Aggregated removal script for 3.6

Doug Way dway at riskmetrics.com
Fri May 2 06:13:23 UTC 2003


(Sorry folks, I've been busy the last few days and am now just catching 
up with the list...)

On Thursday, May 1, 2003, at 08:30 AM, Daniel Vainsencher wrote:

> diegogomezdeck at consultar.com wrote:
>> What about to make a "oficial" call to help?
>>
>> I think a lot of us are able to work in this area and I never know
>> the "door is open".
>
> You're right, I think we should, though I at least would not do it
> *right now*. I think we should have a somewhat clearer idea of how the
> whole process should work (which will come naturally from giving it a
> little time, though we might be a little slower than one might desire
> during this time) and then we'll widen the circle.

Probably a good idea.  Although if we do that, I'd almost be tempted to 
go with the less harvesting-intensive approach of adding all of the 
removals at once.

On the other hand, adding them one at a time might work if we have a 
well-defined process... I wouldn't want to coordinate incorporating 
each one myself one at a time, on an ad-hoc basis, that sounds like a 
lot of work.  Hm, also, if we do one at a time, and make sure each one 
is tested before adding the next, even allowing only 3 days between 
adding each removal, that would mean it would be 30+ days before we 
progress to KCP/MCP/other stuff in the 3.6 plan, which is quite a while.

I would suggest if we try adding them one at a time, we use the regular 
Harvesting Process (defined on the swiki at 
http://minnow.cc.gatech.edu/squeak/3152 ).  Someone could post them all 
to the list, people could comment that the removals (and perhaps the 
re-added packages) work correctly, they'd get approved & then 
incorporated.  They'd probably end up getting incorporated a few at a 
time, depending on when they were approved.  (We might still have a 
harvesting bottleneck, although if enough people comment that the 
removals/additions work, the harvesting should be easy.)  We could set 
an arbitrary goal of trying to get them done within a couple of weeks 
or so.

Or, the other strategy is to just incorporate all 10 removals right 
now, and it's up to the package maintainers to make sure the re-added 
packages work properly and resolve conflicts with other packages.  
Also, we'd have to deal with any problems caused by the 10 removals, 
but those should less of a problem. (at least for these early "easy" 
removals)

I think I'm still leaning toward the "remove all 10 right now" strategy 
since the removals have been out on SqueakMap for awhile now.  We could 
discuss a bit more, but let's decide on a strategy in the next day or 
two.

(I guess the Test packages are not yet available for each package 
addition, which is another issue.  We should probably make it a 
requirement that "Squeak Official" packages must have a matching Test 
package, even if it's just a simple test, along the lines of what 
Marcus suggested in that April 15th message.  Also, we could declare 
Marcus to be the "testing" Volunteer, so perhaps he would be 
responsible/empowered to bug the package owners to add the appropriate 
Test packages, and provide guidance when needed...? :-) )

- Doug Way



More information about the Squeak-dev mailing list