ekatek
f2ebbb1238
more piping errors to stderr
2014-07-30 15:58:12 -07:00
David Glasser
26f9b283a5
Tree hashes of builds now ignore package.json
...
We were finding that npm was inconsistently including various fields (eg
"readme") in package.json, leading to spurious "must update the version
number" errors in publish-release --from-checkout. This should be good
enough, as any actual package change should also have some other file
changed!
2014-07-30 15:44:12 -07:00
ekatek
0902f7e39a
make sure errors go to stderr: cursory pass
2014-07-30 08:07:12 -07:00
ekatek
d316eacab2
check if the package already exists and return a different error code
2014-07-29 21:19:43 -07:00
David Glasser
43b04a12b2
add some buildmessage.assertInJob to PackageClient
2014-07-29 18:57:59 -07:00
David Glasser
70c0ff2c98
put calls to publishPackage in buildmessage
2014-07-29 18:57:59 -07:00
David Glasser
81e4b51cf6
Put --get-ready's PackageLoader.getPackage in job
2014-07-29 18:57:59 -07:00
David Glasser
950a3d9c1b
Use buildmessage in publish-for-arch
...
Also, drop a unipkg.saveToPath that we think is redundant (since
5e72f55eb2
2014-07-29 18:57:59 -07:00
Matthew Arbesfeld
c6ad06fe77
Fix for run --once test. Actually include programs in star.json
2014-07-29 18:42:25 -07:00
David Glasser
30a397cde5
Rename ServerCatalog -> OfficialCatalog
...
Since it it used for catalog.official.
2014-07-29 17:19:31 -07:00
David Glasser
cd57ac3535
isolate different servers' packages
...
also, in self-test, only set $METEOR_PACKAGE_SERVER_URL for the specific
runs that actually want the test server (using a tag) rather than kinda
always by accident
2014-07-28 20:45:08 -07:00
David Glasser
381956d40b
make a test failure not abort the whole self-test
2014-07-28 17:37:20 -07:00
Matthew Arbesfeld
c65a199c3c
Glasser fixes
2014-07-28 17:12:27 -07:00
Matthew Arbesfeld
b8ae210d9b
Add documentation
2014-07-28 17:12:26 -07:00
Matthew Arbesfeld
55fe8a8aae
Fix hot code push client-side reloads.
...
We were overwriting the server directory when a client-side file changed,
which made all process calls fail, such as process.cwd() and fs.*. We
abstracted out some of the builder code so that only the client targets
are "rebuilt" when a client side file changes.
2014-07-28 17:12:26 -07:00
ekatek
3aff2199cc
fixing old cli test
2014-07-28 16:09:19 -07:00
David Glasser
5b69c7c25b
fix report-stats test from release
2014-07-28 16:07:42 -07:00
David Glasser
b7fa7e2452
don't run constraint-solver test from release
2014-07-28 16:07:41 -07:00
ekatek
0e74bdf769
help text for meteor update
2014-07-28 15:07:38 -07:00
Emily Stark
9e65038d08
Update "old cli test" for new 'meteor remove' output
2014-07-28 10:46:15 -07:00
David Glasser
6bb4b4de7b
Help for 'meteor publish-for-arch'
2014-07-25 16:10:08 -07:00
ekatek
a58f2c3bb2
mark some mroe tests as net with comments; first pass on making it ... exit, instead of craash in ensureDeps. Next up, maybe it can do something better than process.exit
2014-07-25 15:34:54 -07:00
ekatek
c622fb4072
label online tests as online
2014-07-25 14:40:43 -07:00
David Glasser
4a231a4c76
improved error handling around bad versions
2014-07-25 14:39:30 -07:00
David Glasser
59d38b28bb
minor cleanup
2014-07-25 14:12:21 -07:00
Avital Oliver
a8e372221d
Correct comment
2014-07-24 23:35:01 -07:00
ekatek
83ac27f9c4
A separate data.json for each package server
2014-07-24 18:36:15 -07:00
ekatek
af8c532411
Fix add-package test with new output.
2014-07-24 18:35:05 -07:00
David Glasser
4428f2b06d
sync tests passes
2014-07-24 17:12:14 -07:00
David Glasser
020824d828
minor commands tweaks
2014-07-24 16:52:23 -07:00
David Glasser
a7cbc3bd87
minor remove improvements
2014-07-24 16:49:12 -07:00
David Glasser
ab7e9366e1
minor 'add' cleanup
2014-07-24 16:43:35 -07:00
David Glasser
7aec7d3801
avoid saying 'version constraint null'
2014-07-24 16:42:07 -07:00
David Glasser
21b8e452a7
get rid of skipPackages
...
- in add, we actually do want to know the added version
- in remove, there's a distinction between truly dropping the package
and just dropping the top level constraint
2014-07-24 16:41:35 -07:00
David Glasser
ea7afca05f
use skipPackages consistently
2014-07-24 16:31:59 -07:00
David Glasser
1591e076e9
spell onDiskPackages consistently
...
(Have not checked to see that the feature actually works)
2014-07-24 16:31:38 -07:00
David Glasser
6850b679e9
fix bug in 'meteor add x@=1.0.0'
...
Used to print a "avoid conflicting dependencies" message. Now doesn't.
Start moving some uses of splitConstraint to parseConstraint
2014-07-24 16:21:25 -07:00
David Glasser
e9d7969da8
various minor error cleanups
2014-07-24 15:47:24 -07:00
David Glasser
5ec341c276
different way to show usage
2014-07-24 15:26:55 -07:00
David Glasser
591fa4efb4
Sort of better errors for nonexistent packages
...
Specifically, it's less good about saying "package doesn't exist" but
does at least include which top level dependency caused the problem
2014-07-24 15:26:55 -07:00
David Glasser
74dea7e1ba
tiny refactor
2014-07-24 15:26:55 -07:00
David Glasser
a5ace22f6c
minor improvement
2014-07-24 15:26:55 -07:00
David Glasser
a548efe175
ignore unknown packages in .meteor/versions
...
Hopefully this isn't ignoring real error cases. The whole
previousSolution data-rewriting hack probably needs to be fixed
anyway. But this seems to work?
2014-07-24 15:26:55 -07:00
ekatek
4205a06ba2
add exact constraints
2014-07-24 15:13:47 -07:00
ekatek
335dc4a312
mark tests as online, refresh catalog to the right server after self-test
2014-07-24 15:13:46 -07:00
Avital Oliver
9afd6f4cf9
A first pass on some package command error messages
2014-07-24 14:02:47 -07:00
Avital Oliver
a6db03f131
Unbreak hot code push test by eliminating references to non-existent spacebars-common package
2014-07-23 20:27:35 -07:00
Avital Oliver
a533f1ed5d
When calling versionsFrom with no track, default to METEOR-CORE
...
Also a small restructuring I thought I'd need for this test, but ended
up not needing. But it's nice anyways.
2014-07-23 20:25:50 -07:00
Emily Stark
053f26da5e
Merge branch 'packaging-experimental-unknown-item' into packaging
2014-07-23 17:57:57 -07:00
Emily Stark
ff7a2fba92
Add catalog.official._refreshFutures back in.
...
This allows `catalog.official.refresh` to be called during another
refresh that has yielded.
`catalog.official.refresh` still cannot be called re-entrantly, so
`recordOrRefresh` checks if a refresh is in progress before starting a
new one.
2014-07-23 17:54:16 -07:00
