Hello there.... 😀
(I have been a lurker here in couple threads and I admire that here are some who actually not only use but also create hardware... I create too)
Here I describe a somewhat different point of view. A view from a non-gamer who still uses old PCs like 486.
ALL is very different if the old machines if to create also hardware, not only software or using the software which requires some special hardware function. 😀 Then the emulators wont help.
I still find very good use for old PCs and actually here I prefer 486 machines becouse of their compatibility with some VGA cards (cirrus logic) which makes some programs to run (Soundclub, Cubic player) and the Gravis Ultrasound cards, demoscene,tracker programs. Of course several things can run faster on the Pentium but then again there are several things which do not work anymore. SO, the next machine again to keep.
Demoscene, trackers and such are why I deal with this and it has been much more accessible to me more than Amiga and others. Well, I had a ZX Spectrum clone (Leningrad version?) but it was more broken than in working order.
About 486. I keep it becouse it is very useful as I have created my own HARDWARE for it, also my programs access hardware and this is really what a emulators cannot do -- give suitable bus for hardware cards what are my own custom made things, MIDI port without latency, and stable timing.
So far I used that 486 with its DOS demos and music programs and trackers, but also hooked up via MIDI port the synths. Thats neither what emulators can cope.
AND lately I began to use more that old hardware like 486 becouse on one day I discovered the more demanding needs to program the chips what the Willem programmer revision did NOT worked as it should or it was not supported -- thus not compatible, even if it claimed to support. I am handy in electronics but neither the provided schematic in order to make expansion board to program 27c800/160 worked but also I had no access to program source. Still the latest versions require another hardware.
So, I just made a ISA card interface and I/O board and also I re-used my own old skills in Turbo pascal and voila! - my own programmer and all things done and working for ANY custom chip! It is later that when I actually realized what I had done really -- I re-used my very old skills, instead of eating through the documentation how to make the drivers in windows environment and to learn new programming tools, but also I used very affordable circuits, whereas USB version or PCI version would be much more expensive and also it saved much of the trouble, but at last, but not least -- I did something on old computer again, what manufacturers would have done and which has "requirement pentium 200000000000 and M$ platform". I saved time, money, nerves. Becouse when times change then the windows platform renders the compilers obsolete and they say that its time to move ahead just becouse they say so... thus it is next to impractical to create any hardware which uses windows as the Intel already abandoned before ISA bus, then PCI buses are neither so much affordable, and lately the LPT port vanished too.
Neither the programming in that windows environment is not my cup of tea becouse lately most of the software developers do not want to compile their programs to support on such old systems as on win98se while there is really no argument that why it would need atleast winXP but only their lazy habits!
Well, also here I skip the story where someone recommended that I can make a I/O bus interface with arduino and can have a speed with 200kB/s with USB, where I actually had my first ISA bus card I/O interface with 900kB/s already for 8bit mode, 2MB/s with 16bit and no hassle about making driver routines becouse it is with direct port access).
It is not the first time when I got to utilize my old programming tools and hardware habits again. with new platform it is always that the everchanging systems are rendering useless the knowledge but also making obsolete the bought or selfcreated hardware.
Also just the Willem programmer programs that one 4Mbit EPROM chip in 4 minutes and blank check is 60 seconds. My SEMF (as I call my programmer which I did) programs it in 36 seconds and blank check is 2 seconds.
The 27C160 chip is programmed in 56 seconds with my SEMF programmer (--the factory specification max speed for this chip!) and blank check is 4..5 seconds. WIth WIllem programmer it is 16 minutes to program that chip.
I am not into games so much and I have used my PCs in a manner than instead to upgrade and to move on the next platform I stick to the old machine and seek the alternative programs which do the same things without upgrading or buying anything at all. Thus I used 2 machines until 2003 and one was 486 and the second was AMD 450MHz Pentium but with win3.x and I managed to do everything on that what the average office work was. I managed to find alternative working thing for DOS or win3.x while others only knew the version working for win2k atleast.
When I watch now what is about the computers then it seems that its ridiculous.
To return to the EPROM burner problem if just solving the problem with money then similar programmer with such high speed would need at first money, for second new computer, newest operating system, and neither i have a good look over the software which is perhaps overly complicated, and thus it is very impractical machine where there does not run any other items what runs fine on my other computers. So far I see that the only things which demand for "upgrade" are the web browsers (so I avoid those pages), which also demand the upgrade for a newer platform, but where does not run anything other practical.
In fact the newest machines are 10 year old in my setup.... 😉