usually it is very important, that your clients liking your product, that gives you a good hint, how your next product should look/work, and of course in that new product you will build in all the popular features of the old product, so the client will like it, and buy it, that is a good business practice, which worked in various part of the industry for very long time. For some reason software industry does not follow that logical and good practice, in case if the get it once right, you could be sure, that the next generation of the same software will be different and with a few exception less use friendly, well that is valid not just for Microsoft products. And also as soon as one learned the required new tricks to get around, there will be new operating system, which has more gimmicks – e.g. touch screen in widows 8 – which are in most of the case annoying, and it turn out very fast that they are even more sensitive to malicious, do not work with the old engineering software and need a new hard ware again, which will be usable after the learning curve for just a short time again. XP was one exception it worked relatively well with software made for previous operating system, did not required a super fast new hardware and it was usable, we got used to it.
Question: it would be not make more sense to "redo XP inside" but leave it the outside ?
I agree with your concerns; in my particular case, all peripheral connections to the PC are USB. I don't require any GPIB or parallel port interfaces. For some OTHER applications, I do use external USB to serial port converters (readily available, as are I think USB-parallel port and USB-GPIB). Because of the number of USB ports (limited on the old XP boxes) I generally use external "universal" port replicators that provide lots of USB ports, an additional Ethernet port, and (most important) another monitor port for dual-monitor configurations. These are dedicated setups used primarily for QA SW testing, not production HW testing.
Actually, you might be surprised -- if you have room for some brackets. My Asrock micro-ATX Motherboard has on-board connectors for serial and parallel ports (but not on the back panel), but unfortunately I can't use them because my 4 slots are used up (2 for video card, 1 for USB 3.0 card, and 1 for Kvaser CAN card).
For industrial use, the Foxconn H61AP motherboard has some interesting features: back panel serial port, back panel parallel port, internal connector for second serial port, and 6 PCI slots. I picked one up last year, and loaded with Windows XP and a whole bunch of PCI cards (mostly CAN cards). It's still available at Newegg for <$60
My guess is that driver support is much better for Win7-32 than Win7-64. VirtualBox has good USB pass-through support and IIRC experimental PCI pass-through support (but no parallel port pass through).
I'm pretty sure you can get parallel port PCI cards, but don't know how well they're supported by software.
My concern with moving to Win 7 for a test environment is the availibility of drivers for legacy hardware. I actually like Win 7, I'm just heasitating because of the old equipment that connects to a parallel port and uses a program that runs under DOS. I notice that new PCs don't seem to come with parallel ports.
Although I make do (we're on Office 2010 at work), I'm not a big fan of the MS ribbon interface.
I strongly recommend LibreOffice over OpenOffice; it's updated much more frequently. And LibreOffice is better than MS Office in some areas; for example, LibreOffice Writer is significantly better than MSWord for writing long technical manuals (I know, I switched from MS Office to Writer for a ~250 page manual because Word was driving me crazy. Writer wasn't perfect, but it was significantly better).
I am mystified that so many folks still subscribe to the notion that WinXP was the most stable OS, and everything after was a step in the wrong direction. I have an engineering lab FULL of WinXP PCs (some LT, some DT) that are running test benches ( vehicle simulators mostly). I can't WAIT to upgrade these all to new HW/Win7Pro because these machines are running 24/7 and are used remotely by teams all over the globe. Typically, I have to do at least a warm reboot every few days, plus a full power cycle cold reboot every week or two to keep them functioning. This is absolutely due to various driver and OS issues that cause test applications (that work fine forever on Win7) to SEEM like they are working but really aren't. I suspect the memory limitations of the 32-bit OS/HW aggravate the situation; all of the replacement machines on order will run 64-bit Win7Pro with at least 8GB RAM and i5 or better CPU. The prototype one I have just keeps on keepin' on!
What are the engineering and design challenges in creating successful IoT devices? These devices are usually small, resource-constrained electronics designed to sense, collect, send, and/or interpret data. Some of the devices need to be smart enough to act upon data in real time, 24/7. Are the design challenges the same as with embedded systems, but with a little developer- and IT-skills added in? What do engineers need to know? Rick Merritt talks with two experts about the tools and best options for designing IoT devices in 2016. Specifically the guests will discuss sensors, security, and lessons from IoT deployments.