titive, so I've changed the output to
a "debug"-priority message that describes what's happening.
Reviewed-By: Nick Lewis
Signed-off-by: Jesse Wolfe
---
Local-branch: ticket/2.7.x/7084
lib/puppet/resource/catalog.rb | 14 +++---
lib/puppet/trans
er as their direct parent", producing a 50% chance of
a malformed containment path in log messages.
Reviewed-By: Jacob Helwig
Signed-off-by: Jesse Wolfe
---
lib/puppet/parser/resource.rb |2 ++
1 files changed, 2 insertions(+), 0 deletions(-)
diff --git a/lib/puppet/parser/resourc
are loaded
using the autoloader instead of using 'require'.
Paired-With: Jacob Helwig
Signed-off-by: Jesse Wolfe
---
Local-branch: maint/next/load-order-bug-with-component
lib/puppet/transaction/event_manager.rb |2 +-
lib/puppet/type.rb |2 +-
2 fil
The watch_file mechanism would refuse to monitor paths to files that
didn't exist. This patch makes it possible to watch a file that hasn't
been created yet, so when it is created, you manifests will get
reparsed.
Backported this change to 2.6.x
Paired-With: Jacob Helwig
Signed-off
The watch_file mechanism would refuse to monitor paths to files that
didn't exist. This patch makes it possible to watch a file that hasn't
been created yet, so when it is created, you manifests will get
reparsed.
Paired-With: Max Martin
Reviewed-By: Jacob Helwig
Signed-off-by: J
the
include function.
Paired-with: Jacob Helwig
Signed-off-by: Jesse Wolfe
---
Local-branch: ticket/next/4576
lib/puppet/parser/compiler.rb | 16 ++--
spec/unit/parser/compiler_spec.rb | 24 +---
test/language/functions.rb|2 +-
3 files ch
Reviewed-By: Paul Berry
Signed-off-by: Jesse Wolfe
---
lib/puppet/type.rb |2 +-
spec/unit/type_spec.rb | 13 -
2 files changed, 13 insertions(+), 2 deletions(-)
diff --git a/lib/puppet/type.rb b/lib/puppet/type.rb
index 205d809..d24cc85 100644
--- a/lib/puppet/type.rb
Reviewed-by: Nick Lewis
Signed-off-by: Jesse Wolfe
---
Local-branch: t/next/6656
app/views/nodes/_inventory_service.html.haml |2 +-
app/views/shared/_node_manager_sidebar.html.haml |2 +-
config/settings.yml.example |3 +++
3 files changed, 5 insertions
ving the bug, and we wouldn't have
to deal the problem of correctly reading and writing all fields.
On Fri, Feb 25, 2011 at 2:00 PM, Luke Kanies wrote:
> On Feb 25, 2011, at 1:55 PM, Sean Millichamp wrote:
>
>> On Fri, 2011-02-25 at 13:32 -0800, Jesse Wolfe wrote:
>>> Sorry
I've merging this patch with one change:
instead of duplicating the IF in all of the defautto blocks, I'm
bubbling it up to :
def retrieve_default_context(property)
+ if @resource[:selinux_ignore_defaults] == :true
+return nil
+ end
On Wed, Dec 29, 2010 at 1:30 PM, Sean Mil
Sorry that I'm late to the party one this one, but I have some questions.
groupadd must also have the problem, is that correct?
And it probably also affects the "pw" provider for BSD?
If so, I think we should remove the getent code entirely and replace
it with passwd and group file parsing at the
The noop option has been suppressing exit statuses. This is
counterintuitive, as per discussion at
http://projects.puppetlabs.com/issues/6322
This patch causes noop runs to return the same exit codes as real runs.
Reviewed-By: Daniel Pittman
Signed-off-by: Jesse Wolfe
---
lib/puppet
m not 100% sure about #1
>
> On Wed, Feb 23, 2011 at 2:16 PM, Jesse Wolfe wrote:
>>>
>>> > Can you comment on the actual implications of this codepath being made
>>> > live Jesse?
>>>
>>> It meant that we were getting audit events for every par
>
>
> > Can you comment on the actual implications of this codepath being made
> > live Jesse?
>
> It meant that we were getting audit events for every parameter of every
> file being recursively managed, even when we weren't intending to audit it.
>
> So, 5 events for every file, and it had to run
avior for this fix.
Reviewed-By: Markus Roberts
Signed-off-by: Jesse Wolfe
---
lib/puppet/type/file/source.rb | 14 --
1 files changed, 0 insertions(+), 14 deletions(-)
diff --git a/lib/puppet/type/file/source.rb b/lib/puppet/type/file/source.rb
index bc464e1..6dda795 100755
--- a/l
Signed-off-by: Jesse Wolfe
---
lib/puppet/application/inspect.rb | 53 +
1 files changed, 53 insertions(+), 0 deletions(-)
diff --git a/lib/puppet/application/inspect.rb
b/lib/puppet/application/inspect.rb
index 77e8476..52ef975 100644
--- a/lib/puppet
I second that git submodules seem to be flaky; I've found it surprisingly
hard to just correctly, consistently update submodules.
And I agree that it can be difficult to safely update code that other
projects depend on - but if we do it right, this could be a forcing function
that causes us to mak
Relax the constraint that a node may only send one report per second
into one "kind" of report per second.
Paired-With: Matt Robinson
Signed-off-by: Jesse Wolfe
---
Local-branch: ticket/next/5863-squash
app/models/report.rb |2 +-
spec/models/report_spec.rb |2 +-
2 fil
Added a boolean to search "all", and changed the default to "latest".
Added new database column and new scopes to facilitate this.
Paired-With: Matt Robinson and Paul Berry
Signed-off-by: Jesse Wolfe
---
Local-branch: ticket/next/5863-squash
app/controllers/r
A preliminary for searching latest inspect reports
Paired-With: Matt Robinson
Signed-off-by: Jesse Wolfe
---
Local-branch: ticket/next/5863-squash
app/helpers/application_helper.rb |2 +-
app/models/node.rb | 16
app/views
This change disambiguates the distinction between a blank search form
and a search that returns no hits.
Paired-With: Paul Berry
Signed-off-by: Jesse Wolfe
---
Local-branch: ticket/next/5864
app/controllers/reports_controller.rb |2 +-
app/views/reports/search.html.haml
No, there's not currently an API presentation of versions other than the
latest - I consider that to be a bug, and I've filed a ticket at
https://projects.puppetlabs.com/issues/5854
~Jesse Wolfe
On Tue, Jan 11, 2011 at 8:50 AM, Thomas Hallgren wrote:
> Hi,
>
> Aside from
> 3) For any file object where the content is in the catalog (i.e. we're
> using content =>) then we should pass along the appropriate type of
> hash in the catalog.
>
>
There's been talk, periodically, of always referring to files in the
compiled catalog using the hash, and pulling the file conten
have more like `apply` reports, and skip
> > properties other than :ensure for absent resources.
> >
> > Reviewed-By: Nick Lewis
> > Signed-off-by: Jesse Wolfe
> > ---
> > lib/puppet/application/inspect.rb |8 ++--
> > lib/puppet/reports/http.rb
I decided to file a ticket for this after we had already written the code.
(Oops.)
It's http://projects.puppetlabs.com/issues/5759
On Mon, Jan 3, 2011 at 4:31 PM, Jesse Wolfe wrote:
> Ruby's default #inspect method can lead to printing factorial-order
> output for large graphs o
uch faster.
Paired-With: Nick Lewis
Signed-off-by: Jesse Wolfe
---
lib/puppet/parser/ast.rb |4
lib/puppet/relationship.rb |4
lib/puppet/resource.rb |4
lib/puppet/resource/type_collection.rb |4
lib/puppet/simpl
Signed-off-by: Jesse Wolfe
---
lib/puppet/application/inspect.rb |8 ++--
lib/puppet/reports/http.rb|2 +-
spec/unit/application/inspect_spec.rb | 25 -
3 files changed, 31 insertions(+), 4 deletions(-)
diff --git a/lib/puppet/application/in
Paired-With: Nick Lewis
Signed-off-by: Jesse Wolfe
---
Local-branch: ticket/next/5174-part2
app/controllers/reports_controller.rb |6 +++---
app/models/report.rb |6 +-
spec/models/report_spec.rb|4 ++--
3 files changed, 6 insertions(+), 10 deletions
Signed-off-by: Jesse Wolfe
---
Local-branch: ticket/next/5174-part2
app/models/report.rb |1 +
spec/models/report_spec.rb |4
2 files changed, 5 insertions(+), 0 deletions(-)
diff --git a/app/models/report.rb b/app/models/report.rb
index 2c38fbb..7c3b77e 100644
--- a/app
Paired-With: Matt Robinson
Signed-off-by: Jesse Wolfe
---
Local-branch: ticket/next/5174-part2
app/controllers/reports_controller.rb |6 ++
app/views/reports/_report.html.haml |4 +++-
config/routes.rb |1 +
3 files changed, 10 insertions(+), 1 deletions
Paired-With: Matt Robinson
Signed-off-by: Jesse Wolfe
---
Local-branch: ticket/next/5174-part2
app/views/reports/_report_status_icon.html.haml |7 ++-
public/images/icons/baseline.png| Bin 0 -> 405 bytes
public/images/icons/inspect.png | Bin 0 -&g
Signed-off-by: Jesse Wolfe
---
Local-branch: ticket/next/5174-part2
app/models/report.rb|4
app/views/reports/_report.html.haml |5 +
2 files changed, 9 insertions(+), 0 deletions(-)
diff --git a/app/models/report.rb b/app/models/report.rb
index 7c3b77e
Paired-With: Matt Robinson
Signed-off-by: Jesse Wolfe
---
Local-branch: ticket/next/5174-part2
lib/puppet/report.rb |4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/lib/puppet/report.rb b/lib/puppet/report.rb
index e4d13f5..12f922d 100644
--- a/lib/puppet/report.rb
You can set a report as baseline by calling #baseline!
Paired-With: Matt Robinson
Signed-off-by: Jesse Wolfe
---
Local-branch: ticket/next/5174-part2
app/models/node.rb |1 +
app/models/report.rb |9
Paired-With: Nick Lewis
Signed-off-by: Jesse Wolfe
---
Local-branch: ticket/next/5174-part2
app/views/reports/_report.html.haml |8 +---
app/views/reports/_report_title.html.haml |6 ++
app/views/reports/diff.html.haml |5 +
app/views/reports
Paired-With: Nick Lewis
Signed-off-by: Jesse Wolfe
---
Local-branch: ticket/next/5174-part2
app/views/reports/diff.html.haml |2 +-
app/views/reports/diff_summary.html.haml |8 ++--
public/stylesheets/tables.css|7 +--
3 files changed, 12 insertions(+), 5
Paired-With: Matt Robinson
Signed-off-by: Jesse Wolfe
---
Local-branch: ticket/next/5174-part2
app/models/report.rb |3 +++
app/views/nodes/show.html.haml | 35 ---
spec/models/report_spec.rb |8
3 files changed, 43 insertions
s to have
our own definition of "string", "number", etc.
2) In places where .pp allows more than one possible literal representation
of the same value, we should pick a canonical representation for pzero; for
example, we might say "all barewords must be replaced with quoted stri
File ctime and mtime are now implemented as read-only properties, so
they can be examined with audit.
Signed-off-by: Jesse Wolfe
---
lib/puppet/type/file.rb |2 ++
lib/puppet/type/file/ctime.rb | 18 ++
lib/puppet/type/file/mtime.rb | 17 +
lib
File ctime and mtime are now implemented as read-only properties, so
they can be examined with audit.
Signed-off-by: Jesse Wolfe
---
lib/puppet/type/file.rb |2 +
lib/puppet/type/file/ctime.rb | 18
lib/puppet/type/file/mtime.rb | 17 +++
spec/unit
ock to get executed.
Signed-off-by: Jesse Wolfe
---
lib/puppet/util/checksums.rb | 11 +++
spec/unit/util/checksums_spec.rb | 10 +-
2 files changed, 20 insertions(+), 1 deletions(-)
diff --git a/lib/puppet/util/checksums.rb b/lib/puppet/util/checksums.rb
index 5aebd83..6
"none", "mtime", and "ctime" aren't real checksums, but they still need
to fully implement the checksum_stream interface, so that the code
inside the supplied block will run.
Signed-off-by: Jesse Wolfe
---
lib/puppet/util/checksums.rb | 11 ++
The previous maintenance patch for this file didn't provide as strong of
a guarantee of loading plugins as the code it replaced.
This patch restores the extremely broad exception catching, but prevents
mocha exceptions from being silently ignored.
Signed-off-by: Jesse Wolfe
---
lib/p
Nick Lewis
Paired-With: Matt Robinson
Signed-off-by: Jesse Wolfe
---
lib/puppet/transaction/change.rb | 74 --
lib/puppet/transaction/event.rb|2 +-
lib/puppet/transaction/resource_harness.rb | 30 ++--
lib/puppet/util/log.rb
This test was succeeding in versions of Mocha before 0.9.10, because the
Mocha expectation exceptions were being incorrectly interpreted as if
they were File Not Found exceptions.
Signed-off-by: Jesse Wolfe
---
spec/unit/provider/service/init_spec.rb |2 ++
1 files changed, 2 insertions
Our code calls Catalog.find multiple times, but this stub did not allow
that. The resulting error was silently ignored on versions of Mocha
before 0.9.10
Paired-With: Nick Lewis
Signed-off-by: Jesse Wolfe
---
spec/unit/configurer_spec.rb |2 +-
1 files changed, 1 insertions(+), 1 deletions
This missing stub was raising an exception that, in versions of Mocha
less than 0.9.10, was coincidentally causing the method under test to
behave as expected.
Paired-With: Nick Lewis
Signed-off-by: Jesse Wolfe
---
spec/unit/file_serving/fileset_spec.rb |1 +
1 files changed, 1 insertions
This test was stubbing Puppet.settings, instead of setting the values
directly. Exceptions thrown by trying to read other settings were
getting swallowed when running a version of Mocha older than 0.9.10.
Paired-With: Nick Lewis
Signed-off-by: Jesse Wolfe
---
spec/unit/indirector
This missing stub was raising an exception that, in versions of Mocha
less than 0.9.10, was mistaken for the exception that was being tested
for.
Paired-With: Nick Lewis
Signed-off-by: Jesse Wolfe
---
spec/unit/type/file_spec.rb |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff
Another missing stub exposed by Mocha 0.9.10
Signed-off-by: Jesse Wolfe
---
spec/unit/type/file/source_spec.rb |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/spec/unit/type/file/source_spec.rb
b/spec/unit/type/file/source_spec.rb
index 522ae1f..00cc2f2 100755
--- a
A test was not testing what it claimed, but the failure exception was
getting swallowed by an unnecessarily broad rescue match.
Paired-With: Nick Lewis
Signed-off-by: Jesse Wolfe
---
lib/puppet/configurer/plugin_handler.rb |4 +---
spec/unit/configurer/plugin_handler_spec.rb |2
OK, I've pushed the squashed patch, without autorequire, into the `next`
branch.
Derek, could you open a new ticket for discussion of adding autorequire for
datasets?
~Jesse Wolfe
On Tue, Dec 7, 2010 at 1:06 PM, deet wrote:
>
> > This code looks correct (as patched). I'd
From: Matt Robinson
There's now a diff method on reports that allows you diff between
another report and returns the properties that have different previous
values. This is really only useful for our new inspect reports for
auditing.
Paired-with: Jesse Wolfe
Signed-off-by: Jesse
This patch implements a rudimentary UI to compare two arbitrary
"inspect" reports.
A more polished interface is planned for ticket #5173
Paired-With: Matt Robinson
Signed-off-by: Jesse Wolfe
---
Local-branch: ticket/next/5174
app/controllers/reports_controller.rb
as Latin-1
This allows all raw binary files to be transmitted losslessly.
Paired-With: Paul Berry
Signed-off-by: Jesse Wolfe
---
lib/puppet/external/pson/pure/generator.rb | 23 +--
spec/unit/util/pson_spec.rb| 15 +++
2 files changed, 16 in
Added a per-page parameter to the pagination system
and created a UI widget to select different page sizes, including an
'all' option.
Signed-off-by: Jesse Wolfe
---
Local-branch: ticket/next/5142
app/helpers/application_helper.rb | 31 +++---
a
Several tables were rendering content that was not correctly wrapped in
table tags.
Signed-off-by: Jesse Wolfe
---
Local-branch: ticket/next/5142
app/views/node_classes/index.html.haml | 13 +++--
app/views/node_groups/index.html.haml | 13 +++--
app/views/nodes
Added a configuration setting that will replace the Puppet Dashboard
logo with any arbitrary image.
Signed-off-by: Jesse Wolfe
---
Local-branch: ticket/next/5117
app/views/shared/_global_nav.html.haml |6 +-
config/settings.yml.example |3 +++
public/stylesheets
self" in the puppet codebase.
There's no official puppet policy on this, as far as I know, but I'm
inclined to say that we should only accept cleanup/refactoring patches if
they are part of a series of changes that also adds or fixes functionality.
Does anyone else have an opinion on
should work. As such, we're going to mark this ticket as "needs design
decision". Ticket #5012 is so closely related that I think we should treat
it as the same feature.
Thanks for the suggestion and the patches, Stefan. If you have any more
thoughts about these issues, please let u
man wrote:
> Jesse Wolfe writes:
>
> > bin/external_node now respects the EXTERNAL_NODE_URL environment
> variable.
> > This means a user can now configure their puppet master to connect to
> > dashboard like so:
> >
> > [puppetmasterd]
> > node_termin
Updated bin/external_node to optionally use environment variables for
all settings, using consistent names with the prefix "PUPPET_"
Signed-off-by: Jesse Wolfe
---
Local-branch: ticket/next/4475
README.markdown | 12 +---
bin/external_node | 12
2 files c
/external_node
Paired-With: Nick Lewis
Signed-off-by: Jesse Wolfe
---
Local-branch: ticket/next/4475
bin/external_node |5 +++--
1 files changed, 3 insertions(+), 2 deletions(-)
diff --git a/bin/external_node b/bin/external_node
index f6ee194..b8db5f9 100755
--- a/bin/external_node
+++ b/bin
The "Daily Run History" graph now respects a "daily_run_history_length"
setting which specifies the number of days to show in the graph.
Paired-With: Paul Berry and Nick Lewis
Signed-off-by: Jesse Wolfe
---
Local-branch: ticket/next/4881
Signed-off-by: Jesse Wolfe
---
Local-branch: ticket/next/4881
spec/views/statuses/_run_failure.html.haml_spec.rb | 29
1 files changed, 29 insertions(+), 0 deletions(-)
create mode 100644 spec/views/statuses/_run_failure.html.haml_spec.rb
diff --git a/spec/views
Signed-off-by: Jesse Wolfe
Local-branch: ticket/next/3435
---
db/migrate/20101109001012_add_status_to_reports.rb |4
1 files changed, 4 insertions(+), 0 deletions(-)
diff --git a/db/migrate/20101109001012_add_status_to_reports.rb
b/db/migrate/20101109001012_add_status_to_reports.rb
nted", so users can truly manage fstab entries without having
puppet try to remount them.
Paired-With: Paul Berry
Signed-off-by: Jesse Wolfe
---
lib/puppet/type/mount.rb |2 +-
spec/unit/type/mount_spec.rb | 32 +++-
2 files changed, 28 insertions(+)
ws the
ActiveSupport version, preventing it from ever being called.
Signed-off-by: Jesse Wolfe
---
lib/puppet/util/monkey_patches.rb |8
1 files changed, 8 insertions(+), 0 deletions(-)
diff --git a/lib/puppet/util/monkey_patches.rb
b/lib/puppet/util/monkey_patches.rb
index 6b5
>
> >> Right now, I wanted to ask a couple of questions about the following
> >> next patch:
> >>
> https://github.com/puppetlabs/puppet/commit/cfe202539018b27c35ff833152f237bc30569a5f
> >
> > Like why all the tests were deleted?
>
>
OK, I've been getting some flak about this one. Clearly we (Paul
> I think the right answer is to get rid of the need for this graph at all by
> having a single graph contain both dependency and containment edges, thus
> not needing the many to many nature. I tried very hard to get this into 2.6
> but couldn't finish it in time. In seeing this multiplier, it's
+1 as a kludge, but I don't know how useful this hook would be, generally.
It seems like you'd usually at least need to have the name of the variable
you're munging.
On Thu, Nov 4, 2010 at 10:39 PM, Markus Roberts wrote:
> The internal format of parameters was changed between 0.25.x and 2.6.x,
>
So isn't this the same as:
>
> +@should.any? { |desired| property_matches?(desired, is) }
>
I never seem to remember that #any? exists, and I keep end up
re-implementing it.
>
> -mode_int = mode ? Puppet::Util::Octal.integerForOctal(mode) : nil
>> +mode_int = mode ? Puppet::Util::Fil
possible to manage the
read and write permissions outside of puppet.
Signed-off-by: Jesse Wolfe
---
lib/puppet/property.rb |7 +-
lib/puppet/type/file.rb |4 +-
lib/puppet/type/file/ensure.rb |2 +-
lib/puppet/type/file/mode.rb |
This patch changes the internal representation of a file's mode to a
string instead of an integer. This simplifies the problem of displaying
the value consistently throughout all of puppet.
Signed-off-by: Jesse Wolfe
---
lib/puppet/type/file.rb|4 +-
lib/puppet
>
> It part of my "cleaning things up for futures" work--it's an internal
> refactor and the question is, to what extent do we expose it?
>
An idea I had for what I might like to have if I was consuming the lookupvar
API is something like:
lookvar('myvar', :undef_as => [])
and to imagine a littl
Two places where the analogy breaks down: some spreadsheets will iterate a
> finite number of times to let loops "settle out" (stumble onto a fixed
> point) and some have a documented default order (generally reading order)
>
That may be true for some spreadsheets, but the industry standard in t
>
> make(1)-style languages, and spreadsheets, also use a similar sort of
> model,
> though they are usually "pull-based" rather than "futures-based". The net
> effect is pretty much identical, however, in that it makes computation as
> lazy
> as possible but no lazier.
>
>
I came to this same con
> As an aside, will the futures work also improve things like:
>
> if defined(Class["foo"])
>
> ie. will those become order independent too? Ditto with tags and ditto
> with accessing the classes list variable in templates?
>
>
Ideally, yes, but since this is a completely different code-path than
Did this code change? It looks exactly the same to me.
On Tue, Oct 19, 2010 at 7:37 PM, Markus Roberts wrote:
> This was a regression, not covered by a test; previously the string
>
>"foo/
>bar"
>
> would be interpreded as "foobar" but this was changed to "foo\\\nbar" in
> 2.6.x with my
> sudo ralsh user jesse ensure=absent
>
> :)
Fortunately, that's not my username.
Maybe I used the wrong jargon, but the case that I'm actually worried about
is when you have a single resource you're trying to query or edit, but you
can't uniquely find it by name. For example, "ssh_authorized_key
tests caught an
unexpected ruby quirk:
"text/".split("/")
and
"text/".split("/", 2)
do not return the same values.
2) File.instances now reproduces the old behavior of listing files in
the root directory. This is now implemented in terms of the existing
file r
>
> I'm for "files in the root directory" if no name is specified and then
> when/if we add parameters we're well positioned to have reasonable
> semantics.
>
>
I can live with that. Updated patch shortly.
--
You received this message because you are subscribed to the Google Groups
"Puppet Devel
> This is really weird. It seems I can say the most absurd things today and
> people will take me quite seriously. If only I could figure out to harness
> this new found power...
>
Are you certain that you're not the one taking my joke seriously?
> That's why I'm suggesting we provide some way
>
> Not practical? Come on, it's one line of code:
>
> `find /`.split(/\n/)
>
>
I started running that command on my VM a few minutes ago, and I'm still
waiting for it to return.
Unfortunately, the RAL resource find function is implemented like this:
res = type(request).instances.find {
>
> puppet resource / recurse=1
>
> and produced a list of the existing files in /.
>
I believe that Dan opened a ticket a couple months ago that we should be
able to pass parameters that way - right now, the command line only works
for setting attributes on a resource that you can find using the
> My guess, then, is that the instances method is set to return [] because
> the system requires it return something (which is probably a bug), and this
> way it just returns an empty set rather than failing.
>
Right, that was my understanding, too.
It looks like the original code, before it star
tests caught an
unexpected ruby quirk:
"text/".split("/")
and
"text/".split("/", 2)
do not return the same values.
Signed-off-by: Jesse Wolfe
---
lib/puppet/indirector/resource/ral.rb |9 +++--
lib/puppet/type/file.rb | 21 ++
> The root problem is that if the resource getting instantiated is a
> user-defined type (hostclass or definition), resource.resource_type is an
> instance of Puppet::Resource::Type, but if the resource getting instantiated
> is a built-in type, then resource.resource_type is actually a Ruby class
> My current answer is "no", just because I'm trying to keep things simple.
I'm leaning toward "no", but let's try this:
class A {
File { owner => "root" }
}
class B
class C inherits A {
file {"foo" :}
class D {
file {"bar" :}
}
}
}
Would you expect "bar" to be owned b
>
> It might require that we support an 'extend'-like concept, where you can
> pull defaults (and maybe variables) from one class into another:
or, alternately, multiple inheritance.
--
You received this message because you are subscribed to the Google Groups
"Puppet Developers" group.
To po
After thinking it through, I agree that #evaluate should be idempotent.
I don't want to lose the tests for #builtin_type? and #virtual? though. I'm
not convinced that it makes sense to fold them into #evaluate - currently
built-in types raise an error, and it would be a behavior change to make
tho
rrd is not installed.
Signed-off-by: Jesse Wolfe
---
lib/puppet/feature/base.rb |3 +-
lib/puppet/util/metric.rb | 47 +++
2 files changed, 40 insertions(+), 10 deletions(-)
diff --git a/lib/puppet/feature/base.rb b/lib/puppet/feature/base.rb
ind
This change just forces a file to be written to the filebucket before we
drop the permissions necessary to do so.
Signed-off-by: Jesse Wolfe
---
lib/puppet/provider/ssh_authorized_key/parsed.rb |7 +++
1 files changed, 7 insertions(+), 0 deletions(-)
diff --git a/lib/puppet/provider
This code was using a constant that might not always be loaded.
Signed-off-by: Jesse Wolfe
---
lib/puppet/type/tidy.rb |1 +
1 files changed, 1 insertions(+), 0 deletions(-)
diff --git a/lib/puppet/type/tidy.rb b/lib/puppet/type/tidy.rb
index 64a7a1a..65cc077 100755
--- a/lib/puppet/type
While writing this proposal, the solution I was describing shrank in size
considerably - some work that had initially appeared to be prerequisites
turned out to actually be more like "it-would-be-nice-if".
I consider this proposal to be the simplest thing that could possibly work,
which is why my "
>
> The real problem here is that run_mode was a broken model when I did it,
> and the reification of it didn't sufficiently fix it.
>
I agree with this statement completely.
> require that applications set it, in the same way that they set things like
> terminus_class. This makes everything si
>
> I propose that instead of checking the run_mode in the indirector, we
> combine these settings into one setting and have the location be
> dependent on run_mode in defaults.rb. I can't think of any reason why
> this wouldn't work and be less confusing. Anyone else?
>
>
Sounds good to me.
We'd
sumed that
unsupported attributes would be ignored with a warning, but the code was
crashing instead (I don't think this would create a user-observable bug,
but I'm not certain.)
Signed-off-by: Jesse Wolfe
---
lib/puppet/type.rb | 16 +---
spec/unit/type/user_spec.rb |
the Provider's responsibility to ignore parameters it doesn't support
rather than the Type's responsibility to query the Provider about its
features.
~Jesse Wolfe
On Sat, Sep 11, 2010 at 8:55 AM, James Cammarata wrote:
>
> This was apparently caused by an earlier commit, per the bug.
I don't think this is excessively heavy-handed, but it's a shame that
apt-listbugs doesn't respect $DEBIAN_FRONTEND - has anyone filed a bug
report with Debian about that?
I say +1, I don't see any way that this could do harm, but is there anyone
with a Debian machine who can test this? - apparent
1 - 100 of 283 matches
Mail list logo
