getting bugs to the right people

Giovanni Giorgi jj at
Fri Jun 4 12:35:46 UTC 2004

goran.krampe at ha scritto in data 04/06/2004 11.34:

> [SNIP]
>Yes, and this is in fact why I thought and still think that TFNR is a
>very important - though currently dormant - project.
>We need to stake out the image and get those areas assigned to one or
>several maintainers. And note - this does NOT mean that the staked out
>parts need to be actually broken out of the image as a separate package.
>But given a method or class we do need to know which package it belongs
I agree but can I shot a small provocation?
I have seen at least two project failed to stake out the image.
The squeak modules project failed, as far as I know.
Stable squeak project seems sleeping or dead.
And TFNR is dormant.

So I think we could start thinking in the **opposite direction**: build 
an image with only the kernel, and start to add things to it.
For example we can start adding kernel and socket, and then SM.
We got a headless image without morphing, but perheaps listening for a 
telnet session on a port, giving to us a command line squeak.
Then we can add all the rest via SM: morphic in primis and so on.

>I am focused on SM right now so I will not rekindle TFNR until SM has
>been moved one more step forward. But it would be just GREAT if someone
>else could pick up the torch in the meantime.
>And yes, perhaps I did fail with TFNR and should step down as a Guide as
>a result, haven't decided yet.
I do not think you failed!
I think we must change our tactics against "the blob image".
SM, BFAV and squeak people are a very big step in the squeak history: we 
must start form there.
If we start with a tiny image, stacking out will be done by the mantainer:

    NewbieGio:"Hey I got my Celeste AddressBook on the tiny image! What
    must I do?"
    Guide: "You must pack it in SM, resolve dependency and ask to the
    telnet squeak to load it :))".

and let's go together!!

 Giovanni Giorgi               ObjectWay SpA - Milano    via Boltraffio, 7 - 20156 Milano

More information about the Squeak-dev mailing list