 Shane Kerr
|
543dbfd90d
Things to do for the boss process.
|
15 years ago |
|
Jelte Jansen
|
17514e12dc
add positional parameters in bigtool; merge back from jelte-datadef rev 342 and 386
|
15 years ago |
|
Jelte Jansen
|
60e8544e60
fixed some en/decoding bugs in the cpp part of the new cc wire format
|
15 years ago |
|
Jelte Jansen
|
4974ce5dc8
some unit tests for Elements, and fixed a few bugs while i was writing them
|
15 years ago |
|
Jeremy C. Reed
|
b59ef35ca1
Also add the install wrappers, so they install too.
|
15 years ago |
|
Jeremy C. Reed
|
0770bad2bd
Add beginning Makefile.am for bigtool, msgq, and bind10.
|
15 years ago |
|
Michael Graff
|
f6b55c6092
add TODO, fix a bug with a name that vanishes/does not exist
|
15 years ago |
|
Michael Graff
|
ff75117590
fix path
|
15 years ago |
|
Michael Graff
|
5fdc00a1ab
rename pymsgq to msgq
|
15 years ago |
|
Michael Graff
|
5c4b0719cd
make msgq_test from msgq_test.in
|
15 years ago |
|
Michael Graff
|
34198282dd
delete old msgq
|
15 years ago |
|
Michael Graff
|
200d42f0d0
disable old msgq
|
15 years ago |
|
Michael Graff
|
d055308d15
search for python, python3, and then python3.1. Require version 3.1. Use @PYTHON@ in all .in files, and add run_bigtool.in. This should cover all script wrappers.
|
15 years ago |
|
Michael Graff
|
342fdf3ded
kill API support for subscription types. 'meonly' is the default. Comment out lines in callers which used 'meonly'
|
15 years ago |
|
Michael Graff
|
f222c66705
Implement directed messages, and a test for it
|
15 years ago |
|
Michael Graff
|
a6823843ad
checkpoint work; Python-based msgq mostly works. Bad input will crash it, which should be fixed, probably by wrapping the entire message processing in a try loop. Gross, but...
|
15 years ago |
|
Michael Graff
|
6383f1fda5
New wire format, which makes things more sane for processing envelope apart from messages. No API changes. The current msgq does not support this, but the pymsgq I'm hoping to finish up tomorrow will.
|
15 years ago |
|
Michael Graff
|
83553f96f1
learning python sockets: msgq, the revenge
|
15 years ago |
|
Jelte Jansen
|
a938e686d0
be lenient on the receiving end whether env["msg"] is a str or a bytearray (until we use the new wireformat i guess)
|
15 years ago |
|
Jelte Jansen
|
52cc97f6b4
Merged to_wire nulltermination bugfix from branches/jelte-datadef revision 297:298
|
15 years ago |
|
Jelte Jansen
|
f49960895f
int type for wireformat in python, ruby and C
|
15 years ago |
|
Jelte Jansen
|
78dabb4433
bool type in cpp cc
|
15 years ago |
|
Jelte Jansen
|
8cf507e89d
this seems safer, when TrueClass/FalseClass
|
15 years ago |
|
Jelte Jansen
|
fcb96d3442
bool type in ruby cc
|
15 years ago |
|
Michael Graff
|
ae6b4206d3
fix; had a duplicate method name
|
15 years ago |
|
Michael Graff
|
7a5fb2613c
require utf8 strings
|
15 years ago |
|
Jelte Jansen
|
bc5d12940d
bool type on wire in python messaging part
|
15 years ago |
|
Michael Graff
|
5b53b4fd6c
add the ideas of UTF8 and BLOB types. Both are decoded into strings, and all strings are encoded into UTF8 for now.
|
15 years ago |
|
Michael Graff
|
ba9a64862e
lame, but at least non-empty, install file
|
15 years ago |
|
Michael Graff
|
735c7b752d
make a difference between binary and string data. This is mostly because strings are utf-8, while binary is, well, binary.
|
15 years ago |
