couchdb-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jan Lehnardt <...@apache.org>
Subject Re: svn commit: r660139 - /incubator/couchdb/trunk/src/couchdb/couch_httpd.erl
Date Mon, 26 May 2008 10:08:33 GMT

On May 26, 2008, at 12:03, Christopher Lenz wrote:

> Hey Jan,
>
> On 26.05.2008, at 11:28, Jan Lehnardt wrote:
>> On May 26, 2008, at 11:20, cmlenz@apache.org wrote:
>>> Author: cmlenz
>>> Date: Mon May 26 02:20:47 2008
>>> New Revision: 660139
>>>
>>> URL: http://svn.apache.org/viewvc?rev=660139&view=rev
>>> Log:
>>> Change error handling on unknown private paths to be more  
>>> consistent with how other errors are handled.
>>
>> This was a deliberate derivation from the error
>> handling to be more user-friendly. That is, directly
>> show the user that they meant "/_utils" when they
>> tried "/_util".
>
> I understand that, but I think this is a slippery slope, and it *is*  
> a deviation :)
>
> First, there are other "private paths" (not sure that's even a good  
> name): /_all_dbs, /_compact, /_search, /_replicate. So even this  
> early in the process we've forgotten to add quite a number of URLs  
> to the list of "known private paths". And as long as the URL mapping  
> is not dynamic/modular, we'll have to remember to keep the list in  
> sync with the URLs actually in use.
>
> Second, where do you stop? You could argue the same about special  
> URLs in the database namespace: /_view, /_design, /_all_docs, etc.
>
> Finally, is it really that hard to remember that it's "/_utils" and  
> not "/_util"? Or just try the former after the latter failed? :P
>
>> with the advent of customizable private paths, I though
>> this a valuable addition.
>
> But that's not there yet. I'd rather wait until it's in place (or we  
> have a better idea about the design) until designing special error  
> messages with that feature in mind.
>
> But of course I can back out the change if you don't agree with any  
> of my arguments, it's not such a big deal anyway :)

I do agree with your arguments ;) but in the special
case of _util and _utils I think it is worth it to go the
extra mile and make the user happy.

If the other's out-vote me, I  shall shut up about this :)

Cheers
Jan
--


> Cheers,
> --
> Christopher Lenz
>  cmlenz at gmx.de
>  http://www.cmlenz.net/
>
>


Mime
View raw message