On 02.08.2012, at 13:38, Colin Putney wrote:

On Thu, Aug 2, 2012 at 1:33 PM, Dale Henrichs dhenrich@vmware.com wrote:

...

I have a small 3 class package (Metacello-Base[1]) that I would hope could be included in the Squeak4.4 release.

+1

It would be nice to have a cross-dialect API for install scripts. Thanks for keeping Squeak support up to date.

Colin

Last time I tried to install something using Metacello it pulled in many unrelated packages (like Gopher). Was that a fault of the package I tried to load, or Metacello's? IOW, will this use Installer to do the actual work or pull in even more stuff?

- Bert -

Makes sense. Thanks for the explanation.

- Bert -

On 02.08.2012, at 17:48, Dale Henrichs wrote:

Bert,

The notion of the Metacello--Base is that the 3 classes are there as a "stub" for the full Metacello system.

If you use the Metacello api, then you will trigger an install of Metacello itself.

I'm only asking that the base image have those 3 classes installed so that the folks who choose to use Metacello will be able to use the Metacello api.

With the Metacello api, the plan is to reduce the public facing dependence upon Gofer, since many Metacello instructions today include a Gofer invocation to get the configuration loaded into the image....

Regarding the implementation of Metacello itself, I have written Metacello against the Gofer api, so Gofer does get loaded into the image in support of Metacello ... Gofer was used because at the time (2009) it was the only cross platform solution for providing a common api to Monticello and frankly at the time I imagined the Gofer would be accepted in the base image for all the platforms, but it has not and I just haven't had the time to changed the architecture ... moving forward I would accept contributions that replaced Gofer with something else on Squeak (I have some basic requirements for functionality that would need to be expressed)...

Using Gofer meant I did't have invent something that works the same on Squeak, but if someone wanted to take on the task of:

a) providing equivalent functionality on Squeak b) helping me rearchitect Metacello to isolate the Gofer dependencies

I'd be supportive of that...

Dale

----- Original Message ----- | From: "Bert Freudenberg" bert@freudenbergs.de | To: "The general-purpose Squeak developers list" squeak-dev@lists.squeakfoundation.org | Sent: Thursday, August 2, 2012 3:37:07 PM | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | | On 02.08.2012, at 13:38, Colin Putney wrote: | | > On Thu, Aug 2, 2012 at 1:33 PM, Dale Henrichs dhenrich@vmware.com | > wrote: | >> I have a small 3 class package (Metacello-Base[1]) that I would | >> hope could be included in the Squeak4.4 release. | > | > +1 | > | > It would be nice to have a cross-dialect API for install scripts. | > Thanks for keeping Squeak support up to date. | > | > Colin | | Last time I tried to install something using Metacello it pulled in | many unrelated packages (like Gopher). Was that a fault of the | package I tried to load, or Metacello's? IOW, will this use | Installer to do the actual work or pull in even more stuff? | | - Bert - | | |

On 2 August 2012 21:49, Chris Muller asqueaker@gmail.com wrote:

If there's little or no danger of any serious regressions elsewhere in the system, then I'm +1. Like Frank, I think it's important for Squeak to continue to put out stable releases.

I need to hack on the CI server first, but if a kind soul wouldn't mind just loading Dale's mcz and running the full suite of tests, I'd be grateful. I don't see how that could cause regressions, but I'm paranoid.

Dale, your efforts on cross-platform are amazing, thanks!

Seconded!

frank

On Thu, Aug 2, 2012 at 3:33 PM, Dale Henrichs dhenrich@vmware.com wrote:

...

I have a small 3 class package (Metacello-Base[1]) that I would hope could be included in the Squeak4.4 release.

Of the 3 classes, the Metacello class is a stub class for the Metacello Scripting Api[2], that will bootstrap Metacello _if_ you use the scripting api. I am promoting the scripting api as the cross platform way to use Metacello. Instead of seeing install scripts fly by that look like this:

Gofer new package: 'ConfigurationOfSeaside30'; squeaksource: 'MetacelloRepository'; load. ((Smalltalk at: #ConfigurationOfSeaside30) version: #stable) load.

(with the caveat that Installer should be used for Squeak)...the install scripts will look like this:

Metacello new configuration: 'Seaside30'; version: #stable; squeaksource: 'MetacelloRepository'; load.

The same script will "do the right thing" on all three supported platforms (Squeak, Pharo, GemStone), presuming that the class Metacello is present in the base images:)

I've recently released a preview version[4] and am actively collecting feedback. I would think that I can finalize the visible portion of the scripting api by September...

I've been ensuring that the metacello preview release is functional on Squeak all along and I am currently running CI builds against Squeak4.3 using travis ci[5]. Moving forward I would add Squeak4.4 to the list of platforms for CI and publish updates to the Metacello-Base pacakge as needed.

The class comments for each of the classes provides basic documentation for using Metacello so I think that Metacello-Base is a clean and small presence for Metacello and will enable the use of "cross-platform install scripts for Metacello."

Thanks,

Dale

[1] http://seaside.gemstone.com/ss/metacello/Metacello-Base-dkh.96.mcz [2] https://github.com/dalehenrich/metacello-work/blob/master/docs/MetacelloUser... [4] http://forum.world.st/Metacello-Preview-release-1-0-beta-32-2-tt4642064.html [5] http://travis-ci.org/#!/dalehenrich/metacello-work/builds/2012331

Frank,

I haven't tried the Metacello preview in Squeak4.4 yet, so I can't be sure that it works there. I've been testing against Squeak4.3...

Give me a couple of days to hook up Squeak4.4 in my travis ci framework and shake out any Metacello bugs related to Squeak4.4 and then you can take it for a spin.

Dale

----- Original Message ----- | From: "Frank Shearar" frank.shearar@gmail.com | To: "ma chris m" ma.chris.m@gmail.com, "The general-purpose Squeak developers list" | squeak-dev@lists.squeakfoundation.org | Sent: Thursday, August 2, 2012 2:05:15 PM | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | | On 2 August 2012 21:49, Chris Muller asqueaker@gmail.com wrote: | > If there's little or no danger of any serious regressions elsewhere | > in | > the system, then I'm +1. Like Frank, I think it's important for | > Squeak to continue to put out stable releases. | | I need to hack on the CI server first, but if a kind soul wouldn't | mind just loading Dale's mcz and running the full suite of tests, I'd | be grateful. I don't see how that could cause regressions, but I'm | paranoid. | | > Dale, your efforts on cross-platform are amazing, thanks! | | Seconded! | | frank | | > On Thu, Aug 2, 2012 at 3:33 PM, Dale Henrichs dhenrich@vmware.com | > wrote: | >> I have a small 3 class package (Metacello-Base[1]) that I would | >> hope could be included in the Squeak4.4 release. | >> | >> Of the 3 classes, the Metacello class is a stub class for the | >> Metacello Scripting Api[2], that will bootstrap Metacello _if_ | >> you use the scripting api. I am promoting the scripting api as | >> the cross platform way to use Metacello. Instead of seeing | >> install scripts fly by that look like this: | >> | >> Gofer new | >> package: 'ConfigurationOfSeaside30'; | >> squeaksource: 'MetacelloRepository'; | >> load. | >> ((Smalltalk at: #ConfigurationOfSeaside30) version: #stable) | >> load. | >> | >> (with the caveat that Installer should be used for Squeak)...the | >> install scripts will look like this: | >> | >> Metacello new | >> configuration: 'Seaside30'; | >> version: #stable; | >> squeaksource: 'MetacelloRepository'; | >> load. | >> | >> The same script will "do the right thing" on all three supported | >> platforms (Squeak, Pharo, GemStone), presuming that the class | >> Metacello is present in the base images:) | >> | >> I've recently released a preview version[4] and am actively | >> collecting feedback. I would think that I can finalize the | >> visible portion of the scripting api by September... | >> | >> I've been ensuring that the metacello preview release is | >> functional on Squeak all along and I am currently running CI | >> builds against Squeak4.3 using travis ci[5]. Moving forward I | >> would add Squeak4.4 to the list of platforms for CI and publish | >> updates to the Metacello-Base pacakge as needed. | >> | >> The class comments for each of the classes provides basic | >> documentation for using Metacello so I think that Metacello-Base | >> is a clean and small presence for Metacello and will enable the | >> use of "cross-platform install scripts for Metacello." | >> | >> Thanks, | >> | >> Dale | >> | >> [1] | >> http://seaside.gemstone.com/ss/metacello/Metacello-Base-dkh.96.mcz | >> [2] | >> https://github.com/dalehenrich/metacello-work/blob/master/docs/MetacelloUser... | >> [4] | >> http://forum.world.st/Metacello-Preview-release-1-0-beta-32-2-tt4642064.html | >> [5] | >> http://travis-ci.org/#!/dalehenrich/metacello-work/builds/2012331 | >> | > | |

On 3 August 2012 17:21, Dale Henrichs dhenrich@vmware.com wrote:

Frank,

Since you're working on the CI stuff, where should I go to get a recent copy of Squeak4.4 .. I have the impression that the release available for ftp[1] is not the one I should use?

I noticed that there are no download artifacts on the CI build site, so perhaps the trunk is the right place after all?

I haven't done the bundle-and-publish part of the CI yet, but https://github.com/frankshearar/squeak-ci has a close-to-up-to-date image you can use in the meantime.

frank

Dale

[1] http://ftp.squeak.org/trunk/ ----- Original Message ----- | From: "Frank Shearar" frank.shearar@gmail.com | To: "ma chris m" ma.chris.m@gmail.com, "The general-purpose Squeak developers list" | squeak-dev@lists.squeakfoundation.org | Sent: Thursday, August 2, 2012 2:05:15 PM | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | | On 2 August 2012 21:49, Chris Muller asqueaker@gmail.com wrote: | > If there's little or no danger of any serious regressions elsewhere | > in | > the system, then I'm +1. Like Frank, I think it's important for | > Squeak to continue to put out stable releases. | | I need to hack on the CI server first, but if a kind soul wouldn't | mind just loading Dale's mcz and running the full suite of tests, I'd | be grateful. I don't see how that could cause regressions, but I'm | paranoid. | | > Dale, your efforts on cross-platform are amazing, thanks! | | Seconded! | | frank | | > On Thu, Aug 2, 2012 at 3:33 PM, Dale Henrichs dhenrich@vmware.com | > wrote: | >> I have a small 3 class package (Metacello-Base[1]) that I would | >> hope could be included in the Squeak4.4 release. | >> | >> Of the 3 classes, the Metacello class is a stub class for the | >> Metacello Scripting Api[2], that will bootstrap Metacello _if_ | >> you use the scripting api. I am promoting the scripting api as | >> the cross platform way to use Metacello. Instead of seeing | >> install scripts fly by that look like this: | >> | >> Gofer new | >> package: 'ConfigurationOfSeaside30'; | >> squeaksource: 'MetacelloRepository'; | >> load. | >> ((Smalltalk at: #ConfigurationOfSeaside30) version: #stable) | >> load. | >> | >> (with the caveat that Installer should be used for Squeak)...the | >> install scripts will look like this: | >> | >> Metacello new | >> configuration: 'Seaside30'; | >> version: #stable; | >> squeaksource: 'MetacelloRepository'; | >> load. | >> | >> The same script will "do the right thing" on all three supported | >> platforms (Squeak, Pharo, GemStone), presuming that the class | >> Metacello is present in the base images:) | >> | >> I've recently released a preview version[4] and am actively | >> collecting feedback. I would think that I can finalize the | >> visible portion of the scripting api by September... | >> | >> I've been ensuring that the metacello preview release is | >> functional on Squeak all along and I am currently running CI | >> builds against Squeak4.3 using travis ci[5]. Moving forward I | >> would add Squeak4.4 to the list of platforms for CI and publish | >> updates to the Metacello-Base pacakge as needed. | >> | >> The class comments for each of the classes provides basic | >> documentation for using Metacello so I think that Metacello-Base | >> is a clean and small presence for Metacello and will enable the | >> use of "cross-platform install scripts for Metacello." | >> | >> Thanks, | >> | >> Dale | >> | >> [1] | >> http://seaside.gemstone.com/ss/metacello/Metacello-Base-dkh.96.mcz | >> [2] | >> https://github.com/dalehenrich/metacello-work/blob/master/docs/MetacelloUser... | >> [4] | >> http://forum.world.st/Metacello-Preview-release-1-0-beta-32-2-tt4642064.html | >> [5] | >> http://travis-ci.org/#!/dalehenrich/metacello-work/builds/2012331 | >> | > | |

Frank,

The Metacello tests add/remove global variables to/from Smalltalk and it looks like Environment does not implement #removeKey:ifAbsent:, so I've got some 400 test failures because I can't remove a global from Smalltalk ...

I can hack this for the time being but I'm curious what the longer term plan will be with regards to removing global variables from Smalltalk or perhaps leaving #removeKey:ifAbsent: out is inadvertent ...

BTW, I'm using update 12154 ...

Thanks,

Dale ----- Original Message ----- | From: "Frank Shearar" frank.shearar@gmail.com | To: "The general-purpose Squeak developers list" squeak-dev@lists.squeakfoundation.org | Sent: Friday, August 3, 2012 9:24:12 AM | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | | On 3 August 2012 17:21, Dale Henrichs dhenrich@vmware.com wrote: | > Frank, | > | > Since you're working on the CI stuff, where should I go to get a | > recent copy of Squeak4.4 .. I have the impression that the release | > available for ftp[1] is not the one I should use? | > | > I noticed that there are no download artifacts on the CI build | > site, so perhaps the trunk is the right place after all? | | I haven't done the bundle-and-publish part of the CI yet, but | https://github.com/frankshearar/squeak-ci has a close-to-up-to-date | image you can use in the meantime. | | frank | | > Dale | > | > [1] http://ftp.squeak.org/trunk/ | > ----- Original Message ----- | > | From: "Frank Shearar" frank.shearar@gmail.com | > | To: "ma chris m" ma.chris.m@gmail.com, "The general-purpose | > | Squeak developers list" | > | squeak-dev@lists.squeakfoundation.org | > | Sent: Thursday, August 2, 2012 2:05:15 PM | > | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | > | | > | On 2 August 2012 21:49, Chris Muller asqueaker@gmail.com wrote: | > | > If there's little or no danger of any serious regressions | > | > elsewhere | > | > in | > | > the system, then I'm +1. Like Frank, I think it's important | > | > for | > | > Squeak to continue to put out stable releases. | > | | > | I need to hack on the CI server first, but if a kind soul | > | wouldn't | > | mind just loading Dale's mcz and running the full suite of tests, | > | I'd | > | be grateful. I don't see how that could cause regressions, but | > | I'm | > | paranoid. | > | | > | > Dale, your efforts on cross-platform are amazing, thanks! | > | | > | Seconded! | > | | > | frank | > | | > | > On Thu, Aug 2, 2012 at 3:33 PM, Dale Henrichs | > | > dhenrich@vmware.com | > | > wrote: | > | >> I have a small 3 class package (Metacello-Base[1]) that I | > | >> would | > | >> hope could be included in the Squeak4.4 release. | > | >> | > | >> Of the 3 classes, the Metacello class is a stub class for the | > | >> Metacello Scripting Api[2], that will bootstrap Metacello _if_ | > | >> you use the scripting api. I am promoting the scripting api as | > | >> the cross platform way to use Metacello. Instead of seeing | > | >> install scripts fly by that look like this: | > | >> | > | >> Gofer new | > | >> package: 'ConfigurationOfSeaside30'; | > | >> squeaksource: 'MetacelloRepository'; | > | >> load. | > | >> ((Smalltalk at: #ConfigurationOfSeaside30) version: #stable) | > | >> load. | > | >> | > | >> (with the caveat that Installer should be used for | > | >> Squeak)...the | > | >> install scripts will look like this: | > | >> | > | >> Metacello new | > | >> configuration: 'Seaside30'; | > | >> version: #stable; | > | >> squeaksource: 'MetacelloRepository'; | > | >> load. | > | >> | > | >> The same script will "do the right thing" on all three | > | >> supported | > | >> platforms (Squeak, Pharo, GemStone), presuming that the class | > | >> Metacello is present in the base images:) | > | >> | > | >> I've recently released a preview version[4] and am actively | > | >> collecting feedback. I would think that I can finalize the | > | >> visible portion of the scripting api by September... | > | >> | > | >> I've been ensuring that the metacello preview release is | > | >> functional on Squeak all along and I am currently running CI | > | >> builds against Squeak4.3 using travis ci[5]. Moving forward I | > | >> would add Squeak4.4 to the list of platforms for CI and | > | >> publish | > | >> updates to the Metacello-Base pacakge as needed. | > | >> | > | >> The class comments for each of the classes provides basic | > | >> documentation for using Metacello so I think that | > | >> Metacello-Base | > | >> is a clean and small presence for Metacello and will enable | > | >> the | > | >> use of "cross-platform install scripts for Metacello." | > | >> | > | >> Thanks, | > | >> | > | >> Dale | > | >> | > | >> [1] | > | >> http://seaside.gemstone.com/ss/metacello/Metacello-Base-dkh.96.mcz | > | >> [2] | > | >> https://github.com/dalehenrich/metacello-work/blob/master/docs/MetacelloUser... | > | >> [4] | > | >> http://forum.world.st/Metacello-Preview-release-1-0-beta-32-2-tt4642064.html | > | >> [5] | > | >> http://travis-ci.org/#!/dalehenrich/metacello-work/builds/2012331 | > | >> | > | > | > | | > | | > | |

On 3 August 2012 21:17, Colin Putney colin@wiresong.com wrote:

On Fri, Aug 3, 2012 at 1:00 PM, Dale Henrichs dhenrich@vmware.com wrote:

...

Frank,

The Metacello tests add/remove global variables to/from Smalltalk and it looks like Environment does not implement #removeKey:ifAbsent:, so I've got some 400 test failures because I can't remove a global from Smalltalk ...

I can hack this for the time being but I'm curious what the longer term plan will be with regards to removing global variables from Smalltalk or perhaps leaving #removeKey:ifAbsent: out is inadvertent ...

BTW, I'm using update 12154 ...

First, Environments is being pushed to Squeak 4.5, so it should be removed from 4.4 shortly.

During the development period for 4.5, Environments will implement #removeKey:ifAbsent:, but eventually that will be deprecated in favour of #undeclare: or something similar.

For the nonce, how about we implement #removeKey:ifAbsent: (see Inbox), so we can at least unblock Dale?

We do need to remove Environments, but Dale's work contributes to ensuring/checking 4.4's stability.

frank

Colin

Sounds fair to me ... I'll patch my current in place and kick off the tests again ...

Thanks guys,

Dale

----- Original Message ----- | From: "Frank Shearar" frank.shearar@gmail.com | To: "The general-purpose Squeak developers list" squeak-dev@lists.squeakfoundation.org | Sent: Friday, August 3, 2012 1:20:57 PM | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | | On 3 August 2012 21:17, Colin Putney colin@wiresong.com wrote: | > | > | > On Fri, Aug 3, 2012 at 1:00 PM, Dale Henrichs dhenrich@vmware.com | > wrote: | >> | >> Frank, | >> | >> The Metacello tests add/remove global variables to/from Smalltalk | >> and it | >> looks like Environment does not implement #removeKey:ifAbsent:, so | >> I've got | >> some 400 test failures because I can't remove a global from | >> Smalltalk ... | >> | >> I can hack this for the time being but I'm curious what the longer | >> term | >> plan will be with regards to removing global variables from | >> Smalltalk or | >> perhaps leaving #removeKey:ifAbsent: out is inadvertent ... | >> | >> BTW, I'm using update 12154 ... | > | > | > First, Environments is being pushed to Squeak 4.5, so it should be | > removed | > from 4.4 shortly. | > | > During the development period for 4.5, Environments will implement | > #removeKey:ifAbsent:, but eventually that will be deprecated in | > favour of | > #undeclare: or something similar. | | For the nonce, how about we implement #removeKey:ifAbsent: (see | Inbox), so we can at least unblock Dale? | | We do need to remove Environments, but Dale's work contributes to | ensuring/checking 4.4's stability. | | frank | | > Colin | > | > | > | > | |

Okay,

Using Frank's 4.4 trunk (update 12154) you can load Metacello-Base-dkh.98 from http://seaside.gemstone.com/ss/metacello and then use the Metacello Preview scripting API[1].

As reported earlier, there are still some test failures in MetacelloPreview 1.0-beta.32.2, but the fixes are to the tests and not the functional code base, so I'd say you can include Metacello-Base-dkh.98 in 4.4 at your leisure.

To load the basic Metacello functionality, you can use the Scripting API to load another configuration or execute the following:

Metacello new configuration: 'MetacelloPreview'; version: #stable; gemsource: 'metacello'; load: 'batch'.

\to load the tests, exedute:

Metacello new configuration: 'MetacelloPreview'; version: #stable; gemsource: 'metacello'; load: 'TravisCI'.

I'll release 1.0-beta.32.3 in a week or so, but I'll just update the #stable definition, so the above expressions should be used to guarantee that you're picking up the latest release...

Dale

[1] https://github.com/dalehenrich/metacello-work/blob/master/docs/MetacelloScri...

----- Original Message ----- | From: "Dale Henrichs" dhenrich@vmware.com | To: "The general-purpose Squeak developers list" squeak-dev@lists.squeakfoundation.org | Sent: Friday, August 3, 2012 4:19:58 PM | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | | After the patch, a reasonable number of test issues: | | 695 run, 691 passes, 0 expected failures, 2 failures, 2 errors, 0 | unexpected passes | | Dale | | ----- Original Message ----- | | From: "Dale Henrichs" dhenrich@vmware.com | | To: "The general-purpose Squeak developers list" | | squeak-dev@lists.squeakfoundation.org | | Sent: Friday, August 3, 2012 1:25:57 PM | | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | | | | Sounds fair to me ... I'll patch my current in place and kick off | | the | | tests again ... | | | | Thanks guys, | | | | Dale | | | | ----- Original Message ----- | | | From: "Frank Shearar" frank.shearar@gmail.com | | | To: "The general-purpose Squeak developers list" | | | squeak-dev@lists.squeakfoundation.org | | | Sent: Friday, August 3, 2012 1:20:57 PM | | | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | | | | | | On 3 August 2012 21:17, Colin Putney colin@wiresong.com wrote: | | | > | | | > | | | > On Fri, Aug 3, 2012 at 1:00 PM, Dale Henrichs | | | > dhenrich@vmware.com | | | > wrote: | | | >> | | | >> Frank, | | | >> | | | >> The Metacello tests add/remove global variables to/from | | | >> Smalltalk | | | >> and it | | | >> looks like Environment does not implement | | | >> #removeKey:ifAbsent:, | | | >> so | | | >> I've got | | | >> some 400 test failures because I can't remove a global from | | | >> Smalltalk ... | | | >> | | | >> I can hack this for the time being but I'm curious what the | | | >> longer | | | >> term | | | >> plan will be with regards to removing global variables from | | | >> Smalltalk or | | | >> perhaps leaving #removeKey:ifAbsent: out is inadvertent ... | | | >> | | | >> BTW, I'm using update 12154 ... | | | > | | | > | | | > First, Environments is being pushed to Squeak 4.5, so it should | | | > be | | | > removed | | | > from 4.4 shortly. | | | > | | | > During the development period for 4.5, Environments will | | | > implement | | | > #removeKey:ifAbsent:, but eventually that will be deprecated in | | | > favour of | | | > #undeclare: or something similar. | | | | | | For the nonce, how about we implement #removeKey:ifAbsent: (see | | | Inbox), so we can at least unblock Dale? | | | | | | We do need to remove Environments, but Dale's work contributes to | | | ensuring/checking 4.4's stability. | | | | | | frank | | | | | | > Colin | | | > | | | > | | | > | | | > | | | | | | | | | | | |

Frank,

----- Original Message ----- | From: "Frank Shearar" frank.shearar@gmail.com | To: "The general-purpose Squeak developers list" squeak-dev@lists.squeakfoundation.org | Sent: Sunday, August 5, 2012 11:07:20 AM | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | | On 4 August 2012 20:00, Dale Henrichs dhenrich@vmware.com wrote: | > Okay, | > | > Using Frank's 4.4 trunk (update 12154) you can load | > Metacello-Base-dkh.98 from | > http://seaside.gemstone.com/ss/metacello and then use the | > Metacello Preview scripting API[1]. | | Hi Dale, | | Now that the infrastructure seems to be settling down, it looks like | a | good source for an up-to-date trunk image is | http://squeakci.org/job/SqueakTrunk/lastSuccessfulBuild/artifact/target/Trun....

Haha, I noticed that and have already used that as the image source for the builderCI work.

| | > As reported earlier, there are still some test failures in | > MetacelloPreview 1.0-beta.32.2, but the fixes are to the tests and | > not the functional code base, so I'd say you can include | > Metacello-Base-dkh.98 in 4.4 at your leisure. | > | > To load the basic Metacello functionality, you can use the | > Scripting API to load another configuration or execute the | > following: | > | > Metacello new | > configuration: 'MetacelloPreview'; | > version: #stable; | > gemsource: 'metacello'; | > load: 'batch'. | > | > \to load the tests, exedute: | > | > Metacello new | > configuration: 'MetacelloPreview'; | > version: #stable; | > gemsource: 'metacello'; | > load: 'TravisCI'. | | Just to get things completely clear in my head, these are | instructions | for Metacello users post-inclusion, right? They're not actions that | we | need to follow other than pushing Metacello-Base into the trunk, | right?

Correct. If you run the above, then Metacello _will_ be booted into the image and that's not what you want to do ... except as part of a test to validate that Metacello boots correctly.

| | frank | | > I'll release 1.0-beta.32.3 in a week or so, but I'll just update | > the #stable definition, so the above expressions should be used to | > guarantee that you're picking up the latest release... | > | > Dale | > | > [1] | > https://github.com/dalehenrich/metacello-work/blob/master/docs/MetacelloScri... | > | > ----- Original Message ----- | > | From: "Dale Henrichs" dhenrich@vmware.com | > | To: "The general-purpose Squeak developers list" | > | squeak-dev@lists.squeakfoundation.org | > | Sent: Friday, August 3, 2012 4:19:58 PM | > | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | > | | > | After the patch, a reasonable number of test issues: | > | | > | 695 run, 691 passes, 0 expected failures, 2 failures, 2 errors, | > | 0 | > | unexpected passes | > | | > | Dale | > | | > | ----- Original Message ----- | > | | From: "Dale Henrichs" dhenrich@vmware.com | > | | To: "The general-purpose Squeak developers list" | > | | squeak-dev@lists.squeakfoundation.org | > | | Sent: Friday, August 3, 2012 1:25:57 PM | > | | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | > | | | > | | Sounds fair to me ... I'll patch my current in place and kick | > | | off | > | | the | > | | tests again ... | > | | | > | | Thanks guys, | > | | | > | | Dale | > | | | > | | ----- Original Message ----- | > | | | From: "Frank Shearar" frank.shearar@gmail.com | > | | | To: "The general-purpose Squeak developers list" | > | | | squeak-dev@lists.squeakfoundation.org | > | | | Sent: Friday, August 3, 2012 1:20:57 PM | > | | | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | > | | | | > | | | On 3 August 2012 21:17, Colin Putney colin@wiresong.com | > | | | wrote: | > | | | > | > | | | > | > | | | > On Fri, Aug 3, 2012 at 1:00 PM, Dale Henrichs | > | | | > dhenrich@vmware.com | > | | | > wrote: | > | | | >> | > | | | >> Frank, | > | | | >> | > | | | >> The Metacello tests add/remove global variables to/from | > | | | >> Smalltalk | > | | | >> and it | > | | | >> looks like Environment does not implement | > | | | >> #removeKey:ifAbsent:, | > | | | >> so | > | | | >> I've got | > | | | >> some 400 test failures because I can't remove a global | > | | | >> from | > | | | >> Smalltalk ... | > | | | >> | > | | | >> I can hack this for the time being but I'm curious what | > | | | >> the | > | | | >> longer | > | | | >> term | > | | | >> plan will be with regards to removing global variables | > | | | >> from | > | | | >> Smalltalk or | > | | | >> perhaps leaving #removeKey:ifAbsent: out is inadvertent | > | | | >> ... | > | | | >> | > | | | >> BTW, I'm using update 12154 ... | > | | | > | > | | | > | > | | | > First, Environments is being pushed to Squeak 4.5, so it | > | | | > should | > | | | > be | > | | | > removed | > | | | > from 4.4 shortly. | > | | | > | > | | | > During the development period for 4.5, Environments will | > | | | > implement | > | | | > #removeKey:ifAbsent:, but eventually that will be | > | | | > deprecated in | > | | | > favour of | > | | | > #undeclare: or something similar. | > | | | | > | | | For the nonce, how about we implement #removeKey:ifAbsent: | > | | | (see | > | | | Inbox), so we can at least unblock Dale? | > | | | | > | | | We do need to remove Environments, but Dale's work | > | | | contributes to | > | | | ensuring/checking 4.4's stability. | > | | | | > | | | frank | > | | | | > | | | > Colin | > | | | > | > | | | > | > | | | > | > | | | > | > | | | | > | | | | > | | | > | | | > | | > | | > | |

On 6 August 2012 05:43, Dale Henrichs dhenrich@vmware.com wrote:

Frank,

FYI: It appears that the Squeak4.4 available from Jenkins does not have the patched Environment loaded ... so travis ci builds fail...

Let me know when the patch is integrated and I'll turn the travis ci builds back on ...

Hi Dale,

The necessary change - Environments-fbs.5 - is sitting in the Inbox waiting for review. If it helps, you could always cherry-pick that?

frank

Dale

----- Original Message ----- | From: "Frank Shearar" frank.shearar@gmail.com | To: "The general-purpose Squeak developers list" squeak-dev@lists.squeakfoundation.org | Sent: Sunday, August 5, 2012 11:07:20 AM | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | | On 4 August 2012 20:00, Dale Henrichs dhenrich@vmware.com wrote: | > Okay, | > | > Using Frank's 4.4 trunk (update 12154) you can load | > Metacello-Base-dkh.98 from | > http://seaside.gemstone.com/ss/metacello and then use the | > Metacello Preview scripting API[1]. | | Hi Dale, | | Now that the infrastructure seems to be settling down, it looks like | a | good source for an up-to-date trunk image is | http://squeakci.org/job/SqueakTrunk/lastSuccessfulBuild/artifact/target/Trun.... | | > As reported earlier, there are still some test failures in | > MetacelloPreview 1.0-beta.32.2, but the fixes are to the tests and | > not the functional code base, so I'd say you can include | > Metacello-Base-dkh.98 in 4.4 at your leisure. | > | > To load the basic Metacello functionality, you can use the | > Scripting API to load another configuration or execute the | > following: | > | > Metacello new | > configuration: 'MetacelloPreview'; | > version: #stable; | > gemsource: 'metacello'; | > load: 'batch'. | > | > \to load the tests, exedute: | > | > Metacello new | > configuration: 'MetacelloPreview'; | > version: #stable; | > gemsource: 'metacello'; | > load: 'TravisCI'. | | Just to get things completely clear in my head, these are | instructions | for Metacello users post-inclusion, right? They're not actions that | we | need to follow other than pushing Metacello-Base into the trunk, | right? | | frank | | > I'll release 1.0-beta.32.3 in a week or so, but I'll just update | > the #stable definition, so the above expressions should be used to | > guarantee that you're picking up the latest release... | > | > Dale | > | > [1] | > https://github.com/dalehenrich/metacello-work/blob/master/docs/MetacelloScri... | > | > ----- Original Message ----- | > | From: "Dale Henrichs" dhenrich@vmware.com | > | To: "The general-purpose Squeak developers list" | > | squeak-dev@lists.squeakfoundation.org | > | Sent: Friday, August 3, 2012 4:19:58 PM | > | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | > | | > | After the patch, a reasonable number of test issues: | > | | > | 695 run, 691 passes, 0 expected failures, 2 failures, 2 errors, | > | 0 | > | unexpected passes | > | | > | Dale | > | | > | ----- Original Message ----- | > | | From: "Dale Henrichs" dhenrich@vmware.com | > | | To: "The general-purpose Squeak developers list" | > | | squeak-dev@lists.squeakfoundation.org | > | | Sent: Friday, August 3, 2012 1:25:57 PM | > | | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | > | | | > | | Sounds fair to me ... I'll patch my current in place and kick | > | | off | > | | the | > | | tests again ... | > | | | > | | Thanks guys, | > | | | > | | Dale | > | | | > | | ----- Original Message ----- | > | | | From: "Frank Shearar" frank.shearar@gmail.com | > | | | To: "The general-purpose Squeak developers list" | > | | | squeak-dev@lists.squeakfoundation.org | > | | | Sent: Friday, August 3, 2012 1:20:57 PM | > | | | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | > | | | | > | | | On 3 August 2012 21:17, Colin Putney colin@wiresong.com | > | | | wrote: | > | | | > | > | | | > | > | | | > On Fri, Aug 3, 2012 at 1:00 PM, Dale Henrichs | > | | | > dhenrich@vmware.com | > | | | > wrote: | > | | | >> | > | | | >> Frank, | > | | | >> | > | | | >> The Metacello tests add/remove global variables to/from | > | | | >> Smalltalk | > | | | >> and it | > | | | >> looks like Environment does not implement | > | | | >> #removeKey:ifAbsent:, | > | | | >> so | > | | | >> I've got | > | | | >> some 400 test failures because I can't remove a global | > | | | >> from | > | | | >> Smalltalk ... | > | | | >> | > | | | >> I can hack this for the time being but I'm curious what | > | | | >> the | > | | | >> longer | > | | | >> term | > | | | >> plan will be with regards to removing global variables | > | | | >> from | > | | | >> Smalltalk or | > | | | >> perhaps leaving #removeKey:ifAbsent: out is inadvertent | > | | | >> ... | > | | | >> | > | | | >> BTW, I'm using update 12154 ... | > | | | > | > | | | > | > | | | > First, Environments is being pushed to Squeak 4.5, so it | > | | | > should | > | | | > be | > | | | > removed | > | | | > from 4.4 shortly. | > | | | > | > | | | > During the development period for 4.5, Environments will | > | | | > implement | > | | | > #removeKey:ifAbsent:, but eventually that will be | > | | | > deprecated in | > | | | > favour of | > | | | > #undeclare: or something similar. | > | | | | > | | | For the nonce, how about we implement #removeKey:ifAbsent: | > | | | (see | > | | | Inbox), so we can at least unblock Dale? | > | | | | > | | | We do need to remove Environments, but Dale's work | > | | | contributes to | > | | | ensuring/checking 4.4's stability. | > | | | | > | | | frank | > | | | | > | | | > Colin | > | | | > | > | | | > | > | | | > | > | | | > | > | | | | > | | | | > | | | > | | | > | | > | | > | |

On 6 August 2012 16:17, Dale Henrichs dhenrich@vmware.com wrote:

Frank,

For the automated builds, I want to use the latestAvailable download as my test bed, because that's the scenario I want to test:

• download image
• execute Metacello

I have ensured that with the patch Metacello runs against Squeak4.4 so I can wait until the patch gets processed and then turn on my automated builds....

I've kind've lost track of what more we need to do, Dale. Shall we shake out the builderCI failure first before dropping Metacello-Base into 4.4, or do you think they're unrelated issues?

frank

Dale

----- Original Message ----- | From: "Frank Shearar" frank.shearar@gmail.com | To: "The general-purpose Squeak developers list" squeak-dev@lists.squeakfoundation.org | Sent: Monday, August 6, 2012 12:29:25 AM | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | | On 6 August 2012 05:43, Dale Henrichs dhenrich@vmware.com wrote: | > Frank, | > | > FYI: It appears that the Squeak4.4 available from Jenkins does not | > have the patched Environment loaded ... so travis ci builds | > fail... | > | > Let me know when the patch is integrated and I'll turn the travis | > ci builds back on ... | | Hi Dale, | | The necessary change - Environments-fbs.5 - is sitting in the Inbox | waiting for review. If it helps, you could always cherry-pick that? | | frank | | > Dale | > | > ----- Original Message ----- | > | From: "Frank Shearar" frank.shearar@gmail.com | > | To: "The general-purpose Squeak developers list" | > | squeak-dev@lists.squeakfoundation.org | > | Sent: Sunday, August 5, 2012 11:07:20 AM | > | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | > | | > | On 4 August 2012 20:00, Dale Henrichs dhenrich@vmware.com | > | wrote: | > | > Okay, | > | > | > | > Using Frank's 4.4 trunk (update 12154) you can load | > | > Metacello-Base-dkh.98 from | > | > http://seaside.gemstone.com/ss/metacello and then use the | > | > Metacello Preview scripting API[1]. | > | | > | Hi Dale, | > | | > | Now that the infrastructure seems to be settling down, it looks | > | like | > | a | > | good source for an up-to-date trunk image is | > | http://squeakci.org/job/SqueakTrunk/lastSuccessfulBuild/artifact/target/Trun.... | > | | > | > As reported earlier, there are still some test failures in | > | > MetacelloPreview 1.0-beta.32.2, but the fixes are to the tests | > | > and | > | > not the functional code base, so I'd say you can include | > | > Metacello-Base-dkh.98 in 4.4 at your leisure. | > | > | > | > To load the basic Metacello functionality, you can use the | > | > Scripting API to load another configuration or execute the | > | > following: | > | > | > | > Metacello new | > | > configuration: 'MetacelloPreview'; | > | > version: #stable; | > | > gemsource: 'metacello'; | > | > load: 'batch'. | > | > | > | > \to load the tests, exedute: | > | > | > | > Metacello new | > | > configuration: 'MetacelloPreview'; | > | > version: #stable; | > | > gemsource: 'metacello'; | > | > load: 'TravisCI'. | > | | > | Just to get things completely clear in my head, these are | > | instructions | > | for Metacello users post-inclusion, right? They're not actions | > | that | > | we | > | need to follow other than pushing Metacello-Base into the trunk, | > | right? | > | | > | frank | > | | > | > I'll release 1.0-beta.32.3 in a week or so, but I'll just | > | > update | > | > the #stable definition, so the above expressions should be used | > | > to | > | > guarantee that you're picking up the latest release... | > | > | > | > Dale | > | > | > | > [1] | > | > https://github.com/dalehenrich/metacello-work/blob/master/docs/MetacelloScri... | > | > | > | > ----- Original Message ----- | > | > | From: "Dale Henrichs" dhenrich@vmware.com | > | > | To: "The general-purpose Squeak developers list" | > | > | squeak-dev@lists.squeakfoundation.org | > | > | Sent: Friday, August 3, 2012 4:19:58 PM | > | > | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | > | > | | > | > | After the patch, a reasonable number of test issues: | > | > | | > | > | 695 run, 691 passes, 0 expected failures, 2 failures, 2 | > | > | errors, | > | > | 0 | > | > | unexpected passes | > | > | | > | > | Dale | > | > | | > | > | ----- Original Message ----- | > | > | | From: "Dale Henrichs" dhenrich@vmware.com | > | > | | To: "The general-purpose Squeak developers list" | > | > | | squeak-dev@lists.squeakfoundation.org | > | > | | Sent: Friday, August 3, 2012 1:25:57 PM | > | > | | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | > | > | | | > | > | | Sounds fair to me ... I'll patch my current in place and | > | > | | kick | > | > | | off | > | > | | the | > | > | | tests again ... | > | > | | | > | > | | Thanks guys, | > | > | | | > | > | | Dale | > | > | | | > | > | | ----- Original Message ----- | > | > | | | From: "Frank Shearar" frank.shearar@gmail.com | > | > | | | To: "The general-purpose Squeak developers list" | > | > | | | squeak-dev@lists.squeakfoundation.org | > | > | | | Sent: Friday, August 3, 2012 1:20:57 PM | > | > | | | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | > | > | | | | > | > | | | On 3 August 2012 21:17, Colin Putney colin@wiresong.com | > | > | | | wrote: | > | > | | | > | > | > | | | > | > | > | | | > On Fri, Aug 3, 2012 at 1:00 PM, Dale Henrichs | > | > | | | > dhenrich@vmware.com | > | > | | | > wrote: | > | > | | | >> | > | > | | | >> Frank, | > | > | | | >> | > | > | | | >> The Metacello tests add/remove global variables | > | > | | | >> to/from | > | > | | | >> Smalltalk | > | > | | | >> and it | > | > | | | >> looks like Environment does not implement | > | > | | | >> #removeKey:ifAbsent:, | > | > | | | >> so | > | > | | | >> I've got | > | > | | | >> some 400 test failures because I can't remove a global | > | > | | | >> from | > | > | | | >> Smalltalk ... | > | > | | | >> | > | > | | | >> I can hack this for the time being but I'm curious | > | > | | | >> what | > | > | | | >> the | > | > | | | >> longer | > | > | | | >> term | > | > | | | >> plan will be with regards to removing global variables | > | > | | | >> from | > | > | | | >> Smalltalk or | > | > | | | >> perhaps leaving #removeKey:ifAbsent: out is | > | > | | | >> inadvertent | > | > | | | >> ... | > | > | | | >> | > | > | | | >> BTW, I'm using update 12154 ... | > | > | | | > | > | > | | | > | > | > | | | > First, Environments is being pushed to Squeak 4.5, so | > | > | | | > it | > | > | | | > should | > | > | | | > be | > | > | | | > removed | > | > | | | > from 4.4 shortly. | > | > | | | > | > | > | | | > During the development period for 4.5, Environments | > | > | | | > will | > | > | | | > implement | > | > | | | > #removeKey:ifAbsent:, but eventually that will be | > | > | | | > deprecated in | > | > | | | > favour of | > | > | | | > #undeclare: or something similar. | > | > | | | | > | > | | | For the nonce, how about we implement | > | > | | | #removeKey:ifAbsent: | > | > | | | (see | > | > | | | Inbox), so we can at least unblock Dale? | > | > | | | | > | > | | | We do need to remove Environments, but Dale's work | > | > | | | contributes to | > | > | | | ensuring/checking 4.4's stability. | > | > | | | | > | > | | | frank | > | > | | | | > | > | | | > Colin | > | > | | | > | > | > | | | > | > | > | | | > | > | > | | | > | > | > | | | | > | > | | | | > | > | | | > | > | | | > | > | | > | > | | > | > | > | | > | | > | |

From my perspective I don't think that the Metacello Preview has seen enough external usage for me to feel comfortable releasing it right now ... I am waiting for feedback from real users of the scripting api ... there are a handful of people using the api, so that's a good sign.

I will feel better when someone reports a nasty bug or complains about one of the features then I will feel that it has gotten some real usage and is ready to go out the door:)

I also want to improve the bootstrap scheme which directly affects the Metacello-Base package...

With all of that said, I am constantly testing the bootstrap code in both pharo and squeak, so I'm not worried as much about bugs and stability as I am that I will decide to change the scripting api based on feedback.

Dale

----- Original Message ----- | From: "H. Hirzel" hannes.hirzel@gmail.com | To: "The general-purpose Squeak developers list" squeak-dev@lists.squeakfoundation.org | Sent: Thursday, September 6, 2012 3:55:20 AM | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | | On 8/23/12, Frank Shearar frank.shearar@gmail.com wrote: | > On 6 August 2012 16:17, Dale Henrichs dhenrich@vmware.com wrote: | >> Frank, | >> | >> For the automated builds, I want to use the latestAvailable | >> download as my | >> test bed, because that's the scenario I want to test: | >> | >> - download image | >> - execute Metacello | >> | >> I have ensured that with the patch Metacello runs against | >> Squeak4.4 so I | >> can wait until the patch gets processed and then turn on my | >> automated | >> builds.... | > | > I've kind've lost track of what more we need to do, Dale. Shall we | > shake out the builderCI failure first before dropping | > Metacello-Base | > into 4.4, or do you think they're unrelated issues? | > | > frank | | Any news about including Metacello for 4.4? | | --Hannes | | | >> | >> ----- Original Message ----- | >> | From: "Frank Shearar" frank.shearar@gmail.com | >> | To: "The general-purpose Squeak developers list" | >> squeak-dev@lists.squeakfoundation.org | >> | Sent: Monday, August 6, 2012 12:29:25 AM | >> | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | >> | | >> | On 6 August 2012 05:43, Dale Henrichs dhenrich@vmware.com | >> | wrote: | >> | > Frank, | >> | > | >> | > FYI: It appears that the Squeak4.4 available from Jenkins does | >> | > not | >> | > have the patched Environment loaded ... so travis ci builds | >> | > fail... | >> | > | >> | > Let me know when the patch is integrated and I'll turn the | >> | > travis | >> | > ci builds back on ... | >> | | >> | Hi Dale, | >> | | >> | The necessary change - Environments-fbs.5 - is sitting in the | >> | Inbox | >> | waiting for review. If it helps, you could always cherry-pick | >> | that? | >> | | >> | frank | >> | | >> | > Dale | >> | > | >> | > ----- Original Message ----- | >> | > | From: "Frank Shearar" frank.shearar@gmail.com | >> | > | To: "The general-purpose Squeak developers list" | >> | > | squeak-dev@lists.squeakfoundation.org | >> | > | Sent: Sunday, August 5, 2012 11:07:20 AM | >> | > | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | >> | > | | >> | > | On 4 August 2012 20:00, Dale Henrichs dhenrich@vmware.com | >> | > | wrote: | >> | > | > Okay, | >> | > | > | >> | > | > Using Frank's 4.4 trunk (update 12154) you can load | >> | > | > Metacello-Base-dkh.98 from | >> | > | > http://seaside.gemstone.com/ss/metacello and then use the | >> | > | > Metacello Preview scripting API[1]. | >> | > | | >> | > | Hi Dale, | >> | > | | >> | > | Now that the infrastructure seems to be settling down, it | >> | > | looks | >> | > | like | >> | > | a | >> | > | good source for an up-to-date trunk image is | >> | > | | >> http://squeakci.org/job/SqueakTrunk/lastSuccessfulBuild/artifact/target/Trun.... | >> | > | | >> | > | > As reported earlier, there are still some test failures in | >> | > | > MetacelloPreview 1.0-beta.32.2, but the fixes are to the | >> | > | > tests | >> | > | > and | >> | > | > not the functional code base, so I'd say you can include | >> | > | > Metacello-Base-dkh.98 in 4.4 at your leisure. | >> | > | > | >> | > | > To load the basic Metacello functionality, you can use the | >> | > | > Scripting API to load another configuration or execute the | >> | > | > following: | >> | > | > | >> | > | > Metacello new | >> | > | > configuration: 'MetacelloPreview'; | >> | > | > version: #stable; | >> | > | > gemsource: 'metacello'; | >> | > | > load: 'batch'. | >> | > | > | >> | > | > \to load the tests, exedute: | >> | > | > | >> | > | > Metacello new | >> | > | > configuration: 'MetacelloPreview'; | >> | > | > version: #stable; | >> | > | > gemsource: 'metacello'; | >> | > | > load: 'TravisCI'. | >> | > | | >> | > | Just to get things completely clear in my head, these are | >> | > | instructions | >> | > | for Metacello users post-inclusion, right? They're not | >> | > | actions | >> | > | that | >> | > | we | >> | > | need to follow other than pushing Metacello-Base into the | >> | > | trunk, | >> | > | right? | >> | > | | >> | > | frank | >> | > | | >> | > | > I'll release 1.0-beta.32.3 in a week or so, but I'll just | >> | > | > update | >> | > | > the #stable definition, so the above expressions should be | >> | > | > used | >> | > | > to | >> | > | > guarantee that you're picking up the latest release... | >> | > | > | >> | > | > Dale | >> | > | > | >> | > | > [1] | >> | > | > | >> https://github.com/dalehenrich/metacello-work/blob/master/docs/MetacelloScri... | >> | > | > | >> | > | > ----- Original Message ----- | >> | > | > | From: "Dale Henrichs" dhenrich@vmware.com | >> | > | > | To: "The general-purpose Squeak developers list" | >> | > | > | squeak-dev@lists.squeakfoundation.org | >> | > | > | Sent: Friday, August 3, 2012 4:19:58 PM | >> | > | > | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | >> | > | > | | >> | > | > | After the patch, a reasonable number of test issues: | >> | > | > | | >> | > | > | 695 run, 691 passes, 0 expected failures, 2 failures, | >> | > | > | 2 | >> | > | > | errors, | >> | > | > | 0 | >> | > | > | unexpected passes | >> | > | > | | >> | > | > | Dale | >> | > | > | | >> | > | > | ----- Original Message ----- | >> | > | > | | From: "Dale Henrichs" dhenrich@vmware.com | >> | > | > | | To: "The general-purpose Squeak developers list" | >> | > | > | | squeak-dev@lists.squeakfoundation.org | >> | > | > | | Sent: Friday, August 3, 2012 1:25:57 PM | >> | > | > | | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | >> | > | > | | | >> | > | > | | Sounds fair to me ... I'll patch my current in place | >> | > | > | | and | >> | > | > | | kick | >> | > | > | | off | >> | > | > | | the | >> | > | > | | tests again ... | >> | > | > | | | >> | > | > | | Thanks guys, | >> | > | > | | | >> | > | > | | Dale | >> | > | > | | | >> | > | > | | ----- Original Message ----- | >> | > | > | | | From: "Frank Shearar" frank.shearar@gmail.com | >> | > | > | | | To: "The general-purpose Squeak developers list" | >> | > | > | | | squeak-dev@lists.squeakfoundation.org | >> | > | > | | | Sent: Friday, August 3, 2012 1:20:57 PM | >> | > | > | | | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | >> | > | > | | | | >> | > | > | | | On 3 August 2012 21:17, Colin Putney | >> | > | > | | | colin@wiresong.com | >> | > | > | | | wrote: | >> | > | > | | | > | >> | > | > | | | > | >> | > | > | | | > On Fri, Aug 3, 2012 at 1:00 PM, Dale Henrichs | >> | > | > | | | > dhenrich@vmware.com | >> | > | > | | | > wrote: | >> | > | > | | | >> | >> | > | > | | | >> Frank, | >> | > | > | | | >> | >> | > | > | | | >> The Metacello tests add/remove global variables | >> | > | > | | | >> to/from | >> | > | > | | | >> Smalltalk | >> | > | > | | | >> and it | >> | > | > | | | >> looks like Environment does not implement | >> | > | > | | | >> #removeKey:ifAbsent:, | >> | > | > | | | >> so | >> | > | > | | | >> I've got | >> | > | > | | | >> some 400 test failures because I can't remove a | >> | > | > | | | >> global | >> | > | > | | | >> from | >> | > | > | | | >> Smalltalk ... | >> | > | > | | | >> | >> | > | > | | | >> I can hack this for the time being but I'm | >> | > | > | | | >> curious | >> | > | > | | | >> what | >> | > | > | | | >> the | >> | > | > | | | >> longer | >> | > | > | | | >> term | >> | > | > | | | >> plan will be with regards to removing global | >> | > | > | | | >> variables | >> | > | > | | | >> from | >> | > | > | | | >> Smalltalk or | >> | > | > | | | >> perhaps leaving #removeKey:ifAbsent: out is | >> | > | > | | | >> inadvertent | >> | > | > | | | >> ... | >> | > | > | | | >> | >> | > | > | | | >> BTW, I'm using update 12154 ... | >> | > | > | | | > | >> | > | > | | | > | >> | > | > | | | > First, Environments is being pushed to Squeak 4.5, | >> | > | > | | | > so | >> | > | > | | | > it | >> | > | > | | | > should | >> | > | > | | | > be | >> | > | > | | | > removed | >> | > | > | | | > from 4.4 shortly. | >> | > | > | | | > | >> | > | > | | | > During the development period for 4.5, | >> | > | > | | | > Environments | >> | > | > | | | > will | >> | > | > | | | > implement | >> | > | > | | | > #removeKey:ifAbsent:, but eventually that will be | >> | > | > | | | > deprecated in | >> | > | > | | | > favour of | >> | > | > | | | > #undeclare: or something similar. | >> | > | > | | | | >> | > | > | | | For the nonce, how about we implement | >> | > | > | | | #removeKey:ifAbsent: | >> | > | > | | | (see | >> | > | > | | | Inbox), so we can at least unblock Dale? | >> | > | > | | | | >> | > | > | | | We do need to remove Environments, but Dale's work | >> | > | > | | | contributes to | >> | > | > | | | ensuring/checking 4.4's stability. | >> | > | > | | | | >> | > | > | | | frank | >> | > | > | | | | >> | > | > | | | > Colin | >> | > | > | | | > | >> | > | > | | | > | >> | > | > | | | > | >> | > | > | | | > | >> | > | > | | | | >> | > | > | | | | >> | > | > | | | >> | > | > | | | >> | > | > | | >> | > | > | | >> | > | > | >> | > | | >> | > | | >> | > | >> | | >> | | >> | > | > | |

Frank,

I want to wait a bit more on the new Metacello ... I haven't received enough feedback to give me comfort that I've got the api right ... putting in early in 4.5 will presumably give me enough feedback and enough time to do something about iany changes that need to be made...

Dale

----- Original Message ----- | From: "Frank Shearar" frank.shearar@gmail.com | To: "The general-purpose Squeak developers list" squeak-dev@lists.squeakfoundation.org | Sent: Monday, October 29, 2012 6:20:58 AM | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | | On 7 September 2012 14:25, Dale Henrichs dhenrich@vmware.com wrote: | > >From my perspective I don't think that the Metacello Preview has | > >seen enough external usage for me to feel comfortable releasing | > >it right now ... I am waiting for feedback from real users of the | > >scripting api ... there are a handful of people using the api, so | > >that's a good sign. | > | > I will feel better when someone reports a nasty bug or complains | > about one of the features then I will feel that it has gotten some | > real usage and is ready to go out the door:) | > | > I also want to improve the bootstrap scheme which directly affects | > the Metacello-Base package... | > | > With all of that said, I am constantly testing the bootstrap code | > in both pharo and squeak, so I'm not worried as much about bugs | > and stability as I am that I will decide to change the scripting | > api based on feedback. | | Dale, is this still your position? You'd rather not put the preview | into the Squeak 4.4 release? It seems like it's not a critical thing | to do, given that you and I have github/travis builds building | against | Squeak 4.4 (and 4.3) right now, so maybe it's not worth adding it (at | the very last minute) to 4.4 Agreed? | | Then if you do change your mind or get feedback or whatever, we can | put a seed Metacello in 4.5 early in the release cycle. | | frank | | > Dale | > | > ----- Original Message ----- | > | From: "H. Hirzel" hannes.hirzel@gmail.com | > | To: "The general-purpose Squeak developers list" | > | squeak-dev@lists.squeakfoundation.org | > | Sent: Thursday, September 6, 2012 3:55:20 AM | > | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | > | | > | On 8/23/12, Frank Shearar frank.shearar@gmail.com wrote: | > | > On 6 August 2012 16:17, Dale Henrichs dhenrich@vmware.com | > | > wrote: | > | >> Frank, | > | >> | > | >> For the automated builds, I want to use the latestAvailable | > | >> download as my | > | >> test bed, because that's the scenario I want to test: | > | >> | > | >> - download image | > | >> - execute Metacello | > | >> | > | >> I have ensured that with the patch Metacello runs against | > | >> Squeak4.4 so I | > | >> can wait until the patch gets processed and then turn on my | > | >> automated | > | >> builds.... | > | > | > | > I've kind've lost track of what more we need to do, Dale. Shall | > | > we | > | > shake out the builderCI failure first before dropping | > | > Metacello-Base | > | > into 4.4, or do you think they're unrelated issues? | > | > | > | > frank | > | | > | Any news about including Metacello for 4.4? | > | | > | --Hannes | > | | > | | > | >> | > | >> ----- Original Message ----- | > | >> | From: "Frank Shearar" frank.shearar@gmail.com | > | >> | To: "The general-purpose Squeak developers list" | > | >> squeak-dev@lists.squeakfoundation.org | > | >> | Sent: Monday, August 6, 2012 12:29:25 AM | > | >> | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | > | >> | | > | >> | On 6 August 2012 05:43, Dale Henrichs dhenrich@vmware.com | > | >> | wrote: | > | >> | > Frank, | > | >> | > | > | >> | > FYI: It appears that the Squeak4.4 available from Jenkins | > | >> | > does | > | >> | > not | > | >> | > have the patched Environment loaded ... so travis ci | > | >> | > builds | > | >> | > fail... | > | >> | > | > | >> | > Let me know when the patch is integrated and I'll turn the | > | >> | > travis | > | >> | > ci builds back on ... | > | >> | | > | >> | Hi Dale, | > | >> | | > | >> | The necessary change - Environments-fbs.5 - is sitting in | > | >> | the | > | >> | Inbox | > | >> | waiting for review. If it helps, you could always | > | >> | cherry-pick | > | >> | that? | > | >> | | > | >> | frank | > | >> | | > | >> | > Dale | > | >> | > | > | >> | > ----- Original Message ----- | > | >> | > | From: "Frank Shearar" frank.shearar@gmail.com | > | >> | > | To: "The general-purpose Squeak developers list" | > | >> | > | squeak-dev@lists.squeakfoundation.org | > | >> | > | Sent: Sunday, August 5, 2012 11:07:20 AM | > | >> | > | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | > | >> | > | | > | >> | > | On 4 August 2012 20:00, Dale Henrichs | > | >> | > | dhenrich@vmware.com | > | >> | > | wrote: | > | >> | > | > Okay, | > | >> | > | > | > | >> | > | > Using Frank's 4.4 trunk (update 12154) you can load | > | >> | > | > Metacello-Base-dkh.98 from | > | >> | > | > http://seaside.gemstone.com/ss/metacello and then use | > | >> | > | > the | > | >> | > | > Metacello Preview scripting API[1]. | > | >> | > | | > | >> | > | Hi Dale, | > | >> | > | | > | >> | > | Now that the infrastructure seems to be settling down, | > | >> | > | it | > | >> | > | looks | > | >> | > | like | > | >> | > | a | > | >> | > | good source for an up-to-date trunk image is | > | >> | > | | > | >> http://squeakci.org/job/SqueakTrunk/lastSuccessfulBuild/artifact/target/Trun.... | > | >> | > | | > | >> | > | > As reported earlier, there are still some test | > | >> | > | > failures in | > | >> | > | > MetacelloPreview 1.0-beta.32.2, but the fixes are to | > | >> | > | > the | > | >> | > | > tests | > | >> | > | > and | > | >> | > | > not the functional code base, so I'd say you can | > | >> | > | > include | > | >> | > | > Metacello-Base-dkh.98 in 4.4 at your leisure. | > | >> | > | > | > | >> | > | > To load the basic Metacello functionality, you can use | > | >> | > | > the | > | >> | > | > Scripting API to load another configuration or execute | > | >> | > | > the | > | >> | > | > following: | > | >> | > | > | > | >> | > | > Metacello new | > | >> | > | > configuration: 'MetacelloPreview'; | > | >> | > | > version: #stable; | > | >> | > | > gemsource: 'metacello'; | > | >> | > | > load: 'batch'. | > | >> | > | > | > | >> | > | > \to load the tests, exedute: | > | >> | > | > | > | >> | > | > Metacello new | > | >> | > | > configuration: 'MetacelloPreview'; | > | >> | > | > version: #stable; | > | >> | > | > gemsource: 'metacello'; | > | >> | > | > load: 'TravisCI'. | > | >> | > | | > | >> | > | Just to get things completely clear in my head, these | > | >> | > | are | > | >> | > | instructions | > | >> | > | for Metacello users post-inclusion, right? They're not | > | >> | > | actions | > | >> | > | that | > | >> | > | we | > | >> | > | need to follow other than pushing Metacello-Base into | > | >> | > | the | > | >> | > | trunk, | > | >> | > | right? | > | >> | > | | > | >> | > | frank | > | >> | > | | > | >> | > | > I'll release 1.0-beta.32.3 in a week or so, but I'll | > | >> | > | > just | > | >> | > | > update | > | >> | > | > the #stable definition, so the above expressions | > | >> | > | > should be | > | >> | > | > used | > | >> | > | > to | > | >> | > | > guarantee that you're picking up the latest release... | > | >> | > | > | > | >> | > | > Dale | > | >> | > | > | > | >> | > | > [1] | > | >> | > | > | > | >> https://github.com/dalehenrich/metacello-work/blob/master/docs/MetacelloScri... | > | >> | > | > | > | >> | > | > ----- Original Message ----- | > | >> | > | > | From: "Dale Henrichs" dhenrich@vmware.com | > | >> | > | > | To: "The general-purpose Squeak developers list" | > | >> | > | > | squeak-dev@lists.squeakfoundation.org | > | >> | > | > | Sent: Friday, August 3, 2012 4:19:58 PM | > | >> | > | > | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | > | >> | > | > | | > | >> | > | > | After the patch, a reasonable number of test issues: | > | >> | > | > | | > | >> | > | > | 695 run, 691 passes, 0 expected failures, 2 | > | >> | > | > | failures, | > | >> | > | > | 2 | > | >> | > | > | errors, | > | >> | > | > | 0 | > | >> | > | > | unexpected passes | > | >> | > | > | | > | >> | > | > | Dale | > | >> | > | > | | > | >> | > | > | ----- Original Message ----- | > | >> | > | > | | From: "Dale Henrichs" dhenrich@vmware.com | > | >> | > | > | | To: "The general-purpose Squeak developers list" | > | >> | > | > | | squeak-dev@lists.squeakfoundation.org | > | >> | > | > | | Sent: Friday, August 3, 2012 1:25:57 PM | > | >> | > | > | | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | > | >> | > | > | | | > | >> | > | > | | Sounds fair to me ... I'll patch my current in | > | >> | > | > | | place | > | >> | > | > | | and | > | >> | > | > | | kick | > | >> | > | > | | off | > | >> | > | > | | the | > | >> | > | > | | tests again ... | > | >> | > | > | | | > | >> | > | > | | Thanks guys, | > | >> | > | > | | | > | >> | > | > | | Dale | > | >> | > | > | | | > | >> | > | > | | ----- Original Message ----- | > | >> | > | > | | | From: "Frank Shearar" frank.shearar@gmail.com | > | >> | > | > | | | To: "The general-purpose Squeak developers list" | > | >> | > | > | | | squeak-dev@lists.squeakfoundation.org | > | >> | > | > | | | Sent: Friday, August 3, 2012 1:20:57 PM | > | >> | > | > | | | Subject: Re: [squeak-dev] Metacello and | > | >> | > | > | | | Squeak4.4 | > | >> | > | > | | | | > | >> | > | > | | | On 3 August 2012 21:17, Colin Putney | > | >> | > | > | | | colin@wiresong.com | > | >> | > | > | | | wrote: | > | >> | > | > | | | > | > | >> | > | > | | | > | > | >> | > | > | | | > On Fri, Aug 3, 2012 at 1:00 PM, Dale Henrichs | > | >> | > | > | | | > dhenrich@vmware.com | > | >> | > | > | | | > wrote: | > | >> | > | > | | | >> | > | >> | > | > | | | >> Frank, | > | >> | > | > | | | >> | > | >> | > | > | | | >> The Metacello tests add/remove global | > | >> | > | > | | | >> variables | > | >> | > | > | | | >> to/from | > | >> | > | > | | | >> Smalltalk | > | >> | > | > | | | >> and it | > | >> | > | > | | | >> looks like Environment does not implement | > | >> | > | > | | | >> #removeKey:ifAbsent:, | > | >> | > | > | | | >> so | > | >> | > | > | | | >> I've got | > | >> | > | > | | | >> some 400 test failures because I can't remove | > | >> | > | > | | | >> a | > | >> | > | > | | | >> global | > | >> | > | > | | | >> from | > | >> | > | > | | | >> Smalltalk ... | > | >> | > | > | | | >> | > | >> | > | > | | | >> I can hack this for the time being but I'm | > | >> | > | > | | | >> curious | > | >> | > | > | | | >> what | > | >> | > | > | | | >> the | > | >> | > | > | | | >> longer | > | >> | > | > | | | >> term | > | >> | > | > | | | >> plan will be with regards to removing global | > | >> | > | > | | | >> variables | > | >> | > | > | | | >> from | > | >> | > | > | | | >> Smalltalk or | > | >> | > | > | | | >> perhaps leaving #removeKey:ifAbsent: out is | > | >> | > | > | | | >> inadvertent | > | >> | > | > | | | >> ... | > | >> | > | > | | | >> | > | >> | > | > | | | >> BTW, I'm using update 12154 ... | > | >> | > | > | | | > | > | >> | > | > | | | > | > | >> | > | > | | | > First, Environments is being pushed to Squeak | > | >> | > | > | | | > 4.5, | > | >> | > | > | | | > so | > | >> | > | > | | | > it | > | >> | > | > | | | > should | > | >> | > | > | | | > be | > | >> | > | > | | | > removed | > | >> | > | > | | | > from 4.4 shortly. | > | >> | > | > | | | > | > | >> | > | > | | | > During the development period for 4.5, | > | >> | > | > | | | > Environments | > | >> | > | > | | | > will | > | >> | > | > | | | > implement | > | >> | > | > | | | > #removeKey:ifAbsent:, but eventually that will | > | >> | > | > | | | > be | > | >> | > | > | | | > deprecated in | > | >> | > | > | | | > favour of | > | >> | > | > | | | > #undeclare: or something similar. | > | >> | > | > | | | | > | >> | > | > | | | For the nonce, how about we implement | > | >> | > | > | | | #removeKey:ifAbsent: | > | >> | > | > | | | (see | > | >> | > | > | | | Inbox), so we can at least unblock Dale? | > | >> | > | > | | | | > | >> | > | > | | | We do need to remove Environments, but Dale's | > | >> | > | > | | | work | > | >> | > | > | | | contributes to | > | >> | > | > | | | ensuring/checking 4.4's stability. | > | >> | > | > | | | | > | >> | > | > | | | frank | > | >> | > | > | | | | > | >> | > | > | | | > Colin | > | >> | > | > | | | > | > | >> | > | > | | | > | > | >> | > | > | | | > | > | >> | > | > | | | > | > | >> | > | > | | | | > | >> | > | > | | | | > | >> | > | > | | | > | >> | > | > | | | > | >> | > | > | | > | >> | > | > | | > | >> | > | > | > | >> | > | | > | >> | > | | > | >> | > | > | >> | | > | >> | | > | >> | > | > | > | > | > | | > | | > | |

I've added Squeak4.4 as one of the travis-ci build options[1].

I've updated the FileTree project to do builds against Squeak4.3 and Squeak4.4[2].

Then as I contemplate adding Squeak4.4 to the Metacello builds, I realized that there is one more addition to Squeak4.4 that would allow for better Metacello support.

If you add the method SmalltalkImage>>metacelloPlatformAttributes, then you can control the default list of Metacello atributes for your platform:

metacelloPlatformAttributes "Returns the tags for the conditional platform loading in Metacello. Pay attention the order is important: from most to least general."

^ #(#squeakCommon #squeak #'squeak4.x' #'squeak4.4.x')

This example is using constant literals to give you and idea of what needs to be defined, but you could create the atributes list dynamically based on the actual version of the release ... if you use constants you'll want to make sure this method is changed when you make a new release...

If you don't implement this method, then I have to add this information in a platform-specific package and right now I don't need a Squeak4.4-specific package.

I'll hold off on doing anything until I hear from you guys, so for the time being the platform attributes will be #squeak.4.3.x for Squeak4.4...

Dale

[1] http://travis-ci.org/#!/dalehenrich/builderCI/builds/2041711 [2] http://travis-ci.org/#!/dalehenrich/filetree/builds/2041755

----- Original Message ----- | From: "Dale Henrichs" dhenrich@vmware.com | To: "The general-purpose Squeak developers list" squeak-dev@lists.squeakfoundation.org | Sent: Saturday, August 4, 2012 12:00:27 PM | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | | Okay, | | Using Frank's 4.4 trunk (update 12154) you can load | Metacello-Base-dkh.98 from http://seaside.gemstone.com/ss/metacello | and then use the Metacello Preview scripting API[1]. | | As reported earlier, there are still some test failures in | MetacelloPreview 1.0-beta.32.2, but the fixes are to the tests and | not the functional code base, so I'd say you can include | Metacello-Base-dkh.98 in 4.4 at your leisure. | | To load the basic Metacello functionality, you can use the Scripting | API to load another configuration or execute the following: | | Metacello new | configuration: 'MetacelloPreview'; | version: #stable; | gemsource: 'metacello'; | load: 'batch'. | | \to load the tests, exedute: | | Metacello new | configuration: 'MetacelloPreview'; | version: #stable; | gemsource: 'metacello'; | load: 'TravisCI'. | | I'll release 1.0-beta.32.3 in a week or so, but I'll just update the | #stable definition, so the above expressions should be used to | guarantee that you're picking up the latest release... | | Dale | | [1] | https://github.com/dalehenrich/metacello-work/blob/master/docs/MetacelloScri... | | ----- Original Message ----- | | From: "Dale Henrichs" dhenrich@vmware.com | | To: "The general-purpose Squeak developers list" | | squeak-dev@lists.squeakfoundation.org | | Sent: Friday, August 3, 2012 4:19:58 PM | | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | | | | After the patch, a reasonable number of test issues: | | | | 695 run, 691 passes, 0 expected failures, 2 failures, 2 errors, 0 | | unexpected passes | | | | Dale | | | | ----- Original Message ----- | | | From: "Dale Henrichs" dhenrich@vmware.com | | | To: "The general-purpose Squeak developers list" | | | squeak-dev@lists.squeakfoundation.org | | | Sent: Friday, August 3, 2012 1:25:57 PM | | | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | | | | | | Sounds fair to me ... I'll patch my current in place and kick off | | | the | | | tests again ... | | | | | | Thanks guys, | | | | | | Dale | | | | | | ----- Original Message ----- | | | | From: "Frank Shearar" frank.shearar@gmail.com | | | | To: "The general-purpose Squeak developers list" | | | | squeak-dev@lists.squeakfoundation.org | | | | Sent: Friday, August 3, 2012 1:20:57 PM | | | | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | | | | | | | | On 3 August 2012 21:17, Colin Putney colin@wiresong.com | | | | wrote: | | | | > | | | | > | | | | > On Fri, Aug 3, 2012 at 1:00 PM, Dale Henrichs | | | | > dhenrich@vmware.com | | | | > wrote: | | | | >> | | | | >> Frank, | | | | >> | | | | >> The Metacello tests add/remove global variables to/from | | | | >> Smalltalk | | | | >> and it | | | | >> looks like Environment does not implement | | | | >> #removeKey:ifAbsent:, | | | | >> so | | | | >> I've got | | | | >> some 400 test failures because I can't remove a global from | | | | >> Smalltalk ... | | | | >> | | | | >> I can hack this for the time being but I'm curious what the | | | | >> longer | | | | >> term | | | | >> plan will be with regards to removing global variables from | | | | >> Smalltalk or | | | | >> perhaps leaving #removeKey:ifAbsent: out is inadvertent ... | | | | >> | | | | >> BTW, I'm using update 12154 ... | | | | > | | | | > | | | | > First, Environments is being pushed to Squeak 4.5, so it | | | | > should | | | | > be | | | | > removed | | | | > from 4.4 shortly. | | | | > | | | | > During the development period for 4.5, Environments will | | | | > implement | | | | > #removeKey:ifAbsent:, but eventually that will be deprecated | | | | > in | | | | > favour of | | | | > #undeclare: or something similar. | | | | | | | | For the nonce, how about we implement #removeKey:ifAbsent: (see | | | | Inbox), so we can at least unblock Dale? | | | | | | | | We do need to remove Environments, but Dale's work contributes | | | | to | | | | ensuring/checking 4.4's stability. | | | | | | | | frank | | | | | | | | > Colin | | | | > | | | | > | | | | > | | | | > | | | | | | | | | | | | | | | | | | | |

+1 from me too.

This looks like a light weight addition with benefit for anyone wanting to use Metacello, and no harm otherwise.

I have not tried unloading Metacello-Base, but it has no dependencies so it should not be a problem. There is an #unloadMetacello method in the base package that presumably performs as advertised.

Dave

On Fri, Aug 03, 2012 at 01:05:24AM +0200, Levente Uzonyi wrote:

+1 from me if it loads/unloads cleanly (i'm pretty sure it does, but it's always better to check it out).

Levente

On Thu, 2 Aug 2012, Dale Henrichs wrote:

...

I have a small 3 class package (Metacello-Base[1]) that I would hope could be included in the Squeak4.4 release.

Of the 3 classes, the Metacello class is a stub class for the Metacello Scripting Api[2], that will bootstrap Metacello _if_ you use the scripting api. I am promoting the scripting api as the cross platform way to use Metacello. Instead of seeing install scripts fly by that look like this:

Gofer new package: 'ConfigurationOfSeaside30'; squeaksource: 'MetacelloRepository'; load. ((Smalltalk at: #ConfigurationOfSeaside30) version: #stable) load.

(with the caveat that Installer should be used for Squeak)...the install scripts will look like this:

Metacello new configuration: 'Seaside30'; version: #stable; squeaksource: 'MetacelloRepository'; load.

The same script will "do the right thing" on all three supported platforms (Squeak, Pharo, GemStone), presuming that the class Metacello is present in the base images:)

I've recently released a preview version[4] and am actively collecting feedback. I would think that I can finalize the visible portion of the scripting api by September...

I've been ensuring that the metacello preview release is functional on Squeak all along and I am currently running CI builds against Squeak4.3 using travis ci[5]. Moving forward I would add Squeak4.4 to the list of platforms for CI and publish updates to the Metacello-Base pacakge as needed.

The class comments for each of the classes provides basic documentation for using Metacello so I think that Metacello-Base is a clean and small presence for Metacello and will enable the use of "cross-platform install scripts for Metacello."

Thanks,

Dale

[1] http://seaside.gemstone.com/ss/metacello/Metacello-Base-dkh.96.mcz [2] https://github.com/dalehenrich/metacello-work/blob/master/docs/MetacelloUser... [4] http://forum.world.st/Metacello-Preview-release-1-0-beta-32-2-tt4642064.html [5] http://travis-ci.org/#!/dalehenrich/metacello-work/builds/2012331

My guess would be that nobody has ever actually unloaded it ;-) But it's nice to know that it can be done.

Dave

On Thu, Aug 02, 2012 at 06:11:28PM -0700, Dale Henrichs wrote:

I admit that I haven't tested the #unloadMetacello method recently as that test would wreak havoc with my test environment, but then I haven't gotten any bug reports on #unloadmetacello either:)

However, I will add an #unloadMetacello test to travisCI to ensure that it functions as advertised moving forward.

Dale ----- Original Message ----- | From: "David T. Lewis" lewis@mail.msen.com | To: "The general-purpose Squeak developers list" squeak-dev@lists.squeakfoundation.org | Sent: Thursday, August 2, 2012 5:00:35 PM | Subject: Re: [squeak-dev] Metacello and Squeak4.4 | | +1 from me too. | | This looks like a light weight addition with benefit for anyone | wanting | to use Metacello, and no harm otherwise. | | I have not tried unloading Metacello-Base, but it has no dependencies | so | it should not be a problem. There is an #unloadMetacello method in | the base | package that presumably performs as advertised. | | Dave | | On Fri, Aug 03, 2012 at 01:05:24AM +0200, Levente Uzonyi wrote: | > +1 from me if it loads/unloads cleanly (i'm pretty sure it does, | > but it's | > always better to check it out). | > | > | > Levente | > | > On Thu, 2 Aug 2012, Dale Henrichs wrote: | > | > >I have a small 3 class package (Metacello-Base[1]) that I would | > >hope could | > >be included in the Squeak4.4 release. | > > | > >Of the 3 classes, the Metacello class is a stub class for the | > >Metacello | > >Scripting Api[2], that will bootstrap Metacello _if_ you use the | > >scripting | > >api. I am promoting the scripting api as the cross platform way to | > >use | > >Metacello. Instead of seeing install scripts fly by that look like | > >this: | > > | > > Gofer new | > > package: 'ConfigurationOfSeaside30'; | > > squeaksource: 'MetacelloRepository'; | > > load. | > > ((Smalltalk at: #ConfigurationOfSeaside30) version: #stable) | > > load. | > > | > >(with the caveat that Installer should be used for Squeak)...the | > >install | > >scripts will look like this: | > > | > > Metacello new | > > configuration: 'Seaside30'; | > > version: #stable; | > > squeaksource: 'MetacelloRepository'; | > > load. | > > | > >The same script will "do the right thing" on all three supported | > >platforms | > >(Squeak, Pharo, GemStone), presuming that the class Metacello is | > >present | > >in the base images:) | > > | > >I've recently released a preview version[4] and am actively | > >collecting | > >feedback. I would think that I can finalize the visible portion of | > >the | > >scripting api by September... | > > | > >I've been ensuring that the metacello preview release is | > >functional on | > >Squeak all along and I am currently running CI builds against | > >Squeak4.3 | > >using travis ci[5]. Moving forward I would add Squeak4.4 to the | > >list of | > >platforms for CI and publish updates to the Metacello-Base pacakge | > >as | > >needed. | > > | > >The class comments for each of the classes provides basic | > >documentation | > >for using Metacello so I think that Metacello-Base is a clean and | > >small | > >presence for Metacello and will enable the use of "cross-platform | > >install | > >scripts for Metacello." | > > | > >Thanks, | > > | > >Dale | > > | > >[1] | > >http://seaside.gemstone.com/ss/metacello/Metacello-Base-dkh.96.mcz | > >[2] | > >https://github.com/dalehenrich/metacello-work/blob/master/docs/MetacelloUser... | > >[4] | > >http://forum.world.st/Metacello-Preview-release-1-0-beta-32-2-tt4642064.html | > >[5] | > >http://travis-ci.org/#!/dalehenrich/metacello-work/builds/2012331 | > > | > > | |