hc-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jon Moore (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HTTPCLIENT-1828) HttpDelete does not extend HttpEntityEnclosingRequestBase
Date Fri, 10 Mar 2017 15:09:04 GMT

    [ https://issues.apache.org/jira/browse/HTTPCLIENT-1828?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15905217#comment-15905217

Jon Moore commented on HTTPCLIENT-1828:

I'll toss my 2c in, hoping to be helpful.

Because new development effort for the project is being driven towards the 5.x release, we
probably want to take a stance that updates to the 4.x series should really only address bugs--things
that are incorrect. A lot of the discussion seems to boil down to whether this is a bug or
an improvement. I believe everyone on the thread so far (including the original requestor!)
seems to acknowledge that bodies on DELETE requests are, at least, very odd. I don't think
this is a common use case we particularly need to worry about, especially when the custom
method workaround is straightforward. If we muck around with class hierarchies, I'm always
wary we might be breaking an existing use. [Side note: Gary - does your patch pass the clirr
binary-compatibility checks?]

Overall, I think a good philosophy for the project is a variant of Postel's Law: we generally
want to be strict in what we send, and permissive in what we accept, for interoperability
reasons. However, sometimes you have to be non-strict in what you send, due to non-permissive
recipients (or recipients who read the spec differently than we do); that's just the real
world. So I think we want something like: the easy/default thing for users to do should generate
strict messages; however it should be possible, through configuration/extension/dependency
injection, to do non-strict things.

In this particular case, my vote would be NOT to implement this request, because:
* The 4.x branch is in maintenance mode.
* It is unclear (given the discussion thread of reasonable people vehemently disagreeing)
whether this is actually a bug.
* There is an easy workaround for the user.
* We believe this use case will be accommodated by the 5.x branch.

> HttpDelete does not extend HttpEntityEnclosingRequestBase
> ---------------------------------------------------------
>                 Key: HTTPCLIENT-1828
>                 URL: https://issues.apache.org/jira/browse/HTTPCLIENT-1828
>             Project: HttpComponents HttpClient
>          Issue Type: Improvement
>    Affects Versions: 4.5.1, 4.5.2, 4.5.3
>            Reporter: Saravanakumar Selvaraj
>             Fix For: 4.5.4, 5.0 Alpha1
> Sending body in DELETE is allowed as the HTTP 1.1 specification does not forbid. Hence
HttpDelete.class should extend from HttpEntityEnclosingRequestBase instead of HttpRequestBase.

This message was sent by Atlassian JIRA

To unsubscribe, e-mail: dev-unsubscribe@hc.apache.org
For additional commands, e-mail: dev-help@hc.apache.org

View raw message