-
What processor are you targeting?
This seems like a perfectly reasonable result to me. Not sure exactly what the relevant standard has to say about it, but I've seen this behavior on several micros in the past. I seem to recall it's one of those undefined or implementation defined behaviors. At any rate, using VC++ or GCC on a PC these days; sizeof(unsigned int) == sizeof(long), on...
2007-04-12 19:07:53 UTC in Small Device C Compiler
-
You should run sphinx3 in a seperate process and use an IPC mechanism to distribute the text stream to your other applications(s).
I've been reviewing the code here recently and I'm pretty sure it's NOT thread safe and probably needs a good bit of work to make it so. Most of the issues I noticed however were in v3-0.4.1 and I haven't verified whether all the issues carried forward into...
2005-01-11 17:16:30 UTC in CMU Sphinx
-
The attached source file attempts to midigate the
defects in prim_type.h by insuring that namespace
collisions will be flagged and the specified types will
be the correct size.
I tested this on Win32 and Cygwin platforms and the
test scripts ran the same as with the original header file.
2005-01-11 00:09:07 UTC in CMU Sphinx
-
I ran make test and got the following output:
./src/tests/regression/test_livepretend.sh
LIVEPRETEND TEST
YOU SHOULD SEE THE RECOGNITION RESULT 'P I T T S B U R G H'
FWDVIT: P I T T S B U R G H (pittsburgh.littleendian)
FWDXCT: pittsburgh.littleendian S 0 T -5792943 A
-4013597 L -1779346 0 -502448 -74100 69 -3949
56 -148286 P 94 -187695 -148286 I 107 -349267 -148286 T
131...
2005-01-10 00:05:13 UTC in CMU Sphinx
-
Logged In: YES
user_id=648746
Sorry, this is duplicate. I hit the back button. This is a
technically a bug in SourceForge! Well, at the very least
it's a design flaw.
2005-01-09 23:42:27 UTC in CMU Sphinx
-
When built in the VC++ environment I got a 'time'
undefined warning. Easily fixed with #include .
2005-01-09 23:39:21 UTC in CMU Sphinx
-
When built in the VC++ environment I got a 'time'
undefined warning. Easily fixed with #include .
2005-01-09 22:48:18 UTC in CMU Sphinx
-
As a recent comment in this header file points out, the
sizes of int8, int16, etc, are entirely dependant on
the compiler implimentation and therefore should either
be fixed to insure that the stated widths are in fact
the final result or they should NOT be used.
The use of int# and float# is also highly likely to
collide with other libraries in many cases as
programmers have reinvented...
2005-01-09 22:45:18 UTC in CMU Sphinx
-
Logged In: YES
user_id=648746
What test cases failed?.
2005-01-09 19:54:15 UTC in CMU Sphinx
-
Logged In: YES
user_id=648746
I think Geoff Kassel (gkassel) has a very good point here.
Any name that is known to be that common, should simply not
be used as a matter of good software engineering principals.
It is widely known that the vast majority of software is
poorly written. Ignoring this problem simply because
everybody else does sloppy work is not good practice.
It is...
2005-01-09 18:42:52 UTC in CMU Sphinx
