Re: [GNUnet-developers] using gdbm for content

[Uli Luckas]

Am Samstag, 4. Mai 2002 22:33 schrieben Sie:
> It's a long patch, so it'll take me a while to check everything, but if I
> can confirm that it works, it will be applied. We may also need a converter
> (old data/content to gdbm database), but that can be written even after CVS
> has the new code in it.
Hi Christian,
this is a big patch and after thinking about it, replacing one data store 
with it's issues with another data store with its own issues might not be the 
right way to go.
I think a better way to go would be to encapsulate all design specific 
details and create an api to access abstract buckets by hash values.
I would vounter to port the "thousend file" data store design as closely as 
posiible to an encapsulated api, step by step and you'd have to look at small 
patches, only.
After that it be easy to write new implementations of the same api in order 
to experiment with different data store layouts.
These new implementations would be a lot easier to verify as they implement 
basic functionality, only.
And maybe people get to configure their data store leyout depending on their 
needs. Some people migth want to store their content in PostgreSQL ;-)

Please let me know what you think and if you are willing to accept a big 
number of small, easy to verify patches.

Thanks
Uli

>
> Great work, thanks a lot!
Fun work. Thanks for starting this project.

Uli