[pypy-dev] Please submit status reports

Laura Creighton lac at strakt.com
Mon May 26 11:24:50 CEST 2003

In a message of Sun, 25 May 2003 15:24:08 EDT, Rocco Moretti writes:
>Jacob Hallén <jacob at strakt.com> wrote:
>>To help the sprinters, we would be grateful if everyone could post a bri
>>status report about the latest things you did on the project and what yo
>>think the next step in this area is. I particular, we would be grateful 
>>Rocco, Dinu, Holger and all the sprinters would provide input.
>  Most of my time on PyPy since the last sprint was involved in setting u
>a testing framework I felt comfortable using (submitted as PyPyTest.py in
>subversion under /user/rocco/tool/). To get the CPython regression tests 
>to work properly, I had to improve exception matching. Once that was 
>working, the missing exec statement functionality was low hanging fruit. 
>Aside from my recent philosophy posts, I haven't done anything more with 
>PyPy due to time constraints.
>Where to next:
>  I'm of the opinion that the acid test for this phase of PyPy developmen
>is the sucessful execution of the CPython regression tests - does PyPy do
>what CPython is supposed to do.  So squashing regression test bugs is hig
>on my priority list. I think some of the bugs will need a slight core 
>redesign in order to fix. So I'd think about the cause of the regression 
>test bugs, and how to redesign the core to fix them.
>  Once that's done, the StdObjSpace needs completion and testing, and if 
>there is time left, adapting/creating an Application Level parser/compile
>And unless I miss my guess, that'll take us to the proposed 0.1.0 
>milestone, at which point the fun can begin.
>In short, empty the issues board. (Assuming "Finish StdObjSpace" and "No 
>PyPy Compiler" get added at some point.)

Right now what we are doing is Retiring Trivial Object Space.  Standard
Object Space is being made useful.  Right now the goal for the end of
the day is to be able to run

    s = "Hello World"
    print len(s)

Right now we are dealing with builtins, including that __builtins__
right now is a module some of the time and sometimes a dict. 

This is true of C python as well, of course ...

>>> __builtins__
<module '__builtin__' (built-in)>
>>> import string
>>> type (string.__builtins__)
<type 'dict'>

We discovered that None is not None as well.  Time for our objects to
get id-s ...

Rocco, can you come to Belgium for a Sprint June 21-24 before EuroPython?

Laura Creighton

More information about the Pypy-dev mailing list