Clojurians Log v2

@slipset Not sure if you saw the latest Clojurists Together survey results, linked in the #announcements channel some days ago: https://www.clojuriststogether.org/news/q4-2018-survey-results/ It mentions "lint tools" and "Eastwood" explicitly as something the people surveyed might be interested in funding improvements of.

I have a full time gig so would not be interested in submitting a proposal for me, and you may be similarly otherwise occupied, too. Just wanted to pass along the mention in case you are interested.

I also have a full time job, so I wouldn’t be able to commit to a grant.

It would, however, be interesting to figure out what the respondents would like to see changed/improved in or added to Eastwood, and figure out a way to satisfy that.

my 2 cents: a release of eastwood that's up to date and prints my actual error when loading a namespace throws, instead of an internal eastwood error in the code handling the original problem - I might find time one of these weekends to figure out how the release process works (2.9 has the fix, using 2.9 via the plugin gives me 2.8, lein install of 3.0 snapshot locally works as expected...) - not making demands here, I could even do this work myself if I found the time / energy for it

I’ll see if I can’t cut a release real soon now. Sorry for the bugs and stuff :/

@noisesmith 0.3.0 is released to Clojars.

@slipset why does it still say "eastwood 0.2.8" when it runs?

Seriously?

Retrieving jonase/eastwood/0.3.0/eastwood-0.3.0.pom from clojars
Retrieving jonase/eastwood/0.3.0/eastwood-0.3.0.jar from clojars
== Eastwood 0.2.8 Clojure 1.9.0 JVM 1.8.0_171
Directories scanned for source files:

is this something in the dep injection machinery?

when I run from a snapshot, it correctly prints the snapshot version iirc

give me a sec.

:thumbsup:

On a fresh lein new app baz

I get

in case it helps, I did verify that the arity exception on error-messages/format-exception is thrown when a namespace is broken on load

21:46 $ cd baz
✔ ~/Documents/github.com/baz
21:46 $ lein eastwood
Retrieving jonase/eastwood/0.3.0/eastwood-0.3.0.pom from clojars
Retrieving jonase/eastwood/0.3.0/eastwood-0.3.0.jar from clojars
== Eastwood 0.3.0 Clojure 1.8.0 JVM 1.8.0_172
Directories scanned for source files:
  src test
== Linting baz.core ==
== Linting baz.core-test ==
== Warnings: 0 (not including reflection warnings)  Exceptions thrown: 0
✔ ~/Documents/github.com/baz
21:46 $

OK - I'll try cleaning then

with profile.clj like

(defproject baz "0.1.0-SNAPSHOT"
  :description "FIXME: write description"
  :url "<http://example.com/FIXME>"
  :license {:name "Eclipse Public License"
            :url "<http://www.eclipse.org/legal/epl-v10.html>"}
  :dependencies [[org.clojure/clojure "1.8.0"]]
  :main ^:skip-aot baz.core
  :target-path "target/%s"
  :plugins [[jonase/eastwood "0.3.0"]]
  :profiles {:uberjar {:aot :all}})

and basically empty ~/.lein/profiles.clj

I'll check my deps plugin-tree too

Do you have eastwood mentioned in your $HOME/.lein/profiles.clj, your project's local project.clj, or both?

both are set to 0.3.0 currently -checking plugins-tree for conflicts

plugin-tree says 0.3.0, profiles.clj says 0.3.0, project.clj says 0.3.0, I get 0.2.8

Currently on my Mac, I have eastwood 0.3.0 in my $HOME/.lein/profiles.clj, and in a project where project.clj and no other file in the project mentions eastwood at all, I see 0.3.0 printed when I run 'lein eastwood'

at this point I'm assuming pebkac and just trying to see where it's happening

OK - it was a module dependency merge issue, seems resolved

so it was all on my side, thanks for the update

Thank you for taking the time to check it out!

Glad you found it. I was about to do a Hail Mary and ask if you had either a file /etc/leiningen/profiles.clj or $HOME/.lein/profiles.d/user.clj on your system, which Leiningen also uses if they exist.

I don't have those names memorized, but the big Eastwood README mentions them in one section of its docs.

nice

Exception thrown during phase :analyze+eval of linting namespace ledger.run

The following form was being processed during the exception:
(/ 1 0)

woot!

the issue was a custom fork of lein modules that lies about :plugin-tree

one weird thing though - after showing that error, I get at the end:

== Warnings: 0 (not including reflection warnings)  Exceptions thrown: 0

Yeah, that’s a regression since 0.2.5

I’ll create an issue on that.

https://github.com/jonase/eastwood/issues/290

:thumbsup:

@slipset I replied on that github PR you tagged me on, looks good except the ANALYZING messages of course

Didn’t get the analyzing part what’s that? British vs Us?

anyways, it’s merged and released with 0.3.1

Happy hacking and thanks for all the help!

excerpt

== Linting ledger.euat-test.stress.kill-zookeeper ==
ANALYZING
== Linting jackdaw.client-extras-test ==
ANALYZING
== Linting ledger.euat-test.traffic ==
ANALYZING
== Linting ledger.euat-test.perf ==
ANALYZING
== Linting ledger.validation-tests ==
ANALYZING
== Linting ledger.event-reverser.unit-test ==
ANALYZING
== Linting ledger.tools.event-ids-test ==
ANALYZING
== Linting ledger.balance-tests ==
ANALYZING
== Linting ledger.euat-test.stress.control ==
ANALYZING
== Linting ledger.balance-loader.drain-tests ==
ANALYZING

not a deal breaker, but probably not intended?

looks like those messages only happen with a local snapshot install

Stange:

$ grep -r ANALYZING *
✘-1 ~/Documents/github.com/eastwood [master|…6]
$ 

any local changes in your local snapshot install, perhaps, noisesmith? I don't see ANALYZING in the published eastwood source, either.

ugh, probably my PEBKAC, right