depot - monorepo for the virus lounge

Age	Commit message (Collapse)	Author	Files	Lines
2004-02-13	* Only include predecessors that are themselves being pushed.	Eelco Dolstra	1	-1/+6
	Otherwise the substitute mechanism can break in subtle ways.
2004-02-10	* Lots of manual stuff. Reference pages for most Nix commands.	Eelco Dolstra	1	-1/+1
	* nix-pull now requires the full url to the manifest, i.e., `/MANIFEST/' is no longer automatically appended. * nix-prefetch-url works again.
2004-01-14	* Synchronous `nix-pull' with `nix-push'.	Eelco Dolstra	1	-1/+1
	* Use curl instead of wget.
2004-01-14	* Improved `nix-push': it now uses HTTP PUT (instead of rsync) to copy	Eelco Dolstra	1	-42/+77
	files. Target location is no longer hard-coded; it accepts a number of URLs on the command line. * `nix-install-package': compatibility fixes.
2003-12-21	* Missing semicolons.	Eelco Dolstra	1	-1/+1

2003-12-01	* Bug fix in nix-push.	Eelco Dolstra	1	-19/+19

2003-12-01	* Use a system name that does not include the OS manufacturer (i.e.,	Eelco Dolstra	1	-1/+1
	"i686-linux" instead of "i686-suse-linux").
2003-11-22	* Fix nix-push.	Eelco Dolstra	1	-36/+32

2003-10-20	* NarPath -> NarName.	Eelco Dolstra	1	-1/+1

2003-10-16	* Get nix-push to work again.	Eelco Dolstra	1	-32/+78
	* Fixed svn:ignore on externals/.
2003-10-07	* Synchronise terminology with the ICSE paper (e.g., slice -> closure,	Eelco Dolstra	1	-2/+2
	fstate -> Nix expression). * Fix src/test.cc.
2003-08-28	* nix-push generated invalid (old-style) slices.	Eelco Dolstra	1	-1/+1
	* nar.sh needs a path.
2003-08-05	* Made nix-push much faster.	Eelco Dolstra	1	-18/+33

2003-07-29	* Use `--query --requisites' and include successors when pushing. Don't	Eelco Dolstra	1	-8/+8
	use `--query --generators' anymore.
2003-07-21	* Syntax fixes.	Eelco Dolstra	1	-10/+20
	* When pushing, put the hash in the file name so that the client can verify (proof-carrying file names?).
2003-07-16	* The write() system call can write less than the requested	Eelco Dolstra	1	-0/+4
	number of bytes, e.g., in case of a signal like SIGSTOP. This caused `nix --dump' to fail sometimes. Note that this bug went unnoticed because the call to `nix --dump' is in a pipeline, and the shell ignores non-zero exit codes from all but the last element in the pipeline. Is there any way to check the result of the initial elements in the pipeline? (In other words, is it at all possible to write reliable shell scripts?)
2003-07-16	* Substitute fixes.	Eelco Dolstra	1	-14/+11

2003-07-13	* Generate the scripts so that we can substitute the prefix	Eelco Dolstra	1	-0/+70
	etc. correctly. * Fixed nix-switch.