Hi Rita
Thank you for your reply. It is good to hear that the Etoy protagonists would be interested in back-porting Etoys improvements to the main Squeak release.
However, do we know if the Squeak release builders want any improvements to morphic - or are they going to cut back to a minimorphic?
I still don't know what questions to ask about how things are managed. I have been looking around more - there have been some changes with lists recently that I didn't know about. (I notice that the lists are not very used as yet.) I still don't have much clue as to where to find things about what goes into the main Squeak release, who controls it, or where to look to see the current state of things. Advice would be welcome.
Yours
Bob
Robert Hawley wrote:
Hi Craig
Would I like to write a proposal? Not at this stage - I am not sufficiently aware of who does what and of how things are done.
Can you answer about is this being managed and/or fixed and about plans, issues and people's motivations? For example - are you
involved
in this topic?
It is a topic that I am interested in, and I have a fair knowledge of how Morphic works. I don't know about the processes of getting
things
into the Squeak releases however. Neither do I know if the current protagonists for the etoys system are wanting to be involved in back-porting their work to Squeak.
Hi Bob,
the Squeakland Foundation, who are the protagonists for the Etoys system, would *love* to be back in sync with the Squeak system. I'm personally interested in exactly what you described in an earlier mail, a "useful route into introducing and teaching Smalltalk". I don't think that we will back-porting Etoys to 3.9 or 3.10, but rather with the next squeak release. This is no short-term goal, but I would like to discuss how a reunion of Squeak and Etoys could be done.
Greetings, Rita Squeakland Foundation www.squeakland.org
Yours
Bob
Hi Bob--
Would you like to write a proposal?-C
-- Craig Latta www.netjam.org next show: 2009-03-13 (www.thishere.org)
On 13.03.2009, at 04:32, Robert Hawley wrote:
Hi Rita
Thank you for your reply. It is good to hear that the Etoy protagonists would be interested in back-porting Etoys improvements to the main Squeak release.
However, do we know if the Squeak release builders want any improvements to morphic - or are they going to cut back to a minimorphic?
Well, I am one of the "Etoys protagonists" and just got re-elected by the community with my best-ever result (http://tinyurl.com/squeak2009) so I take that as a vote pro Etoys. It still will be a major chunk of work, but at least those most opposed to Etoys have their own way of working on a lean Squeak now.
I still don't know what questions to ask about how things are managed. I have been looking around more - there have been some changes with lists recently that I didn't know about. (I notice that the lists are not very used as yet.) I still don't have much clue as to where to find things about what goes into the main Squeak release, who controls it, or where to look to see the current state of things. Advice would be welcome.
You are not alone. The truth is, there never has been a clear process ever since Squeak became community-driven. But it looks like the time is ripe now for change. The community has clearly voted for Andreas whose main campaigning platform was to establish those processes.
Anyway, merging the latest Etoys back into a squeak.org release will probably take a while. Well, unless we find enough volunteers with copious amounts of spare time ;)
If somebody is interested, the first thing to get an overview would be to review the change logs (we started with Etoys 1.0 = Squeak 3.8 and are now at Etoys 4.0):
I'm pretty sure that rather few "system-level" methods were touched. One could start by cherry-picking "interesting" changes/fixes into 3.11.
For a full integration we probably need to packetize Etoys first, so we can at least compare Monticello packages between Etoys and Squeak. Comments welcome.
- Bert -
On 2009-03-13 10:03, Bert Freudenberg wrote:
On 13.03.2009, at 04:32, Robert Hawley wrote:
Hi Rita
Thank you for your reply. It is good to hear that the Etoy protagonists would be interested in back-porting Etoys improvements to the main Squeak release.
However, do we know if the Squeak release builders want any improvements to morphic - or are they going to cut back to a minimorphic?
Well, I am one of the "Etoys protagonists" and just got re-elected by the community with my best-ever result (http://tinyurl.com/squeak2009) so I take that as a vote pro Etoys. It still will be a major chunk of work, but at least those most opposed to Etoys have their own way of working on a lean Squeak now.
I still don't know what questions to ask about how things are managed. I have been looking around more - there have been some changes with lists recently that I didn't know about. (I notice that the lists are not very used as yet.) I still don't have much clue as to where to find things about what goes into the main Squeak release, who controls it, or where to look to see the current state of things. Advice would be welcome.
You are not alone. The truth is, there never has been a clear process ever since Squeak became community-driven. But it looks like the time is ripe now for change. The community has clearly voted for Andreas whose main campaigning platform was to establish those processes.
Anyway, merging the latest Etoys back into a squeak.org release will probably take a while. Well, unless we find enough volunteers with copious amounts of spare time ;)
If somebody is interested, the first thing to get an overview would be to review the change logs (we started with Etoys 1.0 = Squeak 3.8 and are now at Etoys 4.0):
I'm pretty sure that rather few "system-level" methods were touched. One could start by cherry-picking "interesting" changes/fixes into 3.11.
For a full integration we probably need to packetize Etoys first, so we can at least compare Monticello packages between Etoys and Squeak. Comments welcome.
There is Pavels Minimal Morphic with Etoys split out based on Juans packages but those are probably 3.6 or 3.7, but is's a start :-) I'm not sure what is considered Etoys in those packages tho...or what considered Etoys for that part at all.
Karl
On 13.03.2009, at 17:00, Karl Ramberg wrote:
On 2009-03-13 10:03, Bert Freudenberg wrote:
On 13.03.2009, at 04:32, Robert Hawley wrote:
Hi Rita
Thank you for your reply. It is good to hear that the Etoy protagonists would be interested in back-porting Etoys improvements to the main Squeak release.
However, do we know if the Squeak release builders want any improvements to morphic - or are they going to cut back to a minimorphic?
Well, I am one of the "Etoys protagonists" and just got re-elected by the community with my best-ever result (http://tinyurl.com/squeak2009 ) so I take that as a vote pro Etoys. It still will be a major chunk of work, but at least those most opposed to Etoys have their own way of working on a lean Squeak now.
I still don't know what questions to ask about how things are managed. I have been looking around more - there have been some changes with lists recently that I didn't know about. (I notice that the lists are not very used as yet.) I still don't have much clue as to where to find things about what goes into the main Squeak release, who controls it, or where to look to see the current state of things. Advice would be welcome.
You are not alone. The truth is, there never has been a clear process ever since Squeak became community-driven. But it looks like the time is ripe now for change. The community has clearly voted for Andreas whose main campaigning platform was to establish those processes.
Anyway, merging the latest Etoys back into a squeak.org release will probably take a while. Well, unless we find enough volunteers with copious amounts of spare time ;)
If somebody is interested, the first thing to get an overview would be to review the change logs (we started with Etoys 1.0 = Squeak 3.8 and are now at Etoys 4.0):
I'm pretty sure that rather few "system-level" methods were touched. One could start by cherry-picking "interesting" changes/ fixes into 3.11.
For a full integration we probably need to packetize Etoys first, so we can at least compare Monticello packages between Etoys and Squeak. Comments welcome.
There is Pavels Minimal Morphic with Etoys split out based on Juans packages but those are probably 3.6 or 3.7, but is's a start :-) I'm not sure what is considered Etoys in those packages tho...or what considered Etoys for that part at all.
Well, for the sake of this discussion I just care about merging all the improvements and bug fixes we did in the last 2+ years in the Etoys image back into the packages that make up the official Squeak image, and vice versa. This approach worked well back in 3.8 and so I expect it to work again.
For this to succeed we don't really need to define what part of Squeak is Etoys and what part is not.
- Bert -
Sounds ok to me :-) It's a noble task for a brave soul
Karl
On 3/13/09, Bert Freudenberg bert@freudenbergs.de wrote:
On 13.03.2009, at 17:00, Karl Ramberg wrote:
On 2009-03-13 10:03, Bert Freudenberg wrote:
On 13.03.2009, at 04:32, Robert Hawley wrote:
Hi Rita
Thank you for your reply. It is good to hear that the Etoy protagonists would be interested in back-porting Etoys improvements to the main Squeak release.
However, do we know if the Squeak release builders want any improvements to morphic - or are they going to cut back to a minimorphic?
Well, I am one of the "Etoys protagonists" and just got re-elected by the community with my best-ever result (http://tinyurl.com/squeak2009 ) so I take that as a vote pro Etoys. It still will be a major chunk of work, but at least those most opposed to Etoys have their own way of working on a lean Squeak now.
I still don't know what questions to ask about how things are managed. I have been looking around more - there have been some changes with lists recently that I didn't know about. (I notice that the lists are not very used as yet.) I still don't have much clue as to where to find things about what goes into the main Squeak release, who controls it, or where to look to see the current state of things. Advice would be welcome.
You are not alone. The truth is, there never has been a clear process ever since Squeak became community-driven. But it looks like the time is ripe now for change. The community has clearly voted for Andreas whose main campaigning platform was to establish those processes.
Anyway, merging the latest Etoys back into a squeak.org release will probably take a while. Well, unless we find enough volunteers with copious amounts of spare time ;)
If somebody is interested, the first thing to get an overview would be to review the change logs (we started with Etoys 1.0 = Squeak 3.8 and are now at Etoys 4.0):
I'm pretty sure that rather few "system-level" methods were touched. One could start by cherry-picking "interesting" changes/ fixes into 3.11.
For a full integration we probably need to packetize Etoys first, so we can at least compare Monticello packages between Etoys and Squeak. Comments welcome.
There is Pavels Minimal Morphic with Etoys split out based on Juans packages but those are probably 3.6 or 3.7, but is's a start :-) I'm not sure what is considered Etoys in those packages tho...or what considered Etoys for that part at all.
Well, for the sake of this discussion I just care about merging all the improvements and bug fixes we did in the last 2+ years in the Etoys image back into the packages that make up the official Squeak image, and vice versa. This approach worked well back in 3.8 and so I expect it to work again.
For this to succeed we don't really need to define what part of Squeak is Etoys and what part is not.
- Bert -
Bert Freudenberg wrote:
For a full integration we probably need to packetize Etoys first, so we can at least compare Monticello packages between Etoys and Squeak. Comments welcome.
Do you mean package Etoys as a MCZ that can be loaded into Squeak 3.8 or Squeak 3.11 (or 4.0)? This would imply separating the updates into two packages - Etoys package and a package containing patches made to Squeak in the Etoys stream.
Subbu
On 15.03.2009, at 12:55, K. K. Subramaniam wrote:
Bert Freudenberg wrote:
For a full integration we probably need to packetize Etoys first, so we can at least compare Monticello packages between Etoys and Squeak. Comments welcome.
Do you mean package Etoys as a MCZ that can be loaded into Squeak 3.8 or Squeak 3.11 (or 4.0)?
No. What I meant was just to reorganize all classes and methods in the Squeakland image into categories, based on the initial image categorization Andreas originally did for 3.8 (which has been in official use since 3.9). That would allow to create MC packages that could be compared to the versions in 3.11.
- Bert -
Bert Freudenberg wrote on Fri, 13 Mar 2009 10:03:44 +0100
Anyway, merging the latest Etoys back into a squeak.org release will probably take a while. Well, unless we find enough volunteers with copious amounts of spare time ;)
Since we do lack resources to do even this, I have been reluctant to start a discussion about the future of Etoys. But you never know when new resources might become available, so it is good to be prepared.
My own position about Etoys is that it is a very integral part of Squeak and not some application running on top of it. That doesn't mean that I wouldn't be very happy to have images that didn't include it, but I also sometimes need images without SystemBrowser and Compiler or without any networking stuff at all. This is an issue of modularity and that is not where I want this thread to go.
Scratch is so similar to Etoys but is also complementary in many ways. It has been far more successful with younger children or with more casual users (I would have thought that explict loops instead of the "ticking clock" would be harder for these groups, which shows the importance of testing rather than depending on my own opinions).
On the other hand, Elements by Jens Moenig takes the Scratch-style interface closer to regular Smalltalk. I have not tested it myself, but what I have seen seems very reasonable. It makes me think that an Etoys 2 could be possible by borrowing pieces from Scratch, Etoys 1 and Elements.
A very important part of Etoys to me is Kedama. In a very short paper (http://www.vpri.org/pdf/m2008002_massParallel.pdf), Yoshiki speculates about what a Kedama 2 should be like. I have a very strong opinion about that - we should present the parallelism model from APL (see FScript) as the generalization of the current ConnectionMachine model (SIMD). Instead of having the turtles be from different classes, you put them in different groups (with a single turtle able to belong to multiple groups) and then send messages to the groups.
The idea is to reduce the gap between "loose objects" and Kedama, to extend Etoys down to the Scratch level and to extend Etoys up to the full Smalltalk level.
The main incompatibility between Scratch and Etoys that | am aware of is that the former uses drag-n-drop exclusively while the latter builds expressions with tiny menus. Are there other issues? And should I have included other mailing lists in this dicussion?
-- Jecel
squeak-dev@lists.squeakfoundation.org
-
Bert Freudenberg -
Jecel Assumpcao Jr -
K. K. Subramaniam -
karl ramberg -
Karl Ramberg
-
Robert Hawley