Stuart

Section 1.1 of the E/A manual states exactly that. "... there are many fine books available which teach the basics of assembly language use. After you know these basics, this manual gives the details of the TMS9900 assembly language and its application to the TI Home Computer." There's also an assembly language primer on Thiery's site: http://www.unige.ch/medecine/nouspikel/ti99/assembly.htm. My first assembly language program was written using the Minimem LBLA and added two numbers together then returned to the prompt. Had to use EasyBug to check the result in memory.

I'm all for helping people play with hardware. Depending on how dirty you want to get your hands, there are some options available already: -- I've got breadboard designs for a basic TMS9995 (as used in the Cortex) and TMS99xxx systems on my website (www.stuartconner.me.uk), with BASIC and Forth available on EPROM. -- A couple of people have done further development on the 9995 breadboard design. There's now a version with a CF interface for storage and running barebones Unix. Bank switched RAM is going to be added to that, and I suspect it will be committed to a PCB as well. There's a chap in Australia who's done a PCB for the original barebones system, and that is probably going to be expanded as well. -- I know there's at least one other 'experimenters' 9995 design out there, but you would have to re-layout the PCB from photos. The F18A development lets you easily add a VDP with a VGA output - great as you don't have to touch any of that nasty analogue stuff! There's a chap on eBay who sells chip sets for 9900/9995 systems - but not for any particular design. Be aware that getting TMSxxxx chips off eBay at the moment can be a problem - there are lots of fakes from China, and a beginner trying to debug a system with a dodgy chip can be easily put off. My designs have been around for a while and I doubt that more than perhaps 6 people have used them.

Not really sure though that Kickstarter is appropriate. There's little/no money in producing this stuff so they tend to be a labour of love in people's spare time. Would throwing Kickstarter money at it make it happen any quicker?

The point of view sometimes put forward is - if you update it like that, will it still be a TI-99/4A? What would be the market for a TI-99/4A that you've condensed down to just half a dozen chips? 'Real' systems are readily available, relatively cheap and reliable. There won't be much software support for new features unless you can achieve some sort of critical mass of users using it. All IMHO.

Almost certainly not. The browser relies on specific tags in the downloaded text to control the formatting. Heatwave won't be using those tags.

www.stuartconnerdownloads.me.uk/getwebipaddress.php?w=domain_name The line is getting rather long now and approaching the 80 character limit. Might look at increasing that to 100 characters.

Does it only happen to those threads moved between forums?

I haven't seen a diagram of the solder points before. Personally I'd find a sharp knife and do some dissection of the rail in-situ (remove the tape covering and check if it is actually two rails in there). Might be a simple fix, then you can fix it back up with some insulation tape. I'm rather mystified why it was necessary to cut the rail in the first place, rather than just running some longer wires to the switch ...?

I'm currently reading a book all about Atari, and the TI-99/4 gets a mention with a little snippet that I haven't heard before (assuming its true). Apparently a TI salesman shows up at Atari one day to demonstrate a fibre optic cable that has a receiver and transmitter moulded on the ends. One of the problems that early home computer companies (Atari, TI, Apple ...) were having was getting FCC approval for the RF modulator. Atari say that they could use the fibre optic cable to electrically isolate the computer from the RF modulator, then they only need to get FCC approval for the modulator and cable - not the computer as well [not sure why, but ...]. But they think that the FCC wouldn't accept that, so they don't pursue it as they want to stay on good terms with the FCC. The TI salesman goes back to the office and tells the TI home computer group about Atari's suggestion. They apparently *do* design an RF modulator with a fibre optic connecting cable and submit to the FCC for approval ... but the FCC reject it.

If you press <Back> when it's prompting for a URL, it should abort the URL input then the I (and other) keys should work. Let me know if that doesn't work.

ISTR that the 'weird rails' are actually two metal strips with a paper separator, and actually carry both +5V and GND. By trimming it, you've possibly created a short circuit and/or you've only jumpered one side of it. I'd give that a careful inspection.

Got my PHP configuration sorted rather faster than I was expecting, so point your TI browser to ... www.stuartconnerdownloads.me.uk/getweather.php?l=paris,fr The location formats supported for the "?l=" query apparently include: -- city,state (assumes US): "seattle,wa", "seattle,washington","new+york,ny" -- city,state,country: "seattle,wa,us", "seattle,washington,us", "toronto,on,ca", "toronto,ontario,ca" -- city,country: "paris,fr", "london,uk" -- zip code: 98109, 90210 The query also returns the name of a library png to illustrate the weather. It is in theory possible to use this to also show a weather icon (formed from character definitions) in the browser page, but as there are over 100 of them, I'm not going to start drawing them!

The weather may be (mostly) text, but there's still a load of other cr*p on the returned page - at least on the page that Google shows me. There are a couple of weather sites that you can send a request and city name to that return the weather details as XML, and it's easy to call this in PHP (passing in a city name specified in the original URL) then create a 'TI page' from the returned XML data. I'll have a play doing this, but first need to try to get my service provider to tweak the PHP configuration to allow PHP to open a remote page.

Thanks Tursi. I was rather hoping that someone might have some example code. Just tried using DSRLNK and the following seems to do the job. PABADR EQU >1D00 VDP RAM address for PAB. PABSP BYTE >01,>16 Program name length and program name for sub-program in Disk Controller DSR that * sets the number of file buffers. LI R0,PABADR Copy PAB for sub-program to VDP RAM. LI R1,PABSP LI R2,2 BLWP @VMBW MOV R0,@>8356 Set up pointer to sub-program name length byte in VDP RAM. LI R0,>0100 Set number of disk files in the byte required by the sub-program. MOVB R0,@>834C BLWP @DSRLNK Call the sub-program. DATA 10

Tursi: I've added the >FF termination byte and it now loads the FAVS file fine in Classic99, but I've just tried on my TI+NanoPEB and it won't work, just like Kevan is finding. I suspect the problem is this: using bitmap mode on the VDP, you've got to have either the 6K pattern generator table or colour table in the upper half of VDP RAM, from >2000 to >37FF. This is overwriting the disk controller reserved area in VDP RAM from >37D8, and messing up the disk access. Although Classic99 reserves this area, it doesn't actually use it for file access - is that correct? Does that sound like a reasonable reason why accessing the FAVS file works in Classic99 but not on real hardware? Anyone: So what I should be able to do to get round this is call the disk controller DSR sub-program that does a "CALL FILES(1)" - this will move the disk controller reserved area in VDP RAM towards the end of VDP RAM, away from the colour table - correct? How do I call the sub-program - through the PAB? Is it just a standard PAB where everything is ignored apart from the name length and name fields? I then call this with BLWP @DSRLNK followed by DATA 10?

Enter the address www.stuartconnerdownloads.me.uk/tichat.php. This will display any current messages. To add a message, append to the end of that URL "?m=<your_text>" (without the quotes or angle brackets, and with underscores instead of spaces). Remember that the max length of the entire URL is 80 characters.

Kevan, try the attached when you have a moment. This works on Classic99. Vol 10 Backup (Internet Browser).zip

Neat idea to colour code them!

Thanks Tursi. Kevan - I think you said earlier that you have a sector editor. Are you able to add an >FF byte after the last line (</99ml>) of the FAVS file and see if that stops it hanging when you press the I key? And/or create a new FAVS file (<99ml><p>TEST</p></99ml>), check it has the >FF byte terminator, and try that?

Hi Tursi, I've attached a file you can look at. This is a new TIDisk I've just created in Win994a (V3.010), and used the E/A Editor on that to write a single DV80 file 'TEST' that contains a single line of X's. I can open this file again on Win994a for editing, view it in DM1000 and so on, and it works as expected. If I open that TIDisk file in Classic99 (a probably rather old version QI372) and try to view it in DM1000, the Debugger says "Corrupted File - Truncating Read", and it reads a load of 0 byte records after the first line. If I open it in the E/A Editor in Classic99, it loads but shows the correct first line then a load of blank lines - it should be a single line only. When testing yesterday, the behaviour of Classic99 seemed to match the real hardware so I suspect that Classic99 is probably correct, and there is something wrong with Win994a. Or of course, it could just be me! zzzTEST.zip

If I try to open the file in the E/A Editor in Classic99, the Debugger says "corrupted file - truncating read" - so looks to be a problem with the file itself. On Win994a it works fine.

(Thanks Tursi) I've uploaded a new disk image with a fix that should get round the problem. Try it again Kevan when you get the chance. Remember if copying the files between disks that the program looks for the FAVS file on the disk with the disk name BROWSER. It should show an error message on the bottom line of the screen (and not crash) if it can't find it. And if anyone knows why the real hardware and Classic99 don't like a DV80 file written by Win994a, I'd like to hear.

How does the disk DSR detect the end of a variable record length file? Looking at Classic99, the browser is reading the lines from the FAVS file correctly but then sits there reading 0 byte records for evermore. The FAVS file was created using the E/A Editor in Win994a, and I'm not convinced that is writing the file correctly (vaguely seem to recall a problem I had with it before), although within Win994a it works fine.

OK, it seems to work properly on Win994a which I used for testing, but doesn't on an actual machine or in Classic99. *** Classic99 question *** - if I 'insert' a disk as a disk image, I can read files, rename files, but I don't seem able to save files to the disk - is that correct? Get I/O error 3. If I use a disk in FIAD format, how do I name that disk so I can access it by disk name, e.g. DSK.BROWSER.xxx? I've had a quick look at the manual, but can't see anything that helps.

Is that with the default FAVS file on the disk? Try mounting the disk as DSK3 - if you haven't already - and see if that makes any difference. I've only tested it on the Win994A simulator; I'll try and set up the real thing tonight and test on that as well.