It works.

The only annoying thing is it has to read and discard all the
attachment chunks from the beginning of the attachment to the start of
the range. After compaction, the #att contains enough data to avoid
that.

B.

On Mon, Nov 15, 2010 at 12:13 AM, Jan Lehnardt <[email protected]> wrote:
>
> On 18 Oct 2010, at 12:28, Robert Newson wrote:
>
>> upgrade by compaction, same as it ever was.
>>
>> When the new database is written out, the #att.data member becomes
>> [Position, Length] automatically.
>
>
> What happens when I make a Range request against a pre-compcation
> database?
>
> Cheers
> Jan
> --
>
>>
>> B.
>>
>> On Sun, Oct 17, 2010 at 3:26 PM, Jan Lehnardt <[email protected]> wrote:
>>>
>>> On 7 Oct 2010, at 12:11, Robert Newson wrote:
>>>
>>>> I should remind everyone that the ranges support alters the format of
>>>> the #att.data field on compaction (it adds the length of the block so
>>>> we only need read the blocks within the range). While I've tested this
>>>> to my satisfaction, no automated testing exists for verifying upgrades
>>>> at this time.
>>>
>>> For everybody not exactly in the know, what is the expected upgrade
>>> procedure?
>>>
>>> And how do we get tests for it? :)
>>>
>>> Cheers
>>> Jan
>>> --
>>>
>>>>
>>>> B.
>>>>
>>>> On Wed, Oct 6, 2010 at 12:09 AM, sgoto <[email protected]> wrote:
>>>>> On Tue, Oct 5, 2010 at 9:43 AM, Filipe David Manana 
>>>>> <[email protected]>wrote:
>>>>>
>>>>>> On Tue, Oct 5, 2010 at 5:31 PM, Zachary Zolton <[email protected]>
>>>>>> wrote:
>>>>>>> For those of us not following at the commit level, what differences do
>>>>>>> you see between 1.0.2 and 1.1?
>>>>>>> —Zach
>>>>>>
>>>>>> 3 new major features at least:
>>>>>>
>>>>>> - HTTP ranges support
>>>>>> - replicator DB (_replicator)
>>>>>>
>>>>>
>>>>> what does _replicator do ? does it persists the replications settings ?
>>>>>
>>>>>
>>>>>> - SSL support (I have some stuff to add fthere or the replicator)
>>>>>>
>>>>>>>
>>>>>>> On Tue, Oct 5, 2010 at 10:17 AM, Chris Anderson <[email protected]>
>>>>>> wrote:
>>>>>>>> Devs,
>>>>>>>>
>>>>>>>> I think there may be some work to do before 1.1 is released (but maybe
>>>>>> not).
>>>>>>>>
>>>>>>>> We should release it before Paul's externals stuff lands, and also not
>>>>>>>> wait for the Sammy.js Futon or Replicator DB UI.
>>>>>>>>
>>>>>>>> Let's start the round-up of patches and tickets that need to go into
>>>>>>>> 1.1. Big things should wait, but simple fixes and refinements are a
>>>>>>>> good idea to bring up.
>>>>>>>>
>>>>>>>> Cheers,
>>>>>>>> Chris
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> --
>>>>>>>> Chris Anderson
>>>>>>>> http://jchrisa.net
>>>>>>>> http://couch.io
>>>>>>>>
>>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> --
>>>>>> Filipe David Manana,
>>>>>> [email protected], [email protected]
>>>>>>
>>>>>> "Reasonable men adapt themselves to the world.
>>>>>>  Unreasonable men adapt the world to themselves.
>>>>>>  That's why all progress depends on unreasonable men."
>>>>>>
>>>>>
>>>>>
>>>>>
>>>>> --
>>>>> f u cn rd ths u cn b a gd prgmr !
>>>>>
>>>
>>>
>
>

Reply via email to