[elephant-devel] Re: unstable postmodern troubled..

[Ian Eslick]

Fantastic!  Thanks for all the work on this.  Now I have no excuse to  
start the long list of little items necessary to get to a release...

The big areas are:
- Thread safety for all schema-related operations
- Test coverage
- Docs coverage

Some bugs I've seen and haven't been able to fix:
- Leslie's BDB bug on unstable
- BDB on AMD 64

Anything else?

I think Glenn was going to help us out with a little organization of  
this list vis-a-vis documentation?  The missing tests are roughly the  
same as the missing doc sections, so we can come up with a list of  
tests from that.

Anyone want to do a thread-safety review of the schema code?  A fresh  
pair of eyes is usually much better for this than the developer as  
recent e-mails have shown.  However, I'll take a first pass before too  
long.

Thanks,
Ian

On Jun 2, 2008, at 4:47 PM, Alex Mizrahi wrote:

> IE> Sounds like there is an issue with schema coherence (the class  
> object
> IE> cache out of sync with the db in some way?  I had some of these  
> tuning
> IE> the BDB implementation).
>
> i've found the issue -- postmodern backend creates its "system" btrees
> (like root, schema-table etc) in "fancy" way, so it cannot simply
> deserialize
> them if something in DB refers to it, so it depends on object cache.
>
> but some tests call flush-instance-cache, so db-postmodern gets  
> broken.
>
> i'm not sure what is the idea behind "special" btrees.. probably it  
> will
> work fine without this special stuff, but i'd prefer not to touch  
> them :),
> so i'll better specialize flush-instance-cache to ensure system btrees
> are cached after flush.
>
> seems to be passing all tests now, ha!
>
>
>
> _______________________________________________
> elephant-devel site list
> elephant-devel at common-lisp.net
> http://common-lisp.net/mailman/listinfo/elephant-devel