Activity
From 10/09/2009 to 11/07/2009
11/07/2009
-
09:20 PM Bug #159: Flusspferd crashes when calling js function returned from require() within a c++ callback
- Aristid Breitkreuz wrote: > Hello Aaron, > > What does "it crashes" mean? > > Please provide the output of the program. > > > Aristid ...
-
09:10 PM Bug #159 (Feedback): Flusspferd crashes when calling js function returned from require() within a...
- Hello Aaron, What does "it crashes" mean? Please provide the output of the program. Aristid
-
02:49 PM Bug #159: Flusspferd crashes when calling js function returned from require() within a c++ callback
- How to reproduce: http://gist.github.com/228710 Originated from here: http://groups.google.com/group/flusspferd/browse_thread/thread/ff07c6da747b3d3c
11/06/2009
-
08:29 PM Revision 0d3fc9b700dafbd1462119f5b45069fa58102b67: docs: Ubuntu 9.04 -> 9.10
- docs: Ubuntu 9.04 -> 9.10
11/05/2009
-
- doc: mention that ebuilds are not in gentoo repo
11/02/2009
- 03:18 AM Revision 6136c6e12ecfb810aeafe78f9f57e0aabfb1fc00: JS API docs for fs-base module. PDoc wont cur...
- JS API docs for fs-base module. PDoc wont currently parse the namespace (which i'm abusing to mean module) as 'fs-base', so it's been called 'fs_base'
- 03:16 AM Revision 9ba052d723fd9eaddcc649d4efee870ed33482a4: PDoc runner script, will not currently run for...
- PDoc runner script, will not currently run for anyone else
- 03:15 AM Revision 42f0d5fd87ba809eebc3c5a701aed17e445a25e7: Move top level sections into a single pdoc file
- Move top level sections into a single pdoc file
- 02:31 AM Revision ebb090de9567651d910e2d41b9813cca2a693b69: Convert IO docs to pdoc
- Convert IO docs to pdoc
11/01/2009
-
- core/modules: yet more conservative rooting
-
- core/modules: do not use local_root_scope when calling into JS module
-
- curl: trace writefunction_callback
-
- curl: continued implementing options. New internal API. Simple example working
- 02:34 AM Revision 73ce6b5405dcee791b1e585eddca098c4a54d28e: core: Enable zealous GC in debug builds
- core: Enable zealous GC in debug builds
-
- Merge branch 'master' into curl
10/31/2009
-
- test: added test case if function f() is_null
-
- Revert "changed property_attribute to own getter/setter member. (fix #158)" This reverts commit ce9ae3e1934790a200f5880f66c17000b81c3e8a.
-
- new_create: add test for create<function> with void(call_context&) signature
-
- new_create: create<function> is now rudimentarily polymorphic
-
- new_create: return a function in create<NativeFunctionClass>
-
- new_create: want to do overloading for function
-
- Merge branch 'master' into curl
-
- changed property_attribute to own getter/setter member. (fix #158)
-
- The getter/setter member of property_attributes (see source:include/flusspferd/property_attributes.hpp ) is of type boost::optional<function const&>. Thus the function object is not owned by property_attributes and code like this is wrong: <pre...
-
- new_create: rename parameter source->function
-
- new_create: remove some outdated documentation
-
- new_create: move function that is not directly used by anything outside create.hpp into namespace detail
-
- new_create: add flusspferd::param::type<T> for passing types to flusspferd::create()
-
- new_create: move include
-
- new_create: move helper functions into appropriate headers
-
- new_create: adapt tests to ruediger-API
-
- Merge branch 'master' into new_create Conflicts: ChangeLog.txt
-
- Merge branch 'master' into curl
-
- new_create: add create<method> as a (for now) synonym of create<function> flusspferd::method is a tag type
-
- new_create: (publicly) remove create_native_functor_function
-
-
- So ruby is better than java right? ;) I've had a play with "pdoc":http://pdoc.org (for juice) and it produces nice output - see here http://api.prototypejs.org
-
- new_create: move includes to more appropriate places
-
- new_create: move create<NativeFunctionClass> into create/native_function.hpp
-
- new_create: move create<NativeObjectClass> into create/native_object.hpp
-
- new_create: use releate include path
10/30/2009
-
- new_create: add missing include to flusspferd.hpp
-
- new_create: move create<function> into create/function.hpp
-
- new_create: move create<array> into create/array.hpp
-
- new_create: move create<object> into create/object.hpp
-
- new_create: add create<MyFunction>
-
- Rüdiger Sonderfeld wrote: > It doesn't justify a 0.8.1. But if we ship a 0.8.1 anyway we should include this. No. Why should we? It doesn't fix any critical functional bug.
-
- It doesn't justify a 0.8.1. But if we ship a 0.8.1 anyway we should include this.
-
-
- doc: removed reference to lcov
-
- I don't think this is critical enough for 0.8.1.
-
- doc: minor fixes for the ubuntu docs (no lcov and autoconf2.13 required)
-
- ./util/run_tests.sh currently runs test-sqlite3.t.js even if the sqlite3 module wasn't build. As a consequence the test case fails.
-
- doc: updated ubuntu installation tutorial (Ubuntu 9.10)
-
- I don't like 2. Since it would add part of the API exclusively to flusspferd.hpp. You than just couldn't replace #include <flusspferd.hpp> by a finer selection of headers you'd also have to add the aliases yourself.
-
- Updated ChangeLog.txt (fixes #155)
-
- core: changed get_property_attributes to return an optional property_attribute object instead of using call-by-reference (and made it const)
-
- Right now, the convenience header flusspferd.hpp just includes all common flusspferd headers. It could in addition: # include common boost headers (boost.assign, boost.fusion), # set up useful namespace aliases.
-
- new_create: remove create_native_object
-
- core: added get_property_attributes with ID parameter to object (#155)
-
- core: added define_property with ID parameter to object (#155)
-
- new_create: for reference arguments
-
- define_property now takes attrs as const reference
-
- new_create: remove complicated code which existed because i did not understand Fusion
-
- curl: dropped support for older boost releases (see 637ce78350c85afb5a8f79ffb60e05c642772f79)
-
- Merge branch 'master' into curl
-
- dropped support for Boost older than 1.40
-
- new_create: test and fix usage with constructor parameters
-
- doc: say something about what the getter/setter should look like
-
- new_create: add parent and prototype parameters
-
- new_create: add parent object parameter to backend object creation
-
- new_create: create<Class> now works (incompletely) in the parameterised version
-
- core: (typo, compiles again) 431db4dbca4fec076f968281e2c9263422a7e235
-
- use *ById instead of UC strings (#155)
-
- added wrap/get_jsid ( #155 )
-
- Since we only support 1.8.1+ now. We could use the new *ById functions (e.g. "JS_HasPropertyById":https://developer.mozilla.org/en/SpiderMonkey/JSAPI_Reference/JS_HasProperty ) instead of treading @value id@ as a (unicode) string.
-
- curl: minor fixes
-
- new_create: add container and attributes to create<function> parameters
-
- new_create: need BOOST_PARAMETER_MAX_ARITY, not BOOST_PARAMETER_LIMIT
-
- Merge branch 'master' into new_create
-
- curl: continued with options
-
- curl: started implementing options
10/29/2009
-
- commit:90f604d712d171a3f9336735472b3de6fb9bfa51 That was an evil bug. But there, I fixed it. :-)
-
- shell/interpreter: fix #154 attention: if you want to use operator, with phoenix, you need statement/sequence.hpp
-
- curl: version check for boost::get_error_info
-
- Merge branch 'master' into curl
-
- core: fixed the fix for the fix which was fixing the fix... ;-) ( af4d8f389c72435b1211bcd4627f26940e162b6e )
-
- Merge branch 'master' into curl
-
- core: fixed version number check (fixes 2c2c524 )
-
- core/getopt: use fusion instead of tuple (mainly because i want to learn how to use fusion)
-
- ...
-
- core: fix the fix from 7ac34e3 so it works on boost 1.39 too (the boost guys had a breaking API change in 1.39)
-
- curl: removed old setopt stuff
-
- sqlite3: remove unnecessary include
-
- ...
-
- core: make it compile
-
- curl: added opt to trace (fix segfault issue)
-
- core: removed (non working) check for JS_ValueToSource since we dropped support for old spidermonkey releases anyway
-
- curl: use boost::ref instead of workarround
-
- curl: made EasyOpt working
-
- curl: started implementing options object
-
- core.system: remove xFlusspferdVersion (use require('flusspferd').version instead)
-
- curl: tried to change curl to use boost.exception (Warning: BROKEN!)
10/28/2009
- 10:04 PM Flusspferd Packages Revision 4f993acdec78e946fcb8a41c7aa4e04c372ef1b0: Update portfile for 0.8 - checksums and remove...
- Update portfile for 0.8 - checksums and remove xml and curl plugins
-
- new_create: more testing
-
- curl: fixed `what'
-
- Merge branch 'master' into curl
-
- exception: added default `what' (checking for BOOST_EXCEPTION_THROW default infos)
-
- curl: fixed issue, compiles again
-
- curl: tried to improve exception class (not compiling atm!)
-
- new_create: create<NativeObjectClass>() works now, but only without any parameters
-
- Merge branch 'master' into curl
-
- http://code.google.com/p/gpsee/issues/detail?id=12 Talking in IRC with Wes and the problem of how to handle shebang lines, the following idea came up: ...
10/27/2009
-
- new_create: improve macro formatting
-
- new_create: remove create_function
-
- new_create: add create<function>
-
- new_create: allow adding created stuff to be added to container objects
-
- May be obsoleted by #147.
-
- define_properties, too.
-
- For setting / defining many properties at once, (in the case of define) all with the same property_attributes.
10/26/2009
-
- new_create: boost parameter limits to 10
-
- new_create: remove create_object
-
- new_create: implement create<object>
-
- new_create: remove unused code
-
- new_create: allow multiple parameters (how fancy)
-
- new_create: ChangeLog entry
-
- new_create: remove create_array
-
- Merge branch 'master' into new_create
-
- Merge branch 'maint'
-
- shell: fix rooting error
-
- new_create: flusspferd::create<array>(...) works now
-
- Done.
-
- Done for the 0.8 docs. I can't touch @/var/www/0.7/docs@, because it is @uid ruediger@. So this is something that Ash has to fix (the user rights).
-
- Fixed in master. commit:673d309fbf223430a111d42e883a67344515c750
-
- doc: fix outdated link (fixes #150)
-
- Locally, people don't have mod_rewrite or anything similar.
-
- http://flusspferd.org/0.8/docs/docs/index.html <- error 404
-
- http://flusspferd.org/0.8/docs/index.html http://flusspferd.org/0.7/docs/index.html
-
10/25/2009
-
- new_create: initial stub
-
- An example of the named parameters (in a few of the varieties that come to mind): ...
-
- commit:4a8983ad723b638bb07e84e064e7a53ede2c8a79 I _always_ forget to actually close the fixed bugs.
-
- An issue that has to be considered, too, is the passing of constructor parameters to native object types. Example (in the old way): ...
-
- This one is a biggie. But also very high on the "coolness" scale, so it gets priority High. ;-) I would like to have flusspferd::create<T>(NamedParameters), with the following requirements: # flusspferd::create<object>() creates an empty obj...
-
- But I'm wondering whether there should actually be a whole new framework for flusspferd::create*.
-
- c++api: create_array accepts an input range for creation (fixes #145)
-
- In order to implement this, I will remove the possibility to create an array with pre-defined size, for now.
-
- Merge branch 'new_exception' Conflicts: ChangeLog.txt
-
- readded empty as a wrapper for is_js_exception but marked deprecated.
-
-
- commit:358b47b3e412fb8538b6d6769914896adbc65d90
-
- c++api: Adapt array::set_element to object::set_property interface (fixes #146)
-
- all: use array::push
-
- Merge branch 'master' into new_exception
-
- set_property is more flexible, set_element should get the same level of flexibility.
-
- commit:a47ce59921e209887fb025914c80682a48e898ac
-
- c++api: Add push for flusspferd::array (fixes #144)
-
- For example by passing a "boost::assign::list_of":http://www.boost.org/doc/libs/1_40_0/libs/assign/doc/index.html#list_of.
-
- While there, other array methods should probably be added: pop,shift,unshift etc.
-
- So call("push", ...) does not need to be done every time.
-
- minor cleanup
-
- changed empty to is_js_exception (internal)
-
- changed the initialisation of exception. init shouldn't be part of the API.
-
- changed parameter of exception_message (this should be faster. see c++next blog)
-
- exception_message shouldn't be part of the API
-
- changed exception::empty into more meaningful exception::is_js_exception
-
- Ash Berlin wrote: > This is now possible by doing: Actually, it isn't. The load function was envisioned NOT to do any sandboxing with an exports object or anything. But good to know, I hadn't noticed that require() now has this nice feature.
-
- This is now possible by doing: ...
-
- Having the same API/protocol as V8's debugger mode might be a good idea: http://code.google.com/p/v8/wiki/DebuggerProtocol
-
- To close and go to 100%, we need MSVC support.
-
- Probably unsafe, rejecting.
-
-
-
-
- Probably not the only problem of that module.
-
- Like gdb, just for Flusspferd. ;) Also it should support "debugger protocols", so it can be integrated in common debugging tools.
-
- Can you please please try this again with Flusspferd 0.8? For me, Flusspferd 0.8 works with Spidermonkey 1.8rc1.
-
- Definitely. Could it be even used to make a debugger?
-
- We should provide debug features. Spidermonkey provides an (pretty much undocumented) API for this http://developer.mozilla.org/en/JSDBGAPI_Reference This issue can probably be split into several smaller tasks.
-
- ChangeLog: update
-
- commit:3b51f57214e27f571a19557ccc23edb66c71f3c5
-
- core: add default toString and toSource for native classes (fixes #139)
-
- commit:d3a9b8ed330e70fe870dd90aa5360f2f530c3897
-
- doc: document == and != for value
-
- c++api: add == for value (fixes #141)
-
- value should support @==@, which should behave like @===@ in Javascript.
-
- added support for boost.exception
-
- core: context::get_prototype<native_object_base> now returns a REAL prototype
10/24/2009
-
- CONTRIBUTORS: add contributor of set_jit.diff
-
-
- commit:0fd65deebae09ccc2282b16936ed8f4b6403696d Did not add you to CONTRIBUTORS because Outmin is not your real name.
-
- apply set_jit.diff (fixes #140)
-
11:44 PM Bug #140: JIT is not enabled
- Flusspferd doesn't enable JIT (Tracemonkey), so I added a new method to context, shamelessly ripped from the set_strict() method. It will not be enabled until you call it. Patch is attached.
-
- Since the REPL now calls JS_ValueToSource (or what ever it is) this menas that native objects are now printed in a most unhelpfully fashion: ...
- 10:20 PM Revision 0127e0e09076447f2f46ea97d6da0bdc693ce662: Merge branch 'maint'
- Merge branch 'maint'
- 10:18 PM Revision 337dbd808ddc4c458771f3ce9b4807ec1ad2d567: build: build and install plugin modules proper...
- build: build and install plugin modules properly on windows
- 10:18 PM Revision f406e5e985f9751727c17782768de08533594fec: build: change sqlite cmake section to work on ...
- build: change sqlite cmake section to work on windows (no pkg_check)
-
- configure: slightly improved error message when spidermonkey without config.h is found
-
- commit:775e87e6be507ac8960f8ae7b399349641049ac0
-
- c++-api: root_* constructor is now explicit (fixes #124)
-
- curl: started adding exception stuff
-
-
-
- I'm not sure if I understand this bug. Please check if this problem is solved in commit:b960e17157af55f39b72c085435f9b95d036ed80.
-
- core: clean up Date ctor check (fixes #127 ?)
-
- core: remove unneeded include
-
- all: remove support for old spidermonkey releases
-
- doc: mention need to install iconv
-
- I think how function params are treated needs a bit of tidying: # Optional params should use @boost::optional@, otherwise they are required # The number of params passed to a function must be at least as many as the required params #* This w...
-
- commit:81da597939687c3523fcf1eb6ba26fa9eb4f2a40
- 01:25 PM Revision bb785a948b28b4967f4be24397647dc60ca8097b: core/encodings: make error messages more under...
- core/encodings: make error messages more understandables (fixes #137)
-
- I can't address the "Value is no object" and the "needs 2 arguments" as part of this bug, because they are part of a more general issue.
-
- ...
-
- * flusspferd_version would be nice, but cmake now accepts -DFLUSSPFERD_VERSION=0.8 which works for releases * sys.env is now enumerable.
10/23/2009
-
- I think it would be great if it was easier (or possible, it's hard to grasp because the code extensively uses templates and boost) if users of the C++ API could extend detail::convert to support user-defined types. For instance, in my project I...
10/22/2009
-
- False alarm - this was infact a cross-compile endian issue in the spidermonkey.
-
- Discovered by @require('io').File('foo','w')@ which yields an error of "File.open: mode '' not supported". But to make things weirder it behaves fine if the string is longer than 1 character long: @File('foo','ww')@ -> "File.open: mode 'ww' not...
10/21/2009
-
- curl: Easy::perfom now throws an exception in case of error instead of returning the error code
-
- curl: minor changes and finished writefunction
-
- curl: continued with setopt
-
- curl: global_init now throws an exception instead of returning the error code
-
- curl: started implementing setopt (curl_easy_setopt)
-
-
- Merge branch 'master' into curl
-
- doc: fixed typo
-
- Because of course the headers are NOT always installed in a @js/@ directory. To quote myself: > Spidermonkey is a mess.
-
- curl: easy_* functions are now methods
-
- curl: started from scratch again
-
- Merge branch 'master' into curl
-
- core: create_native_object now accepts void* parameters. Fixes #133
10/20/2009
-
- create_native_object ( source:include/flusspferd/create.hpp ) ignores (ending) void* parameters due to the usage of "boost::enable_if":http://www.boost.org/doc/libs/1_40_0/libs/utility/enable_if.html#htoc5 Changing create_native_object to use e...
-
- "Download Flusspferd from Sourceforge.":http://sourceforge.net/projects/flusspferd/files/ This version is hot on the heels of 0.7 (relatively speaking) due to a lot of work Ash Berlin put in place to get features for the first release of "Juice...
-
- Merge branch 'master' into curl Conflicts: CMakeLists.txt
- 09:41 PM Revision 3595512a12fd3966de662eb4eaa34a1788b79ee8: Release prep
- Release prep
-
- Moving to next release. Not sure how easy this is actually to fix. I guess some Cmake config check magic might be possible
-
- This seems to be fixed now - guessing it got done in the SQLite refactor ...
-
- Permissions not done, but everything else is. Good enough for now!
- 09:29 PM Revision 728c6060f0943be3fe608c969dc3a02fa4b721c3: Alias fs-base to filesystem-base
- Alias fs-base to filesystem-base
- 09:28 PM Revision 37e071a8f4cfbe5a41c18907112b3c056fe586d4: Re-implement require.alias (top-level ids only)
- Re-implement require.alias (top-level ids only)
-
-
- Done
- 09:17 PM Revision 9e7782f0ca70c1d3872508fa95c2f4f25b51d00e: Fix issue #115 - error messages for File ops o...
- Fix issue #115 - error messages for File ops on Win32
-
-
- Fixed
- 08:32 PM Revision dc48f737cbb6483c0c86843c19511b99169d56c3: Fix issue #130 - trying to File.open a dir giv...
- Fix issue #130 - trying to File.open a dir gives buffer underflow Now says ``can't open x - its a dir you fool'' but more polite
-
- build: fix version detection
-
- flusspferd module: installPrefix is always exported
-
- The actual issue is that spidermonkey's ABI changes depending on if its built with debug or not - and we need to set the flag to match. Perhaps a compile test would be better way of doing this then tying to the Release/Debug build type of fluss...
-
- fixed typos in the ReleaseNotes.txt
-
- help: fixed small inconsistency in manpage template
-
- Confirmed fixed, and tested by this: ...
-
- 100% == Closed. Silly redmine.
-
- Done! Tested on OSX, Windows, and i think I tested it on Linux too
-
- Need to use ncurses directly to get round this issue I think >_< Not happening in time for 0.8 (which is today)
-
-
- If Flusspferd's latest trunk is built without building in release mode, it will error out at this point: ...
- 09:09 AM Revision 7daeea643fafb4e130544884e21dadb0f6f9a199: Release notes for upcoming 0.8 release
- Release notes for upcoming 0.8 release
10/19/2009
- 11:16 PM Revision c1ff1ad4546c6d0fcb688cc38be9755ba92f8aa3: core: fix bug in relocatable include path in a...
- core: fix bug in relocatable include path in a dir named flusspferd
- 11:15 PM Revision 471e0f22483c1f91f1c4ca3dbbd5daca1aa2d766: build: Install most things on windows (slightl...
- build: Install most things on windows (slightly less under MSVC)
- 10:39 PM Revision 18c0daa9b77505429007004293b53b59fa539e77: core/io/fs-base: On windows canonicalize to in...
- core/io/fs-base: On windows canonicalize to include the drive letter.
- 10:32 PM Revision deaeb9377dde64fbb040371c534e2403fc11cc34: core/system: Fix enumeration of environment va...
- core/system: Fix enumeration of environment variables (portably) on Windows.
- 10:31 PM Revision a312878fdc4142932eea041bcb62e7f17faf06db: Fix compile issues on MinGW
- Fix compile issues on MinGW
10/18/2009
- 09:45 PM Revision 3417f1578f95e824f15202f3512748008081a41f: Use an explicit byte order in the MBCS encodin...
- Use an explicit byte order in the MBCS encoding tests
- 08:54 PM Revision 04c9dbaa13c0e00a6426fdb8451bfbdfbbe5f1d8: build: libflusspferd has no transitive link re...
- build: libflusspferd has no transitive link requirements.
- 08:52 PM Revision 0eb1780a3e57651837fc6e7ae9f01d913d2980c2: build: Optionally link sqlite3 against static lib
- build: Optionally link sqlite3 against static lib
- 08:51 PM Revision 9a24d7551d8f2f5a497effa6dfdecca0a5ca6ccc: build: don't make all libraries link against l...
- build: don't make all libraries link against libmozjs - they don't need to.
10/17/2009
- 04:04 PM Revision 60c55331b3ecbe9650f0c1563c61b211735c6188: Extra tests for require behaviour (const var l...
- Extra tests for require behaviour (const var leaking, and require.id)
- 03:57 PM Revision e9c8ca7d24bafcda9190bac99d552def32133038: Move handling of require.main into modules.cpp
- Move handling of require.main into modules.cpp
- 03:51 PM Revision 11788bb7d7fefc40607f42d33bc1c8509711c1f1: run modules.t.js test stand alone to test diff...
- run modules.t.js test stand alone to test different code path of REPL
- 03:51 PM Revision 40978b06413cbbe7909ea53c59aa289cd33b5a66: repl: '--interactive' now gives REPL when quit...
- repl: '--interactive' now gives REPL when quit() is called
- 03:51 PM Revision 2393627304cc098a9bff16aacf1e1a1687298dc2: Use include style in tests to be relative (sup...
- Use include style in tests to be relative (support to out of tree test runs)
-
- Currently the @is_native< Class >(obj)@ and @get_native<>()@ funcitons only work with native objects, not native functions. Ideally these should work for functions and objects. Also, while we are in that area, @native_object_base::is_object_nat...
- 12:56 AM Revision dbaba2173df2b01c155939496f36d8e581ad3a46: Better handling of dot files in prove
- Better handling of dot files in prove
- 12:49 AM Revision 285db6fba44dfd4db12b0059c948a22f12d6c972: Fix bug in fs-base.canonical (with tests of so...
- Fix bug in fs-base.canonical (with tests of sorts)
- 12:16 AM Revision 498b1853956fd12ef700157ca80c1aec8c67d528: Make relocatable native modules work properly ...
- Make relocatable native modules work properly under OSX
- 12:06 AM Revision 38abd90d0d7f65aee825205641d7ba1ee8bbed63: have run_tests.sh print a final pass/fail message
- have run_tests.sh print a final pass/fail message
- 12:05 AM Revision 028359f145fe2c501443452da7d2d50e79eaadc2: Make tests pass (context setup needed fixing)
- Make tests pass (context setup needed fixing)
10/14/2009
10/12/2009
- 09:15 PM Revision 055a20a10247ebb1ea110061215f4868ab51c771: repl: remove debug print
- repl: remove debug print
- 09:14 PM Revision fddb07cb795c0ef4e982c2b54913568fab375c6e: Fix missing slash in prelude concatination
- Fix missing slash in prelude concatination
- 08:18 PM Revision e8770d83aceff7191bc4be0f86c737f9b01a6db0: Merge branch 'master' into relocatable
- Merge branch 'master' into relocatable
-
- core: fix compile error
10/11/2009
- 10:38 PM Revision 5ed76b5424174e1fc37cb44ab229908233435e06: core/io: Fix bugs in File.open w.r.t creation ...
- core/io: Fix bugs in File.open w.r.t creation 1) using a string open mode was always erroneously setting create 2) Missing perms arg to open() caused 000 files to be made
- 10:35 PM Revision f7b7caa1bfd364495b75d24f228b049dcc78789d: Make jsrepl.js use relocatable info (and find ...
- Make jsrepl.js use relocatable info (and find it using the relocatable path)
-
- There is no a flusspferd module which has an @executableName@ property containing the canonical location of the binary (or as best as we can make out, anyway). Now we just need to built the right jsrepl.js file
- 06:17 PM Revision 2b626543a4026324be9a353059473e432d8c9dae: Pass argv[0] into the flusspferd module and fa...
- Pass argv[0] into the flusspferd module and fallback to trying to work out exec name from that
- 05:47 PM Revision 7b65c8a9f7621c5f8716a0b75ac8e31f272fd2b6: System specific ways of getting hold of real b...
- System specific ways of getting hold of real binary name
- 04:02 PM Revision 444a9b68104cda227f8b70c60632e52cb3aab10b: Create a flusspferd module to expose info abou...
- Create a flusspferd module to expose info about interpreter to JS
- 02:59 PM Revision a7c169726b5f4d20d10328fca6a87ba04568feb7: build: various platform specific checks for ge...
- build: various platform specific checks for getting hold of exe name Using methods details in http://stackoverflow.com/questions/1023306/finding-current-executables-path-without-proc-self-exe
- 12:34 PM Revision bd936ef2430659c06b78d24947a158e9fcad7a18: Merge branch 'master' into relocatable
- Merge branch 'master' into relocatable
- 12:16 PM Revision 4f9a7d79a2584b3fe06a168e05a2fbce03676630: core/encodings: Fix error in convertFromString...
- core/encodings: Fix error in convertFromString "Error: Invalid multibyte sequence at the end of input" was being thrown before.
- 11:56 AM Revision 5820e22c76efb9befe0fd2ab3bce20c35dc04f72: Fix style nits.
- Fix style nits.
- 11:55 AM Revision a97c38618c673305d977f2f5b1c96a29d66ca3a4: flusspferd::execute uses the same behaviour as...
- flusspferd::execute uses the same behaviour as require to get source text (fixes utf8 literals in source)
- 11:16 AM Revision bc3ee3d58d79e8ce6a92e2802168d5966be3e115: tests: add asserts.matches function
- tests: add asserts.matches function
- 11:15 AM Revision 573b0dd57697f36c75b62e3ab9bc678b4716ad9c: Add tests for encodings.convertFromString
- Add tests for encodings.convertFromString
10/10/2009
- 09:55 PM Revision 1db62923bba7dcdb0fd422ecfadc9f891fb576c0: core/js: remove debug prints from prove function
- core/js: remove debug prints from prove function
-
- Done.
- 09:55 PM Revision 88f0511b0492e9b162ae9d85e79fe809cec92a7b: tests: fix encodings test (was bug in test. Fi...
- tests: fix encodings test (was bug in test. Fixes #118)
-
- So run_test.sh now runs JS tests. Coverage figures would be nice, but we don't produce them for the C++ anymore either
- 08:55 PM Revision ea570d9d7774d006752a42118eaf7e40c7bcc3b9: util: run JS tests as part of run_tests.sh scr...
- util: run JS tests as part of run_tests.sh script (Fixes #4)
- 08:34 PM Revision f4aa92d7035fd0bc924e37ac546f73b6e7528763: core/js: build a prove function that will run ...
- core/js: build a prove function that will run tests
- 08:07 PM Revision a12a7374a7228074eddd359f9369d912b7bf09d4: tests: Rename javascript test files to .t.js
- tests: Rename javascript test files to .t.js
- 08:07 PM Revision fc472abdb2cb6bd906f28a943283031862d93b6d: core/require: Don't replace extension, add to it.
- core/require: Don't replace extension, add to it.
- 08:06 PM Revision 2cb549c214a2edcc2b129e3e93dabda61c1ae8f0: core/io: Fix but in canonicalization of files ...
- core/io: Fix but in canonicalization of files They got a '/' appended to them
-
- This is now done as part of the filesystem-base module (which is a CommonJS spec): ...
- 07:48 PM Revision 476c4494d7b956bf0d43a450115d51f440d5783f: core/modules: JS files are now UTF-8 (rather t...
- core/modules: JS files are now UTF-8 (rather than ascii/latin-1)
- 04:06 PM Revision 2c200628a94f2aed627fde168619f08b3b0f401d: core/modules: tidy up includes + namespaces
- core/modules: tidy up includes + namespaces
- 04:03 PM Revision 180876cc8a0a56587e3c01feb91b855abe0ce7e8: core: Use canonicalize from fs-base
- core: Use canonicalize from fs-base
- 03:44 PM Revision 03eae0c88509efe08508ade5f50bd50de7fdcecc: Merge branch 'master' into require-refactor
- Merge branch 'master' into require-refactor
- 02:31 PM Revision 5f409d2f876e11edc7fff0cc3db29ba0017ee40f: Changelog filesystem-base module addition
- Changelog filesystem-base module addition
- 02:27 PM Revision 775a23110dface20ca549b599dbb55d70151cb28: core/io: rename file-0 to filesystem-base
- core/io: rename file-0 to filesystem-base
- 01:04 PM Revision f37780b62db1f431f724413089066c161aa11762: core/io: Make optional args in stream.read and...
- core/io: Make optional args in stream.read and readBinary behave
- 12:55 PM Revision 5e4aac2832d26353919a81924783c2488c57b05e: core/file-0: add security checks
- core/file-0: add security checks
- 10:28 AM Revision 91f28d775a568587f39bf8a5fc02bd7097f4d921: core/io/file: support create flag (to ::open a...
- core/io/file: support create flag (to ::open and security)
10/09/2009
-
- Probably worth doing after Feature #44
-
- I've implemented most of this spec, apart from the permissions part as this is still in flux. Not yet merged branch back to master as i need to add in security checks
- 10:22 PM Revision eab16a4e259e1f95923de8218796597493991a14: core/file-0: rawOpen - mostly complete file-{0...
- core/file-0: rawOpen - mostly complete file-{0,engine,base} spec now
- 10:22 PM Revision 42d1a786e0fcc13ca635a0462c7666b69c8993a8: core/io/file: A second constructor taking expl...
- core/io/file: A second constructor taking explicit params.
- 08:52 PM Revision 54545a67906c5af662e2d954d0bfa232b8d2b636: core/file-0: Add more open modes to io.File
- core/file-0: Add more open modes to io.File
-
02:57 PM Bug #129: Flusspferd doesn't build with spidermonkey 1.8.0-rc1
- When trying to build Flusspferd with Spidermonkey 1.8.0-rc1 on Ubuntu 9.04 as described in http://flusspferd.org/docs/ubuntu.html I got the following build error: Linking CXX executable bin/flusspferd lib/libflusspferd.so: undefined reference ...
Also available in: Atom