[Glass] Jade 1.059: package extension naming ...

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
7 messages Options
Reply | Threaded
Open this post in threaded view
|

[Glass] Jade 1.059: package extension naming ...

GLASS mailing list
Something has changed in the behaviour of recognizing, that an extension
has changed.

If I have a package named "Hallo" and "HalloExtension" and I add a
method category "*HalloExtension" to a class defined in "Hallo" and add
a method to that category, the system will not make any package "black"
(= changed).

If I add a category "*halloextension" the system works as in in earlier
releases: the extension is "blacked".

The problem for me is: I have lots of extensions with upper/lower cased
conventions and I assume if I add methods to these already defined
extension method categories in various classes the system will NOW not
recognize, that the message belongs to an extension.

Perhaps it may even get lost ... because it is not exported at all ???

And another wish I have since I use Squeak/Pharo/Jade: is there any hope
to get rid of this Squeak/Pharo-introduced brain damaged extension
management in the system browser ?

Or another question: what is the correct way to have an overview of the
content of an extension package ?

How it could be done better: look at the application browser of
VASmalltalk. They did it right ...



Marten




--
Marten Feldtmann
_______________________________________________
Glass mailing list
[hidden email]
http://lists.gemtalksystems.com/mailman/listinfo/glass
Reply | Threaded
Open this post in threaded view
|

Re: [Glass] Jade 1.059: package extension naming ...

GLASS mailing list
Actually it seems, that there is more broken in this area ...

http://www.schrievkrom.de/extfiles/mutter/JadePackageStateError/JadeError.html

Marten

--
Marten Feldtmann
_______________________________________________
Glass mailing list
[hidden email]
http://lists.gemtalksystems.com/mailman/listinfo/glass
Reply | Threaded
Open this post in threaded view
|

Re: [Glass] Jade 1.059: package extension naming ...

GLASS mailing list
In reply to this post by GLASS mailing list


On Wed, Nov 26, 2014 at 10:36 PM, [hidden email] via Glass <[hidden email]> wrote:
... 
And another wish I have since I use Squeak/Pharo/Jade: is there any hope
to get rid of this Squeak/Pharo-introduced brain damaged extension
management in the system browser ?

Yes there is hope ... The current plan is to add Cypress support to the base system for 3.3 which means that we will revisit package handling and it is likely that we will go away from overloading class and method categories...

Dale

_______________________________________________
Glass mailing list
[hidden email]
http://lists.gemtalksystems.com/mailman/listinfo/glass
Reply | Threaded
Open this post in threaded view
|

Re: [Glass] Jade 1.059: package extension naming ...

GLASS mailing list
In reply to this post by GLASS mailing list
I can anyone confirm this ? I'm working here further on with this
version, but get not a very good feeling.

If I add a method to an extension method category, this method is stored
in the base package but also in the extension package. This at least can
be seen in the Monticello packages created via Jade.

I looked at the source code within Dolphin, but actually on a first
quikck view I simply did not understand how the stuff is working.


Marten

Am 27.11.2014 um 07:36 schrieb [hidden email] via Glass:

> Something has changed in the behaviour of recognizing, that an extension
> has changed.
>
> If I have a package named "Hallo" and "HalloExtension" and I add a
> method category "*HalloExtension" to a class defined in "Hallo" and add
> a method to that category, the system will not make any package "black"
> (= changed).
>
> If I add a category "*halloextension" the system works as in in earlier
> releases: the extension is "blacked".
>


--
Marten Feldtmann
_______________________________________________
Glass mailing list
[hidden email]
http://lists.gemtalksystems.com/mailman/listinfo/glass
Reply | Threaded
Open this post in threaded view
|

Re: [Glass] Jade 1.059: package extension naming ...

GLASS mailing list
Marten,

Thank you for this report. I think that it is related to the method used to compile code. I made a change to get back warnings as well as errors and it appears that this low-level method does not set the package properly. I will investigate and report back.

James

> On Dec 2, 2014, at 1:41 AM, [hidden email] via Glass <[hidden email]> wrote:
>
> I can anyone confirm this ? I'm working here further on with this
> version, but get not a very good feeling.
>
> If I add a method to an extension method category, this method is stored
> in the base package but also in the extension package. This at least can
> be seen in the Monticello packages created via Jade.
>
> I looked at the source code within Dolphin, but actually on a first
> quikck view I simply did not understand how the stuff is working.
>
>
> Marten
>
>> Am 27.11.2014 um 07:36 schrieb [hidden email] via Glass:
>> Something has changed in the behaviour of recognizing, that an extension
>> has changed.
>>
>> If I have a package named "Hallo" and "HalloExtension" and I add a
>> method category "*HalloExtension" to a class defined in "Hallo" and add
>> a method to that category, the system will not make any package "black"
>> (= changed).
>>
>> If I add a category "*halloextension" the system works as in in earlier
>> releases: the extension is "blacked".
>>
>
>
> --
> Marten Feldtmann
> _______________________________________________
> Glass mailing list
> [hidden email]
> http://lists.gemtalksystems.com/mailman/listinfo/glass
_______________________________________________
Glass mailing list
[hidden email]
http://lists.gemtalksystems.com/mailman/listinfo/glass
Reply | Threaded
Open this post in threaded view
|

Re: [Glass] Jade 1.059: package extension naming ...

GLASS mailing list
The GLASS extensions and modifications to base classes update the package as modified if a newly added method is used. If previously existing methods are used the package is not updated. Unfortunately, the new method does not report compile errors. What I’m doing is calling both methods now and it seems to work. Let me know if this is fixed and if you see other issues.

James

> On Dec 2, 2014, at 6:15 AM, James Foster <[hidden email]> wrote:
>
> Marten,
>
> Thank you for this report. I think that it is related to the method used to compile code. I made a change to get back warnings as well as errors and it appears that this low-level method does not set the package properly. I will investigate and report back.
>
> James
>
>> On Dec 2, 2014, at 1:41 AM, [hidden email] via Glass <[hidden email]> wrote:
>>
>> I can anyone confirm this ? I'm working here further on with this
>> version, but get not a very good feeling.
>>
>> If I add a method to an extension method category, this method is stored
>> in the base package but also in the extension package. This at least can
>> be seen in the Monticello packages created via Jade.
>>
>> I looked at the source code within Dolphin, but actually on a first
>> quikck view I simply did not understand how the stuff is working.
>>
>>
>> Marten
>>
>>> Am 27.11.2014 um 07:36 schrieb [hidden email] via Glass:
>>> Something has changed in the behaviour of recognizing, that an extension
>>> has changed.
>>>
>>> If I have a package named "Hallo" and "HalloExtension" and I add a
>>> method category "*HalloExtension" to a class defined in "Hallo" and add
>>> a method to that category, the system will not make any package "black"
>>> (= changed).
>>>
>>> If I add a category "*halloextension" the system works as in in earlier
>>> releases: the extension is "blacked".
>>>
>>
>>
>> --
>> Marten Feldtmann
>> _______________________________________________
>> Glass mailing list
>> [hidden email]
>> http://lists.gemtalksystems.com/mailman/listinfo/glass

_______________________________________________
Glass mailing list
[hidden email]
http://lists.gemtalksystems.com/mailman/listinfo/glass
Reply | Threaded
Open this post in threaded view
|

Re: [Glass] Jade 1.059: package extension naming ...

GLASS mailing list
Yes, the entry points for tODE and GemTools compilation is intended to mimic the Pharo/Squeak API so if Jade is not making the same calls then the behavior isn't likely to be the same either ...

Dale

On Sun, Dec 7, 2014 at 3:23 PM, James Foster via Glass <[hidden email]> wrote:
The GLASS extensions and modifications to base classes update the package as modified if a newly added method is used. If previously existing methods are used the package is not updated. Unfortunately, the new method does not report compile errors. What I’m doing is calling both methods now and it seems to work. Let me know if this is fixed and if you see other issues.

James

> On Dec 2, 2014, at 6:15 AM, James Foster <[hidden email]> wrote:
>
> Marten,
>
> Thank you for this report. I think that it is related to the method used to compile code. I made a change to get back warnings as well as errors and it appears that this low-level method does not set the package properly. I will investigate and report back.
>
> James
>
>> On Dec 2, 2014, at 1:41 AM, [hidden email] via Glass <[hidden email]> wrote:
>>
>> I can anyone confirm this ? I'm working here further on with this
>> version, but get not a very good feeling.
>>
>> If I add a method to an extension method category, this method is stored
>> in the base package but also in the extension package. This at least can
>> be seen in the Monticello packages created via Jade.
>>
>> I looked at the source code within Dolphin, but actually on a first
>> quikck view I simply did not understand how the stuff is working.
>>
>>
>> Marten
>>
>>> Am 27.11.2014 um 07:36 schrieb [hidden email] via Glass:
>>> Something has changed in the behaviour of recognizing, that an extension
>>> has changed.
>>>
>>> If I have a package named "Hallo" and "HalloExtension" and I add a
>>> method category "*HalloExtension" to a class defined in "Hallo" and add
>>> a method to that category, the system will not make any package "black"
>>> (= changed).
>>>
>>> If I add a category "*halloextension" the system works as in in earlier
>>> releases: the extension is "blacked".
>>>
>>
>>
>> --
>> Marten Feldtmann
>> _______________________________________________
>> Glass mailing list
>> [hidden email]
>> http://lists.gemtalksystems.com/mailman/listinfo/glass

_______________________________________________
Glass mailing list
[hidden email]
http://lists.gemtalksystems.com/mailman/listinfo/glass


_______________________________________________
Glass mailing list
[hidden email]
http://lists.gemtalksystems.com/mailman/listinfo/glass