Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2007-05-01 07:21:23 -0700 (Tue, 01 May 2007)
Revision: 10578
Log message:
Merge to rev 10570.
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2007-05-01 11:12:04 -0700 (Tue, 01 May 2007)
Revision: 10579
Log message:
Add .STATIC rules. This is different from a "static. =" section.
- .STATIC rules are delayed; they are evaluated only if the value is needed.
- .STATIC rules allow dependencies, and explicit variables.
- The values are stored in .omakedb
* Basic usage:
.STATIC:
println(foo)
X = 1
Y = $X
The variable X is exported. The rule is evaluated if the value for $X is needed,
but it is lazy. The definition of Y does not force evaluation.
* With dependencies:
.STATIC: x.input
X = $(expensive-function x.input)
This is be evaluated if x.input changes and X is forced.
* Exporting variables
.STATIC: X: x.input
Y = 1
X = $Y
Y is not exported from the section.
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2007-05-01 14:53:27 -0700 (Tue, 01 May 2007)
Revision: 10583
Log message:
Use reference cells for the delayed values.
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2007-05-01 16:05:54 -0700 (Tue, 01 May 2007)
Revision: 10585
Log message:
Merge to rev 10583.
Changes by: Jason Hickey (jyh at cs.caltech.edu)
Date: 2007-05-01 16:32:38 -0700 (Tue, 01 May 2007)
Revision: 10586
Log message:
No need for all the virtual_var definitions.
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-01 20:18:45 -0700 (Tue, 01 May 2007)
Revision: 10589
Log message:
Backported the export hoisting. This is something that never worked,
but should have.
Z. =
x = 1
f() =
x = 2
export
Z.f()
echo $(Z.x)
# Prints "2"
This works with arbitrary levels of nesting.
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-02 09:19:09 -0700 (Wed, 02 May 2007)
Revision: 10592
Log message:
0.9.8.2 is not being released just yet :-(
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-02 10:37:50 -0700 (Wed, 02 May 2007)
Revision: 10593
Log message:
Merge to rev 10598.
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-02 11:33:32 -0700 (Wed, 02 May 2007)
Revision: 10594
Log message:
Export rules and phonies correctly.
| Changes | Path |
| +3 -5 | omake-branches/0.9.8.x/src/env/omake_env.ml |
| +4 -4 | omake-branches/0.9.8.x/src/eval/omake_eval.ml |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-02 11:41:35 -0700 (Wed, 02 May 2007)
Revision: 10595
Log message:
Merge to rev 10594.
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-02 16:11:43 -0700 (Wed, 02 May 2007)
Revision: 10600
Log message:
Added the path definitions. This fixes bug #580.
X. =
Y. =
Z. =
x = 1
X.Y.Z.y = 2
X.Y.Z.f() =
value $(add $x, $y)
echo $(X.Y.Z.f)
# prints "3"
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-02 16:50:49 -0700 (Wed, 02 May 2007)
Revision: 10601
Log message:
Merge to 10600.
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-02 19:32:00 -0700 (Wed, 02 May 2007)
Revision: 10607
Log message:
Be more careful about specifying field operations explicitly (some field
operations will require checking). No visible change.
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-02 19:46:20 -0700 (Wed, 02 May 2007)
Revision: 10608
Log message:
Rename the eval_find_field_* functions.
| Changes | Path |
| +2 -2 | omake-branches/0.9.8.x/src/build/omake_builtin_util.ml |
| +43 -43 | omake-branches/0.9.8.x/src/eval/omake_eval.ml |
| +3 -2 | omake-branches/0.9.8.x/src/eval/omake_eval.mli |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-02 20:47:19 -0700 (Wed, 02 May 2007)
Revision: 10610
Log message:
Merge to rev 10608.
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-03 10:17:49 -0700 (Thu, 03 May 2007)
Revision: 10613
Log message:
Added the array arguments, bug #645. The syntax is:
- ... for a normal body argument
- [...] for an array body argument
X =
file([...])
a
b
c
- : <array
/Users/jyh/projects/omake/0.9.8.x/a : File
/Users/jyh/projects/omake/0.9.8.x/b : File
/Users/jyh/projects/omake/0.9.8.x/c : File>
: Array
If an application has an elision, you can omit the colon
at the end in interactive mode.
This can still be improved, especially with keyword
arguments. When there are keyword arguments, the argument
lines can get long. It might be nice to allow using
a line break instead of the , separator.
f(X[] =
a
b
c
foo
Y = 1)
Or perhaps allow mixed notation.
f(X = 1
Y = 2
a, b, c)
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-03 10:36:01 -0700 (Thu, 03 May 2007)
Revision: 10615
Log message:
OMake follow-up to the LibMojave commit rev. 10614.
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-03 10:49:05 -0700 (Thu, 03 May 2007)
Revision: 10616
Log message:
Rename StringBody to be ArrayBody to be more accurate.
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-03 11:37:15 -0700 (Thu, 03 May 2007)
Revision: 10618
Log message:
Use the new "dual equality" cons hashing to preserve filename case even when
faced with a case-insensitive filesystem (or OS). Not tested much just yet.
| Changes | Path |
| +111 -30 | omake-branches/0.9.8.2/src/ir/omake_node.ml |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-03 11:49:31 -0700 (Thu, 03 May 2007)
Revision: 10619
Log message:
Follow-up to LibMojave commit rev 10614 (introduction of Lm_hash_sig),
so we can continue to compile.
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-03 16:10:46 -0700 (Thu, 03 May 2007)
Revision: 10621
Log message:
Re-naming to match rev 10620.
I've been thinking about the new scheme. I like the basic idea, but I am
not sure about it.
The entire reason to use cons-hashing is to improve performance of the
compare function. In the original implementation, if the hash is perfect,
comparisons take time O(1). We go through great pains to get good hashes.
I haven't checked recently, but the original implementation showed 2
hash collisions (out of about 1e6 comparisons IIRC) for a MetaPRL compile.
The problem with the new implementation is that the normal operation, the
coarse comparison, now gets a lot of collisions *by definition*.
There is a solution, which is a vertical split, where each fine
node has the hash-consed coarse node as a component. Comparisons
should be based on the coarse node, but the full hash-consing would
be case-sensitive.
To be more explicit, something like this:
type coarse_dir =
DirCoarseRoot of Lm_filename_util.root
| DirCoarseSub of FileCase.t * DirCoarseHash.t
type dir =
DirRoot of Lm_filename_util.root
| DirSub of string * DirHash.t * DirCoarseHash.t
| Changes | Path |
| +107 -52 | omake-branches/0.9.8.x/src/ir/omake_node.ml |
| Properties | omake-branches/0.9.8.x/src/libmojave/ |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-03 17:03:57 -0700 (Thu, 03 May 2007)
Revision: 10622
Log message:
Added some consistency with the 0.9.8.x branch.
| Changes | Path |
| Replaced | omake-branches/0.9.8.2/src/ir/omake_node.ml |
| Copied | omake-branches/0.9.8.2/src/ir/omake_node.ml (from rev 10621, omake-branches/0.9.8.x/src/ir/omake_node.ml) |
| Properties | omake-branches/0.9.8.2/src/libmojave/ |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-03 17:11:25 -0700 (Thu, 03 May 2007)
Revision: 10623
Log message:
svn merge -r 10551:10622 svn+ssh://svn.metaprl.org/svnroot/mojave/omake-branches/0.9.8.2 .
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-03 18:58:34 -0700 (Thu, 03 May 2007)
Revision: 10627
Log message:
Corrected the sense of "fine" in the Node hashing.
Collision performance is good. Here are stats on a build of OMake.
Hash statistics:
Node: reintern = 21, compare = 1054325, collisions = 0
DirList: reintern = 0, compare = 5371, collisions = 0
Dir: reintern = 112, compare = 1122245, collisions = 0
Symbol: reintern = 658, compare = 2861235, collisions = 0
| Changes | Path |
| +3 -3 | omake-branches/0.9.8.x/src/ir/omake_node.ml |
| +12 -2 | omake-branches/0.9.8.x/src/main/omake_main.ml |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-03 19:34:43 -0700 (Thu, 03 May 2007)
Revision: 10628
Log message:
Merge to rev 10627. Here is what I get for hash-cons, I'll try to
see why we're getting collisions.
Hash statistics:
virtual_var: reintern = 206, compare = 357091, collisions = 0
virtual_id: reintern = 127, compare = 645, collisions = 0
Node: reintern = 1946, compare = 1765036, collisions = 31114
DirList: reintern = 0, compare = 439, collisions = 0
Dir: reintern = 695, compare = 894933, collisions = 0
StringHash: reintern = 21, compare = 316, collisions = 0
Symbol: reintern = 933, compare = 135440, collisions = 0
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-03 20:05:13 -0700 (Thu, 03 May 2007)
Revision: 10629
Log message:
Hmmm, the reason why I wasn't seeing collisions was because I
wasn't using ocamldep -modules. Here are the actual stats.
Hash statistics:
Node: reintern = 1868, compare = 1664094, collisions = 28543
DirList: reintern = 0, compare = 277, collisions = 0
Dir: reintern = 817, compare = 903508, collisions = 0
Symbol: reintern = 588, compare = 737854, collisions = 0
Although I stated I was cracked, I take it back:(
The problem is that two nodes with the same "coarse" name will in
general not be pointer-equal. Thus, hash-consing does no good.
We have to hash-cons the coarse version completely separately.
| Changes | Path |
| +5 -5 | omake-branches/0.9.8.x/src/ir/omake_node.ml |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-04 09:01:34 -0700 (Fri, 04 May 2007)
Revision: 10634
Log message:
Microscopic performance improvement in omake_node.
Mainly, the code is a little easier to read without the
conditionals.
| Changes | Path |
| +146 -126 | omake-branches/0.9.8.x/src/ir/omake_node.ml |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-04 10:48:22 -0700 (Fri, 04 May 2007)
Revision: 10637
Log message:
Here is a proposal for fixing the case-sensitive problem.
The basic idea is that we should enforce case-sensitivity
even on insensitive filesystems. If that is too strong, then
we should at least enforce it during scanning. The reason is
that when you port your case-insensitive project to some
other system, you want the build to work the same.
We can't just enforce case-sensitivity by turning off the detection.
The proper solution requires figuring out the actual filenames.
This commit adds Node.real_tail and related functions that
returns the actual filename for the node, which it uses to
implement a strict (case-sensitive) form of target-is-buildable. Let
me know what you think.
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-05 09:33:47 -0700 (Sat, 05 May 2007)
Revision: 10647
Log message:
Revert to rev 10634 from 10637. I think the strict Omake_cache.stat is the
right way to go, but need some feedback from Aleksey.
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-05 10:55:55 -0700 (Sat, 05 May 2007)
Revision: 10648
Log message:
Implemented the strictly case-sensitive Omake_cache.stat, and use only case-sensitive
comparisons on nodes. This should be considered a proposal, to be reviewed.
Potential problems:
- People might wonder why OMake is case-sensitive on Win32/MacOS etc.
Cleanup, if we decide to go this way:
- Nodes use only the fine comparison, so the code can be substantially
cleaned up.
- The case-detection code should be moved to Omake_cache.
| Changes | Path |
| +167 -90 | omake-branches/0.9.8.x/src/ir/omake_cache.ml |
| +8 -3 | omake-branches/0.9.8.x/src/ir/omake_node.ml |
| +6 -0 | omake-branches/0.9.8.x/src/ir/omake_node_sig.ml |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-05 14:20:58 -0700 (Sat, 05 May 2007)
Revision: 10649
Log message:
Split the Omake_cache.stat code into a separate file.
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-05 15:03:34 -0700 (Sat, 05 May 2007)
Revision: 10650
Log message:
Removed canonical names from Omake_node. Add the function "realpath"
to translate name case.
The -Wcase option warn for case errors, but does not terminate the build.
The --no-case-check turns off case checking.
Please remember, this is a proposal, to be reviewed.
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-05 18:38:33 -0700 (Sat, 05 May 2007)
Revision: 10652
Log message:
Merge to rev 10650.
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-05 19:28:40 -0700 (Sat, 05 May 2007)
Revision: 10653
Log message:
If the stat succeeds, and the entry is not in the cache, rescan.
| Changes | Path |
| +21 -11 | omake-branches/0.9.8.x/src/ir/omake_cache_stat.ml |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-06 20:13:08 -0700 (Sun, 06 May 2007)
Revision: 10655
Log message:
Added a simple syntax that is a little more like a standard
programming language. The normal $-style expressions are
always allowed, but in PL mode, identifiers stand for variables,
function application is the f(e1, ..., e2) form, and there
are the standard infix operators.
Note, shell commands and rules never use PL syntax, except
within function arguments.
************************************************************************
*** syntax depends on keyword
Add "program" syntax. This provides a more standard programming
language, where strings must be explicit, and variables represent
applications.
The outer syntax is normal; the program syntax is an ast->ast
translation. The translation is turned on with the command
"program-syntax", which is scoped like "export".
Here is an example:
#!/usr/bin/env osh
program-syntax
f(x) =
return x + 1
println(f(f(1)))
I'll have to document the syntax, but it is pretty normal,
with C-style infix operators and function application.
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-07 12:12:06 -0700 (Mon, 07 May 2007)
Revision: 10658
Log message:
A little cleanup on the syntax implementation. After parsing, flatten
strings as much as possible (for efficiency).
However, keep white-space and normal strings separate. We should probably
use this to improve the Omake_eval string parsers. We have the following
invariants:
- ValString s : s has no whitespace
- ValWhite s : s is only whitespace
- ConstString s : s has not whitespace
- WhiteString s : s is only whitespace
This means we can simplify the string parser, because we don't have to
scan the actual strings for whitespace.
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-07 13:05:57 -0700 (Mon, 07 May 2007)
Revision: 10660
Log message:
Merge dll2 + syntax -> dll-syntax
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-07 13:45:26 -0700 (Mon, 07 May 2007)
Revision: 10661
Log message:
Let's try wrapping fuse.
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-07 13:48:41 -0700 (Mon, 07 May 2007)
Revision: 10662
Log message:
Template from gtk.
| Changes | Path |
| Copied | omake-jumbo-branches/dll-fuse/tests/dll/fuse/ (from rev 10661, omake-jumbo-branches/dll-fuse/tests/dll/gtk) |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-07 13:51:54 -0700 (Mon, 07 May 2007)
Revision: 10663
Log message:
Rename the files.
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-07 20:06:05 -0700 (Mon, 07 May 2007)
Revision: 10670
Log message:
Change the parsing code a little to reflect changes in exports.
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-08 15:13:20 -0700 (Tue, 08 May 2007)
Revision: 10681
Log message:
Make the scope qualifier more uniform.
open build/C
C::CFLAGS += -g
println($(C::CFLAGS))
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-08 15:29:07 -0700 (Tue, 08 May 2007)
Revision: 10682
Log message:
Collapsing open->var3. The scope synatx is really a requirement.
| Changes | Path |
| Deleted | omake-jumbo-branches/var3/ |
| Copied | omake-jumbo-branches/var3.2/ (from rev 10681, omake-jumbo-branches/var3) |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-08 15:29:33 -0700 (Tue, 08 May 2007)
Revision: 10683
Log message:
Retiring var4.
| Changes | Path |
| Deleted | omake-jumbo-branches/var4/ |
| Copied | omake-jumbo-closed/var4/ (from rev 10682, omake-jumbo-branches/var4) |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-08 15:29:47 -0700 (Tue, 08 May 2007)
Revision: 10684
Log message:
Retiring export jumb branch.
| Changes | Path |
| Deleted | omake-jumbo-branches/export/ |
| Copied | omake-jumbo-closed/export/ (from rev 10683, omake-jumbo-branches/export) |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-08 15:30:32 -0700 (Tue, 08 May 2007)
Revision: 10685
Log message:
Renaming open->var3 jumbo branch.
| Changes | Path |
| Deleted | omake-jumbo-branches/open/ |
| Copied | omake-jumbo-branches/var3/ (from rev 10684, omake-jumbo-branches/open) |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-08 15:34:25 -0700 (Tue, 08 May 2007)
Revision: 10686
Log message:
Put var3.2 in the wrong place.
| Changes | Path |
| Deleted | omake-jumbo-branches/var3.2/ |
| Copied | omake-jumbo-closed/var3.2/ (from rev 10685, omake-jumbo-branches/var3.2) |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-08 15:37:14 -0700 (Tue, 08 May 2007)
Revision: 10687
Log message:
Collect the merge:
rename open -> var3
remove var4
remove export
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-09 12:01:28 -0700 (Wed, 09 May 2007)
Revision: 10695
Log message:
Added fuse bindings.
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-10 12:03:02 -0700 (Thu, 10 May 2007)
Revision: 10699
Log message:
Implemented a "hello world" filesystem.
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-11 10:42:45 -0700 (Fri, 11 May 2007)
Revision: 10707
Log message:
Had the sense inverted in the case-senitivity test--I never saw it on OSX because of the "optimal" test.
We were printing root directories wrong. For example:
osh> file(/foo)
//foo : File
| Changes | Path |
| +4 -4 | omake-branches/0.9.8.x/src/ir/omake_cache_stat.ml |
| +2 -1 | omake-branches/0.9.8.x/src/ir/omake_node.ml |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-11 10:49:34 -0700 (Fri, 11 May 2007)
Revision: 10708
Log message:
OMake version of the "hello" filesystem.
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-12 17:56:01 -0700 (Sat, 12 May 2007)
Revision: 10723
Log message:
Add a "build" function, so that builds can be performed from
osh scripts (the function may be called only from osh).
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-13 09:36:03 -0700 (Sun, 13 May 2007)
Revision: 10727
Log message:
Added the "OMake filesystem".
It is pretty simple. When you ask for a file, omake tries
to build it, and if successful, returns a symlink to the file.
Phony targets work too, but the symlink is junk (I should probably
change it so you can at least cat the file).
<jyh@jaoquin 1226>ls -l /tmp/fuse/dir1
total 0
lrwxrwxrwx 1 root wheel 0 Dec 31 1969 OMakefile@ -> /Users/jyh/projects/omake/git/dll-fuse/tests/dll/fuse/omakefs/dir1/OMakefile
lrwxrwxrwx 1 root wheel 0 Dec 31 1969 OMakefile.omc@ -> /Users/jyh/projects/omake/git/dll-fuse/tests/dll/fuse/omakefs/dir1/OMakefile.omc
<jyh@jaoquin 1227>cat /tmp/fuse/dir1/hello.txt
cat: /tmp/fuse/dir1/hello.txt: No such file or directory
Exit 1
<jyh@jaoquin 1228>ls -l /tmp/fuse/dir1
total 0
lrwxrwxrwx 1 root wheel 0 Dec 31 1969 OMakefile@ -> /Users/jyh/projects/omake/git/dll-fuse/tests/dll/fuse/omakefs/dir1/OMakefile
lrwxrwxrwx 1 root wheel 0 Dec 31 1969 OMakefile.omc@ -> /Users/jyh/projects/omake/git/dll-fuse/tests/dll/fuse/omakefs/dir1/OMakefile.omc
<jyh@jaoquin 1229>cat /tmp/fuse/dir1/hello.txt
hello
<jyh@jaoquin 1230>ls -l /tmp/fuse/dir1
total 0
lrwxrwxrwx 1 root wheel 0 Dec 31 1969 OMakefile@ -> /Users/jyh/projects/omake/git/dll-fuse/tests/dll/fuse/omakefs/dir1/OMakefile
lrwxrwxrwx 1 root wheel 0 Dec 31 1969 OMakefile.omc@ -> /Users/jyh/projects/omake/git/dll-fuse/tests/dll/fuse/omakefs/dir1/OMakefile.omc
lrwxrwxrwx 1 root wheel 0 Dec 31 1969 hello.txt@ -> /Users/jyh/projects/omake/git/dll-fuse/tests/dll/fuse/omakefs/dir1/hello.txt
<jyh@jaoquin 1231>ls /tmp/fuse/clean
/tmp/fuse/clean: No such file or directory.
<jyh@jaoquin 1232>ls -l /tmp/fuse/dir1
total 0
lrwxrwxrwx 1 root wheel 0 Dec 31 1969 OMakefile@ -> /Users/jyh/projects/omake/git/dll-fuse/tests/dll/fuse/omakefs/dir1/OMakefile
lrwxrwxrwx 1 root wheel 0 Dec 31 1969 OMakefile.omc@ -> /Users/jyh/projects/omake/git/dll-fuse/tests/dll/fuse/omakefs/dir1/OMakefile.omc
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-13 20:16:59 -0700 (Sun, 13 May 2007)
Revision: 10734
Log message:
Change the handling of .PHONY nodes, so that they appear like real files.
Reading the file builds that target.
<jyh@jaoquin ~/links/omake/svn 1314>ls -l /tmp/fuse
total 7
-r--r--r-- 1 root wheel 1024 Dec 31 1969 .OCAMLLINK
-r--r--r-- 1 root wheel 1024 Dec 31 1969 .PHONY
lrwxrwxrwx 1 root wheel 0 Dec 31 1969 .omakedb@ -> /Users/jyh/projects/omake...
lrwxrwxrwx 1 root wheel 0 Dec 31 1969 .omakedb.lock@ -> /Users/jyh/projects/...
-r--r--r-- 1 root wheel 1024 Dec 31 1969 CGeneratedFilesTarget
-r--r--r-- 1 root wheel 1024 Dec 31 1969 OCamlGeneratedFilesTarget
lrwxrwxrwx 1 root wheel 0 Dec 31 1969 OMakefile@ -> /Users/jyh/projects/omak...
lrwxrwxrwx 1 root wheel 0 Dec 31 1969 OMakefile.omc@ -> /Users/jyh/projects/...
lrwxrwxrwx 1 root wheel 0 Dec 31 1969 OMakeroot@ -> /Users/jyh/projects/omak...
lrwxrwxrwx 1 root wheel 0 Dec 31 1969 OMakeroot.omc@ -> /Users/jyh/projects/...
-r--r--r-- 1 root wheel 1024 Dec 31 1969 TeXGeneratedFilesTarget
-r--r--r-- 1 root wheel 1024 Dec 31 1969 clean
drwxrwxr-x 4 jyh jyh 136 Apr 25 1939 dir1/
lrwxrwxrwx 1 root wheel 0 Dec 31 1969 omakefs@ -> /Users/jyh/projects/omake/...
lrwxrwxrwx 1 root wheel 0 Dec 31 1969 omakefs.omc@ -> /Users/jyh/projects/om...
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-14 09:38:18 -0700 (Mon, 14 May 2007)
Revision: 10738
Log message:
Minor: configure/ncurses was opened twice.
| Changes | Path |
| +1 -2 | omake-branches/0.9.8.x/lib/configure/readline.om |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-14 09:44:56 -0700 (Mon, 14 May 2007)
Revision: 10739
Log message:
Minor: static_sym was defined twice.
| Changes | Path |
| +0 -1 | omake-branches/0.9.8.x/src/ir/omake_symbol.ml |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-14 10:05:28 -0700 (Mon, 14 May 2007)
Revision: 10740
Log message:
A temporary hack of 0.9.8.2 only: in AST-TO-IR conversion, recognize the
".STATIC:" section header as if it was "static. =". This is a HACK that is
meant to add a limited form of forward-compatibility with the upcoming
0.9.8.3+ syntax.
| Changes | Path |
| +14 -1 | omake-branches/0.9.8.2/src/env/omake_ir_ast.ml |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-14 10:12:33 -0700 (Mon, 14 May 2007)
Revision: 10741
Log message:
EMPTY and EMPTY_ARRAY need to be declared?
| Changes | Path |
| +2 -0 | omake-branches/0.9.8.x/lib/Pervasives.om |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-14 10:13:06 -0700 (Mon, 14 May 2007)
Revision: 10742
Log message:
Small follow-up changes to case sensitivily, as discussed on omake-devel.
| Changes | Path |
| +1 -0 | omake-branches/0.9.8.x/src/ir/omake_cache.ml |
| +6 -26 | omake-branches/0.9.8.x/src/ir/omake_cache_stat.ml |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-14 10:27:37 -0700 (Mon, 14 May 2007)
Revision: 10743
Log message:
Merging the case-[in]sensitivity changes from 0.9.8.x:
svn merge -r 10623:10707 svn+ssh://svn.metaprl.org/svnroot/mojave/omake-branches/0.9.8.x .
svn merge -r 10738:10742 svn+ssh://svn.metaprl.org/svnroot/mojave/omake-branches/0.9.8.x .
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-14 11:45:05 -0700 (Mon, 14 May 2007)
Revision: 10744
Log message:
dir_is_sensitive must always return "false" on Windows, not "true"!
| Changes | Path |
| +1 -1 | omake-branches/0.9.8.x/src/ir/omake_cache_stat.ml |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-14 11:45:09 -0700 (Mon, 14 May 2007)
Revision: 10745
Log message:
dir_is_sensitive must always return "false" on Windows, not "true"!
| Changes | Path |
| +1 -1 | omake-branches/0.9.8.2/src/ir/omake_cache_stat.ml |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-14 11:56:41 -0700 (Mon, 14 May 2007)
Revision: 10746
Log message:
Add nativeint, int32, and int64 values.
Use Nativeint.t instead of int in the Dll interface. For example, the dates are now
accurate.
-r--r--r-- 1 root wheel 1024 Dec 31 1969 .OCAMLLINK
-r--r--r-- 1 root wheel 1024 Dec 31 1969 .PHONY
lrwxrwxrwx 1 root wheel 0 Dec 31 1969 .omakedb@ -> /Users/jyh/projects/...
lrwxrwxrwx 1 root wheel 0 Dec 31 1969 .omakedb.lock@ -> /Users/jyh/proj...
-r--r--r-- 1 root wheel 1024 Dec 31 1969 CGeneratedFilesTarget
-r--r--r-- 1 root wheel 1024 Dec 31 1969 OCamlGeneratedFilesTarget
lrwxrwxrwx 1 root wheel 0 Dec 31 1969 OMakefile@ -> /Users/jyh/projects...
lrwxrwxrwx 1 root wheel 0 Dec 31 1969 OMakefile.omc@ -> /Users/jyh/proj...
lrwxrwxrwx 1 root wheel 0 Dec 31 1969 OMakeroot@ -> /Users/jyh/projects...
lrwxrwxrwx 1 root wheel 0 Dec 31 1969 OMakeroot.omc@ -> /Users/jyh/proj...
-r--r--r-- 1 root wheel 1024 Dec 31 1969 TeXGeneratedFilesTarget
-r--r--r-- 1 root wheel 1024 Dec 31 1969 clean
drwxrwxr-x 4 jyh jyh 136 May 14 11:28 dir1/
lrwxrwxrwx 1 root wheel 0 Dec 31 1969 omakefs@ -> /Users/jyh/projects/o...
lrwxrwxrwx 1 root wheel 0 Dec 31 1969 omakefs.omc@ -> /Users/jyh/projec...
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-14 13:26:26 -0700 (Mon, 14 May 2007)
Revision: 10749
Log message:
For now, reverting the download page to its 0.9.8.1 state.
| Changes | Path |
| +7 -16 | omake-branches/0.9.8.x/doc/html/download.html |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-14 14:20:35 -0700 (Mon, 14 May 2007)
Revision: 10750
Log message:
Going back to the case-insensitive case-preserving nodes on case-insensitive
filesystems. Namely, I reverted 10744:10741 and 10650:10636, while making sure
that the rev 10707 fix is copied from omake_cache_stat back to omake_node...
Note that this have killed the "realpath" built-in function. Once the node
stuff is stable, we might want to implement it again.
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-14 14:25:36 -0700 (Mon, 14 May 2007)
Revision: 10751
Log message:
Minor documentation fix.
| Changes | Path |
| +3 -3 | omake-branches/0.9.8.x/src/builtin/omake_builtin_file.ml |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-14 16:38:15 -0700 (Mon, 14 May 2007)
Revision: 10753
Log message:
Implemented the find-ocaml-targets-in-path-optional function (not fully tested
yet).
| Changes | Path |
| +2 -8 | omake-branches/0.9.8.x/lib/build/OCaml.om |
| +46 -24 | omake-branches/0.9.8.x/src/builtin/omake_builtin_file.ml |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-14 16:59:26 -0700 (Mon, 14 May 2007)
Revision: 10754
Log message:
When Omake_value_type was created in rev. 10570, the omake_shell_sys_win32.ml was not updated accordingly.
| Changes | Path |
| +1 -1 | omake-branches/0.9.8.x/src/shell/omake_shell_sys_win32.ml |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-15 08:09:25 -0700 (Tue, 15 May 2007)
Revision: 10755
Log message:
NOTE: absname(/) was always ".", fixed. Need to propagate back to 0.9.8.
Some minor changes to allow mounting from within omake.
It would be nice to have more flexibility in mounting.
/vol/omake:
# Blocking: build requests are through the filesystem.
mount($@)
But this is too hard, because it is a recursive call to the build loop.
OTOH, we could have the mount complete immediately, and serve it
during the wait loop (the -P loop).
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-15 09:10:30 -0700 (Tue, 15 May 2007)
Revision: 10756
Log message:
Prepare to split the int32 etc. into a separate branch.
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-15 09:15:48 -0700 (Tue, 15 May 2007)
Revision: 10757
Log message:
Push back some of the Dll changes so that all branches compile again.
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-15 09:28:25 -0700 (Tue, 15 May 2007)
Revision: 10759
Log message:
Reverting back to case-insensitive case-preserving implementation and adding a
special function for searching for OCaml targets. This is a merge off the
0.9.8.x branch (revs 10749:10753).
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-15 09:52:45 -0700 (Tue, 15 May 2007)
Revision: 10761
Log message:
Solving the absname(/) problem using the "impossible root" approach.
| Changes | Path |
| +7 -2 | omake-branches/0.9.8.x/src/ir/omake_node.ml |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-15 12:03:03 -0700 (Tue, 15 May 2007)
Revision: 10762
Log message:
- When the executable is already a node, we still need to check whether the .exe (or similar) suffix needs to be added. I have no idea why this worked before and how it got broken, but it did have to be fixed, otherwise even self-builds would break under Windows...
- When omake_shell_sys_create_process (omake_shell_sys.c, Win32) fails to create the process, try including the string representation of the error in the failure exception.
- Couple more warnings need to be disabled in order to be able to compile with Visual Studio 7.0 (AKA 2002 AKA .NET 1.0). Not sure we still care (7.1/2003/.NET 1.1 and 8.0/2005/.NET 2.0 compile fine anyway), but made these changes anyway.
| Changes | Path |
| +27 -6 | omake-branches/0.9.8.x/src/clib/omake_shell_sys.c |
| +1 -1 | omake-branches/0.9.8.x/src/clib/readline.c |
| +4 -1 | omake-branches/0.9.8.x/src/shell/omake_shell_job.ml |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-15 12:05:40 -0700 (Tue, 15 May 2007)
Revision: 10764
Log message:
Merged revs 10760:10762 from the 0.9.8.x branch.
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-15 15:28:25 -0700 (Tue, 15 May 2007)
Revision: 10766
Log message:
Adding ODBC.
| Changes | Path |
| Copied | omake-jumbo-branches/dll-odbc/ (from rev 10765, omake-jumbo-branches/dll-syntax) |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-15 15:32:54 -0700 (Tue, 15 May 2007)
Revision: 10767
Log message:
Initial ODBC binding.
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-15 16:11:29 -0700 (Tue, 15 May 2007)
Revision: 10768
Log message:
Complete query example (idea from David's mythfs query).
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-16 17:22:47 -0700 (Wed, 16 May 2007)
Revision: 10771
Log message:
JDBC-style interface for omake.
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-17 15:43:46 -0700 (Thu, 17 May 2007)
Revision: 10772
Log message:
See about so-called "clean" canonical nodes.
| Changes | Path |
| Copied | omake-branches/0.9.8.x.case/ (from rev 10749, omake-branches/0.9.8.x) |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-20 18:01:20 -0700 (Sun, 20 May 2007)
Revision: 10779
Log message:
Removing "case" branch. I'll agree that the reorg is a hassle,
so for now we go with the old version.
| Changes | Path |
| Deleted | omake-branches/0.9.8.x.case/ |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-21 09:48:00 -0700 (Mon, 21 May 2007)
Revision: 10782
Log message:
"make boot" -> "make bootstrap"
Under Windows, "nmake boot" confuses the phony target "boot" with the "boot"
directory. On Unix this is not a problem, but it is nice to be consistent.
| Changes | Path |
| +2 -2 | omake-branches/0.9.8.x/INSTALL |
| +5 -5 | omake-branches/0.9.8.x/Makefile |
| +1 -1 | omake-branches/0.9.8.x/Makefile.nt |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-21 09:49:31 -0700 (Mon, 21 May 2007)
Revision: 10783
Log message:
"make boot" -> "make bootstrap"
Under Windows, "nmake boot" confuses the phony target "boot" with the "boot"
directory. On Unix this is not a problem, but it is nice to be consistent.
| Changes | Path |
| +2 -2 | omake-branches/0.9.8.2/INSTALL |
| +5 -5 | omake-branches/0.9.8.2/Makefile |
| +1 -1 | omake-branches/0.9.8.2/Makefile.nt |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-21 18:05:02 -0700 (Mon, 21 May 2007)
Revision: 10786
Log message:
Correctly parse the output of OCaml 3.10's "ocamldep -modules"
| Changes | Path |
| +1 -1 | omake-branches/0.9.8.2/lib/build/OCaml.om |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-21 18:05:09 -0700 (Mon, 21 May 2007)
Revision: 10787
Log message:
Correctly parse the output of OCaml 3.10's "ocamldep -modules"
| Changes | Path |
| +1 -1 | omake-branches/0.9.8.x/lib/build/OCaml.om |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-21 19:58:56 -0700 (Mon, 21 May 2007)
Revision: 10789
Log message:
Merge to rev 10787.
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-22 15:19:32 -0700 (Tue, 22 May 2007)
Revision: 10790
Log message:
Preparing for the 0.9.8.2-2 release:
- Bumped the version number to 0.9.8.2-2
- Filled in the release date for 0.9.8.2 (tomorrow - May 23, 2007)
- Regenerated the documentation (had to make a small fix in
omake_builtin_file.ml)
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-22 15:21:54 -0700 (Tue, 22 May 2007)
Revision: 10791
Log message:
Merging in the "pre-release" changes from the 0.9.8.2 branch (except for the
"downloads" page - this one I will update when the downloads actually become
available).
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-22 15:29:21 -0700 (Tue, 22 May 2007)
Revision: 10792
Log message:
Minor update.
| Changes | Path |
| +1 -5 | omake-branches/0.9.8.2/omake.aip |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-22 15:30:08 -0700 (Tue, 22 May 2007)
Revision: 10793
Log message:
Rev. 10792 merge from 0.9.8.2
| Changes | Path |
| +1 -5 | omake-branches/0.9.8.x/omake.aip |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-22 15:40:01 -0700 (Tue, 22 May 2007)
Revision: 10794
Log message:
Use 0.9.8.2, not 0.9.8.x!
| Changes | Path |
| +1 -1 | omake-branches/0.9.8.2/make_checkout |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-22 15:52:04 -0700 (Tue, 22 May 2007)
Revision: 10795
Log message:
Include version number in the product name.
| Changes | Path |
| +1 -1 | omake-branches/0.9.8.2/omake.aip |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-22 15:56:08 -0700 (Tue, 22 May 2007)
Revision: 10796
Log message:
Include version number in the product name.
| Changes | Path |
| +1 -1 | omake-branches/0.9.8.x/omake.aip |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-22 16:24:08 -0700 (Tue, 22 May 2007)
Revision: 10797
Log message:
We now require OCaml >= 3.09.2
| Changes | Path |
| +1 -1 | omake-branches/0.9.8.2/omake.spec |
| +1 -1 | omake-branches/0.9.8.2/omake.spec.in |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-22 16:24:56 -0700 (Tue, 22 May 2007)
Revision: 10798
Log message:
We now require OCaml >= 3.09.2
| Changes | Path |
| +1 -1 | omake-branches/0.9.8.x/omake.spec |
| +1 -1 | omake-branches/0.9.8.x/omake.spec.in |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-22 17:06:40 -0700 (Tue, 22 May 2007)
Revision: 10800
Log message:
Rev 10795 follow-up: now that we have the version number in the "Product Name" field, we should not be using "[ProductName]" shortcut where the "OMake" is meant. Similarly, for the "[Manufacturer]" key where "MetaPRL" is expected.
| Changes | Path |
| +8 -8 | omake-branches/0.9.8.2/omake.aip |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-22 17:07:58 -0700 (Tue, 22 May 2007)
Revision: 10801
Log message:
Copying off the 0.9.8.2 branch.
| Changes | Path |
| Replaced | omake-branches/0.9.8.x/omake.aip |
| Copied | omake-branches/0.9.8.x/omake.aip (from rev 10800, omake-branches/0.9.8.2/omake.aip) |
| +1 -1 | omake-branches/0.9.8.x/omake.aip (from rev 10800, omake-branches/0.9.8.2/omake.aip) |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-24 09:51:50 -0700 (Thu, 24 May 2007)
Revision: 10809
Log message:
Be a bit more careful not to do
try
...;
Unix.close fd;
...
with exn ->
close fd;
raise exn
since this may result in closing the same fd twice in case the second "..."
raises an exception.
| Changes | Path |
| +8 -6 | omake-branches/0.9.8.2/src/builtin/omake_builtin_file.ml |
| +5 -3 | omake-branches/0.9.8.2/src/ir/omake_cache.ml |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-24 10:02:30 -0700 (Thu, 24 May 2007)
Revision: 10811
Log message:
Rename a "local" protected variable to avoid clashing with a commonly used one :-(
| Changes | Path |
| +5 -5 | omake-branches/0.9.8.2/lib/build/OCaml.om |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-24 12:47:11 -0700 (Thu, 24 May 2007)
Revision: 10812
Log message:
Fixed the "close(): bad file descriptor" problem.
The issue was that the Omake_shell_sys.set_close_on_exec was called,
but Unix.close was used to close the descriptor. Omake_shell_sys.close_fd
should be used instead, to delete the descriptor from the fork table too.
| Changes | Path |
| +8 -2 | omake-branches/0.9.8.2/src/build/omake_build.ml |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-24 12:47:44 -0700 (Thu, 24 May 2007)
Revision: 10813
Log message:
Oops, reversed the sense of a comment.
| Changes | Path |
| +1 -1 | omake-branches/0.9.8.2/src/build/omake_build.ml |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-24 14:57:51 -0700 (Thu, 24 May 2007)
Revision: 10814
Log message:
The directory got lost in commit 10241.
| Changes | Path |
| +3 -2 | omake-branches/0.9.8.2/src/build/omake_build.ml |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-25 11:03:13 -0700 (Fri, 25 May 2007)
Revision: 10815
Log message:
Various minor changes, mainly adding a few stubs.
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-27 12:00:58 -0700 (Sun, 27 May 2007)
Revision: 10816
Log message:
Backport the sort function.
| Changes | Path |
| +62 -1 | omake-branches/0.9.8.x/src/builtin/omake_builtin_object.ml |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-27 18:32:38 -0700 (Sun, 27 May 2007)
Revision: 10818
Log message:
Added "replace-nth" function.
| Changes | Path |
| +28 -0 | omake-branches/0.9.8.x/src/builtin/omake_builtin_base.ml |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-29 09:43:51 -0700 (Tue, 29 May 2007)
Revision: 10825
Log message:
- Added the replace_nth function (merges rev 10818 from 0.9.8.x)
- Include the function name in the "index is out of bounds" messages for nth
and replace-nth.
| Changes | Path |
| +29 -1 | omake-branches/0.9.8.2/src/builtin/omake_builtin_base.ml |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-29 09:45:12 -0700 (Tue, 29 May 2007)
Revision: 10826
Log message:
Include the function name in the "index is out of bounds" messages for the
*nth* functions.
| Changes | Path |
| +4 -4 | omake-branches/0.9.8.x/src/builtin/omake_builtin_base.ml |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-29 09:46:06 -0700 (Tue, 29 May 2007)
Revision: 10827
Log message:
Include the function name in the "index is out of bounds" messages for the
nth-{hd,tl} functions (merges rev. 10826 from the 0.9.8.x).
| Changes | Path |
| +2 -2 | omake-branches/0.9.8.2/src/builtin/omake_builtin_base.ml |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-29 10:30:26 -0700 (Tue, 29 May 2007)
Revision: 10829
Log message:
Fix the documentation for "sort", and add the methods to the Sequence object.
| Changes | Path |
| +31 -5 | omake-branches/0.9.8.x/lib/Pervasives.om |
| +12 -12 | omake-branches/0.9.8.x/src/builtin/omake_builtin_object.ml |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-29 15:37:26 -0700 (Tue, 29 May 2007)
Revision: 10830
Log message:
The static section indexing resulted in duplicates.
This fixes bug #669.
This is already done correctly in 0.9.8.x, so I just hacked it
the easiest way possible, and used a ref cell for the static
index. Do not propagate this hack.
| Changes | Path |
| +11 -6 | omake-branches/0.9.8.2/src/env/omake_ir_ast.ml |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-30 14:51:54 -0700 (Wed, 30 May 2007)
Revision: 10832
Log message:
Use the same debug_db variable in Omake_env and Lm_db.
| Changes | Path |
| +3 -8 | omake-branches/0.9.8.2/src/env/omake_env.ml |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-30 15:06:34 -0700 (Wed, 30 May 2007)
Revision: 10833
Log message:
Use the same debug_db variable in Omake_env and Lm_db. (Merges rev. 10832 off
the 0.9.8.2 branch).
| Changes | Path |
| +1 -6 | omake-branches/0.9.8.x/src/env/omake_env.ml |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-30 19:53:21 -0700 (Wed, 30 May 2007)
Revision: 10835
Log message:
Merge to rev 10833.
I had a NFS server failure partway through, and got a fatally-corrupt
git repository. Bleh, but I believe I was able to recover from some
backup copies I had. The failed repo is dll-syntax.
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-30 20:09:54 -0700 (Wed, 30 May 2007)
Revision: 10836
Log message:
Create a potentially-final branch, "everything".
| Changes | Path |
| Copied | omake-jumbo-branches/everything/ (from rev 10835, omake-jumbo-branches/dll-odbc) |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-30 20:19:22 -0700 (Wed, 30 May 2007)
Revision: 10837
Log message:
Had a case-insensitive error here, Test.om and test.om.
| Changes | Path |
| Deleted | omake-jumbo-branches/everything/tests/dll/gtk/Test.om |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-30 20:24:34 -0700 (Wed, 30 May 2007)
Revision: 10838
Log message:
Merge "everything".
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-31 13:00:35 -0700 (Thu, 31 May 2007)
Revision: 10839
Log message:
Instead of using the OCAMLDEP_PRESERVE_TARGETS variable (which causes scoping
mismatch issues in the "stable" 0.9.8.2 branch), pass the value explicitly.
This fixed the problem Mike Furr pointed out.
P.S. The "semi-stable" 0.9.8.x branch does this correctly, so this change
should _not_ be merged there.
| Changes | Path |
| +10 -14 | omake-branches/0.9.8.2/lib/build/OCaml.om |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-31 13:15:22 -0700 (Thu, 31 May 2007)
Revision: 10840
Log message:
When passing a command line as a command line argument (e.g. with menhir's
--ocamldep and --ocamlc options), we need to use $(quote-argv ...) instead of
just $"..." to make sure that special symbols that are already quoted in the
argument, get "double-quoted" appropriately.
| Changes | Path |
| +2 -2 | omake-branches/0.9.8.2/lib/build/OCaml.om |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-31 13:23:43 -0700 (Thu, 31 May 2007)
Revision: 10841
Log message:
- When passing a command line as a command line argument (e.g. with menhir's
--ocamldep and --ocamlc options), we need to use $(quote-argv ...) instead
of just $"..." to make sure that special symbols that are already quoted in
the argument, get "double-quoted" appropriately.
This merges the r10840 off the 0.9.8.2 "stable" branch.
- Changed the default value of OCAMLDEP_MODULES_ENABLED to
$(OCAMLDEP_MODULES_AVAILABLE) instead of false and updated the documentation
accordingly.
| Changes | Path |
| +12 -15 | omake-branches/0.9.8.x/lib/build/OCaml.om |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-31 13:25:40 -0700 (Thu, 31 May 2007)
Revision: 10842
Log message:
Minor documentation change - OCaml 3.10 was released and the "ocamldep
-modules" is available there.
This was already included in r10841 on the "semi-stable" 0.9.8.x branch, and
should not be re-merged there.
| Changes | Path |
| +2 -3 | omake-branches/0.9.8.2/lib/build/OCaml.om |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-31 14:01:12 -0700 (Thu, 31 May 2007)
Revision: 10843
Log message:
On the second thought, let's set
OCAMLDEP_MODULES_ENABLED = $(OCAMLDEP_MODULES_AVAILABLE)
on the "stable" branch as well.
| Changes | Path |
| +8 -10 | omake-branches/0.9.8.2/lib/build/OCaml.om |
| +4 -4 | omake-branches/0.9.8.2/mk/make_config |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-31 14:02:01 -0700 (Thu, 31 May 2007)
Revision: 10844
Log message:
Comment update (same as r10843 on "stable" 0.9.8.2 branch).
| Changes | Path |
| +4 -4 | omake-branches/0.9.8.x/mk/make_config |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-31 15:01:42 -0700 (Thu, 31 May 2007)
Revision: 10845
Log message:
Preparing the 0.9.8.3 release:
- renamed the branch
- updated the version number to 0.9.8.3-1
- updated the 0.9.8.x version number to indicate it's at least 0.9.8.3+
- updated the changelog and the download page
- regenerated the documentation
- copied the above changes to changelog and documentation (but not download
page) to 0.9.8.x
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-31 15:01:42 -0700 (Thu, 31 May 2007)
Revision: 10845
Log message:
Preparing the 0.9.8.3 release:
- renamed the branch
- updated the version number to 0.9.8.3-1
- updated the 0.9.8.x version number to indicate it's at least 0.9.8.3+
- updated the changelog and the download page
- regenerated the documentation
- copied the above changes to changelog and documentation (but not download
page) to 0.9.8.x
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-31 15:04:13 -0700 (Thu, 31 May 2007)
Revision: 10846
Log message:
Use the right branch
| Changes | Path |
| +1 -1 | omake-branches/0.9.8.3/make_checkout |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-31 15:15:06 -0700 (Thu, 31 May 2007)
Revision: 10847
Log message:
Built 0.9.8.3-1 on Windows.
| Changes | Path |
| +3 -3 | omake-branches/0.9.8.3/omake.aip |
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-31 15:23:57 -0700 (Thu, 31 May 2007)
Revision: 10848
Log message:
Preserve the venv during shell calls. This fixes the problem with
Shell.cd not having any effect.
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-31 17:15:11 -0700 (Thu, 31 May 2007)
Revision: 10850
Log message:
When evaluation is forced on a .STATIC variable that is bound to a lazy value,
the lazy value should be forced as well (of corse, the .STATIC variable will
remain bound to the original lazy value).
| Changes | Path |
| +2 -2 | omake-branches/0.9.8.x/src/eval/omake_eval.ml |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-31 17:16:11 -0700 (Thu, 31 May 2007)
Revision: 10851
Log message:
Automatic update based on version.txt
| Changes | Path |
| +1 -1 | omake-branches/0.9.8.x/omake.spec |
Changes by: Aleksey Nogin (nogin at metaprl.org)
Date: 2007-05-31 17:18:04 -0700 (Thu, 31 May 2007)
Revision: 10852
Log message:
Rearranged the .STATIC sections to make things lazier.
Basically, the issue was that things of the sort
.STATIC:
XYZ_AVAILABLE = ...
XYZ_ENABLED = $(XYZ_AVAILABLE)
force the static section to be evaluated right away. My solution is to do
.STATIC:
XYZ_AVAILABLE =
XYZ_ENABLED = $(XYZ_AVAILABLE)
instead.
Jason, is this reasonable?
Changes by: Jason J. Hickey (jyh at cs.caltech.edu)
Date: 2007-05-31 18:44:52 -0700 (Thu, 31 May 2007)
Revision: 10853
Log message:
Merge revs 10812-10814.
- fixed close descriptor error.
- use the proper directory for targets.
| Changes | Path |
| +1 -1 | omake-branches/0.9.8.x/src/Makefile |
| +1 -1 | omake-branches/0.9.8.x/src/Makefile.nt |
| +9 -1 | omake-branches/0.9.8.x/src/build/omake_build.ml |