Squeak-dev May 2001

squeak-dev@lists.squeakfoundation.org

143 participants
597 discussions

Re: What happened to the conversion method prototypes?
by Jesse Welton 01 May '01

01 May '01

Ned Konz wrote: > > Using CS 3960, I am filing out a change set that changes the shape of a > class. I have my Preferences set to notify me of conversion methods. > I am prompted for conversion methods when I try to file out. > However, when I choose "Write a conversion method by editing a prototype" I > don't get a prototype to edit. It just tells me to "remember to file out > again after editing these methods". > > Which methods? > > Didn't there used to be a method generated with the proper signature that > then could be edited? The old conversion methods have been completely supplanted by #convertToCurrentVersion:refStream:. Comments explaining the change are available in SmartRefStream>>aComment. Unfortunately, there are still a few references the the old-style conversion mechanism, like those hanging around in the ChangeSorter. -Jesse

1 0

Re: Sharing with you my enthousiasm
by Richard A. O'Keefe 01 May '01

01 May '01

Bob Arning <arning(a)charm.net> wrote: What I like about the different colors: - I can quickly find windows I am looking for even when only a corner is visible under a pile of other stuff. - It's a quick visual clue of what this window is about. Others have said the same. But the same benefits could be obtained from coloured borders, especially if there were a new kind of corner. The new kind of corner is rounded on the inside like rounded corners, square on the outside like square corners. This gives you a little "gusset" of border colour, quite noticeable. "Pretty-print with color" looks rather ugh against a hospital-green background.

2 1

[BUG] newest updates to 3.1alpha seem to break all browsers if whisker 0.92 installed
by txporter＠mindspring.com 01 May '01

01 May '01

here is the debug log zip file produced while filing in whisker-0.92 into 3.1alpha after the newest updates. Sorry for reporting this in drips and drabs. Tom Porter

2 1

Comments on new changes to Preferences
by Ned Konz 01 May '01

01 May '01

I like the idea of notification on a preference change. In fact, I posted a change set some time ago on the list that would do this. But I think there are some problems with Preference as it stands in 3960: * There is a single, hard-wired informee. Why only one? This makes it hard for user code, or added code, or whatever to hear about appropriate preferences changes. * Why duplicate already-existing notification schemes? There are already two publish/subscribe mechanisms in the image: the stock MVC as well as the event stuff. I'd just add a myEvents instance variable and accessor and use the stock event mechanism to provide notification. This would allow multiple objects to be notified on a preference change, and may reduce the amount of code in Preference. -- Ned Konz currently: Stanwood, WA email: ned(a)bike-nomad.com homepage: http://bike-nomad.com

1 0

Re: [updates] 36 for 3.1alpha
by Dan Ingalls 01 May '01

01 May '01

I wrote: >> Herewith a bunch of updates from SqC over the last month... Henrik Gedenryd <Henrik.Gedenryd(a)lucs.lu.se> replied privately... >Note that I didn't receive the note about 3911-3923. Sending it out again >might probably be a good idea, I assume no one got it. This is the identified >problem with the listserv dropping some postings. In this particular case, it's not. I never sent such a message. But for completeness, here is the note about 3911-3923: Updates 3911-3923 are missing from the 3.1alpha update stream. We have moved to a system where the internal and external update streams share the same files, and only the indices (updates.list) are different. This allows us to do experimental stuff without endangering the whole community, but to release updates that have been tested by just changing the index file. At the time we made this change, the numbering of the internal and external streams differed by 13, with external 3910 being the same as internal 3923. Therefore, when I released the last set of updates, 13 sequence numbers were skipped in order to reconcile the two numbering systems. - Dan

1 0

RE: Squeak for 3-year-olds
by Pennell, David 01 May '01

01 May '01

What student to teacher ratio is needed for this style to be successful? -david > -----Original Message----- > From: Alan Kay [mailto:Alan.Kay@disney.com] > Sent: Monday, April 30, 2001 1:16 AM > To: squeak(a)cs.uiuc.edu > Subject: Re: Squeak for 3-year-olds > > > Oh, but you are missing the entire point of that part of that > project. It was specifically made to be difficult to control so the > kids would say "wow, this is difficult to control", and we say "yeah, > that's because you made a race car. There are no gears between the > steering wheel and the front wheels of your car". And they say > "what's a gear?" and we say, "see that little triangle after the word > 'heading', click on it". They do and then for the first time in their > lives, they get to see why division might be useful by dividing the > number coming out of 'heading' by 3 or 4". > > The whole idea of good learning environments is to avoid as many > gratuitous difficulties as possible, but to specifically introduce > just the difficulties that will motivate appropriate learning. > > This is why we have worked with literally thousands of children now > since 1973 or so. BTW, the "drive a car" etoy is specifically aimed > at children from about 8 years old and older. There are numerous > projects suitable for younger children, but they have different aims. > Each developmental stage of children allows them to do different > kinds of thinking, and you have to take that into account when you > design etoys for them (and when you bring children to them). There > are a number of examples on Squeakland.org, and there will be many > more over the next few weeks and months. > > For those of you who have very young children, you are much better > off having them start to learn to build things with their hands. Most > computers are not equipped with enough tactile feedback to be great > learning environments for very young children. > > A great book to read for those who are interested is "Towards a > Theory of Instruction" by Jerome Bruner. It doesn't talk about > computers but it is one of the best books about designing good > educational environments for children of various ages. Last year I > wrote a chapter for a book about kindergarten that discusses the > suitability of computers for 3 and 4 year olds. This will also go > online sometime this summer. > > Cheers, > > Alan > > ----- > > At 4:51 AM +0100 4/30/01, John Hinsley wrote: > >Lex Spoon wrote: > >> > >> Steve Wart <thecows(a)home.com> wrote: > >> > Had an interesting session with my son today. Played > around in the > >> > scripting world for over an hour - I think that's the > longest he's ever > >> > sat still at the computer. > >> > > >> > He loves the car, but steering is pretty hopeless (I > can't get the hang > >> > of it either :). He liked it when I changed the colors. > Usually he wants > >> > to do things himself, but I think he realized changing > colors was too > >> > finicky for him. The color picker is pretty though. It > would be nice to > >> > see more like this. > >> > >> Have you tried "gearing down" the stearing wheel: > dividing the amount in > >> the "turn by" command by 3 (or so) ? It makes it much easier. :) > > > >Yes! I think only an 8 year old with the reactions and skill > of the late > >great Ayrton Senna could keep the car on the track with its original > >settings. > > > >I think one of the greatest things we can do to improve our programs > >(and our programming skills) might be to have young children > (and aged > >relatives) act as human programming interfaces. I'm sure most adults > >(and most "real" programmers) get too carried away with the beauty of > >the code to adequately consider the underlying metaphors: we > really need > >"untainted" brains to show us the way. (Not, incidentally, that I > >consider myself a "real" programmer -- an adult, maybe.) > > > >Maybe Steve could try my Rolodex Tutorial out on his son in > 2 or 3 years > >time? > > > >Cheers > > > >John > >> > >> Overall, thanks for sharing your experiences -- I've found them an > >> interesting read! > >> > >> -Lex > > > >-- > >************************************************************* > ***************** > >Marx: "Why do Anarchists only drink herbal tea?" > >Proudhon: "Because all proper tea is theft." > >************************************************************* > ***************** >

2 1

[BUG] cannot load BasicTextExample.ggd in GENIE
by Jörn Eyrich 01 May '01

01 May '01

I just downloaded the BasicTextExample gesture dictionary from Nathanael's web site, tried to load it into Squeak (#3960), and got a walkback. SqueakDebug.log is attached. Seems the StandardFileStream underlying the ReferenceStream used has a fileID of nil. If I do a "ReferenceStream oldFileNamed: <path>" in a workspace, the fileID's there. If I try to slip in a manually created referenceStream in the debugger and proceed, it runs without error but nothing happens... Joern

1 0

Re: [BUG] newest updates to 3.1alpha seem to break all browsers if whisker 0.92 installed
by Bob Arning 01 May '01

01 May '01

On Tue, 01 May 2001 09:10:00 -0400 txporter(a)mindspring.com wrote: >BTW, is there a way to capture the top pane of the debug window/error handler >window to a file, since my mail debug report seems not to work as I am behind >a firewall? Tom, The top pane is essentially what is written to your default directory as SqueakDebug.log. Cheers, Bob

1 0

Re: newest alpha [BUG]updates seem to break all browsers if Whisker 0.92 installed
by Karl Ramberg 01 May '01

01 May '01

txporter(a)mindspring.com wrote: > > Hello folks, > > I upgraded to the newest alpha updates on both Linux and WIn98, starting from both a stock 3.0final and a 3.1alpha3910b set of changes and images. > > All combinations seem to break all browsers if I have Whisker 0.92 installed before the updates. If I install Whisker afterwards, I get the same debug error during the update: > > The Debug reports: > False(Object)>>doesNotUnderstand: > Preferences class>>addNonExistingPreferences:catagory:default:ballonHelp: > WhiskerBrowser class>>initializePreferences > WhiskerBrowser class>>initialize > UndefinedObject>>DoIt > > BTW, is there a way to capture the top pane of the debug window/error handler window to a file, since my mail debug report seems not to work as I am behind a firewall? > > Thanks, > > Tom P. Control click (blue ?) the predebug or debugger window, select debug -> inspect model. The instance variable contextStack seem to hold the info you want. Karl ps Didn't the debugger use to have a file-out-stack-to-disk menu option ?

1 0

RE: Different Forks for Different Folks
by Jarvis, Robert P. (Contingent) 01 May '01

01 May '01

> From: Richard A. O'Keefe [mailto:ok@atlas.otago.ac.nz] > Sometimes a computer is just an executive status symbol. Too true. One of my favorites was something I watched happen every day for over two years: At a former employer I worked in the Information System Department (ISD). The VP of that department was proud of how computer-and-business-saavy he was, and how *his* choices for computers were The Best! His choice? Buy IBM! As long as it was made by Big Blue, he'd buy it! We had an IBM Token-Ring network, with IBM servers, and loads-n-loads of PS/2's on desktops. It didn't matter that they cost twice what compatible machines from other vendors cost - IBM good! Other BA-A-A-AD!! The problem was that said VP didn't *personally* like IBM machines - he wanted a Mac. So on his desktop he had the only Mac in the company. Of course, it couldn't talk to the IBM servers, but that was OK with him. The kicker to all this is that in order to get his email he had his secretary print out all his emails every day from HER computer. He'd mark up the email, hand-writing replies on the printouts she gave him. Then he'd give this pile of emails back to her, she'd enter the replies to the emails and send them, schedule meetings, etc. And this bozo couldn't figure out why people hated the email system. His attitude was "*I* never seem to have a problem with it - it must be user error!". I understand that he was eventually walked to the door by a nice man from the security department... Bob Jarvis Compuware @ Timken

1 0

Jump to page: