[
https://issues.apache.org/jira/browse/COUCHDB-1446?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13233218#comment-13233218
]
Paul Joseph Davis commented on COUCHDB-1446:
--------------------------------------------
Also, while I do agree this is a good idea, the setting that borked is a bug
with how that value is used when it's found. Regardless of how strict you are
on type information it'll always be possible to throw a wrench in the gears
with an unexpected value in the config. But having some sort of internal
tooling to handle the parsing and presentation of these values that is separate
from the code where the value is used seems like it'd be quite beneficial from
an engineering perspective.
> Config settings input validation
> --------------------------------
>
> Key: COUCHDB-1446
> URL: https://issues.apache.org/jira/browse/COUCHDB-1446
> Project: CouchDB
> Issue Type: Bug
> Components: Database Core
> Affects Versions: 1.2
> Reporter: Jason Smith
> Assignee: Jason Smith
> Priority: Minor
> Labels: config
>
> Today I received a bug report from a user who thought CouchDB might be
> relaxing and evaluate an arithmetic expression in the config. (That makes
> sense, since it seems to evaluate erlang terms elsewhere.)
> https://getsatisfaction.com/iriscouch/topics/my_couchdb_is_broken_after_putting_a_badarg_for_a_configuration_value
> It would be nice for CouchDB to validate config input, particularly when bad
> values permanently take it offline. (In this case, it returns 500 for all
> subsequent requests.)
> IMO, a good bang-for-buck is to have three basic value types:
> 1. String
> 2. Erlang tuple
> 3. Integer
> And each setting knows what type it must be.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
