- Dev / test cycles are one use case for the "running a local copy of
the client" instructions, but simply running bleeding edge Certbot is
another
- So edit the docs to once again explain how to just run bleeding edge
Certbot, without (say) always getting staging certs.
* Support "certbot-auto --no-bootstrap"
* Tell people about --no-bootstrap?
* Document new certbot-auto flag in its cli help
* Rebuild
* Less variables is less variability
* Alphabetize help
* Make it extra clear we only take one branch
* Add --no-bootstrap message to experimentalbootstrap exit
* Begin fixing incorrect defaults
* Fix more defaults
* Make more defaults correct
* Update cli-help.txt
(To show what this PR does)
* Lint
* Extend argparse rather than vendoring it
* lint
* Move sample User Agent generation into the same module as UA generation
* Revert cli-help.txt to previous release version
* Slightly more consistent linebreaks
* Start reorganising -h output
* Fix the --debug flag
- Currently exceptions are often caught and burried in log files, even
if this flag is provided!
* Explain the insanity
* Parallalelise nosetests from tox (#3836)
* Parallalelise nosetests from tox
* Parallelise even more things, break even more things
* Now unbreak all the tests that aren't ready for ||ism
* Try to pass tests!
- Remove non-working hack in reporter_test
- also be selective about ||ism in the cover environment
* Try again
* certbot-apache tests also work, given enough time
* Nginx may need more time in Travis's cloud
* Unbreak reporter_test under ||ism
* More timeout
* Working again?
* This goes way faster
* Another big win
* Split a couple more large test suites
* A last improvement
* More ||ism!
* ||ise lint too
* Allow nosetests to figure out how many cores to use
* simplify merge
* Mark the new CLI tests as ||izable
* Simplify reporter_test changes
* Rationalise ||ism flags
* Re-up coverage
* Clean up reporter tests
* Stop modifying testdata during tests
* remove unused os
* Improve the "certbot certificates" output (#3846)
* Begin making "certbot certificates" future safe
* Handle the case where a renewal conf file has no "server" entry
* Improvements, tweaks
* Capitalise on things
* Print the command summary for -h and -h all, but not otherwise
Also, update nginx not installed CLI hint
* Add a "certificates" help section
* Clean up usage string construction
* Greatly improve "certbot -h TOPIC"
- subcommands now get their own usage headings if they want them
- added "certbot -h commands"
* A few more cli formatting tests
* Auto-populate the verb subgroups from the docs
* Show the new help output
* Lint, tweak
* More lint, and cleanup
* Infinite lint
* Add rename to command summary; sort "-h commands" output
* Use fancy string formatting
* More space
* Implement --help manage
Also, implement a general mechanism for documenting subcommands within
topics
* Remove one comma
* Only create weird parser structures if -h is provided :)
* Update sample cli out
* Lint
* Revert cli-help.txt to previous release version
* Grammar & style
* Fix the --debug flag
- Currently exceptions are often caught and burried in log files, even
if this flag is provided!
* Explain the insanity
* Make things slightly nicer
* Parallalelise nosetests from tox
* Parallelise even more things, break even more things
* Now unbreak all the tests that aren't ready for ||ism
* Try to pass tests!
- Remove non-working hack in reporter_test
- also be selective about ||ism in the cover environment
* Try again
* certbot-apache tests also work, given enough time
* Nginx may need more time in Travis's cloud
* Unbreak reporter_test under ||ism
* More timeout
* Working again?
* This goes way faster
* Another big win
* Split a couple more large test suites
* A last improvement
* More ||ism!
* ||ise lint too
* Allow nosetests to figure out how many cores to use
* simplify merge
* Mark the new CLI tests as ||izable
* Simplify reporter_test changes
* Rationalise ||ism flags
* Re-up coverage
* Clean up reporter tests
* Stop modifying testdata during tests
* remove unused os
Test farm tests should test the version of letsencrypt-auto that's in the git tree, not the one from the previous release.
* Test the new leauto, not the previously released one
* Ensure tests pass with openssl 1.1
A bunch of the acme.standalone and acme.crypto_util tests were using
weak crypto that is now prohibited :/
* lint
* lintlint
* Fix symlink
* Document the Nginx plugin release
* Tweak
* Remove mrueg nginx instructions for now?
* Shipped -> included
* keep order of plugin descriptions consistent with the table
* Document the Nginx plugin release
* Tweak
* Remove mrueg nginx instructions for now?
* Shipped -> included
* keep order of plugin descriptions consistent with the table
One more tiny tweak, placing the slower sudo environments back up the list. I expect this should save us another 10-20 seconds.
- there's about a 50 second delay in starting apacheconftest, so move
it back up the priority queue of jobs to start
Re-do the fix for #3497 to ensure it works in all cases.
* If lineages are in an inconsistent (non-deployed) state, deploy them
* Test new _handle_identical_cert case
* Move lineage.has_pending_deployment() check up to _auth_from_domains
Less conceptually nice, but in the "renew" verb case it wasn't being called :(
* Swap _auth_from_domains return type
* It now matches _treat_as_renewal & _handle_identical_cert_request etc
* Revert "Move lineage.has_pending_deployment() check up to _auth_from_domains"
This reverts commit a7fe734d73.
* Move test back to handle_identical_cert_request
* We need to check for non-deployment on two separate code paths
- Once high up in "renew" (because failure to be deployed stops us from
divind down the stack)
- Once way down in _handle_identical_cert_request (because that's where it
makes the most sense for run / certonly)
- So refactor that work into storage.py
* We don't necessarily reinstall
* Tweak for Travis performance
- merge cover and py27 BOULDER_INTEGRATION into one matrix entry
- re-order to put the fastest environments last, improving average
case parallelism
* Also put the things most likely to fail at the top
