Anybody know what this error means? This is from version 1.0 of the emptool source compiled under LinuxPPC.

-Mike

[mcomb@groveg4 emptool]$ ./emptool
Scanning Ethernet for an empegcar...
P socket.cpp 103 ( 409): Binding to 0.0.0.0:8300empegcar 'empegcar' (ID number 0) found at 192.168.1.107
Connection is 0x10073b58
P socket.cpp 79 ( 409): Socket::CloseP #W connection_unix.cpp 625 ( 409): poll on socket timed out (0ms)
P connection_unix.cpp 531 ( 409): Flushed receive buffer
2000 Checking connection [ 0%]P #W connection_unix.cpp 625 ( 409): poll on socket timed out (0ms)
2000 Checking connection [Done]P #W connection_unix.cpp 625 ( 409): poll on socket timed out (0ms)
Protocol version of emptool (4) is too old for your player (1024)
Please obtain an updated emptool

Hmmm, no responses. Well I finally had a few minutes to poke around and it seems that the PROTOCOL_VERSION_MAJOR is defined as 4 in packets.h, but my empeg is returning 1024 or, its response is getting mangled by emptool. Changing the value in packets.h (trying for a quick fix) resolved that error, but leaves me with a never ending stream of...


[mcomb@macgeek emptool]$ ./emptool 192.168.1.200
P #W connection_unix.cpp 625 ( 2784): poll on socket timed out (0ms)
P connection_unix.cpp 531 ( 2784): Flushed receive buffer
2000 Checking connection [ 0%]P #W connection_unix.cpp 625 ( 2784): poll on socket timed out (0ms)
2000 Checking connection [Done]P #W connection_unix.cpp 625 ( 2784): poll on socket timed out (0ms)

2121 Downloading
2121 Starting download [Done]
2122 Retrieving tags index [ 0%]P #W connection_unix.cpp 625 ( 2784): poll on socket timed out (0ms)
P protocolclient.cpp 1360 ( 2784): EREMOTENAK_DROPOUT
P #W connection_unix.cpp 625 ( 2784): poll on socket timed out (0ms)
P #W connection_unix.cpp 625 ( 2784): poll on socket timed out (0ms)
P protocolclient.cpp 1360 ( 2784): EREMOTENAK_DROPOUT
P #W connection_unix.cpp 625 ( 2784): poll on socket timed out (0ms)
P #W connection_unix.cpp 625 ( 2784): poll on socket timed out (0ms)
P protocolclient.cpp 1360 ( 2784): EREMOTENAK_DROPOUT
P #W connection_unix.cpp 625 ( 2784): poll on socket timed out (0ms)
P #W connection_unix.cpp 625 ( 2784): poll on socket timed out (0ms)
P protocolclient.cpp 1360 ( 2784): EREMOTENAK_DROPOUT
P #W connection_unix.cpp 625 ( 2784): poll on socket timed out (0ms)
P #W connection_unix.cpp 625 ( 2784): poll on socket timed out (0ms)


Looks like I need to dig a little deeper later (anybody know what a 'EREMOTENAK_DROPOUT' is?). I think we just had an earthquake here (somewhere near Dublin, CA) and I want to find out what happened.

-Mike

You appear to be running into endian-ness issues. The empeg/x86 are little-endian, the PPC is big-endian.

Basically this means that any structures that are read/written over the connection have the wrong byte ordering - a short which is value 4 will appear as 1024, and vice-versa.

This will probably take quiet a bit of work - I suspect we might try and attack it ourselves at some point.

Hugo

Yeah, I was afraid you where going to say that. I don't suppose you have any idea when you might 'attack it'. I know you guys don't like to give dates, but I would hate to put a lot of effort into something and then have you guys do it better the following week.

Actually that brings up an interesting point. If someone outside empeg does any significant work to emptool do you guys want/plan to take contributions back to the emptool source tree so that we don't have to maintain a seperate version forever?

-Mike

Emptool is GPL'd - so in fact you HAVE to make available any modifications that you make if you want to distribute your own version. Anything that's neat and fits in with our stuff may well be included in our own distribution.

If you go off on a complete tangent then it will probably work the other way - you'll have to reintegrate our modifications (if you want them) into your distribution.

Hopefully everyone will be able to work consistantly so we all benefit from each others work in a seemless way.

Rob

As the problems are in the protocol libraries, which have been through an overhaul even since they were released - yes, we're actively working on them and will do another release when appropriate (I suspect it'll be at 1.1 time). I can't guarantee that endianness issues will have been addressed in 1.1, but such things are not good to have hanging around.

Hugo

Thanks Hugo. Guess I will play with something else for the time being then.

1.1 is really sounding like a yummy update.

-Mike