[squeak-dev] Its a shame

Keith Hodges keith_hodges at yahoo.co.uk
Sat May 31 19:42:24 UTC 2008


Dear Edgar,

My own personal goal with my contributions to 3.11 is to encourage 
teamwork, the contributions of others and the potential for integration 
of the various existing streams of squeak past present and future. My 
other goal is to continue to have a solid basis on which to build 
commercial applications. I believe that Matthew and others share these 
goals, and we also recognize that it is not going to be easy to achieve.

Having professed this goal a number of months ago, I find it very 
disappointing and somewhat ironic we now have 3 different planned routes 
forward for 3.10+. Especially since the fundamentals of the way all 
three branches have proposed moving forward are essentially the same.

For you to say there has been no progress in 8 months indicates to me 
that you haven't looked at the progress that IS being made. I think 
there has been lots of progress.

Example 1
========
If we take a goal of 3.10 and 3.11, that being to unload packages from 
the image all the way back to a kernel image and make them loadable 
again. Kernel images dont have a UI, so if we do not have an effective 
tool for loading things back in again, there is no point in taking 
things out in the first place. Hopefully Sake/Packages is that tool.

So with your infinite knowledge of loading lots of things into squeak to 
build FunSqueak, there is a great deal that you can contribute to the 
community effort of 3.11. You could help fill Sake/Packages with your 
knowledge of how to load and unload packages into 3.10.

Example 2
========
Some that know better than I have stated that Atomic Loading is needed 
for effectively managing loading/unloading crucial packages in the 
kernel. Monticello 1.5 has supported Atomic Loading for a while, however 
there have been some System Editor bugs to iron out, which we now think 
has been done. So in the past 8 months the possibility of Atomic loading 
has moved from coded and guaranteed to crash your image to actually 
usable. This is progress, because without it we know that the release 
team would expect real pain the closer the KernelImage ideal is reached.

Example 3
========
One major hindrance to the different streams working  together has been 
the traits issue. Thanks to Matthew's UnloadTraits script this is no 
longer as big an issue as it was.

The above three examples demonstrate that there has indeed been 
significant progress in the past 8 months, and these are not the only 
examples.

The current situation is that we have some already overworked people 
working on different things, and there is always room for more 
contributions. However we know that our sub-projects will eventually 
work together for good. None of us feel we need to take over and "do it 
all ourselves". There are lots and lots of smaller sub-projects to which 
you could contribute, to the team effort.

so how about it?

Keith




More information about the Squeak-dev mailing list