Re: [ANN] Persistor, Transparent Persistence for Lua data!

[Coda Highland <chighland@...>]

On Thu, Jun 6, 2013 at 11:36 AM, Javier Guerra Giraldez
<javier@guerrag.com> wrote:
> On Thu, Jun 6, 2013 at 6:07 PM, Rena <hyperhacker@gmail.com> wrote:
>> On Thu, Jun 6, 2013 at 2:01 PM, Petite Abeille <petite.abeille@gmail.com>
>> wrote:
>>>
>>>
>>> On Jun 6, 2013, at 7:27 PM, Rena <hyperhacker@gmail.com> wrote:
>>>
>>> > That can't
>>> > be done with SQLite though since only one process can access the
>>> > database
>>> > at the time.
>>>
>>> Eh?
>>>
>>> http://www.sqlite.org/wal.html
>>>
>>
>> Interesting, I hadn't seen that before.
>
>
> in fact, SQLite has always had robust concurrency capabilities.  But,
> to make it fast and not too complex or system-dependent, the lock
> strategies were more global than most server-based databases.
>
> in many cases, applications keep a write lock even during user
> interaction, making the serialization of tasks visible to the user.
> But that's an application choice, not a hard SQLite limitation.
>
> WAL doesn't bring new features, just reduces the window of exclusivity
> to levels similar to other RDBMS implementations.
>
> --
> Javier
>

Prior to WAL, the statement that only one process can access the
database at a time was true, BECAUSE of the global lock. Multiple
processes could have it OPEN but there's no actual concurrency.

/s/ Adam