program . It would essentially be the equivalent of an entry level program for a student.
typically accessing files (even relational databases) would constitute maybe a quarter or less of a typical program. The rest would be logic.
They announce that they’re going to break with the past, release something that points to a legacy-free future.
Then a chap from marketing reminds them that if they do that, then they’re going to have to build market share from scratch.
And then the whole idea gets binned … for another five years.
Microsoft is trapped in a legacy hellhole they cannot escape. As time marches on Windows gets ever more bloated because of this captivity. Apple has the luxury of just telling the legacy luddites to fuck off. Every Apple blog and every Apple discussion forum is loaded with outraged users livid because their legacy hardware and software no longer perform. The dropping of 32bit compatibility is just one example. Microsoft could never do that. Apple simply tells the legacy crowd too bad, so sad. The kicker is all those outraged legacy types stay with the platform anyway.
That may be one of the reasons why MacOS is an "also ran".
Support for older hardware is one of the strengths Windows has. We saw that with WIndows 7 as well as with IE: Microsoft wanted to move on but its users, particularly corporate users had too much invested in those so called "legacy systems". Currently the 14 year old Thinkpad I use for financial work is running Windows 8.1. But, when I get a break from yard work and tutoring my grandson, I plan to stick in an SSD and upgrade it Windows 10. Why not? It runs fine, the upgrade will cost almost nothing and will stop me from having to sink money into a new machine for a few more years.
Another example is COBOL, the business language the proliferated in 30-40 years ago. Today many businesses still run on business critical systems developed with it.
The truth is: while hardware can continue to move forward, it is the software that businesses rely on and where the investment lies. They aren't going to walk away from that investment quickly or easily. And, in some cases, like COBOL, they can't. The resources to replace it aren't available.
See my reply to "crowley" about legacy software and security holes.
While many businesses may still run COBOL programs, I'm pretty certain that IBM has upgraded their mainframe hardware and OS, is it still called 360? I'm also pretty certain that the COBOL compiler of old (70's, 80's) isn't the same as the COBOL compiler of today which runs on much more modern architecture.
Think of SQL Server 7 meant to run on Windows NT systems and SQL Server 2019 meant to run on todays's modern systems. At a most basic level they're just DBMS's but I can assure you, they're quite different.
As an IT person I can tell you it is irresponsible of IT departments hanging on to older versions of their software. But often we do get trumped by accountants and budgets and we find ourselves scrambling to make massive upgrades from SQL Server 2008 --> 2019 or Oracle 10g --> 19c. And the more bean counters wait, the more it costs.
Yeh, I get the "newer is better" stuff.
But, those old systems were built like battleships to be solid, reliable and to dependably get the the job done. And many of them still do. It's not about "accountants". It's about executives who want systems that "just work" -- year, after year, after year.... And, in a mission critical system, that trumps "new" -- even if it is "better".
And, your example of going from a crappy DBS to another crappy DBS shows you likely have likely never seen a quality application system.
What is crappy about SQL Server or Oracle DBMS’s? Or are you one of those mainframe is king DB2 guys? Also keeping up with modern versions of OS’s and software isn’t just change for the sake of change.
It’s also about security.
Would you recommend to your users to stick to with the old 32 bit versions of their software because they still work? Would you recommend that they stick with Windows NT? Lotus Notes?
You missed the point: Those old COBOL programs are still around because they were well built, still doing the job and, in most cases, cannot feasibly or easily be replaced because often they are supporting mission critical systems, nobody can gamble trying to replace them.
Typically, their biggest problem is finding somebody who knows the language and is able to support and maintain them.
They announce that they’re going to break with the past, release something that points to a legacy-free future.
Then a chap from marketing reminds them that if they do that, then they’re going to have to build market share from scratch.
And then the whole idea gets binned … for another five years.
Microsoft is trapped in a legacy hellhole they cannot escape. As time marches on Windows gets ever more bloated because of this captivity. Apple has the luxury of just telling the legacy luddites to fuck off. Every Apple blog and every Apple discussion forum is loaded with outraged users livid because their legacy hardware and software no longer perform. The dropping of 32bit compatibility is just one example. Microsoft could never do that. Apple simply tells the legacy crowd too bad, so sad. The kicker is all those outraged legacy types stay with the platform anyway.
That may be one of the reasons why MacOS is an "also ran".
Support for older hardware is one of the strengths Windows has. We saw that with WIndows 7 as well as with IE: Microsoft wanted to move on but its users, particularly corporate users had too much invested in those so called "legacy systems". Currently the 14 year old Thinkpad I use for financial work is running Windows 8.1. But, when I get a break from yard work and tutoring my grandson, I plan to stick in an SSD and upgrade it Windows 10. Why not? It runs fine, the upgrade will cost almost nothing and will stop me from having to sink money into a new machine for a few more years.
Another example is COBOL, the business language the proliferated in 30-40 years ago. Today many businesses still run on business critical systems developed with it.
The truth is: while hardware can continue to move forward, it is the software that businesses rely on and where the investment lies. They aren't going to walk away from that investment quickly or easily. And, in some cases, like COBOL, they can't. The resources to replace it aren't available.
See my reply to "crowley" about legacy software and security holes.
While many businesses may still run COBOL programs, I'm pretty certain that IBM has upgraded their mainframe hardware and OS, is it still called 360? I'm also pretty certain that the COBOL compiler of old (70's, 80's) isn't the same as the COBOL compiler of today which runs on much more modern architecture.
Think of SQL Server 7 meant to run on Windows NT systems and SQL Server 2019 meant to run on todays's modern systems. At a most basic level they're just DBMS's but I can assure you, they're quite different.
As an IT person I can tell you it is irresponsible of IT departments hanging on to older versions of their software. But often we do get trumped by accountants and budgets and we find ourselves scrambling to make massive upgrades from SQL Server 2008 --> 2019 or Oracle 10g --> 19c. And the more bean counters wait, the more it costs.
Yeh, I get the "newer is better" stuff.
But, those old systems were built like battleships to be solid, reliable and to dependably get the the job done. And many of them still do. It's not about "accountants". It's about executives who want systems that "just work" -- year, after year, after year.... And, in a mission critical system, that trumps "new" -- even if it is "better".
And, your example of going from a crappy DBS to another crappy DBS shows you likely have likely never seen a quality application system.
What is crappy about SQL Server or Oracle DBMS’s? Or are you one of those mainframe is king DB2 guys? Also keeping up with modern versions of OS’s and software isn’t just change for the sake of change.
It’s also about security.
Would you recommend to your users to stick to with the old 32 bit versions of their software because they still work? Would you recommend that they stick with Windows NT? Lotus Notes?
You missed the point: Those old COBOL programs are still around because they were well built, still doing the job and, in most cases, cannot feasibly or easily be replaced because often they are supporting mission critical systems, nobody can gamble trying to replace them.
Typically, their biggest problem is finding somebody who knows the language and is able to support and maintain them.
The extent of legacy Cobol being mission critical probably cannot be disputed. But we've seen major hacks against financial mission critical systems -- in my mind, this means these complex never updated Cobol systems. Many upgrades to these legacy systems have been akin to lipstick on a pig -- hide unchangeable Cobol code behind a GUI interface, and some fancy REST API.
Many years ago now, Michael Stonebraker, etc al, wrote a books on Migrating Legacy Systems. Hiding legacy code behind new GUI and REST API would be the first step in upgrading legacy code -- hiding the changes behind the APIs protects systems dependent on the API from changes to the legacy code.
They announce that they’re going to break with the past, release something that points to a legacy-free future.
Then a chap from marketing reminds them that if they do that, then they’re going to have to build market share from scratch.
And then the whole idea gets binned … for another five years.
Microsoft is trapped in a legacy hellhole they cannot escape. As time marches on Windows gets ever more bloated because of this captivity. Apple has the luxury of just telling the legacy luddites to fuck off. Every Apple blog and every Apple discussion forum is loaded with outraged users livid because their legacy hardware and software no longer perform. The dropping of 32bit compatibility is just one example. Microsoft could never do that. Apple simply tells the legacy crowd too bad, so sad. The kicker is all those outraged legacy types stay with the platform anyway.
That may be one of the reasons why MacOS is an "also ran".
Support for older hardware is one of the strengths Windows has. We saw that with WIndows 7 as well as with IE: Microsoft wanted to move on but its users, particularly corporate users had too much invested in those so called "legacy systems". Currently the 14 year old Thinkpad I use for financial work is running Windows 8.1. But, when I get a break from yard work and tutoring my grandson, I plan to stick in an SSD and upgrade it Windows 10. Why not? It runs fine, the upgrade will cost almost nothing and will stop me from having to sink money into a new machine for a few more years.
Another example is COBOL, the business language the proliferated in 30-40 years ago. Today many businesses still run on business critical systems developed with it.
The truth is: while hardware can continue to move forward, it is the software that businesses rely on and where the investment lies. They aren't going to walk away from that investment quickly or easily. And, in some cases, like COBOL, they can't. The resources to replace it aren't available.
See my reply to "crowley" about legacy software and security holes.
While many businesses may still run COBOL programs, I'm pretty certain that IBM has upgraded their mainframe hardware and OS, is it still called 360? I'm also pretty certain that the COBOL compiler of old (70's, 80's) isn't the same as the COBOL compiler of today which runs on much more modern architecture.
Think of SQL Server 7 meant to run on Windows NT systems and SQL Server 2019 meant to run on todays's modern systems. At a most basic level they're just DBMS's but I can assure you, they're quite different.
As an IT person I can tell you it is irresponsible of IT departments hanging on to older versions of their software. But often we do get trumped by accountants and budgets and we find ourselves scrambling to make massive upgrades from SQL Server 2008 --> 2019 or Oracle 10g --> 19c. And the more bean counters wait, the more it costs.
Yeh, I get the "newer is better" stuff.
But, those old systems were built like battleships to be solid, reliable and to dependably get the the job done. And many of them still do. It's not about "accountants". It's about executives who want systems that "just work" -- year, after year, after year.... And, in a mission critical system, that trumps "new" -- even if it is "better".
And, your example of going from a crappy DBS to another crappy DBS shows you likely have likely never seen a quality application system.
What is crappy about SQL Server or Oracle DBMS’s? Or are you one of those mainframe is king DB2 guys? Also keeping up with modern versions of OS’s and software isn’t just change for the sake of change.
It’s also about security.
Would you recommend to your users to stick to with the old 32 bit versions of their software because they still work? Would you recommend that they stick with Windows NT? Lotus Notes?
You missed the point: Those old COBOL programs are still around because they were well built, still doing the job and, in most cases, cannot feasibly or easily be replaced because often they are supporting mission critical systems, nobody can gamble trying to replace them.
Typically, their biggest problem is finding somebody who knows the language and is able to support and maintain them.
The extent of legacy Cobol being mission critical probably cannot be disputed. But we've seen major hacks against financial mission critical systems -- in my mind, this means these complex never updated Cobol systems. Many upgrades to these legacy systems have been akin to lipstick on a pig -- hide unchangeable Cobol code behind a GUI interface, and some fancy REST API.
Many years ago now, Michael Stonebraker, etc al, wrote a books on Migrating Legacy Systems. Hiding legacy code behind new GUI and REST API would be the first step in upgrading legacy code -- hiding the changes behind the APIs protects systems dependent on the API from changes to the legacy code.
I'm not sure where you come up with the claim that " major hacks against financial mission critical systems" are against systems running COBOL, The reality would tend to be the opposite since they would be least likely to be accessible by those hackers (not because COBOL is inherently secure, no computer language is) but because those systems would tend to be strictly internal with the least public access. And besides, the hackers would have to know COBOL to change the code!
Microsoft have always been cowards when it comes to taking even the tiniest risks, and moving things forward, and dropping legacy. Its why the newest surface laptops still come with USB-A ports. They're just so fucking terrified of losing customers if they drop them, and thus they're slowing down the full adoption of USB-C and extending the transition indefinitely. Thats how little confidence they have in the merits of their products. They lean on ancient tech as some kind of competitive advantage.
It's not cowardice, it's commitment to backwards compatibility. There are files and software created decades ago that still work on modern Windows machines. That's admirable in a sense, though it does limit their ability to deliver something new that rocks the boat. But it means that business and customers are assured that things won't break if they stick with Microsoft, which is a large part of why they're so successful, they've willingly tied themselves to legacy.
Having different priorities to Apple is not a weakness, it's just different.
Also, I would like a USB-A port on my Mac. It'd be very useful.
They are numerous cheap dongles on Amazon that have your USB-A port plus a lot of other ports. All under $30 and some under $20
Comments
That legacy software isn't always being updated with security patches and even if it is, not at the same frequency as the more modern versions of said software.
As an IT professional, I will not recommend that as soon as a service pack or update comes along we immediately apply it but within a period of 6-12 months we have to start looking at staying with the times.
See my reply to "crowley" about legacy software and security holes.
While many businesses may still run COBOL programs, I'm pretty certain that IBM has upgraded their mainframe hardware and OS, is it still called 360?
I'm also pretty certain that the COBOL compiler of old (70's, 80's) isn't the same as the COBOL compiler of today which runs on much more modern architecture.
Think of SQL Server 7 meant to run on Windows NT systems and SQL Server 2019 meant to run on todays's modern systems.
At a most basic level they're just DBMS's but I can assure you, they're quite different.
As an IT person I can tell you it is irresponsible of IT departments hanging on to older versions of their software.
But often we do get trumped by accountants and budgets and we find ourselves scrambling to make massive upgrades from SQL Server 2008 --> 2019 or Oracle 10g --> 19c. And the more bean counters wait, the more it costs.
And, in a mission critical system, that trumps "new" -- even if it is "better".
Cobol came into being at about the same time as Fortran, Lisp, and Algol.
The truth about Cobol is not the language, but the legacy managers. The Cobol language has incorporated many newer features in addition to having DB2 relational database the backend. But the old Cobol programs were written against card readers and mag tapes, and they still today treat DB2 as a dumb sequential record system. And managers won't allow, much less require, the use of the newer more robust Cobol language features.
It's legacy mindsets, not legacy programs, hardware, etc. Businesses don't have to walk away from Cobol -- they can take the relatively trivial step to refactor their Cobol programs to use the new Cobol features, and SQL against the DB2 database. The result would be more reliability and half the existing Cobol programs could be each be replaced by a single SQL query.
Also keeping up with modern versions of OS’s and software isn’t just change for the sake of change.
Would you recommend that they stick with Windows NT? Lotus Notes?
Many years ago now, Michael Stonebraker, etc al, wrote a books on Migrating Legacy Systems. Hiding legacy code behind new GUI and REST API would be the first step in upgrading legacy code -- hiding the changes behind the APIs protects systems dependent on the API from changes to the legacy code.
I'm not sure where you come up with the claim that " major hacks against financial mission critical systems" are against systems running COBOL, The reality would tend to be the opposite since they would be least likely to be accessible by those hackers (not because COBOL is inherently secure, no computer language is) but because those systems would tend to be strictly internal with the least public access. And besides, the hackers would have to know COBOL to change the code!