depot - monorepo for the virus lounge

Age	Commit message (Collapse)	Author	Files	Lines
2010-04-21	* Don't use an ostringstream in unparseDerivation(), because it's	Eelco Dolstra	1	-38/+39
	slow. A `nix-env -qa --drv-path \*' on Nixpkgs was spending 40% of its time in unparseDerivation() because of this (now 11%).
2010-04-21	* Update the expected test output (no longer an ATerm).	Eelco Dolstra	35	-39/+38

2010-04-21	* Fix the interpretation of ''\<character> in indented strings.	Eelco Dolstra	1	-1/+1

2010-04-21	* Because --parse-only no longer produces an ATerm, don't check the	Eelco Dolstra	6	-10/+1
	output. Whether it parses at all should be enough.
2010-04-21	* Fixed builtins.genericClosure.	Eelco Dolstra	3	-26/+62

2010-04-21	* Store user environment manifests as a Nix expression in	Eelco Dolstra	11	-183/+205
	$out/manifest.nix rather than as an ATerm. (Hm, I thought I committed this two days ago...)
2010-04-19	* Drop the dependency on the ATerm library.	Eelco Dolstra	21	-413/+32

2010-04-19	* Don't use the ATerm library for parsing/printing .drv files.	Eelco Dolstra	14	-330/+174

2010-04-19	* Added parsing of manifests in ATerm format.	Eelco Dolstra	3	-13/+165

2010-04-19	* Refactoring: move the user environment stuff into its own module.	Eelco Dolstra	4	-20/+47

2010-04-16	* Updated some more primops.	Eelco Dolstra	3	-58/+37

2010-04-16	* _combineChannels shouldn't be an integer.	Eelco Dolstra	1	-1/+1

2010-04-16	* Improve sharing a bit.	Eelco Dolstra	2	-5/+8

2010-04-16	* Remove some redundant tests.	Eelco Dolstra	1	-3/+3

2010-04-16	* In an nested `with' where the inner with is a variable (`with ...;	Eelco Dolstra	3	-7/+15
	with someVar; ...'), the contents of the variable would be clobbered. (The attributes in the outer `with' were added to the variable.)
2010-04-15	* Store lists as lists of pointers to values rather than as lists of	Eelco Dolstra	7	-34/+41
	values. This improves sharing and gives another speed up. Evaluation of the NixOS system attribute is now almost 7 times faster than the old evaluator.
2010-04-14	* Better stats.	Eelco Dolstra	2	-5/+16

2010-04-14	* Implemented inherit.	Eelco Dolstra	4	-29/+34

2010-04-14	* Fix builtins.	Eelco Dolstra	4	-28/+30

2010-04-14	* Refactoring: move variable uses to a separate class.	Eelco Dolstra	4	-35/+31

2010-04-14	* Implemented withs.	Eelco Dolstra	4	-15/+38

2010-04-14	* After parsing, compute level/displacement pairs for each variable	Eelco Dolstra	7	-135/+250
	use site, allowing environments to be stores as vectors of values rather than maps. This should speed up evaluation and reduce the number of allocations.
2010-04-14	* Remove more obsolete code.	Eelco Dolstra	5	-503/+2

2010-04-14	* Get nix-env to compile again.	Eelco Dolstra	1	-29/+14

2010-04-14	* Change the semantics of "with" so that inner "withs" take	Eelco Dolstra	2	-25/+2
	precedence, i.e. `with {x=1;}; with {x=2;}; x' evaluates to 2'. This has a simpler implementation and seems more natural. There doesn't seem to be any code in Nixpkgs or NixOS that relies on the old behaviour.
2010-04-13	* Use std::tr1::unordered_set instead of std::set for the symbol	Eelco Dolstra	2	-1/+8
	table. This gives a 10% speed increase on `nix-instantiate /etc/nixos/nixos -A system --readonly-mode'.
2010-04-13	* Evaluate lets directly (i.e. without desugaring to `rec { attrs...;	Eelco Dolstra	5	-1/+45
	<let-body> = e; }.<let-body>). This prevents the unnecessary allocation of an attribute set.
2010-04-13	* Use a symbol table to represent identifiers and attribute names	Eelco Dolstra	15	-101/+228
	efficiently. The symbol table ensures that there is only one copy of each symbol, thus allowing symbols to be compared efficiently using a pointer equality test.
2010-04-12	* Finished the ATerm-less parser.	Eelco Dolstra	6	-102/+68

2010-04-12	* Remove some obsolete functions.	Eelco Dolstra	2	-52/+0

2010-04-12	* Indented strings.	Eelco Dolstra	7	-70/+65

2010-04-12	* More missing constructs.	Eelco Dolstra	6	-137/+160

2010-04-12	* Don't use ATerms for the abstract syntax trees anymore. Not	Eelco Dolstra	19	-579/+693
	finished yet.
2010-04-12	* Don't use ATerms to represent integers in the lexer.	Eelco Dolstra	2	-3/+5

2010-04-12	* Hack to support builderDefs expressions.	Eelco Dolstra	1	-1/+7

2010-04-12	* Doh.	Eelco Dolstra	1	-1/+1

2010-04-12	(no commit message)	Eelco Dolstra	1	-55/+58

2010-04-09	* Keep more statistics about stack space usage.	Eelco Dolstra	2	-43/+78
	* Reduce stack space usage.
2010-04-08	* Remove a lot of dead code.	Eelco Dolstra	4	-763/+10

2010-04-08	* Fix blackholing. If evaluation fails due to an assertion failure,	Eelco Dolstra	1	-2/+8
	then the blackhole has to be removed to ensure that repeated evaluation of the same value gives an assertion failure again rather than an "infinite recursion" error.
2010-04-07	* Update autoCallFunction() and findAlongAttrPath().	Eelco Dolstra	10	-188/+120

2010-04-07	* expr-to-xml -> value-to-xml.	Eelco Dolstra	5	-10/+15

2010-04-07	* Implemented the primops necessary for generating the NixOS manual.	Eelco Dolstra	8	-145/+159

2010-04-06	* In eval(), don't use the target value `v' as a temporary.	Eelco Dolstra	1	-13/+17
	Overwriting `v' breaks when the expression evaluation to an assertion failure or throw.
2010-04-01	* Removed the `~' operator.	Eelco Dolstra	3	-3/+1

2010-04-01	* Quick hack to make coerceToString work more or less correctly on	Eelco Dolstra	1	-1/+4
	nested lists. `nix-instantiate' can now evaluate the NixOS system derivation attribute correctly (in 2.1s on my laptop vs. 6.2s for the trunk).
2010-04-01	* Improve sharing.	Eelco Dolstra	3	-19/+25

2010-04-01	* Functions are incomparable.	Eelco Dolstra	1	-6/+13

2010-04-01	* Make `derivation' lazy again for performance. It also turns out	Eelco Dolstra	3	-37/+17
	that there are some places in Nixpkgs (php_configurable / composableDerivation, it seems) that call `derivation' with incorrect arguments (namely, the `name' attribute missing) but get away with it because of laziness.
2010-03-31	* Fixed the trace primop and path comparison.	Eelco Dolstra	2	-41/+11
	* Removed exprToString and stringToExpr because there is no ATerm representation to work on anymore (and exposing the internals of the evaluator like this is not a good idea anyway).