Does anyone know what constraints apply to POST connections in terms of their content length (and/or response length) on the following phones:
6310i, 3650 (or I guess, S60 phones), 7210, 3590.
I'm noticing inconsistent behaviour lately and I'm not sure whether it's my code (which had been pretty stable and none of the networking stuff changed, except the length of the content of the Post requests) and/or using a session id
On the 6310i, I was getting Http status code 502 (Unknown Response or Gateway Error) for all Post requests whose content length was ~ 160 characters; I'm not sure what the cutoff was as I haven't had time to experiement with the maximum length. This was occuring on the TMobile network in the US although it didn't appear to be happening on the Vodafone network in Ireland. (I say "didn't appear" because I didn't test it personally; someone else did and said they had no problems). The post request was always sent to my server which returned a response, but the response never made it to the phone (Http status 502).
Since shortening the content of the POST request, I'm retrieving some content but getting a LOT of java.io.IOException errors - this could be the network but seems to happen mainly when the response is long. Does the 6310i have a problem reassembling packets? I'm also seeing this problem in the last few days on the 7210, where I never saw that problem before. I had the same problems also on the 3590 (also limited to 30K apps) on AT&T's network.
On the 3650, the networking seems to be more stable but occasionally results in a "App. Closed jes0d0-ConnectionsEventServer@1? " error - anyone know what causes this? There's a posting on this site referring to this error being caused by URLs being too long but that's not the case here as my URL length is constant, only the content of the POST request and response varies.
If anyone has some experience or generalizations on these issues (is it the phone? is it the network? is it - yikes - my code?)I'd love to hear from them,
thx S

Reply With Quote

