[ZODB-Dev] zodbpickle claim (but OS X is not unix)

[Christian Tismer]

On 04.07.13 17:52, Stephan Richter wrote:
> On Thursday, July 04, 2013 05:47:39 AM Christian Tismer wrote:
>> there is now a pull request, completely tested on OS X,
>> ready to be applied with no other work involved.
> Yeah, I saw that. thanks.
>
>> It would be nice if that version could be uploaded, soon, so
>> that I can close this issue and move on. ;-)
> Yeah, I just want Tres and/or Jim to weigh in, since they did the last
> iterations on this code. And you know, we have Independence Day today, so this
> week might be a little bit tough. ;-)
>
>> Thanks guys, and keep up the good work
> Thank you for using/testing zodbpickle and sending us the patch. BTW, are you
> using zodbpickle by itself to create Py2/3 compatible code?
>

Hi Stephan,

I am reviving Durus right now as my "super pickle for the pocket",
and don't want to stay stuck on protocol 2, incompatibility with the
python version etc. .
Personally, I have moved my projects to Py3.3, but a database is
a different thing that should really not suffer from that.

After some hacking, I realized that the problem is not so trivial, and
fortunately found zodbpickle.

So I thought that is the way to go, contribute a bit and use it.

I'm working on BTree forests for a versioned, read-only database,
and those versions come every two weeks, but I want to keep them all
in the same database without keeping redundant data.

That got me to the forest idea.

First thing I was looking into was the B+Tree impl. of Zodb, but that
was too much for me to change just for a prototype, because of all
the optimized C code.
And also the bucket pointers of B+Tree are disturbing a bit, because
every bucket/subtree can be part of many trees, so I have to think
how that should be.

But in the end I agree that Zodb is the real thing, and I will eventually
move there, when my forests prove useful and working.

Oh, back on the question:
Yes! We will use zodbpickle for all persistence stuff.

And I want python.org to incorporate these patches, because I think
that would help everyone. Why don't they want that small change?

cheers - Chris

-- 
Christian Tismer             :^)   <mailto:tismer at stackless.com>
Software Consulting          :     Have a break! Take a ride on Python's
Karl-Liebknecht-Str. 121     :    *Starship* http://starship.python.net/
14482 Potsdam                :     PGP key -> http://pgp.uni-mainz.de
phone +49 173 24 18 776  fax +49 (30) 700143-0023
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
       whom do you want to sponsor today?   http://www.stackless.com/