True, I'd have to assume actual usage degrades the longevity of some electronics (and number of hot/cold cycles). But there is still some natural shelf-life decay (maybe moisture in the air can find its way, even in what should be sealed components).
> I am completely ignorant to the operation of the 5100 but can't you just > dump that memory when the system is on? Or is there some Jon Titor type > reason behind it? Some of it, yes. Enough that a viable 5110 emulator was made (5100 emulator still in work). But there is some code in the PALM processor itself (about a dozen "tin cans" on there, related to how it actually executed instructions), the Base IO card, display card, comm. cards. Like for the display card, and how its character set is generated (there should be similar code like was shown for the DP). So there are some limits to the emulator, such as MARKing a tape/disk and doing both a read/write of tape/disk images (there is some limited read capability). So, true, baring some of the I/O features and some precise timing of the processor, a reasonable replica could be made (a good replica of the font is in Corti's X11 parts of the emulator - one of the sets at least). But that's an emulation, not a replica :D On Sun, Nov 13, 2022 at 3:00 PM Sellam Abraham via cctalk < [email protected]> wrote: > Steve said: > > > I recall a talk from one of the early 1980s Commodore engineer, where he > was amazed ANY C64 was still > > running since the components were truly not designed to last more than a > few years. > > Me too, to be honest. But then, they seem to just randomly die at any > moment, so maybe the C64 components are only good for so many hours of use, > like a light bulb. The fact that so many of them were made (upwards of 25 > million) is probably why some can still be found working. Those are still > somewhere on the vertical of the bathtub curve on the backend, where their > useful life is fast coming to an end. We will all probably witness the > last working Commodore 64's in our lifetimes :D > > > But, in extracting the data on those TTLs, it seems like a modern replica > of a DP2200 would be possible. > > Can't say the same for the 5100 because apparently nobody left on the > planet understands those MOSFET > > silver cans (and how to extract the 6KB of content from them). > > I am completely ignorant to the operation of the 5100 but can't you just > dump that memory when the system is on? Or is there some Jon Titor type > reason behind it? > > Sellam > > > On Sun, Nov 13, 2022 at 12:15 PM Steve Lewis via cctalk < > [email protected]> wrote: > > > Thanks Jos, I hadn't realized how similar the DP1100 is. > > > > This brochure has a great image of the font right on the front page > (80x12 > > text): > > > http://www.bitsavers.org/pdf/datapoint/1100/Dataform_1100_Brochure_1974.pdf > > > > And it's probably a safe bet that it's the same font as in the 1972 > > models. Would be neat to see the entire character set. In the photo, > the > > screen looks fairly inset -- like maybe an inch? That's good for keeping > > glare off the screen. > > > > I see there was a Cassette 1100 and Disk 1100 (by '75): > > http://www.bitsavers.org/pdf/datapoint/1100/60259_1100_Brochure_1975.pdf > > > > Then I came across a DP2200 emulator, except -- it was apparently made in > > 1973 and ran on a DP2200! (ACM link, but click the PDF, it's freely > > available) > > https://dl.acm.org/doi/10.1145/800192.805722 > > > > > > What a neat system. In an old IBM 5110, I replaced its power supply > with > > modern components. From the DP2200 manual, it looks like it needs -5 > > -12 +5 +12 and +24V? There is a "trick" in the modern buck-boost voltage > > converters to get negative voltage (the IBM PSU needs -5 -12 +5 +12 > > and +8.5V). I put notes about it here: > > https://voidstar.blog/ibm-5100-power-supply/ > > > > Maybe something similar can be done for the old DP's? I understand for > > authentic/historical perspective all original components is prefered, but > > using a substitute PSU is reasonable for checking out the rest of the > > system. > > > > Were there any contemporary complaints about the DP PSU in the mid-1970s? > > Like was it noisy, ran hot, cause any fires? I recall a talk from one > of > > the early 1980s Commodore engineer, where he was amazed ANY C64 was still > > running since the components were truly not designed to last more than a > > few years. > > > > > > What an amazing system those Datapoints were, for their time. The > > chicken-farm story in the DP2200 book is really fun - these farmers being > > savvy enough to code up what they needed, and the systems compact enough > to > > fit in the farms and using modems even to sync up data (pre-1975). > > > > The IBM 5100: 64x16 screen (instead of 80x12 used in DP), and a > > slightlyBut, in extracting the data on those TTLs, it seems like a modern > > replica > > of a DP2200 would be possible. Can't say the same for the 5100 because > > apparently nobody left on the planet understands those MOSFET silver cans > > (and how to extract the 6KB of content from them). > > larger "box"(case) that had a "horn" inside for better airflow over all > the > > components (not an audible horn, but a thing that channel air from the > PSU > > fan to distribute over all the electronic cards and display circuits). > > Plus the 5100 supported the external BNC video (I'm not sure if any of > the > > DP systems had an external video connector? I didn't see it mentioned in > > the DP2200 manual) - I've put 3x extra CRT's chained up to the IBM 5100, > in > > the manual I think it says it can go up to 16 (not sure what the limiting > > factor of that signal is). I'm not sure if quality-wise the IBM PSU was > > "better" (it takes about 3/4th of the back half of the case, the other > > 1/4th for the fan) - other than to say quite a few 5100's are still > running > > in the world. Maybe all that altogether makes it (the 5100) a more > > "portable" system (construction sites, forward edge battlespace, etc -- > > i.e. being more robust to handle outside heat). Also it had a minimum of > > 8K. The APL stuff made the 5100 expensive, but the base BASIC model was > > ~$9K (I think even with the single QIC tape for 207KB storage; but that > > price didn't include async/comm cards). Weren't base DPs $5K-$7K (all > > throughout 72-75) ? > > > > > > But, in extracting the data on those TTLs, it seems like a modern replica > > of a DP2200 would be possible. Can't say the same for the 5100 because > > apparently nobody left on the planet understands those MOSFET silver cans > > (and how to extract the 6KB of content from them). > > > > > > Sorry for the tangent:) I really was just curious about the DP2200 font, > > and possibly seeing where it came from (just based on its style). The DP > > has a better "0" (zero) font than the 5100 :) (IMO) > > > > > > -Steve > > > > > > > > On Sun, Nov 13, 2022 at 3:45 AM jos via cctalk <[email protected]> > > wrote: > > > > > On 13.11.22 07:13, Steve Lewis via cctalk wrote: > > > > I've been looking for a video or image that shows what font the > > original > > > > Datapoint 2200 used. > > > > > > > > It's not shown in the manual. There is one vintage image with the > > > office > > > > lady and the DP2200 on the desk- but the font isn't very clear in > that. > > > > > > > > In any modern video about the DP2200, none of them seem to power it > on > > -- > > > > which is certainly understandable. From what I've read, the power > > > supply > > > > of that system is prone to failure. Also, the system is hard-coded > to > > > load > > > > from Tape 1 -- which means both the tape drive, and tape media, still > > > needs > > > > to be in good working order (which would be pretty rare after this > > time). > > > > > > > > In "the" DP2200 book, it only briefly mentions that the original tape > > > > software was developed "on an HP system" (without any elaboration > that > > I > > > > could tell on which HP system that was). > > > > > > > > Nothing in the manual suggests the original DP2200 could "program > > itself" > > > > (i.e. no built in machine code monitor -- those TTL chips had one > > strict > > > > boot up sequence: load from tape 1). If there was a read error or > no > > > tape > > > > available, I'm curious if any message showed on the CRT. > > > > > > > > So, I was just wondering if there was any known pre-1973 Datapoint > > 2200's > > > > that are still working? (and/or if any HD video of them powered on > and > > > > legible font can be seen) Or any other more current system that we > > know > > > > for sure used the same font? > > > > > > > > Thanks! > > > > -Steve > > > > > > > > > Not only is the powersupply prone to failure, it is also the mostBut, > > in extracting the data on those TTLs, it seems like a modern replica > > of a DP2200 would be possible. Can't say the same for the 5100 because > > apparently nobody left on the planet understands those MOSFET silver cans > > (and how to extract the 6KB of content from them). > > > dangerous I have ever seen, and I am hesitant on working it. Primary > and > > > secondary sides not separated, isolation between the two almost > > > nonexistant, many primary nodes exposed. Would never pass modern safety > > > checks. > > > > > > But here is a picture of my DP1100, a DP2200 derivative, while it was > > > running a memory selftest, for a short time in 2021, before the > > powersupply > > > blew again : > > > > > > > > > > > > https://forum.vcfed.org/index.php?threads/its-alive-my-datapoint-2200-1100.1222197/#post-1222197 > > > > > > While the DP2200 is hardcoded to start from tape, the DP1100, otherwise > > > identical, boots from a ROM. This ROM also contains a minimal > machinecode > > > monitor. I recovered & disasembled the ROM and Gordon Peterson, from > > > Datapoint, commented it out : > > > > > > http://www.bitsavers.org/pdf/datapoint/1100/DisketteBootDisassemblyGEP2.txt > > > > > > Note that there are multiple videoboard options : the later DP2200, my > > > DP1100, and the DP5500 share the same videoboard. This relies on a > > > programmable characterset. In the disassembly mentioned above above, > > > starting at line 3660 you will see a load of gobldecook, these are > > actually > > > fondsets to be loaded into the machine. > > > > > > The fontset has a very particular "look" to it. How much is due to > > > fontdefinition, and how much is due to the diddlescan, that I dont > know. > > > Diddlescan is where they scan each character in full, before proceding > to > > > the next. > > > > > > Note that a ROM based bootboard for a DP2200 would be a trivial > > > undertaking, and only involve changing the cassette reader board for > the > > > ROM board. > > > > > > > > > Jos > > > > > > > > > > > >
