Amstrad CPC WiFi

[number-one]

A friend assembled tcp.s for me.


Always these strange characters same as symtel.
(see the picture attached)

[Duke]

A friend assembled tcp.s for me.


Always these strange characters same as symtel.
(see the picture attached)

Maybe it's the character encoding of the telnet daemon that isn't plain ascii?
If possible you could try to run wireshark and capture the packets going to the M4, by setting filter HOST <ip num of M4/CPC>

[arnoldemu]

Maybe it's the character encoding of the telnet daemon that isn't plain ascii?
If possible you could try to run wireshark and capture the packets going to the M4, by setting filter HOST <ip num of M4/CPC>

Looking at the bytes it's 'telnet negotiation'.

http://www.faqs.org/rfcs/rfc855.html
https://tools.ietf.org/html/rfc854

quite a lot of rfcs explain various additions:
https://tools.ietf.org/html/rfc652
https://tools.ietf.org/html/rfc1073

in relation to some other telnet server:

http://mud-dev.wikidot.com/telnet:negotiation

ff,fd is the first two bytes. It's not a BOM and it's not UTF-8 encoding, so looks like telnet negotiation data to me.

symtel needs to be updated to support it or ignore it.

first two bytes are "IAC" and "DO".

[arnoldemu]

Hi, I am experimenting with calling a server rapidly to see if it is viable to cater for multiplayer arcade gaming on the CPC.  At the moment though my server seems to service lots of clients consistently, but only really at about 5 calls per second each client.  For some games this is likely ok, but then I'm wondering... with a CPC's general lower framerate than PCs... a PC that might have 20 calls per second to service a game, would 5 be ok on a CPC?  I am not even certain I can consistently get 5 calls per second from the CPC, but I think it should be possible.  For now i have a very rudimentary synchronous loop that counts missed call opportunities. I have also created a JS POC of the same which gives terrible results on firefox, but good results on chrome.  I am using an OK ISP (not top of the range, but not budget either).  Localhosted WAMP I can consistently get 10 calls per second serviced. but then my actual server starts to degrade beyond that (additional calls with multiple clients at the same time).

(I wasn't sure if this should be posted here, or a separate thread in programming since I am using the M4 from the CPC).

Julian

How much data is transferred for each call?
Are you handling the data on each interrupt on the cpc?
Are you using a central PC server and all the CPCs are clients or is it peer-to-peer?
Is it non-blocking ports and are you using threads to handle the requests?

Networking gaming is involved, packets drop or arrive out of order and there is quite a lot of logic to make sure all the clients agree on the same values. It's easier with local networking because often you don't suffer these problems. If you use a PC as a server it can take all the logic and tell the CPCs the information and all they need to do is pass their inputs to the PC and let it decide what they should do.
The CPC then runs the graphics, plays the sound and accepts the servers word for what is right or not.

[zhulien]

How much data is transferred for each call?
Are you handling the data on each interrupt on the cpc?
Are you using a central PC server and all the CPCs are clients or is it peer-to-peer?
Is it non-blocking ports and are you using threads to handle the requests?

Networking gaming is involved, packets drop or arrive out of order and there is quite a lot of logic to make sure all the clients agree on the same values. It's easier with local networking because often you don't suffer these problems. If you use a PC as a server it can take all the logic and tell the CPCs the information and all they need to do is pass their inputs to the PC and let it decide what they should do.
The CPC then runs the graphics, plays the sound and accepts the servers word for what is right or not.

At this moment brute force from CPC RSX commands, no interrupts.  M4 though ultimately can fetch in the background so I am hoping that it will take almost no CPC time to fetch the statuses from the server, only to play the game itself (actually more like a full-screen protocol of commands so the CPC is like a gfx terminal).  Not peer to peer.  The server is PHP-based but non-blocking.  I am not using Node.js as it is not supported on most hosting providers here at least. MMOL (as many as a server can handle and a CPC I guess )... 4?  At this moment I am experimenting with minimal transmission of data, under 50 bytes back and forth.  Depending on the type of game and type data going, coordination challenges change - I doubt any decent CPC internet game will be an FPS with player movement prediction will be required - but... even if my slow server can handle 4 players at 5 data updates per second, it is not too bad for some types of arcade games.  I'd like my server to handle more than 4 players only though too.  So, after some more CPC side performance tests,  back to the server.  Seems my current ISP caps the requests at 5 per second from Melbourne no matter how many clients (tested up to 4 as a mix of CPC and browser simulating calls).  From Philippines and Indonesia they get exactly 1 call per second - you can simulate my server performance from your location here:  http://8bitology.net/itp/ FPS: frames per second, IPS: Game Loop Iterations per second, CPS: Server Calls per second.


At the minute a lot of assumptions and possibilities, but no definites.  If a hosted server on my budget is too bad for any purpose except in a local city, then... it may only be good in a LAN environment with a PC running WAMP, however... my test with WAMP is worst (in a different way).  I can get 10 calls per second in my LAN with 1 client, but add a second, they both get about 3 to 4 calls per second, so my PC sucks as a server using 'WAMP'.  Of course the server is all PHP code and LAMP likely will work better, but lets see what happens... If it can be hosted that's better as then people can join the lobby and join games with players instead of having to create a LAN session with just CPCers.

[XeNoMoRPH]

Hi , Duke !!! .. any new firmware update in a near future .. ? 

[Duke]

Hi , Duke !!! .. any new firmware update in a near future .. ? 

Sorry, not in the near future. But I will be back

[Gryzor]

Watched @chinnyhill10 's vid last night on this board. I agreed with all he said and I was amazed once again at its magnificence Pity he didn't remember the absolutely great price.

[ComSoft6128]

What vid?

[Gryzor]

[ComSoft6128]

Cheers for that. I recently purchased the M4 but haven't had the time to look at it. This will be useful.

[Gryzor]

Oh, do watch it, it's a great intro to the device!

Sent from my HTC 10 using Tapatalk

[chinnyhill10]

Oh, do watch it, it's a great intro to the device!

Sent from my HTC 10 using Tapatalk

Even if I claim the main chip is an FPGA when it is not and I have now found out there is cart support in Beta which I could have shown. Bah!

[Gryzor]

I still enjoyed the video very much, even though I've used it extensively [emoji3]

Yes, the FPGA comment made me go "hmm?", but it didn't affect much, so...

Sent from my HTC 10 using Tapatalk

[chinnyhill10]

Yes, the FPGA comment made me go "hmm?", but it didn't affect much, so...

Always a danger when I'm dealing with modern stuff. I can poke around inside old machines and generally have a good idea of what is what. New stuff is all magic dust and rainbows to me! 

[pelrun]

Easiest way to tell is: if the chip doesn't say Xilinx, Altera or Lattice, it's not an FPGA (there are a couple of others, but you will probably never see them). Every other big chip is most likely a microcontroller.

[Targhan]

I had success in transfering many SNAs, but for some reason, this one can't. Saved with Winape, "format 2", 64k. I tried format 1, 3 too, the CPC displays "error occurred" every time. If it's possible for you, can you quickly check what is wrong? Thanks a lot!

[Duke]

I had success in transfering many SNAs, but for some reason, this one can't. Saved with Winape, "format 2", 64k. I tried format 1, 3 too, the CPC displays "error occurred" every time. If it's possible for you, can you quickly check what is wrong? Thanks a lot!

There's a bug when using remote .sna start if using > 8.3 filenames. Krusty had that issue recently, it would start from the CPC itself though. Will check it tomorrow, if that's not the issue.

EDIT: tested it. Yes that seems to be the problem. When I issue |sna,"PlayerAT1.sna" from the CPC itself it works. Renaming it to "PlayAT1.sna" and launching it remotely it works too.
So for now, please use 8.3 filenames, when wanting to remote start SNA's. Will fix in next upgrade.

[roudoudou]

Hi Duke
Is there a new batch going on this year?
I'd love to have one precious 

[Duke]

Hi Duke
Is there a new batch going on this year?
I'd love to have one precious 

Sorry I've gotten quite a bit behind (unfortunate events).
I am working my way down the list.  9 more and I get to yours, I will run out of some components soon, but new ones are on their way, so hopefully in about 2 weeks.

[Targhan]

Thanks for the help!!

[Jungsi]

Is there a chance to play Chibi Akuma 2 (four disks) with the M4?

[Duke]

Is there a chance to play Chibi Akuma 2 (four disks) with the M4?

Sure, just extract them all to one directory. IIRC, that's what I did when I played it.

[George Schizas]

This forum post has become a bit awkward, but I don't see any other way to do this...


My WiFi expansion board hasn't able to start for a while (it has worked in the past), and I'm afraid it's something to do with the hardware, most probably the extension port of my 6128. When I move the WiFi board about, sometimes one light turns on, sometimes the other one, sometimes all three, but even in those rare cases, the M4 ROM doesn't load.


Is there something that can be done for this? Both to make it work, and to make it stable?


The hardware version is 2.1, it's using the edge connector, and I think I've put the latest non-beta release of the firmware (certainly a 2.x edition).


Just to be clear: I wouldn't even mind getting a second WiFi expansion board - just the chance to demonstrate a 30 year old computer being able to connect to the Internet just tickles my fancy way too much. I'm also open to having my Amstrad go a preventative repair (the bottom of the mainboard has began to bubble, and I'm sure that a lot of the connectors do have some corrosion on them).

[Duke]

My WiFi expansion board hasn't able to start for a while (it has worked in the past), and I'm afraid it's something to do with the hardware, most probably the extension port of my 6128. When I move the WiFi board about, sometimes one light turns on, sometimes the other one, sometimes all three, but even in those rare cases, the M4 ROM doesn't load.


Is there something that can be done for this? Both to make it work, and to make it stable?

Imo. it all points to bad contacts on your edge connector of the CPC.
The blue led with the text "ON", should be on at all time when the CPC is powered, if it isn't there's a bad contact on the power line of the expansion port and likely many other signals.
So first advice is to clean the expansion port, ie. using an eraser and IPA. In all likelyhood it will work fine after that.
The 2nd led with DD, should only light briefly when the CPC is loading or at boot time.
Anyway try cleaning the connector thoroughly first. And tell me the behaviour of those two leds, if it still fails.

(the bottom of the mainboard has began to bubble, and I'm sure that a lot of the connectors do have some corrosion on them).

That sounds bad, could you take a picture so we can see what it looks like ?