> Unless the card happens to contain a dsp, or interface chip whose > specifications are public you might as well give up.
True. Especially with the DMCA. > This is a similar problem to having an ethernet card and a driver, but > expecting a different card with a different chip on it to just work with > your driver. (Obviously you totally missed out on this sort of fun in the > early days of Linux/FreeBSD) Nope, I remember Yggdrisl Linux, Linux kernel .97, and FreeBSD 2.2.1. > Having hardware I would say is not even half the battle - whether it is > commercial or custom, often the designs are based closely on the > manufacturer's app notes, so designs using the same core chip are often > interchangable. Having software that makes the hardware work for your > application is the hard part since often there is no reference design at > all for this from the manufacturer, let alone one that will work with Linux > or FreeBSD. Right, but why reinvent the wheel. If you need a card to do something and there is something on the market, you can write drivers for it and use it. There is a fast ethernet card for the Silicon Graphics Indigo^2 that is made by Phobos. Wait, actually it is an off the shelf 3com EISA fast ethernet card with Phobos drivers, and a reprogrammed identifier... because SGI boxes aren't as popular, you end up paying a couple grand for the driver. > Over the summer Atheros (makes the radio modules in the dlink and linksys > wireless stuff) took the groundbreaking step to release a "sort of" open > source driver for their hardware, but this is not the norm at all. Take > another example of ATI vs NVidia and compare the driver availability. Yea I was looking at using the dlink USB radio tuner on a different project of mine. Speaking of, we are running 10 PCI sound cards in a single FreeBSD machine... people on here have mentioned this... we use a PICMG SBC on a backplane with 19 PCI slots. Got two fo them for $125 off of ebay. So there are solutions for tons of PCI FXO cards, but a channel bank would be cleaner. pictures: http://users.757.org/~ethan/pics/geek/soundcard_champion/ > So in summary, unless you happen to have some pipeline of information > coming from a winmodem manufacturer, making it into a linmodem let alone > another specialized telephony device is anything but trivial, unless it > happens to be based on exactly the same chips and reference designs as the > software you have is for. Hmmmm _______________________________________________ Asterisk-Users mailing list [EMAIL PROTECTED] http://lists.digium.com/mailman/listinfo/asterisk-users