This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.
Messages - z80man
Pages: 1 ... 23 24 [25] 26 27 ... 62
361
« on: April 20, 2011, 01:50:41 am »
Yeah so basically the idea is to post the requirements for working at TI on either the nspire or the 83+ series. 1. Optimizations are key. When adding 9 or 15 use bcall_HLplus9 or bcall_HLplus15 instead of their unoptimized add counterparts. 2. You must know less than 10 asm instructions. If you know more than that then you might try to optimize the code 3. f**k user friendliness. We're here to prevent those d**n programmers from playing their games. 4. The teacher is your boss. 5. A certificate of high school dropout is necessary to become one of our coders. 6. When quoting our inspiration for the CX's color screen say, "There was none, we were the original creators" then hide the Prizm you were playing minesweeper on
362
« on: April 20, 2011, 01:34:26 am »
I've had a Prizm since the day it came out. Even though drawing in BASIC is slow I love how it is backwards compatible with all previous Casio graphing calcs. I also like how it is much more acceptive of asm code than the nspire is.
363
« on: April 20, 2011, 01:30:07 am »
I've been trying to find a purpose for the SYNCO instruction that Casio used in the DirectDraw routine and I found something interesting. Now I haven't had enough time to fully disassemble the routine to find the true purpose, but what I found is that SYNCO is good for self modifying code. Reason being is the separate instruction and data caches meaning that if you modify some code that is close to the PC, the changes may not result when the instruction is executed. What SYNCO will do though is ensure that the pipeline is reorganized ensuring that the changes will take place. Now it does seem strange for Casio to use something as advanced as self-modifying code, but they just might of for this routine.
364
« on: April 20, 2011, 12:23:23 am »
I'm also rather new to ubuntu too, but I actually have a rather powerful computer. I did download the 64 bit version partly because I wanted to start learning x86-64 asm code, but now I'm wondering if that was a smart choice or not because there are many driver and software issues now. I also need to get used to the new terminal which seems to be rather similar to the cmd on windows systems.
365
« on: April 20, 2011, 12:17:06 am »
Yeah it would be nice to have an actual shell that can run ASM/C files, to save space. Doors CS would be a nice shell, although since there are some people who dislike the mouse cursor Kerm would have to ensure that HomeRun is possible on the Prizm.
Other shells are welcome too, as long as they're compatible with each others. Just for the sake of user preference.
About a month ago Qwerty drew up a proposed file format for Prizm asm programs that I'm now implementing in a shell with only minor modifications. What I like about his format is its clean packet layout that allows for near unlimited future additions, backwards compatibility, and cross-compatibility. I'm also making a big deal of cross-compatibility in allowing libraries to be treated as individual files that can be imported to the shell during runtime. As a note this is much easier than it sounds because the Prizm memory follows the FAT32 protocol allowing it to act like a standard usb drive when attached to a computer and also negating the need for TI-connect like software.
366
« on: April 19, 2011, 03:20:14 pm »
The problem being with apps/g3a is that 1. That they run from flash which isn't all that bad, but prevents the use of self modifying code 2. The header size is 29 kb 3. There are extensive formatting and checksums necessary in the header 4. Startup code is required to run apps
367
« on: April 19, 2011, 03:10:35 pm »
Even with all of the new hardware we still have some of the same old issues with calc gaming. This being most notably hardware restraints. Even though these two new calcs have powerful procs and much more ram compared to the 83+ we also need to account for the greater need of memory and proc power due to the new color, high-resolution screens. Just as an example the Prizm can only fit about 5 16-bit images in its ram due to its limited 2Mb size. So we aren't really dealing with high power devices yet like ubergraphX, but we are getting closer.
368
« on: April 19, 2011, 02:22:59 pm »
369
« on: April 19, 2011, 02:15:44 am »
I don't know what will happen to the processor (other than higher heat production), but given that the feature is supported by the manufacturer, I'd hope that the appropriate cooling systems exist.
* z80man decides to add a liquid cooling system to his Prizm
As can be told from this quote from the early days of Prizm hacking, there were plans to install a liquid cooling system to prevent the catastrophic meltdown that was experienced earlier today. Unfortunately the coolant had not arrived in time for the LAN party which therefore caused the excessive heat and fumes. Luckily the shipment is expected to arrive by next week allowing all future overclocking endeavors to be done in a much safer manner. I'm happy to report that the only harm done to me was a large fluctuation in my nasal hair. My Prizm on the other hand required a new processor due to the extreme heat damage. As of now it has been replaced with the new SH5 proc to replace the aging SH4 version that previously inhabited the Prizm. Overclocking tests on the prototype Prizm emulator have shown to produce similar fumes.
370
« on: April 18, 2011, 02:11:23 pm »
The screen receiver seems to work in BASIC progs only when there is little action. For example when the calc is is a repeat getkey loop then it will send an image, but if there are many drawing operations going on then the screen receiver will wait till the drawing calms down. I'm not sure if it will work when there are math only operations going on.
371
« on: April 18, 2011, 12:55:44 am »
The BSoD is even better if you extreme overclock. Once I pushed my cpu above 4.0 Ghz with high voltage and everything. Once the cpu reached 80 C I saw a random colorful rainbow and then everything shut down. At first the computer would not respond to me turning it on so I had to reset the cmos pin before the computer would work again.
372
« on: April 18, 2011, 12:46:51 am »
Oh dear Microsoft must of hired TI to code Vista. There's probably a hidden z80 somewhere in your laptop and that must be what Vista is running on. It wold also explain the bloated and unoptimized code
373
« on: April 18, 2011, 12:37:50 am »
Yeah I agree your signature is a bit too large. You should put that video in spoiler tags or just put a link.
As for Halo I used to play almost every night on XBL, but then I got bored because I am terrible at FPSes, plus sometimes the people I played with were annoying so I lost interest. My favorite game was Halo 3 personally, but I didn't play Reach much.
I suck at action FPSes, but I'm pretty good at half life single player, 007, tom clancy rainbow six, etc., since it requires precision rather than speed.
I suck at FPS's only when I'm using a console like the 360 or PS3. On a PC though I'm pretty good. So because Microsoft is very slow with the PC releases of Halo I don't get that much action there. I do believe that the Halo 3 PC release is coming out this year though
374
« on: April 18, 2011, 12:22:46 am »
The worst exception is division, because that can take something like 96 cycles to do.
Well in that case you need to be careful. Of course speed optimizations can be done by rolling out the for loop. But even better is try to avoid division it in the first place. The best thing to do is bit shifts for the cases of when you are dividing by powers of 2. And the SH4 is just brilliant at bit shifts
375
« on: April 18, 2011, 12:17:21 am »
I hope they didn't go the same way as TI and use a 1024 or even 2048 bit RSA private key...
At least we can still run ASM/C code easily, though.
Not really positive, but as a note the SH4 has a built-in hardware device that specializes in the fast calculation of RSA keys. It doesn't mean that Casio used an RSA key, but they could easily if they wanted to
Pages: 1 ... 23 24 [25] 26 27 ... 62
|