I don't know who will toss cookies if you try to send a message that is larger the the defined size. The sender may behave rationally and reject the premise, or the receiver could overrun it's buffer and crash. YMMV. MS
Alban Hessler wrote: > Most of the radio chips have a 128 bytes buffer, which means that the > whole packet cannot be bigger than 128 bytes, overhead included. > > So, if you want to send a large packet, you need a transport layer > which will fragment it before sending it over the air. AFAIK, there is > no core tinyos library doing that, but there have been contributions > from universities doing that. So your 256 bytes datagram would > actually be 3 or more packets over the air. > > Also, the DATA_LENGTH cannot be 128, but has to be less, because of > the overhead of the layers below your app, look on the mailing list > for the maximum value. > > Remember that you will experience larger packet loss if you increase > the payload size, and there is a point where it's better to send two > small packets than one big. > > HTH, > > Alban > > 2009/5/19 Nahr Elk <[email protected]>: >> Hi! >> Ok, so the best solution in my case (the worst case for RAM) is by compiling >> my app with the largest DATA_LENGTH (128) and then it will change the data >> payload size if necessary. >> >> But, please just a clarification: >> >> Suppose that I kept DATA_LENGTH to the default value and the payload size >> will not exceed this value >> However at the Send command level I do something like that : >> send(0, MyMsg, 256); >> >> Could this be done?? >> >> >> Thank you! >> >> Best regards, >> Nahr Elk >> >> 2009/5/19 Razvan Musaloiu-E. <[email protected]> >>> Hi! >>> >>> On Mon, 18 May 2009, Michael Schippling wrote: >>> >>>> Unless Things Have Changed(TM) in T2, only the length number of bytes >>>> is transmitted -- at least over the radio -- so you're not losing >>>> bandwidth with a larger DATA_LENGTH. You do waste RAM because buffers >>>> are allocated for the max size, which is one reason it's a compile >>>> time setting, but that may not matter in your system. >>>> >>> Thinks haven't changed so what Michael said is correct. The only penalty >>> is in RAM. The packets in the air have variable length based on the length >>> specified in AMSend.send: >>> >>> command error_t send(am_addr_t addr, message_t* msg, uint8_t >>> len); >>> >>> -- >>> Razvan ME >>> >>>> Nahr Elk wrote: >>>>> Hi, >>>>> Thank you Razvan, >>>>> I would like to change the data payload size dynamically according to my >>>>> application exigences. >>>>> And I am not so convinced that setPayloadLength(message_t* msg, uint8_t >>>>> len) is the right solution >>>>> >>>>> Best regards, >>>>> Nahr Elk >>>>> >>>>> 2009/5/18 Razvan Musaloiu-E. <[email protected]> >>>>> >>>>> Hi! >>>>> >>>>> >>>>> On Sun, 17 May 2009, Nahr Elk wrote: >>>>> >>>>> Hi >>>>> Is there any way to change DTOSH_DATA_LENGTH at run time instead >>>>> of at >>>>> compile time ? :) >>>>> >>>>> >>>>> No. TOSH_DATA_LENGTH is used to allocate the size of the data field >>>>> from message_t (types/message.h) so you cannot alter it at during >>>>> runtime. May I ask why do you want to do this? >>>>> >>>>> -- >>>>> Razvan ME >>>>> >>>>> >>>>> >> >> _______________________________________________ >> Tinyos-help mailing list >> [email protected] >> https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help >> > > _______________________________________________ > Tinyos-help mailing list > [email protected] > https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help _______________________________________________ Tinyos-help mailing list [email protected] https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
