Hi togehter, due to the fact that android itself has only the HTTP4 client and I don't want to access some REST services I decided to replace my old code (http4 client with manual json deserialisation) with the android restlet version.
today i spend a lot of time with the profiler to figure out, why the restlet approach is so extremly slower than my old approach. I figured out, that in contrast to my own solution, restlet reads, sets and transfers header parameters in the requests and responses. There are some fields containing dates such as the field "Expires". Anyway, it seems to be a known bug in android that all SimpleDateFormat method calls are very slow. To give one example: There is a method called ClientAdapter.toSpecific() which calles HeaderUtils.addGeneralHeaders() which calles DateWriter.write()... The stack ends in the call SimpleDateFormat.format(). The execution now takes always some seconds, which is about 10 times slower than the http4 aproach. Some numbers: 02-16 23:43:38.775: INFO/app(12413): sendLocation with httpClient took 874 ms 02-16 23:43:42.015: INFO/app(12413): sendLocation with restlet took 3204 ms 02-16 23:46:20.665: INFO/app(12413): sendLocation with httpClient took 655 ms 02-16 23:46:24.005: INFO/app(12413): sendLocation with restlet took 3301 ms 02-16 23:50:40.825: INFO/app(12413): sendLocation with httpClient took 656 ms 02-16 23:50:44.855: INFO/app(12413): sendLocation with restlet took 4003 ms 02-16 23:51:06.775: INFO/app(12413): sendLocation with httpClient took 612 ms 02-16 23:51:10.005: INFO/app(12413): sendLocation with restlet took 3192 ms Because the header modification stuff is somewhere in the deep of restlet, it is not easy to override. Any ideas how to handle this problem? -- View this message in context: http://restlet-discuss.1400322.n2.nabble.com/Reading-setting-date-header-fields-is-extreme-slow-on-android-tp6033977p6033977.html Sent from the Restlet Discuss mailing list archive at Nabble.com. ------------------------------------------------------ http://restlet.tigris.org/ds/viewMessage.do?dsForumId=4447&dsMessageId=2704886
