Can somebody say where we are in the release process of 3dot9?
As far as I can see we are somewhere near the finish of 3dot9 except:
1) There are reports of the sources being corrupted.
2) The need for some really big changes are coming out of the woodwork. -- ffi removal and the consequences of that. -- source file limitations
3) Some problems with condensed sources have become apparent to some. (see 1) -- how are these affecting the release?
4) What is the method/process for accepting the final 3dot9?
Accepting is a customer function rather than a developer function.
-- Who is playing the customer here?
I figure:
1) 3dot9 is done and I am just not aware of it.
2) Its not done but the resources for the doing of it have been exausted.
3) Its some where near done.
4) We've just discovered what the real problems are. There will be a considerable delay in its doneness.
----
and either:
1) Its acceptable.
2) Its not acceptable.
So what's the official word(s)?
Yours in service and curiosity, -- Jerome Peace
__________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com
Peace Jerome wrote:
- There are reports of the sources being corrupted.
I don't know where this stands but I would recommend just issuing another update that fixes those methods that are broken instead of redoing the entire condenseSources process.
- The need for some really big changes are coming out
of the woodwork.
Not sure what you mean by that.
-- ffi removal and the consequences of that.
The only consequence is that packages who need it will have to ensure that the FFI is loaded first. If there are any other issues I'm unaware of them.
-- source file limitations
Not sure what these limitations are - one of the reasons for making a new sources file was precisely to address some of these issues before they become a problem.
- Some problems with condensed sources have become
apparent to some. (see 1) -- how are these affecting the release?
It seems to me that there may be a problem in condenseSources but since we can simply reissue an updated with the methods in question fixed, this is by no means a show-stopper.
Cheers, - Andreas
We are trying to fix the condenseChanges bugs but this is not easy. I spent some times to try to arrive to a smaller test fixture that would show the problem. But so far I have been too busy (esug is approaching and the academic year too).
I thought that it would be better to ship 3.9 without the bug but I will discuss with adrian since he should be back from the military service to see. What we do.
On 27 août 06, at 00:17, Peace Jerome wrote:
Can somebody say where we are in the release process of 3dot9?
As far as I can see we are somewhere near the finish of 3dot9 except:
- There are reports of the sources being corrupted.
yes
- The need for some really big changes are coming out
of the woodwork. -- ffi removal and the consequences of that. -- source file limitations
FFI has been removed There is no source file limitation since we condenseChanges and condenseSources in 7054
- Some problems with condensed sources have become
apparent to some. (see 1) -- how are these affecting the release?
See above
- What is the method/process for accepting the final
3dot9?
Normally for us we will either find the bug and fix it and ship or ship 3.9
Accepting is a customer function rather than a developer function.
? sorry but where is the money of the customer?
-- Who is playing the customer here?
I figure:
- 3dot9 is done and I am just not aware of it.
you have 7054 which is gamma
- Its not done but the resources for the doing of it
have been exausted.
Not totally but nearly
Its some where near done.
We've just discovered what the real problems are.
There will be a considerable delay in its doneness.
Stef
squeak-dev@lists.squeakfoundation.org
-
Andreas Raab -
Peace Jerome -
stéphane ducasse