Hello Larry,
Am 13.03.2023 um 13:17 schrieb Larry DrFranken Bolhuis <midrange@xxxxxxxxxxxx>:
I could regale you with many stories where this mentality hurt the company, and nearly none where it helped.
That's my point.
VERY often someone rather high in I.T., perhaps a CTO or even CEO gets fed the biggest lie in all of I.T.: "IBM i is just AS/400 with a coat of black paint. It's old, incapable, and needs to go away."
That sounds pretty familiar. The cases I've witnessed were often kicked off by external consultants, just knowing about Windows and maybe a bit of this mysterious Linux thing. This situation is worsened by the gold guy retiring next year and finding an AS/400 guy as replacement proves almost impossible. They're virtually not existent.
This paints the IBM i team into a corner putting them on the back foot, in a defensive position. Then sadly, all too often, it's discovered that the IBM i team has been treating the platform like it IS an AS/400! They continue using record level I/O, building new subfile programs, and writing in fixed form RPG/400 with SEU. They've given the boss all the ammunition he needs.
This *might* also stem from a boss not giving the team the necessary resources to learn and exploit the newer techniques available with newer releases of the platform. If those guys are kept busy with implementing yet another specialized but rarely needed report, I don't wonder. Additionally, if the boss is like "what's the benefit compared to the status quo", the guys are also in the corner, because it works (almost) perfectly. Why fix what it is (apparently) not broken? (Note: This is not my argument!)
Also it's frequently discovered that the server itself was sold as a 'one purpose machine.' It was configure to run just an ERP package.
Yeah, but that's pretty much how things worked back in the day. Machines were specialized. The AS/400 was for ERP. The NT Server was for file sharing. The Linux Box was doing the intranet website. It is just another case of using the "right" tool for the job. "Right" being a relative truth, depending on the skills and resources of the people involved in establishing that function.
I observed that most (German) firms aren't really interested in brands or ideological discussions about the best platform. They usually have a problem to solve. They give you the money if charges are reasonable, and you provide the solution. If you only sell a product and leave them with that, they're unhappy and sooner or later will seek someone else's consultancy, not only selling them a product but provide implementation into the existing infrastructure, and service (updates, support, etc.).
This gives a lot of freedom to the implementors to choose their preferred product. The customer doesn't care as long as the pricing is reasonable and the thing does the job. Since IBM i is a niche product in here, it rarely is considered to be a target platform for new stuff being put on, unless there are people involved knowing the platform good enough to include it in the planning process.
Again, my observations are mainly about small to mid-sized companies.
No room for growth, no ability to add additional partitions (Virtual Machines), no ability to use advanced technologies like remote copy, flash copy, live partition migration, and such.
Of course not. IBM pricing usually dictated to buy the machine with just enough resources to provide a reasonable computing environment.
This is somehow a chicken-egg problem. If you don't put more (reasonable) workload on the machine, it's hard to justify expenses for it. If you don't, users start complaining about "the AS400 being slow again".
As a consequence even if the boss throws the IBM i team a bone: "Hey, can't that server run Linux? I have an app that we need that might fit well there."
If the boss is technically interested, this might happen.
Still the team might refuse because "we have no clue about that Linux thing and already more work to do that we can handle".
But the server has no HMC, no spare memory, no available disk and acquiring them will take took long, so "Just put it on VMWare."
Yes. From a overall economic viewpoint, this is often the cheapest solution.
And as a consequence, all too often: "We have met the enemy, and he is us!"
Yes. And for that reason I think it's important to build bridges by not doing things in the backroom (local hosts file) but seek communication to the other team asking for a dns entry, bothering them, and making them aware that in the end, there is *one* network connecting all machines.
Building islands has been a bad idea ever since. But it's a natural thing for humans to do.
:wq! PoC
midrange.com
