OSSP CVS Repository
OSSP CVS Repository |
|
|
| Date | Version | Description |
|---|---|---|
| 1.39 | Check-in [6015]: remember bugfix By rse. (diff) | |
| 1.38 | Check-in
[5978]:
Fix __VA_COPY_USE_{CSP,CPP} fallback macros for va_copy.
Submitted by: Mark A. Lindner <mark.a.lindner@gmail.com> By rse. (diff) | |
| 1.37 | Check-in [5684]: upgrade to autoconf 2.61 By rse. (diff) | |
| 1.36 | Check-in [5683]: prepare for release By rse. (diff) | |
| 1.35 | Check-in [5682]: add jslint copyright header and add jspack script By rse. (diff) | |
| 1.34 | Check-in [5681]: remember upgrades By rse. (diff) | |
| 1.33 | Check-in [5656]: *** empty log message *** By rse. (diff) | |
| 1.32 | Check-in [5653]: Upgrade to upstream sources as of 2006-08-20. By rse. (diff) | |
| 1.31 | Check-in [5652]: Add support for UTF-8 C strings in the API and internally by adding a --with-utf8 build-time options which builds the code with JS_C_STRINGS_ARE_UTF8. By rse. (diff) | |
| 1.30 | Check-in [5636]: bump version before release By rse. (diff) | |
| 1.29 | Check-in
[5635]:
Add optional Dynamic Shared Object (DSO) support (see src/jsdso.[ch]).
This comes in two flavors:
1. Provide two public C API functions... JSBool JS_DSOLoad (JSContext *cx, int *id, const char *filename); JSBool JS_DSOUnload (JSContext *cx, int id); ...as an ultra-thin wrapper around the POSIX dlopen(3) API. It especially mimics the BSD RTLD behaviour of calling pre-defined functions (mandatory "js_DSO_load" and optional "js_DSO_unload") inside the DSO after/before the dlopen/dlclose calls. This allows the DSOs to perform their init/shutdown actions. 2. Provide a small JavaScript global object "DSO" which binds the two public DSO C API functions into the JavaScript language as "DSO.load" and "DSO.unload". The "DSO" object can be created by the friend C API function js_InitDSOClass(). This function is used by the command-linne interface "js" by default. As a result the OSSP Mozilla JavaScript engine is now able to dynamically load C extensions similar to what other programming languages provide since a longer time. By rse. (diff) | |
| 1.28 | Check-in [5634]: Be more clean and replace weak "#if JS_HAS_FILE_OBJECT" constructs with the stronger "#if defined(JS_HAS_FILE_OBJECT) && (JS_HAS_FILE_OBJECT - 0)" as the JS_HAS_FILE_OBJECT can be not defined at all (in contrast to other JS_HAS_XXXX defines which are all listed in src/jsconfig.h) By rse. (diff) | |
| 1.27 | Check-in [5633]: Fix a bunch of compiler warnings. By rse. (diff) | |
| 1.26 | Check-in [5632]: Change from -O2 to -O0 and without -Wshadow under --enable-debug. By rse. (diff) | |
| 1.25 | Check-in [5623]: Fix "make distclean": src/perlconnect/Makefile.PL is always generated and hence has to be always deleted. By rse. (diff) | |
| 1.24 | Check-in [5622]: Fix "devtool dist" command: "make man" does not exist. By rse. (diff) | |
| 1.23 | Check-in [5621]: bump before release By rse. (diff) | |
| 1.22 | Check-in
[5620]:
Fix "make install": the executable "js" was not installed via GNU
libtool and this way under --enable-shared the wrapper script was
installed only.
Submitted by: Alfred Reibenschuh <alfred.reibenschuh@it-austria.com> By rse. (diff) | |
| 1.21 | Check-in [5619]: Add a src/perlconnect/MANIFEST file to make ExtUtils::MakeMaker more happy. By rse. (diff) | |
| 1.20 | Check-in
[5618]:
Fix "make install": js-config.1 was not installed.
Submitted by: Alfred Reibenschuh <alfred.reibenschuh@it-austria.com> By rse. (diff) | |
| 1.19 | Check-in
[5610]:
Use Autoconf macros AC_CANONICAL_BUILD and AC_CANONICAL_HOST to
make sure that the host identification is really available.
Submitted by: Alfred Reibenschuh <alfred.reibenschuh@it-austria.com> By rse. (diff) | |
| 1.18 | Check-in [5547]: remember the upgrade shift By rse. (diff) | |
| 1.17 | Check-in [5546]: Add support for Windows CygWin and MinGW environments. By rse. (diff) | |
| 1.16 | Check-in [5544]: change address on owner's request By rse. (diff) | |
| 1.15 | Check-in [5539]: Upgrade to upstream sources as of 2006-07-29 By rse. (diff) | |
| 1.14 | Check-in [5538]: Upgrade to upstream sources as of 2006-07-28. By rse. (diff) | |
| 1.13 | Check-in [5537]: use ctime By rse. (diff) | |
| 1.12 | Check-in [5536]: Changed GNU libtool shared library versioning from 0:0 to 1:6 By rse. (diff) | |
| 1.11 | Check-in [5535]: Increase portability by gracefully downgrading the stat(2) use of st_birthtime to st_birthtimensec or even st_mtime. By rse. | |
| 1.10 | Check-in [5526]: Apply a few more upstream fixes to jsfile.c. By rse. (diff) | |
| 1.9 | Check-in [5520]: bump version before release By rse. (diff) | |
| 1.8 | Check-in [5518]: Apply multiple code cleanups and bugfixes. By rse. (diff) | |
| 1.7 | Check-in [5517]: remember change By rse. (diff) | |
| 1.6 | Check-in [5515]: Consistently mark all OSSP bugfixes with a "/* BUGFIX */" tag on the "#ifdef OSSP" line. This way one can more easily see what are specific OSSP specific changes what should be taken over by upstream vendor. By rse. (diff) | |
| 1.5 | Check-in [5514]: Fix an incorrect argument type bug in jsfile.c related to a call to js_InflateString(). A "size_t *" has to be passed, not a "size_t". By rse. (diff) | |
| 1.4 | Check-in [5513]: Remember upgrade. By rse. (diff) | |
| 1.3 | Check-in [5511]: Added jslint (see ¤http://www.jslint.com/) together with some home-brewn option parser. The result is installed as a stand-alone "jslint" program. By rse. (diff) | |
| 1.2 | Check-in
[5510]:
o Install all src/js*.h headers during "make install" to allow
applications to at least optionally poke around in the internals.
This also allows an application to call js_InitFileClass() from
<jsfile.h> without us having to really call this function in the
standard API function JS_InitStandardClasses().
o Change the default of the --with-file/--without from --without-file to --with-file (enable the File object by default), but activate it only in the CLI (where it doesn't hurt and where it is actually really required to get something useful running there in practice) but explicitly NOT in JS_InitStandardClasses() (as this is what would hurt the security in applications which assume that JS_InitStandardClasses() initialized only really the standard classes). By rse. (diff) | |
| 1.1 | Check-in [5472]: more files, better versioning By rse. |