Jump to content


  • Content Count

  • Joined

  • Last visited

Everything posted by wierd_w

  1. Since I have the dehydrator (kinda) working, I can do long prints with nylon now, without the filament absorbing water like a depends undergarment. As such, I have made some modifications to the previous design, specifically to stiffen the shell up a bit for use with the smurf-blue weed trimmer nylon. Once I test the changes, and am happy with them, I will upload them. There is nothing wrong with the prior models; this is explicitly changes made specifically to add the extra stiffness needed for the nylon.
  2. It is highly unlikely that a USB based FDC is going to be able to handle having a tape drive laid on it. Instead, I would look for a vintage computer/motherboard, and use it to get at your data.
  3. No. Thermal printers use heat to cause a chemical reaction in the "paper". There is no mechanism to make them work with ordinary typing paper, unless you are also a closet chemist, and want to apply the thermal sensitive chemical coating to said paper first. (I have made my own photopaper in the past using unflavored gelatin, would not recommend. Not an easy task.) Depending on how hot the thermal element gets, the thermal printer might be able to do some interesting things for you though. There are interesting thermal-sensitive foils intended for scrapbookers. Using some scotch tape, and and some folded over corners, it might be possible to coax your thermal printer to selectively heat the foil against cardstock to get pretty gold, silver, et--- transfer images. Take for instance, this stuff. Since you can do this with a laser printer, it should be doable with your thermal printer too. Just sandwich your foil with whatever paper you want to transfer to, then run it through. See what happens.
  4. Just well known retail games is greater than 5000+ I know that with certainty.
  5. I have recieved my unit, but have not installed it yet. Been busy.
  6. Yes, that is exactly the situation I see also, (also, why not interested in purchase at this time.) Ultimately, I see this thing defaulting to some kind of pre-baked "idle" loop with an exit condition, with a mapper chip exposing 4k of RAM to the TI. The mapper chip should be controlled by the ARM cpu, but since the ARM cpu has no program to run (other than the idle loop), the page just stays there. The TI is then able to write 4k of IPL to this page, and then do a write-on-rom to set the loop's exit condition, which then jumps execution of the ARM cpu to the exposed page. The IPL loads, and the ARM CPU can then move its window (because it has the software to run the mapper chip now), and can accept a larger program from the TI. This would make the strangecart fully configurable by whatever software tries to handle it, because that software controls the ARM chip's boot cycle. Whatever personality or featureset would be determined by this initial bootstrap, and whatever code gets shuttled to the ARM chip. Again, I would want to use it to supply zram-like compressed 4k pages, using write-on-rom switching. Depending on the data you want to put in the cart, the compression could yield a lot of storage there, and some routines for other useful functions (since the ARM cpu actually owns all that memory flatly, and just moves the TI's window around) that manipulate that memory without the TI's involvement, so some pretty sophisticated games could be made that leverage the cpu there, even with the compressed ram function. I would approach it as a software programmer's multitool, rather than a consumer device at this stage in the game-- As programmers play with the toy, and find ways to use it, then it would become useful for consumers who want to use that software. Right now though? No-- It's a development powertoy.
  7. Since you have I2S, you could also put a simple UART on there. Why is that potentially neat? There are UART->Wifi solutions for retro computers. A simple I2S UART module could be slaved with such a module (and maybe even still fit inside the cart shell?), and permit the TI to do some additional tasks. Since the UART and WIFI would be controlled by the ARM processor, the whole thing could be software controlled from the TI just by writing to and reading from some memory addresses. Imagine a terminal cartridge, that is a fully capable terminal. Another interesting use could be an I2S bluetooth module (which do in fact exist), which could let you pair your TI with bluetooth speakers, mice, and bluetooth file storage capable devices. Rather than integrate a sound DAC there, I would leave it as a 90 degree 4 wire header, and give a retention clip of some kind, so that a module could be added by the user later.
  8. Dosbox allows commandline argument invocation with the -c option. As such, you can create direct shortcuts that launch your games right from windows. For example, you can make a shortcut to dosbox like this: C:\dosbox\dosbox.exe -c "mount c c:\dosgames" -c "c:" -c "cd doom" -c "doom.exe" -c "exit" Which will process dosbox's autoexec.bat section, then do all the -c lines in order-- in this case mounting a dosgames folder as the emulated c drive, switching to it, moving into the doom folder, running doom, and then when you quit, killing dosbox. The shortcut can use whatever icon you set for it and live either on your desktop or in your start menu. It also works fine on linux, just use appropriate paths. Linux program launcher works fine with it.
  9. It's a naked ARM processor. The instruction set is well documented and there are loads of compiler toolchains. I am not terribly interested in purchase at this time, but I already mentioned what I would try with it-- Adapting linux's ZRAM code to supply compressed 4k pages on the TI, to make that ~128kb of memory act more like 384k.
  10. Dont be a stodge! There were all manner of accessories for sound on the PC: Disney Soundsource / Covox speech thing Adlib Soundblaster MPU401+external midi synth Microsoft sound system ESS Audiodrive etc. Also, VGA was alive and well in the 90s. Trident 1mb ISA cards were very popular, and later 2 to 4mb versions for PCI were likewise hot sellers. The platform hosted such memorable classics as Doom, Heretic, Quake, Starsiege tribes, EF2000, and numerous others. (my favorites were mystic towers, from an australian publisher about a crochety old bastard who has to fix the mess his ancestor left behind, and Hocus Pocus-- a VGA platformer with a quirky premise about a junior wizard.) Dont be so quick to write it off.
  11. I would suggest that the audio DAC and pals, either route audio out through an RCA port, OR-- To accompany the device with a sidecar adapter board. (Has a cardslot, and then an IDC cable for the other signal wires, including the audio out line.) That way it could be installed in either configuration.
  12. Ah, did not see that you already said you knew it was bad; Yeah, I see what you mean about finding a 640pF tantalum. Given the teeny capacitance difference, I would go with the 650pF as a replacement, and hope for the best.
  13. IN the bottom picture, the one on the left looks like it has cracked. I would check that one.
  14. After exhausting my remnant supply of smurf-blue weed-eater filament a few days ago, I ordered 2 new 5lb spools. Since the stuff is REALLY intended for, you know, CUTTING DOWN WEEDS, it is not packaged in anything even resembling a hermetically sealed bag-- (which basically all quality printer filaments are shipped in), so I need to dry the stuff. Sadly, nylon is so hygroscopic, that it sucks up water out of the air up to 3% of its weight, and does so VERY fast. This means that long-duration print times (like those PEB enclosure shells), suffer terribly, even though nylon would be a nearly ideal material for them. SO-- I have thrown together a super ghetto hot-box out of a cardboard icecream treat box (Great value variety pack of icecream treats is ideally sized to hold a hand-wrapped spool of the trimmer line!), that has a hole cut for filament to come out, and is getting actively roasted with a hair dryer on high. The side of the box is open, with the filament roll on the roll holder (which has been pushed through a slot I cut in the side of the box), and the hair dryer is aimed into the box (several inches distant, because I do not want the dryer to overheat). This seems to be giving "Sufficient" hot-box active drying while I print the actual subject of this post-- (Not really dry enough, but sufficient to not be completely just opaque blue fuzz. It has some fuzziness still though. It will probably get better over the duration of the print, since the box holds heat from the hair dryer, and gets actively circulated. A few hours into the print and it should clear up. I only have 2 days off, and used yesterday getting DHE's shells shipped. I want this thing printing ASAP.) I have a rather large Oster brand food dehydrator that would be IDEAL for holding the mammoth sized spools this stuff comes on, and for functioning as a much safer hotbox active dryer for long prints. It is some 13.6" in diameter. I have designed a set of custom shelves for the thing that will be capable of suspending and feeding the huge rolls this stuff comes on. I am making *FULL* use of the large print volume of this new chiron. According to Cura, the printing process will take an ENTIRE week. If this works, I might print another set of the shelves, and send them to Arcade Shopper. (Actual drying units cost between 60$ and 300$, and do not have anywhere close to the capacity that this relatively inexpensive (30$) dehydrator has. Since I am printing the shelves out of nylon, they will have very good heat tolerance, so cranking that dehydrator up all the way wont hurt the new shelf. Having an active dryer could be helpful for his printing exploits.)
  15. Not really? See for instance, the zstd compression algo used by ZRAM. It is able to pretty reasonably get at least 3:1 compression on normal RAM contents. (worst case is that you load the entire page with uncompressable data of course.) Here you see ~101mb of swap allocated in the zram pool, packed into ~21mb of space, for a ~5:1 compression rate. zstd packs that stuff hard. Instead of packing swap though, we would be packing normal memory pages. (the zram algo already works as arbitrary pages, btw.) Since memory accesses are heavily hobbled by waitstates in the TI, you have multiple execution cycles of the 9900 in which to do the needful to: 1) Catch the bankswitch trigger 2) write the current uncompressed page into its appropriate compressed page. 3) Read the requested compressed virtual page, and uncompress it into the uncompressed page inside the window. From the 9900's point of view, it looks as if a normal bankswitch occurred. (it has no clue.) However the MIPS processor is doing the heavy lifting to implement somethig akin to zram's algorithm, with zstd compression, to make the small amount of RAM it has onboard the strangecart look and feel like it is many many times larger, by leveraging the fact that it is many many times faster than the TMS9900.
  16. One potential application for such a device, is compressed (or even procedural generated) content being made available JustInTime(tm) as if it were bank swapped memory. The data is stored in compressed form (Say, using xz ?), and the MIPS chip is fed a decoding / "bankswitching" program when the accompanying program is executed. (So it can say, emulate Write on ROM behavior to catch a bank switch). Since the returned "pages" are just decompressed data from the compressed data fed into the cart's actual memory, they would be virtual pages. (The MIPS processor overwrites a single page of actual RAM with the uncompressed contents of the compressed page data in its other pages, and does not actual move the window. It just loads the appropriate page into the decompressed page area as requested.) If procedural content is done instead, the content generation routine could be loaded into the cartridge's memory along with the procedural generation algorithm, and then the cartridge itself generates all the data the program needs, rather then trying to load it from diskette. (Such as say, rules-based routines to generate mazes and collision definitions-- or generate random overworld maps for say, a zelda-like, using a random seed.) Another interesting use would be as a compiler accelerator module. (say, a re-write of the BASIC compiler, that leverages the MIPS processor in this cartridge.) There are a LOT of things this kind of gadget could be doing. (I rather like the idea of something like Linux's ZRAM; The strangecart could be fed a program to create compressed virtual pages, and thus act like it can hold way more data than it actually can, and do so fairly transparently, because it can do the compression/decompression fast enough that the TI has no clue.)
  17. During the peak of the DOS/Win9x era, I was working computer repair at a Mom&Pop and got an employee discount on hardware purchased from my employer. As such, I was the proud owner of first a 486 and then some early pentium systems. I even owned a VFX-1 at one point. (but quickly learned that my astigmatism was so pronounced that I could not use it for more than 10 minutes at a time. Sad.)
  18. They made a sequel based on the Quake engine sometime later also. (has an original CD still.)
  19. wow. Lots of kit. Nothing I truly desperately need though.
  20. Looking at the pinout of the expansion port... Not much? It has the address and data lines, MEMEN and CRUOUT/CRUCLK. So.. Build one, and write some software to use it?
  21. If they werent such boat anchors, that cost more to ship than they are worth... I would give one of the . (ahem) small collection in my upstairs storage.
  22. You want the June 2010 offline redistributable. Newer iterations of 9c dropped support for XP.
  23. The newest that can be installed on XP is DirectX9c which version are you trying to install?
  • Create New...