IBM killing mainframe coding kit for PCs this year IBM is killing off a mainframe coding toolkit for PCs and withdrawing all support, directing developers to instead use a cloud-hosted environment for dev and test purposes. In a notice issued on September 30, Big Blue informed customers that it will cease marketing the System z Personal Development Tool (zPDT) as a standalone …
I can't answer the lower- vs upper-case thing, but from experience, ZD&T and zPDT are effectively identical in the emulator part of the product. ZD&T has extra pieces to enable multiple instances to be provisioned (within the allowed licence count) and software-based licensing (no hardware USB dongles that are supported by zPDT).
There is a slight difference in the machine model type which is almost too boring to mention (type 1090 vs 1091), but from the individual z/OS user perspective, zPDT and ZD&T are basically the same.
I really hope they don't kill ZD&T....it's a real shame IBM make it so hard to get people familiar with z/OS (and z/VM) and Z hardware.
It's a real problem. I deal frequently with customers where the divide between the client (Windows & various Unices) and the mainframe side is vast, with neither apparently being aware of the differences.
I'm fortunate to have spent most of my 40 years working with mainframes on the "shrink wrap" software business where I got to play with everything from the base operating systems to lots of the middleware, and at the right time to pick up enough of all of them to be a jack of all trades. You simply do not get that experience working in the much more controlled environments as used by major banks and insurance companies, so you really need something that you can afford to trash and restart ad hoc.
The big problem is that most of the people who do have that experience is that we're all getting on now, and retirement is on the horizon. Finding people with a serious systems software background now is like looking for hen's teeth. It appears that IBM are not helping with this, and seem to be subcontracting a fair number of what used to be core functions (SDSF, JES, etc.) to 3rd parties, such as Phoenix & Rocket, presumably because they still have some old people who can still maintain them. Withdrawing the z/PDT doesn't help.
Some of IBM's own dev teams will effectively be dropping support for Z because they can't access the partner program and can't get the budget for Z either.
By that I mean Z access reliable enough for production build and test.
Even using other teams Z always ended up 'too expensive' and famed Z reliability was generally in the two rusty tin cans and a length of damp string category when you weren't in the same TZ.
Won't that be fun.
A little over $20k/user/os/year
They probably are scared that the emulator is about as fast as the real machine.
Also... if you are targeting z/vm it was only cost effective if you had like one developer. For z/os, though, that's like one cpu second of time on a real machine.
I think you might be misunderstanding what's happening. This is for folks not at a mainframe shop, but who want to develop mainframe-quality software. The ability to do so on one's own terms is a crucial factor, and avoids the need to concern oneself with data sovereignty, security and other exposures that exist when one relies on other parties at other locations.
