Etoys image also have a stack morph
Karl
On Wed, Feb 27, 2013 at 11:24 AM, Jecel Assumpcao Jr. jecel@merlintec.comwrote:
While the DynaBookJr (I always misread "DBJr" as "DataBase Jr") is indeed a good modern replacement for HyperCard, isn't the normal BookMorph already available in Squeak a reasonable approximation?
-- Jecel
Ted Kaehler has done two implementations of DBJr AFAIK, one in Squeak and one in Javascript.
Here's the javascript version (for Lively Kernel)
http://forum.world.st/Try-DBJr-HyperCard-like-stacks-in-Lively-Kernel-td4599...
I had to dig to find the Moshi image, but here's an actual Squeak image -- and I think this was the right image -- note that this was a prototype and that some things (e.g. saving) aren't implemented. Also, I can't get it to open anymore, but I think that might be because I don't have a non-Cog VM now? Not sure which .sources file you want but I'm betting it's V3. The file of interest is "Text Field for LObject.zip"
http://tinlizzie.org/lesserphic2/
I was really fascinated with both the approach and the presentation he did here. Active essays are cool.
Meantime I have to figure out why I can't open the image. I thought Cog could open older images, but interpreters can't open Cog images. Hmmm.
On Wed, Feb 27, 2013 at 10:12 AM, karl ramberg karlramberg@gmail.comwrote:
Etoys image also have a stack morph
Karl
On Wed, Feb 27, 2013 at 11:24 AM, Jecel Assumpcao Jr. <jecel@merlintec.com
wrote:
While the DynaBookJr (I always misread "DBJr" as "DataBase Jr") is indeed a good modern replacement for HyperCard, isn't the normal BookMorph already available in Squeak a reasonable approximation?
-- Jecel
On Wed, Feb 27, 2013 at 04:31:36PM -0800, Casey Ransberger wrote:
Ted Kaehler has done two implementations of DBJr AFAIK, one in Squeak and one in Javascript.
Here's the javascript version (for Lively Kernel)
http://forum.world.st/Try-DBJr-HyperCard-like-stacks-in-Lively-Kernel-td4599...
I had to dig to find the Moshi image, but here's an actual Squeak image -- and I think this was the right image -- note that this was a prototype and that some things (e.g. saving) aren't implemented. Also, I can't get it to open anymore, but I think that might be because I don't have a non-Cog VM now? Not sure which .sources file you want but I'm betting it's V3. The file of interest is "Text Field for LObject.zip"
http://tinlizzie.org/lesserphic2/
I was really fascinated with both the approach and the presentation he did here. Active essays are cool.
Meantime I have to figure out why I can't open the image. I thought Cog could open older images, but interpreters can't open Cog images. Hmmm.
Casey,
Try an interpreter VM from squeakvm.org. It will run Cog images as well as older Squeak image formats back to at least Squeak 3.6. It runs the image attp://tinlizzie.org/lesserphic2/ without any problem.
Dave
Thank you Casey and Dave for the references to get a working DynaBook Junior image.
It works fine.
A thing which might not be obvious it to download the example stack http://tinlizzie.org/lesserphic2/aboutDBJr%205-Aug.morph in addition, locate it with the file browser and then choose 'load as morph'
--Hannes
On 2/28/13, David T. Lewis lewis@mail.msen.com wrote:
On Wed, Feb 27, 2013 at 04:31:36PM -0800, Casey Ransberger wrote:
Ted Kaehler has done two implementations of DBJr AFAIK, one in Squeak and one in Javascript.
Here's the javascript version (for Lively Kernel)
http://forum.world.st/Try-DBJr-HyperCard-like-stacks-in-Lively-Kernel-td4599...
I had to dig to find the Moshi image, but here's an actual Squeak image
and I think this was the right image -- note that this was a prototype and that some things (e.g. saving) aren't implemented. Also, I can't get it to open anymore, but I think that might be because I don't have a non-Cog VM now? Not sure which .sources file you want but I'm betting it's V3. The file of interest is "Text Field for LObject.zip"
http://tinlizzie.org/lesserphic2/
I was really fascinated with both the approach and the presentation he did here. Active essays are cool.
Meantime I have to figure out why I can't open the image. I thought Cog could open older images, but interpreters can't open Cog images. Hmmm.
Casey,
Try an interpreter VM from squeakvm.org. It will run Cog images as well as older Squeak image formats back to at least Squeak 3.6. It runs the image attp://tinlizzie.org/lesserphic2/ without any problem.
Dave
On Thu, Feb 28, 2013 at 02:04:13AM +0000, H. Hirzel wrote:
Thank you Casey and Dave for the references to get a working DynaBook Junior image.
It works fine.
A thing which might not be obvious it to download the example stack http://tinlizzie.org/lesserphic2/aboutDBJr%205-Aug.morph in addition, locate it with the file browser and then choose 'load as morph'
Hannes,
Thanks! That was not obvious to me at all, thanks for the tip :)
Dave
On 2/28/13, David T. Lewis lewis@mail.msen.com wrote:
On Wed, Feb 27, 2013 at 04:31:36PM -0800, Casey Ransberger wrote:
Ted Kaehler has done two implementations of DBJr AFAIK, one in Squeak and one in Javascript.
Here's the javascript version (for Lively Kernel)
http://forum.world.st/Try-DBJr-HyperCard-like-stacks-in-Lively-Kernel-td4599...
I had to dig to find the Moshi image, but here's an actual Squeak image
and I think this was the right image -- note that this was a prototype and that some things (e.g. saving) aren't implemented. Also, I can't get it to open anymore, but I think that might be because I don't have a non-Cog VM now? Not sure which .sources file you want but I'm betting it's V3. The file of interest is "Text Field for LObject.zip"
http://tinlizzie.org/lesserphic2/
I was really fascinated with both the approach and the presentation he did here. Active essays are cool.
Meantime I have to figure out why I can't open the image. I thought Cog could open older images, but interpreters can't open Cog images. Hmmm.
Casey,
Try an interpreter VM from squeakvm.org. It will run Cog images as well as older Squeak image formats back to at least Squeak 3.6. It runs the image attp://tinlizzie.org/lesserphic2/ without any problem.
Dave
It is not that we at Viewpoints are trying to be secretive, but we do have a newer system (or systems). Hopefully we can put some code out when our report is done.
(Sorry for keeping people guessing.)
Sometimes there can be too many cooks in the kitchen. I think most people probably do understand this.
I'm very glad to read the papers and follow whatever artifacts that make it out of the lab. You folks may try something that doesn't turn out to be the right approach for Frank, but might be otherwise good and really useful for other folks with different goals. This is why I'm so grateful for the VPRI bits that *do* hit the net.
So: just keep blazing trails and don't stop! We'll hear about it when we hear about it:)
Casey
On Feb 27, 2013, at 8:09 PM, Yoshiki Ohshima Yoshiki.Ohshima@acm.org wrote:
It is not that we at Viewpoints are trying to be secretive, but we do have a newer system (or systems). Hopefully we can put some code out when our report is done.
(Sorry for keeping people guessing.)
-- -- Yoshiki
On 2/28/13 2:09 AM, "Yoshiki Ohshima" Yoshiki.Ohshima@acm.org wrote:
It is not that we at Viewpoints are trying to be secretive, but we do have a newer system (or systems). Hopefully we can put some code out when our report is done.
(Sorry for keeping people guessing.)
-- -- Yoshiki
The files at http://tinlizzie.org/dbjr/ could be used ? If I guess how to have DBJr (kind of) into the coming FunSqueakRos4dot4 (in development) is legal ?
First step is load aboutDBJr 5-Aug.morph into the system ...
Edgar
P.S. Thanks to all who participate in this thread, the info was very useful
To run the latest and greatest FONC system the Gezire plugin is necessary. I tried to get the Gezira plug in to compile on Windows but I could not get the tool chain right and got lost in all the quirky stuff. I would be nice if there where compiled plugins for all platforms for download somewhere.
Karl
On Thu, Feb 28, 2013 at 5:09 AM, Yoshiki Ohshima Yoshiki.Ohshima@acm.orgwrote:
It is not that we at Viewpoints are trying to be secretive, but we do have a newer system (or systems). Hopefully we can put some code out when our report is done.
(Sorry for keeping people guessing.)
-- -- Yoshiki
Let's keep cool and wait. Or, just take the Gezira code that's leaked out so far and then run as fast as we can with it. I have a feeling, though, that waiting might end up working out better than a lot of folks expect.
The results of these experiments may confront us with new challenges.
Who could ask for anything more?
Let's keep our heads and just pay close attention, okay? I think that's our best plan, don't you?
Even if what they're doing isn't a "product" or intended that way, because it's really a big complicated science experiment, I think we will get the best output in terms of understanding the work. If we can understand the work well, we can repeat it without much effort, no?
Let us be patient, as we may entreat angels unawares. (Bad quote! I got it all wrong!)
:)
C
On Feb 28, 2013, at 2:02 AM, karl ramberg karlramberg@gmail.com wrote:
To run the latest and greatest FONC system the Gezire plugin is necessary. I tried to get the Gezira plug in to compile on Windows but I could not get the tool chain right and got lost in all the quirky stuff. I would be nice if there where compiled plugins for all platforms for download somewhere.
Karl
On Thu, Feb 28, 2013 at 5:09 AM, Yoshiki Ohshima Yoshiki.Ohshima@acm.org wrote: It is not that we at Viewpoints are trying to be secretive, but we do have a newer system (or systems). Hopefully we can put some code out when our report is done.
(Sorry for keeping people guessing.)
-- -- Yoshiki
Thank you, Yoshiki, for the clarification.
I understand that you are doing research prototypes at VPRI and not supported products and you do many of them.
But I have the impression that the VPRI and you are a bit too modest in terms of advertising the results.
On 2/28/13, Yoshiki Ohshima Yoshiki.Ohshima@acm.org wrote:
It is not that we at Viewpoints are trying to be secretive, but we do have a newer system (or systems).
Do you think that the existing code (the HyperCard stack implementation) in http://tinlizzie.org/lesserphic2/ is under MIT license, thus compatible with Squeak 4.0 + ?
I'd like to look into porting it to a newer release of Squeak and/or Cuis?
It looks like the "Moshi image" is an internal fork you use at VPRI? Do you see any possible problem areas when porting it to a newer version / different fork?
Hopefully we can put some code out when our report is done.
Great. And a few appendices with README.md and release information type of info.
(Sorry for keeping people guessing.)
That is fine. You wrote the mail and so stopped us from doing guess work ;-)
-- -- Yoshiki
--Hannes
We have an internal image based on the trunk actually and the active development is being done in that image. (That's why there are occasional bug reports from the VPRI folks other than Bert^^;)
The Lesserphic version of DBJr is considered to be obsolete. It uses the "Etoys-style" uniclasses for pages but we later figured out a simpler way of implementing it.
On Mon, Mar 4, 2013 at 10:19 AM, H. Hirzel hannes.hirzel@gmail.com wrote:
Thank you, Yoshiki, for the clarification.
I understand that you are doing research prototypes at VPRI and not supported products and you do many of them.
But I have the impression that the VPRI and you are a bit too modest in terms of advertising the results.
On 2/28/13, Yoshiki Ohshima Yoshiki.Ohshima@acm.org wrote:
It is not that we at Viewpoints are trying to be secretive, but we do have a newer system (or systems).
Do you think that the existing code (the HyperCard stack implementation) in http://tinlizzie.org/lesserphic2/ is under MIT license, thus compatible with Squeak 4.0 + ?
I'd like to look into porting it to a newer release of Squeak and/or Cuis?
It looks like the "Moshi image" is an internal fork you use at VPRI? Do you see any possible problem areas when porting it to a newer version / different fork?
Hopefully we can put some code out when our report is done.
Great. And a few appendices with README.md and release information type of info.
(Sorry for keeping people guessing.)
That is fine. You wrote the mail and so stopped us from doing guess work ;-)
-- -- Yoshiki
--Hannes
On 2/28/13, Yoshiki Ohshima Yoshiki.Ohshima@acm.org wrote:
It is not that we at Viewpoints are trying to be secretive, but we do have a newer system (or systems). Hopefully we can put some code out when our report is done.
(Sorry for keeping people guessing.)
-- -- Yoshiki
Hello Yoshiki,
are there any news on this? Or a screen shot maybe?
Regards Hannes
On Wed, May 1, 2013 at 11:14 AM, H. Hirzel hannes.hirzel@gmail.com wrote:
On 2/28/13, Yoshiki Ohshima Yoshiki.Ohshima@acm.org wrote:
It is not that we at Viewpoints are trying to be secretive, but we do have a newer system (or systems). Hopefully we can put some code out when our report is done.
(Sorry for keeping people guessing.)
-- -- Yoshiki
Hello Yoshiki,
are there any news on this? Or a screen shot maybe?
A new research report is indeed up:
http://www.vpri.org/pdf/m2013003_ksapps.pdf
and the (final) annual report is getting ready but it'll take a bit more time. Unfortunately, neither has the actual details of the new DBJr organization.
I am intending to release the code, once we do necessary clean up.
Thank you for being interested.
-- -- Yoshiki
Thank you, Yoshiki, for the interesting research report 'Making applications in KSWorld' and the equally interesting report of October 2012 'A Report on KScript and KSWorld' http://www.vpri.org/html/writings.php
Chapter 6 of the the October report talks about the hosting environment which happens to be Squeak
<citation> Hosting Environment: Squeak As of this writing, the KSWorld is hosted on top of Squeak Smalltalk [10]. A special Morphic widget called KSMorph is created under RectangleMorph. </citation>
I assume 'KS' in 'KSWorld' just means 'KScript'? And the 'K' should be associated with 'CoffeeScript'?
The scripting language for this Hypercard like environment (but with uniform objects down to the character level) is a variant of CoffeeScript?
--Hannes
On 5/1/13, Yoshiki Ohshima Yoshiki.Ohshima@acm.org wrote:
On Wed, May 1, 2013 at 11:14 AM, H. Hirzel hannes.hirzel@gmail.com wrote:
On 2/28/13, Yoshiki Ohshima Yoshiki.Ohshima@acm.org wrote:
It is not that we at Viewpoints are trying to be secretive, but we do have a newer system (or systems). Hopefully we can put some code out when our report is done.
(Sorry for keeping people guessing.)
-- -- Yoshiki
Hello Yoshiki,
are there any news on this? Or a screen shot maybe?
A new research report is indeed up:
http://www.vpri.org/pdf/m2013003_ksapps.pdf
and the (final) annual report is getting ready but it'll take a bit more time. Unfortunately, neither has the actual details of the new DBJr organization.
I am intending to release the code, once we do necessary clean up.
Thank you for being interested.
-- -- Yoshiki
On Wed, May 1, 2013 at 11:51 AM, H. Hirzel hannes.hirzel@gmail.com wrote:
Thank you, Yoshiki, for the interesting research report 'Making applications in KSWorld' and the equally interesting report of October 2012 'A Report on KScript and KSWorld' http://www.vpri.org/html/writings.php
Chapter 6 of the the October report talks about the hosting environment which happens to be Squeak
<citation> Hosting Environment: Squeak As of this writing, the KSWorld is hosted on top of Squeak Smalltalk [10]. A special Morphic widget called KSMorph is created under RectangleMorph. </citation>
I assume 'KS' in 'KSWorld' just means 'KScript'? And the 'K' should be associated with 'CoffeeScript'?
Well, not all these characters have meanings. It could mean anything.
The scripting language for this Hypercard like environment (but with uniform objects down to the character level) is a variant of CoffeeScript?
KScript, yes. The syntax of it has been changed since October. So looking at the new one would be good. In the new document, Section 7.1 discusses the "Document Model" and Section 8.9 says it took about 200 lines to implement.
-- -- Yoshiki
On 27-02-2013, at 6:04 PM, H. Hirzel hannes.hirzel@gmail.com wrote:
Thank you Casey and Dave for the references to get a working DynaBook Junior image.
It works fine.
A thing which might not be obvious it to download the example stack http://tinlizzie.org/lesserphic2/aboutDBJr%205-Aug.morph in addition, locate it with the file browser and then choose 'load as morph'
It sorta works on my Pi; rather slowly but that seems to be an oldish looking image. It also fail when trying to dreatee a log file of some sort because the ancient AcornFileDirectory code isn't handling /Users/tim/Dowlnloads very well. Not too surprisingly, to be honest. Quite where it got the idea that the default directory is a UnixFileDirectory on /Users/tim/Downloads I'm not sure. Oh, wait; something in the FilreDirectroy startUp failed because of an invalid utf8 input detected. Holey Kao - that UTF8TExtConverter>nextFromStream: method is scary.
But bugs can be fixed and this would make a neat project to extend and improve and make into a usable application.
tim -- tim Rowledge; tim@rowledge.org; http://www.rowledge.org/tim Strange OpCodes: LTT: Lose Timing Track
Optimization was not a goal of the research. I did notice that the image was a bit slow even on my Mac. I didn't bother to profile.
On Thu, Feb 28, 2013 at 6:52 PM, tim Rowledge tim@rowledge.org wrote:
On 27-02-2013, at 6:04 PM, H. Hirzel hannes.hirzel@gmail.com wrote:
Thank you Casey and Dave for the references to get a working DynaBook Junior image.
It works fine.
A thing which might not be obvious it to download the example stack http://tinlizzie.org/lesserphic2/aboutDBJr%205-Aug.morph in addition, locate it with the file browser and then choose 'load as
morph'
It sorta works on my Pi; rather slowly but that seems to be an oldish looking image. It also fail when trying to dreatee a log file of some sort because the ancient AcornFileDirectory code isn't handling /Users/tim/Dowlnloads very well. Not too surprisingly, to be honest. Quite where it got the idea that the default directory is a UnixFileDirectory on /Users/tim/Downloads I'm not sure. Oh, wait; something in the FilreDirectroy startUp failed because of an invalid utf8 input detected. Holey Kao - that UTF8TExtConverter>nextFromStream: method is scary.
But bugs can be fixed and this would make a neat project to extend and improve and make into a usable application.
tim
tim Rowledge; tim@rowledge.org; http://www.rowledge.org/tim Strange OpCodes: LTT: Lose Timing Track
On Feb 27, 2013, at 4:31 PM, Casey Ransberger casey.obrien.r@gmail.com wrote:
Ted Kaehler has done two implementations of DBJr AFAIK, one in Squeak and one in Javascript.
Here's the javascript version (for Lively Kernel)
http://forum.world.st/Try-DBJr-HyperCard-like-stacks-in-Lively-Kernel-td4599...
I had to dig to find the Moshi image, but here's an actual Squeak image -- and I think this was the right image -- note that this was a prototype and that some things (e.g. saving) aren't implemented. Also, I can't get it to open anymore, but I think that might be because I don't have a non-Cog VM now? Not sure which .sources file you want but I'm betting it's V3. The file of interest is "Text Field for LObject.zip"
http://tinlizzie.org/lesserphic2/
I was really fascinated with both the approach and the presentation he did here. Active essays are cool.
Meantime I have to figure out why I can't open the image. I thought Cog could open older images, but interpreters can't open Cog images. Hmmm.
Depends on how old. Cog (& StackVM) can only run closure images. Eliot (phone)
On Wed, Feb 27, 2013 at 10:12 AM, karl ramberg karlramberg@gmail.com wrote:
Etoys image also have a stack morph
Karl
On Wed, Feb 27, 2013 at 11:24 AM, Jecel Assumpcao Jr. jecel@merlintec.com wrote:
While the DynaBookJr (I always misread "DBJr" as "DataBase Jr") is indeed a good modern replacement for HyperCard, isn't the normal BookMorph already available in Squeak a reasonable approximation?
-- Jecel
-- Casey Ransberger
On 2/28/13 10:39 AM, "Eliot Miranda" eliot.miranda@gmail.com wrote:
Meantime I have to figure out why I can't open the image. I thought Cog could open older images, but interpreters can't open Cog images. Hmmm.
Squeak 4.2.5beta1U opens it fine in Mac 10.6.8
Edgar
Below.
On Feb 28, 2013, at 5:39 AM, Eliot Miranda eliot.miranda@gmail.com wrote:
On Feb 27, 2013, at 4:31 PM, Casey Ransberger casey.obrien.r@gmail.com wrote:
Ted Kaehler has done two implementations of DBJr AFAIK, one in Squeak and one in Javascript.
Here's the javascript version (for Lively Kernel)
http://forum.world.st/Try-DBJr-HyperCard-like-stacks-in-Lively-Kernel-td4599...
I had to dig to find the Moshi image, but here's an actual Squeak image -- and I think this was the right image -- note that this was a prototype and that some things (e.g. saving) aren't implemented. Also, I can't get it to open anymore, but I think that might be because I don't have a non-Cog VM now? Not sure which .sources file you want but I'm betting it's V3. The file of interest is "Text Field for LObject.zip"
http://tinlizzie.org/lesserphic2/
I was really fascinated with both the approach and the presentation he did here. Active essays are cool.
Meantime I have to figure out why I can't open the image. I thought Cog could open older images, but interpreters can't open Cog images. Hmmm.
Depends on how old. Cog (& StackVM) can only run closure images. Eliot (phone)
I bet that's exactly it; I think the image is based on an early 3x which is well before you introduced closure semantics to the existing not-quite-closure blocks.
I am having too much fun with that image and Morph - thanks Casey and Hannes (and Ted of course)
I bet that's exactly it; I think the image is based on an early 3x which is well before you introduced closure semantics to the existing not-quite-closure blocks.
It is based on a 2007 OLPC image. If I am not mistaken, that is the image that became Etoys 3.0
I think it isn't quite the Moshi image that the demo card stack was intended for, but it mostly works well in an interpreter VM.
Ideally the image would have a parts bin containing back and forward buttons, but I was able to copy ThreePhaseButtonMorphs from another substack to make my own stack.
I love the script browser, and the reusable message list browser (which reminds me of Amber and Newspeak browsers). (Are there any reusable browsers in Squeak 4.4 ?) I also love the cmd-f cmd-g search through the text in the stack.
There is too much meta programming behind it for me to quickly grok. I couldn't see how the page turning buttons worked, as it seemed to that the #previousPage:mouseUp: method was the card's event handling script for the page turning button, but it didn't contain any code, just a comment, yet the button worked. I know that in Smalltalk "everything happens somewhere else" but this time I couldn't find the 'somewhere else'. Could someone explain it for me?
I think this could be the basis of a nicer way to have a high ceiling scripting environment than the 'show as script' button in Etoys, but I wait to see what else VPRI has up its collective sleeve :)
Have fun! David
On 28 February 2013 17:55, David Corking lists@dcorking.com wrote:
I am having too much fun with that image and Morph - thanks Casey and Hannes (and Ted of course)
I bet that's exactly it; I think the image is based on an early 3x which is well before you introduced closure semantics to the existing not-quite-closure blocks.
It is based on a 2007 OLPC image. If I am not mistaken, that is the image that became Etoys 3.0
I think it isn't quite the Moshi image that the demo card stack was intended for, but it mostly works well in an interpreter VM.
Ideally the image would have a parts bin containing back and forward buttons, but I was able to copy ThreePhaseButtonMorphs from another substack to make my own stack.
I love the script browser, and the reusable message list browser (which reminds me of Amber and Newspeak browsers). (Are there any reusable browsers in Squeak 4.4 ?) I also love the cmd-f cmd-g search through the text in the stack.
What does "reusable" mean here? If you mean that navigating to something doesn't spawn a whole new window, then we _sometimes_ have reusable browsers. For instance, in a Browser, selecting a class name and pressing <alt>-<shift>-b will, within the same Browser, jump to the new class. There are some ways of doing something similar with MessageLists, but I forget how to trigger the behaviour. I _think_ it's when you view the senders of some message in the currently viewed method, you get the senders spliced into the viewed list, indented a bit.
frank
Hi
I _think_ it's when you view the senders of some message in the currently viewed method, you get the senders spliced into the viewed list, indented a bit.
to get this behaviour the Preference "trace messages" needs to be enabled. checking.... Right now this only works for implementors, senders spawns a new window. 4.4 #12332
Cheers
Herbert
On 28 February 2013 18:46, Herbert König herbertkoenig@gmx.net wrote:
Hi
I _think_ it's when you view the senders of some message in the currently viewed method, you get the senders spliced into the viewed list, indented a bit.
to get this behaviour the Preference "trace messages" needs to be enabled. checking.... Right now this only works for implementors, senders spawns a new window. 4.4 #12332
Mm, that's actually a 4.5 image - 4.4-12327 was the last 4.4 image :)
But thanks for the extra info. I'll have to look at the preference later on.
frank
squeak-dev@lists.squeakfoundation.org
-
Casey Ransberger -
David Corking -
David T. Lewis -
Edgar J. De Cleene -
Eliot Miranda -
Frank Shearar -
H. Hirzel -
Herbert König -
karl ramberg -
tim Rowledge -
Yoshiki Ohshima