[ruby-dbi-users] ADO adapter.

Erik Hollensbe erik at hollensbe.org
Mon Mar 10 08:56:49 EDT 2008

Tim Uckun wrote:
>>  Very soon we will not be supporting the Windows-based DBD's, but we're
>>  looking for someone to manage these. If you have any interest, let me
>>  know privately.
> I am disappointed to hear that.  i don't know if I want the
> responsibility of maintaining the codebase but I will think about it.
> Thanks for the git URL.
> BTW does this mean you are not interested in the patches

Yes, we *are* interested in the patches.

The core issue (that I should have explained better earlier, but I just 
got back from the bar :) is that we're eventually breaking the DBDs out 
into separate gems for separate maintainership, similar to how Perl 
manages DBI and its related DBDs. Chris and myself have no access to 
windows testing resources and the databases in particular, and we both 
consider testing to be crucial to how development takes place. To top 
this, neither of us are Windows developers, making low-level work 
painful and error-prone (which doesn't benefit anyone, really).

So, since we can't really *test* ADO and SQLServer and other 
Windows-based databases, we're unfortunately unable to support them 
directly. What would be our ideal situation is that these DBDs do not 
disappear but are picked up by someone that has the same interest in 
testing that we do and enough drive to maintain these against a DBI/DBD 
standard that evolves with time. And hey, that person gets a free test 
suite to accompany the DBD. ;)

Right now I can almost predict, considering the state of other DBDs that 
I've worked on, that when the current spec is followed, it's followed 
poorly. We're looking to limit the burden of supporting all these 
databases by allowing others to maintain DBDs that they can test against 
so we can focus on making DBI and DBDs better as a whole by allowing 
more people to work on them, write tests for them, release them 

I HTH. It's not that we don't want to support them, it's that it's 
really not practical to do so for a number of reasons. We are looking 
for a windows person that can help us with the project as a lead 
developer and hope to find that person before our first feature release.


More information about the ruby-dbi-users mailing list