From mboxrd@z Thu Jan 1 00:00:00 1970 From: Bernard Pidoux Subject: Re: BUGs into libax25 Date: Mon, 02 Jun 2008 18:28:41 +0200 Message-ID: <48441FB9.6000405@ccr.jussieu.fr> References: <1K36Sn-0OkOu00@fwd26.aul.t-online.de> <20080602104149.GS3700@mea-ext.zmailer.org> Mime-Version: 1.0 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: In-Reply-To: <20080602104149.GS3700@mea-ext.zmailer.org> Sender: linux-hams-owner@vger.kernel.org List-ID: Content-Type: text/plain; charset="iso-8859-1"; format="flowed" To: Matti Aarnio Cc: DL5DI , linux-hams@vger.kernel.org Hi, Sorry, I wrongly wrote 6 letters callsigns where I wanted to say 6=20 characters. Not the same of course. My typo leaded to an interesting discussion. Somedays I dreamt that we could increase the SSID number up to 31. But the necessary bytes are already taken by DAMA protocol and EAX25 fl= ags. Back to the bug. With previous loop limit (<6) into the code, a callsign like KD4YAL-8=20 would not be correctly translated into shifted ASCII AX25 code. Actually, with this 6 characters callsign, SSID was truncated and an=20 error message sent complaining about SSID not being correctly entered. I had seen it very often in FPAC. I am glad I found the error while=20 investigating on other bugs. 73 de Bernard, f6bvp Matti Aarnio a =E9crit : > On Mon, Jun 02, 2008 at 11:36:25AM +0200, DL5DI wrote: > =20 >> This special event station callsign that we sometimes find like >> DQ2008ANYEVENT will never work in packet radio. >> The protocol header of ax25 frames only allows 6 bytes of callsign a= nd 1 >> halfbyte of SSID. >> Not only for the destiantion, also for all digipeaters inbetween. >> We would need to change the whole protocol and all digipeaters and s= tations >> using it worldwide to get that covered. >> I expect that packet radio will be dead before that happens. :-) >> =20 > > Arbitrary long callsigns... Only OSI stack can handle that, and eve= n it > has some upper limits on address lengths. > > On IPv6 the addresses are 16 bytes long, and low 8 bytes (64 bits) of= them > are reserved for "link local" use. Considering that.. one can enc= ode > any callsign with 38 characters, add couple codes for things like > "no character here", and "next is ssid", that would be 40 codes. > Using 6 bits per character one could encode 10 characters + 4 bits on= IPv6- > address "local half". With arithmetic encoding (see APRS mic-e) mayb= e > 17 code bytes. > > Would that be AX.25 ? Definitely not. That would be IPv6. > > On IPv6 the basic assumption is that link local area is /64 size slic= e, > and thus "AX.25 space" would be something smaller.. like /32. > > How could we implement smooth transition from AX.25 to IPv6 on data l= inks ? > Very good question... old legacy junk will want to live a long time. > (we are way too willing to recycle obsolete hardware...) > > =20 >> 73 de Hans, DL5DI >> =20 > > 73 de Matti, OH2MQK > -- > To unsubscribe from this list: send the line "unsubscribe linux-hams"= in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html > > > =20 -- To unsubscribe from this list: send the line "unsubscribe linux-hams" i= n the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html