Re: [Ql-Users] Developer assistance request.
Any other takers? The proposed spec of the machine has improved a little and it will be a lot faster than a SGC, with 64x the memory, too... What would you do for developer access to new hardware? :) Dave On Wed, Jan 22, 2014 at 10:21 PM, Alexandre Souza alexandre.tabaj...@gmail.com wrote: May I add 2 cents to the subject? With the W5300 on-chip implementation, you depend on one specific manufacturer. If the TCP/IP stack runs on the QL side, one is more flexible to use different hardware. In my case, the NE2000 or Realtek on the Q40 and Q60, but the CP2200 on the Q68. Even TCP/IP over SER instead of ethernet is possible. Wizchip devices are around for about 10 years. I've used that with lots of products, with no troubles and cheap cost (facing a stronger processor to put an entire tcp stack into). Of course, anything is possible. But with wizchip, you need only a SPI connection and no stack whatsoever. In my case, it means a socket library, since existing internet applications were written in C, not S*BASIC. An extension which interfaces to S*BASIC is doable, just had low priority for me. I vote for basic, since wizchip takes most of the work. I'd even expand the capabilities of the old (and outdated) MCS-48 processor on QL, and put an AVR on its place, creating expansions like a dedicated I2C/SPI/TWI interface to the outside world. Maybe I need to get more into QL workings... :o) As always, I can not promise when I can finish something. So if you have already implemented a lot of things around the W5300 and plan to finish soon, don't let my info regarding the CP2200 hinder you. Sure! By the way, does the W5300 include a unique ethernet address? IEEE registration can become an expensive issue if you need to do it yourself. Not that expensive. But the MAC number can always be reused from an old (and broken) ethernet card, or like. ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm -- Dave Park Sandy Electronics, LLC d...@sinclairql.com ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm
Re: [Ql-Users] Developer assistance request.
May I add 2 cents to the subject? With the W5300 on-chip implementation, you depend on one specific manufacturer. If the TCP/IP stack runs on the QL side, one is more flexible to use different hardware. In my case, the NE2000 or Realtek on the Q40 and Q60, but the CP2200 on the Q68. Even TCP/IP over SER instead of ethernet is possible. Wizchip devices are around for about 10 years. I've used that with lots of products, with no troubles and cheap cost (facing a stronger processor to put an entire tcp stack into). Of course, anything is possible. But with wizchip, you need only a SPI connection and no stack whatsoever. In my case, it means a socket library, since existing internet applications were written in C, not S*BASIC. An extension which interfaces to S*BASIC is doable, just had low priority for me. I vote for basic, since wizchip takes most of the work. I'd even expand the capabilities of the old (and outdated) MCS-48 processor on QL, and put an AVR on its place, creating expansions like a dedicated I2C/SPI/TWI interface to the outside world. Maybe I need to get more into QL workings... :o) As always, I can not promise when I can finish something. So if you have already implemented a lot of things around the W5300 and plan to finish soon, don't let my info regarding the CP2200 hinder you. Sure! By the way, does the W5300 include a unique ethernet address? IEEE registration can become an expensive issue if you need to do it yourself. Not that expensive. But the MAC number can always be reused from an old (and broken) ethernet card, or like. ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm
Re: [Ql-Users] Developer assistance request.
On 20 Jan 2014 at 20:05, Dave Park wrote: We are also looking for anyone interested in working on development of an ethernet driver. We're looking closely at the WizNet 5300. The datasheet is here: http://www.wiznet.co.kr/UpLoad_Files/ReferenceFiles/W5300_DS_V128E.pdf For your information: The Q68 already has a prototyped and finished PCB using the CP2200, communication was successfully tested on packet level. http://www.silabs.com/Support%20Documents/TechnicalDocs/CP2200.pdf The CP2200 has only 48 pins and a simple 8-bit bus suitable for QL purposes. It is 5V tolerant. Using the CP2200 would allow similar or even identical drivers. Peter ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm
Re: [Ql-Users] Developer assistance request.
Hi Peter, In all cases, I would like to go for compatibility with other systems. Drivers are a huge obstacle, so any kind of progress is a head start. I did a quick read of the datasheet, and I can see that this device does have a couple of advantages over the WS5300 - the onboard MAC eliminates a serial EEPROM. Obviously, we're at the very earliest stages of development, so it's quite easy to change direction at this time. Are you interested in open sourcing the work you've done? Dave On Tue, Jan 21, 2014 at 3:12 AM, Peter pg...@q40.de wrote: On 20 Jan 2014 at 20:05, Dave Park wrote: We are also looking for anyone interested in working on development of an ethernet driver. We're looking closely at the WizNet 5300. The datasheet is here: http://www.wiznet.co.kr/UpLoad_Files/ReferenceFiles/W5300_DS_V128E.pdf For your information: The Q68 already has a prototyped and finished PCB using the CP2200, communication was successfully tested on packet level. http://www.silabs.com/Support%20Documents/TechnicalDocs/CP2200.pdf The CP2200 has only 48 pins and a simple 8-bit bus suitable for QL purposes. It is 5V tolerant. Using the CP2200 would allow similar or even identical drivers. Peter ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm -- Dave Park Sandy Electronics, LLC d...@sinclairql.com ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm
Re: [Ql-Users] Developer assistance request.
Nice thing about the W5300 is that it has a TCP/IP stack implemented on the chip which makes the driver *much* simpler to implement. One interesting question regarding this whole conversation is: what is meant by a driver here? Does driver in this context mean a SuperBasic device interface? I.e. something along the lines of open #3,tcp_www.bbc.co.uk_80 Or does it mean a POSIX compliant socket library? Or something else? On Tue, Jan 21, 2014 at 3:22 PM, Dave Park d...@sinclairql.com wrote: Hi Peter, In all cases, I would like to go for compatibility with other systems. Drivers are a huge obstacle, so any kind of progress is a head start. I did a quick read of the datasheet, and I can see that this device does have a couple of advantages over the WS5300 - the onboard MAC eliminates a serial EEPROM. Obviously, we're at the very earliest stages of development, so it's quite easy to change direction at this time. Are you interested in open sourcing the work you've done? Dave On Tue, Jan 21, 2014 at 3:12 AM, Peter pg...@q40.de wrote: On 20 Jan 2014 at 20:05, Dave Park wrote: We are also looking for anyone interested in working on development of an ethernet driver. We're looking closely at the WizNet 5300. The datasheet is here: http://www.wiznet.co.kr/UpLoad_Files/ReferenceFiles/W5300_DS_V128E.pdf For your information: The Q68 already has a prototyped and finished PCB using the CP2200, communication was successfully tested on packet level. http://www.silabs.com/Support%20Documents/TechnicalDocs/CP2200.pdf The CP2200 has only 48 pins and a simple 8-bit bus suitable for QL purposes. It is 5V tolerant. Using the CP2200 would allow similar or even identical drivers. Peter ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm -- Dave Park Sandy Electronics, LLC d...@sinclairql.com ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm
Re: [Ql-Users] Developer assistance request.
Good questions, Petri. Given the structure of C, SuperBASIC and Minerva/QDOS in general, I think it's important to have at least a functional subset of the POSIX socket API. However, while that would bring functionality it would be strictly in the realm of the experienced programmer. The great thing about the QL is it's a wonderful tinkerer's machine. Being accessible from C and SuperBASIC would make use of ethernet not just for the elite, but a widespread pass time. We'd see a glut of multi-user games, web browsers, mail programs, web servers all easily accessible and modifiable. At that point, anyone can contribute. The addition of ethernet to the QL is as much about community development as it is about functionality of hardware. This is so important that if we reach a point where there's a workable driver I would build and sell ethernet cards for black box QLs at a loss, just to get them into as wide use as possible. However, until there's a meaningful expansion ROM image, it's a no-go. On Tue, Jan 21, 2014 at 7:39 AM, Petri Pellinen p...@iki.fi wrote: Nice thing about the W5300 is that it has a TCP/IP stack implemented on the chip which makes the driver *much* simpler to implement. One interesting question regarding this whole conversation is: what is meant by a driver here? Does driver in this context mean a SuperBasic device interface? I.e. something along the lines of open #3,tcp_www.bbc.co.uk_80 Or does it mean a POSIX compliant socket library? Or something else? On Tue, Jan 21, 2014 at 3:22 PM, Dave Park d...@sinclairql.com wrote: Hi Peter, In all cases, I would like to go for compatibility with other systems. Drivers are a huge obstacle, so any kind of progress is a head start. I did a quick read of the datasheet, and I can see that this device does have a couple of advantages over the WS5300 - the onboard MAC eliminates a serial EEPROM. Obviously, we're at the very earliest stages of development, so it's quite easy to change direction at this time. Are you interested in open sourcing the work you've done? Dave On Tue, Jan 21, 2014 at 3:12 AM, Peter pg...@q40.de wrote: On 20 Jan 2014 at 20:05, Dave Park wrote: We are also looking for anyone interested in working on development of an ethernet driver. We're looking closely at the WizNet 5300. The datasheet is here: http://www.wiznet.co.kr/UpLoad_Files/ReferenceFiles/W5300_DS_V128E.pdf For your information: The Q68 already has a prototyped and finished PCB using the CP2200, communication was successfully tested on packet level. http://www.silabs.com/Support%20Documents/TechnicalDocs/CP2200.pdf The CP2200 has only 48 pins and a simple 8-bit bus suitable for QL purposes. It is 5V tolerant. Using the CP2200 would allow similar or even identical drivers. Peter ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm -- Dave Park Sandy Electronics, LLC d...@sinclairql.com ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm -- Dave Park Sandy Electronics, LLC d...@sinclairql.com ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm
Re: [Ql-Users] Developer assistance request.
Richard Zidlicky's document QDOS TCP/IP and socket functionality at http://www.dilwyn.me.uk/docs/manuals/socket.html seems to propose that most of the POSIX-type functionality be implemented as an extension to standard Trap #3 calls with D0 values $50-$7B inclusive mapped to socket API calls. Does anyone here on the mailing list know of possible conflicts with other drivers if this proposed scheme was used? On Tue, Jan 21, 2014 at 4:05 PM, Dave Park d...@sinclairql.com wrote: Good questions, Petri. Given the structure of C, SuperBASIC and Minerva/QDOS in general, I think it's important to have at least a functional subset of the POSIX socket API. However, while that would bring functionality it would be strictly in the realm of the experienced programmer. The great thing about the QL is it's a wonderful tinkerer's machine. Being accessible from C and SuperBASIC would make use of ethernet not just for the elite, but a widespread pass time. We'd see a glut of multi-user games, web browsers, mail programs, web servers all easily accessible and modifiable. At that point, anyone can contribute. The addition of ethernet to the QL is as much about community development as it is about functionality of hardware. This is so important that if we reach a point where there's a workable driver I would build and sell ethernet cards for black box QLs at a loss, just to get them into as wide use as possible. However, until there's a meaningful expansion ROM image, it's a no-go. On Tue, Jan 21, 2014 at 7:39 AM, Petri Pellinen p...@iki.fi wrote: Nice thing about the W5300 is that it has a TCP/IP stack implemented on the chip which makes the driver *much* simpler to implement. One interesting question regarding this whole conversation is: what is meant by a driver here? Does driver in this context mean a SuperBasic device interface? I.e. something along the lines of open #3,tcp_www.bbc.co.uk_80 Or does it mean a POSIX compliant socket library? Or something else? On Tue, Jan 21, 2014 at 3:22 PM, Dave Park d...@sinclairql.com wrote: Hi Peter, In all cases, I would like to go for compatibility with other systems. Drivers are a huge obstacle, so any kind of progress is a head start. I did a quick read of the datasheet, and I can see that this device does have a couple of advantages over the WS5300 - the onboard MAC eliminates a serial EEPROM. Obviously, we're at the very earliest stages of development, so it's quite easy to change direction at this time. Are you interested in open sourcing the work you've done? Dave On Tue, Jan 21, 2014 at 3:12 AM, Peter pg...@q40.de wrote: On 20 Jan 2014 at 20:05, Dave Park wrote: We are also looking for anyone interested in working on development of an ethernet driver. We're looking closely at the WizNet 5300. The datasheet is here: http://www.wiznet.co.kr/UpLoad_Files/ReferenceFiles/W5300_DS_V128E.pdf For your information: The Q68 already has a prototyped and finished PCB using the CP2200, communication was successfully tested on packet level. http://www.silabs.com/Support%20Documents/TechnicalDocs/CP2200.pdf The CP2200 has only 48 pins and a simple 8-bit bus suitable for QL purposes. It is 5V tolerant. Using the CP2200 would allow similar or even identical drivers. Peter ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm -- Dave Park Sandy Electronics, LLC d...@sinclairql.com ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm -- Dave Park Sandy Electronics, LLC d...@sinclairql.com ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm
Re: [Ql-Users] Developer assistance request.
On 21 Jan 2014, at 15:41, Petri Pellinen p...@iki.fi wrote: Richard Zidlicky's document QDOS TCP/IP and socket functionality at http://www.dilwyn.me.uk/docs/manuals/socket.html seems to propose that most of the POSIX-type functionality be implemented as an extension to standard Trap #3 calls with D0 values $50-$7B inclusive mapped to socket API calls. Does anyone here on the mailing list know of possible conflicts with other drivers if this proposed scheme was used? Trap#3 $50 to $5E, $60, $61 and $68 are used for GD2 colours. George ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm
Re: [Ql-Users] Developer assistance request.
Petri Pellinen wrote: Nice thing about the W5300 is that it has a TCP/IP stack implemented on the chip which makes the driver *much* simpler to implement. I had a working native TCP/IP stack plus applications a decade ago, so in my case, the easier way is trying to reuse that. With the W5300 on-chip implementation, you depend on one specific manufacturer. If the TCP/IP stack runs on the QL side, one is more flexible to use different hardware. In my case, the NE2000 or Realtek on the Q40 and Q60, but the CP2200 on the Q68. Even TCP/IP over SER instead of ethernet is possible. However, with a slow CPU, your approach has a performance advantage. One interesting question regarding this whole conversation is: what is meant by a driver here? Does driver in this context mean a SuperBasic device interface? I.e. something along the lines of open #3,tcp_www.bbc.co.uk_80 Or does it mean a POSIX compliant socket library? Or something else? In my case, it means a socket library, since existing internet applications were written in C, not S*BASIC. An extension which interfaces to S*BASIC is doable, just had low priority for me. Should I run out of time, I could also imagine something else. As always, I can not promise when I can finish something. So if you have already implemented a lot of things around the W5300 and plan to finish soon, don't let my info regarding the CP2200 hinder you. Apart from compatibility to the Q68, the CP2200 may be a nice chip for QL use in general. By the way, does the W5300 include a unique ethernet address? IEEE registration can become an expensive issue if you need to do it yourself. Peter ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm
Re: [Ql-Users] Developer assistance request.
The WS5300 is not recommended for new designs, so I would need to buy a lifetime supply, yes, which is a drawback. It, like the CP2200 can be configured to use any MAC address but does have one configured in its internal NVRAM, whereas the WS5300 requires an external EEPROM. Given the cost difference, availability and functionality, it does look like the CP2200 has many advantages. I have forwarded the details to Nasta so he can read the datasheet and make a decision from the hardware side. It will be several months before he works on that part of the design, anyway. So, amend the first post! Anyone interested in driver development for the CP2200 *or* the WS5300... On Tue, Jan 21, 2014 at 10:12 AM, Peter pg...@q40.de wrote: Petri Pellinen wrote: Nice thing about the W5300 is that it has a TCP/IP stack implemented on the chip which makes the driver *much* simpler to implement. I had a working native TCP/IP stack plus applications a decade ago, so in my case, the easier way is trying to reuse that. With the W5300 on-chip implementation, you depend on one specific manufacturer. If the TCP/IP stack runs on the QL side, one is more flexible to use different hardware. In my case, the NE2000 or Realtek on the Q40 and Q60, but the CP2200 on the Q68. Even TCP/IP over SER instead of ethernet is possible. However, with a slow CPU, your approach has a performance advantage. One interesting question regarding this whole conversation is: what is meant by a driver here? Does driver in this context mean a SuperBasic device interface? I.e. something along the lines of open #3,tcp_www.bbc.co.uk_80 Or does it mean a POSIX compliant socket library? Or something else? In my case, it means a socket library, since existing internet applications were written in C, not S*BASIC. An extension which interfaces to S*BASIC is doable, just had low priority for me. Should I run out of time, I could also imagine something else. As always, I can not promise when I can finish something. So if you have already implemented a lot of things around the W5300 and plan to finish soon, don't let my info regarding the CP2200 hinder you. Apart from compatibility to the Q68, the CP2200 may be a nice chip for QL use in general. By the way, does the W5300 include a unique ethernet address? IEEE registration can become an expensive issue if you need to do it yourself. Peter ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm -- Dave Park Sandy Electronics, LLC d...@sinclairql.com ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm
Re: [Ql-Users] Developer assistance request.
Thank you George, that's an interesting piece of information. Kind regards, Petri On Tue, Jan 21, 2014 at 5:51 PM, George Gwilt grggw...@gmail.com wrote: On 21 Jan 2014, at 15:41, Petri Pellinen p...@iki.fi wrote: Richard Zidlicky's document QDOS TCP/IP and socket functionality at http://www.dilwyn.me.uk/docs/manuals/socket.html seems to propose that most of the POSIX-type functionality be implemented as an extension to standard Trap #3 calls with D0 values $50-$7B inclusive mapped to socket API calls. Does anyone here on the mailing list know of possible conflicts with other drivers if this proposed scheme was used? Trap#3 $50 to $5E, $60, $61 and $68 are used for GD2 colours. George ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm
Re: [Ql-Users] Developer assistance request.
On Tue, Jan 21, 2014 at 6:12 PM, Peter pg...@q40.de wrote: Petri Pellinen wrote: Nice thing about the W5300 is that it has a TCP/IP stack implemented on the chip which makes the driver *much* simpler to implement. I had a working native TCP/IP stack plus applications a decade ago, so in my case, the easier way is trying to reuse that. That's very impressive. It's a shame that the solution was not more widespread. As always, I can not promise when I can finish something. So if you have already implemented a lot of things around the W5300 and plan to finish soon, don't let my info regarding the CP2200 hinder you. Ok... Based on that comment it seems that for clarity's sake I have to state the following explicitly: I am *not* associated in any way with Dave's project. I had two motivations for posting my reply to Dave's message: 1) To point out a reason I personally chose the W5300 chip for a home project - this difference between the discussed chips was not stated in earlier emails so I thought it might add information to the discussion 2) To express my general curiosity about what a driver means in this case since this was not 100% clear to me. Also, I am NOT advocating the use of any specific chip for any purpose whatsoever. And following your cue, I'm also not making any promises or offering any implementations around any chipset. ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm
Re: [Ql-Users] Developer assistance request.
In this case, it would be interesting to know how QPC can implement both TCP/IP and GD2 high-color? My QPC manual says: This manual won't go into the details of the interface, however, as the interface is mostly compatible to the uQLx implementation Maybe Marcel can shed some light on that? Regards, Tobias -Original-Nachricht- Betreff: Re: [Ql-Users] Developer assistance request. Datum: Tue, 21 Jan 2014 16:51:42 +0100 Von: George Gwilt grggw...@gmail.com An: ql-us...@q-v-d.com On 21 Jan 2014, at 15:41, Petri Pellinen p...@iki.fi wrote: Richard Zidlicky's document QDOS TCP/IP and socket functionality at http://www.dilwyn.me.uk/docs/manuals/socket.html seems to propose that most of the POSIX-type functionality be implemented as an extension to standard Trap #3 calls with D0 values $50-$7B inclusive mapped to socket API calls. Does anyone here on the mailing list know of possible conflicts with other drivers if this proposed scheme was used? Trap#3 $50 to $5E, $60, $61 and $68 are used for GD2 colours. George ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm
Re: [Ql-Users] Developer assistance request.
On second thought, maybe not. Obviosly no one will want to set a color on a TCP channel nor will anyone want to send a packet to a high-res screen channel. -The overlap doesn't hurt. Tobias -Original-Nachricht- Betreff: Re: [Ql-Users] Developer assistance request. Datum: Tue, 21 Jan 2014 18:37:49 +0100 Von: tobias.froesc...@t-online.de tobias.froesc...@t-online.de An: ql-us...@q-v-d.com ql-us...@q-v-d.com In this case, it would be interesting to know how QPC can implement both TCP/IP and GD2 high-color? My QPC manual says: This manual won't go into the details of the interface, however, as the interface is mostly compatible to the uQLx implementation Maybe Marcel can shed some light on that? Regards, Tobias -Original-Nachricht- Betreff: Re: [Ql-Users] Developer assistance request. Datum: Tue, 21 Jan 2014 16:51:42 +0100 Von: George Gwilt grggw...@gmail.com An: ql-us...@q-v-d.com On 21 Jan 2014, at 15:41, Petri Pellinen p...@iki.fi wrote: Richard Zidlicky's document QDOS TCP/IP and socket functionality at http://www.dilwyn.me.uk/docs/manuals/socket.html seems to propose that most of the POSIX-type functionality be implemented as an extension to standard Trap #3 calls with D0 values $50-$7B inclusive mapped to socket API calls. Does anyone here on the mailing list know of possible conflicts with other drivers if this proposed scheme was used? Trap#3 $50 to $5E, $60, $61 and $68 are used for GD2 colours. George ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm
Re: [Ql-Users] Developer assistance request.
Hi Tobias, that's entirely right, thanks for pointing that out! On Tue, Jan 21, 2014 at 7:48 PM, tobias.froesc...@t-online.de tobias.froesc...@t-online.de wrote: On second thought, maybe not. Obviosly no one will want to set a color on a TCP channel nor will anyone want to send a packet to a high-res screen channel. -The overlap doesn't hurt. Tobias -Original-Nachricht- Betreff: Re: [Ql-Users] Developer assistance request. Datum: Tue, 21 Jan 2014 18:37:49 +0100 Von: tobias.froesc...@t-online.de tobias.froesc...@t-online.de An: ql-us...@q-v-d.com ql-us...@q-v-d.com In this case, it would be interesting to know how QPC can implement both TCP/IP and GD2 high-color? My QPC manual says: This manual won't go into the details of the interface, however, as the interface is mostly compatible to the uQLx implementation Maybe Marcel can shed some light on that? Regards, Tobias -Original-Nachricht- Betreff: Re: [Ql-Users] Developer assistance request. Datum: Tue, 21 Jan 2014 16:51:42 +0100 Von: George Gwilt grggw...@gmail.com An: ql-us...@q-v-d.com On 21 Jan 2014, at 15:41, Petri Pellinen p...@iki.fi wrote: Richard Zidlicky's document QDOS TCP/IP and socket functionality at http://www.dilwyn.me.uk/docs/manuals/socket.html seems to propose that most of the POSIX-type functionality be implemented as an extension to standard Trap #3 calls with D0 values $50-$7B inclusive mapped to socket API calls. Does anyone here on the mailing list know of possible conflicts with other drivers if this proposed scheme was used? Trap#3 $50 to $5E, $60, $61 and $68 are used for GD2 colours. George ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm
Re: [Ql-Users] Developer assistance request.
You never know what Crazy Dave might do! On Tue, Jan 21, 2014 at 12:28 PM, Petri Pellinen p...@iki.fi wrote: Hi Tobias, that's entirely right, thanks for pointing that out! On Tue, Jan 21, 2014 at 7:48 PM, tobias.froesc...@t-online.de tobias.froesc...@t-online.de wrote: On second thought, maybe not. Obviosly no one will want to set a color on a TCP channel nor will anyone want to send a packet to a high-res screen channel. -The overlap doesn't hurt. Tobias -Original-Nachricht- Betreff: Re: [Ql-Users] Developer assistance request. Datum: Tue, 21 Jan 2014 18:37:49 +0100 Von: tobias.froesc...@t-online.de tobias.froesc...@t-online.de An: ql-us...@q-v-d.com ql-us...@q-v-d.com In this case, it would be interesting to know how QPC can implement both TCP/IP and GD2 high-color? My QPC manual says: This manual won't go into the details of the interface, however, as the interface is mostly compatible to the uQLx implementation Maybe Marcel can shed some light on that? Regards, Tobias -Original-Nachricht- Betreff: Re: [Ql-Users] Developer assistance request. Datum: Tue, 21 Jan 2014 16:51:42 +0100 Von: George Gwilt grggw...@gmail.com An: ql-us...@q-v-d.com On 21 Jan 2014, at 15:41, Petri Pellinen p...@iki.fi wrote: Richard Zidlicky's document QDOS TCP/IP and socket functionality at http://www.dilwyn.me.uk/docs/manuals/socket.html seems to propose that most of the POSIX-type functionality be implemented as an extension to standard Trap #3 calls with D0 values $50-$7B inclusive mapped to socket API calls. Does anyone here on the mailing list know of possible conflicts with other drivers if this proposed scheme was used? Trap#3 $50 to $5E, $60, $61 and $68 are used for GD2 colours. George ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm -- Dave Park Sandy Electronics, LLC d...@sinclairql.com ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm
[Ql-Users] Developer assistance request.
Hi all, We are looking for someone with experience developing custom yet highly compatible versions of Minerva. The revised version of Minerva may have features like a very large amount of physical RAM, and exist on system-writable flash memory. This provides an exciting opportunity for Minerva to undergo active development on a new hardware platform, where it can be updated using a software update system that needs to be safe, secure and reliable. We are also looking for anyone interested in working on development of an ethernet driver. We're looking closely at the WizNet 5300. The datasheet is here: http://www.wiznet.co.kr/UpLoad_Files/ReferenceFiles/W5300_DS_V128E.pdf We would like the driver to be well integrated with Minerva. We would also like to discuss the option of some BASIC extensions to make the new interface as accessible as possible. ROM size no limit! If anyone is interested and would like to know more, please contact me privately by email. Compensation includes subsidized early access to hardware. -- Dave Park Sandy Electronics, LLC d...@sinclairql.com ___ QL-Users Mailing List http://www.q-v-d.demon.co.uk/smsqe.htm
