Thursday, March 20, 2008

new testing device

I got a RIM blackberry pearl. It does j2me, and I like it. My recording application tries to do AMR, but fails. The problem may be that the audio gets saved to the j2me "record store" mechanism. I have a j2me penpal who has advised that the record store is often unreliable.

I think I will put my record and play code together in a MIDP and play audio straight from the memory object instead of messing with record store. I hope to get time to code again tonight.

Saturday, March 8, 2008

umm, partial success or something

Well, I am quite proud :) I took some code that I found on the net, a program to record audio, and another that plays audio. I had to do some tweaking to get my development environment to build the applications correctly (at least I guess it was correctly). There are no emulators currently that support MMAPI fully, so I could not test it on my machine. I succeeded in uploading the code to an LG phone of some sort. Upon running the record program, the phone no longer had audio until I rebooted it :) So I am meddling enough to break things! A really great sign!

The play application seemed happy, but since the record part was crashing, I do not know yet. However, playing audio is part of MIDP 2.0, so I have confidence in that part.

This week I will test on a couple of windows mobile phones and see what happens. In the meantime, maybe I will google for LG MMAPI issues...

Friday, March 7, 2008

putting it together

Since the success of running my program on the blackberry, I also tested it on some windows mobile devices. Of course at that point it was just a splash screen :) However at this time, I have started pulling sample code together and actually writing I am having some trouble tying things together in my noggin, since midlet code is really the tip of the j2me iceberg. There is a lot of magic happening that doesn't make sense reading code, so I made a little shell script to make comment-free copies of the j2me sources, and I put them all in one directory. So now I can just type "grep thing *" and find any place "thing" appears in the j2me sources. I also put the sources on my home web server so I can browse them if I am away from my development MacBook. Hopefully this will help me tie the "magical" parts of j2me code together in my head.