[pmwiki-users] RFC- Database access via cookbook - agnostic? Re: SelectQuery development: database records as pages?

[Pico]

Crisses wrote:
> 
> On Sep 29, 2006, at 8:42 PM, Kathryn Andersen wrote:
> 
>> But having something that can act as an easy portal to a database (even
>>
>> better if it is database-agnostic, since I tend to use SQLite rather
>>
>> than MySQL for website databases) that could use the power of the
>>
>> database for selecting and sorting, but the convenience of pagelist
>>
>> templates for formatting its records, that would be great!
>>
> 
> Ben Wilson and I have discussed (and agreed on the benefits of) using 
> ADODB for creating a database-agnostic version of AuthUserDbase that 
> includes options for the features I added to create self-sign-up, email 
> validation, etc.
> 
> I'm not sure when that's going to happen as he's working on a degree, 
> and I'm working on paying my rent.  If one additional person in my 
> household gets a job, I'll be able to put more effort towards PmWiki 
> modules (justifiable because nothing looks so good in my portfolio as my 
> involvement in this community LOL).
> 
> I think it's a good idea to use a tool like ADODB, however, for 
> something like SelectQuery as well.  I'd be more than willing to talk to 
> the package maintainer(s) about it.
> 
> As long as we can all agree on one call to include ADODB it should all 
> be good ;)  And that would allow you to use SQLite instead of 
> mysql/postgresql, etc. since that's one of the databases that ADODB 
> supports.
> 
> Does anyone object to the use of a tool like ADODB for some recipes?  1) 
> same license as PmWiki 2) 2.7 MB installed (currently) 3) one of the 
> better open-source database abstraction layers 4) supports MANY 
> databases: http://phplens.com/adodb/supported.databases.html (some only 
> on Windows, or with a plug-in)  (may even include LDAP, but they're not 
> very clear on that one)

How does ADOdb compare with Pear db?

While the temptation is to focus on issues like speed and functionality 
(in which case I gather from some quick google searches that ADOdb may 
have the advantage), I suggest that equally, if not more important, in 
this context are issues of availability, installation, and size.

Specifically, the most important issue, in my view, would be which tool 
is more commonly available on shared hosting servers.  I gather from 
past posts that both of you, including Ben W. ;-), maintain, or have 
access to your own servers and, thus I thought it was important to raise 
this issue explicitly.

Most shared hosting servers provide php and mySQL, but I do not know the 
extent to which ADOdb or Pear db are included and, if not, whether they 
must be installed by the hosting provider (since shared hosting accounts 
typically have no command line access and ability to install executables).

> 
> So essentially I'm asking cookbook authors if we can all agree on one 
> database abstraction layer for PmWiki -- for use when we're going to 
> deal with external databases.  It would be quite helpful if we all 
> agreed, and greatly extend the cookbook's support for a variety of data 
> sources.
> 

That sounds like a great idea.

Pico

-- 

         __  /
        /   /
       /___/ _/  ___/  __  /
      /      /  /     /   /
    _/     _/  ____/ ____/

 >>>===pmwiki at ben-amotz.com===>