[squeak-dev] Etoys developer chat log

Bert Freudenberg bert at freudenbergs.de
Tue Apr 27 00:09:45 UTC 2010


Hi all,

we had another nice chat today. Topics included

* what does @HELP achieve in the #etoys channel
* how to fix up the ancestry of Monticello packages we create
* and what to do about remaining packaging differences
* how to get more reliable #etoys logs (perhaps same as #squeak?)
* how to call next release? The update stream uses "etoys4.1" for now, the bug tracker "etoys 2010". "Etoys 5" would be an option too. Discussion welcome.
* lack of documentation on extending Etoys, or even of its overall design
* progress on User Manual
* discussing Ricardo's GSoC project (Randy will set up a Google Group)
* Maybe move all Squeakland mailing lists and forums to Google? would get rid of the forum spam issues

Next chat will be next Monday, same time and place. But the channel is open 24/7, you can try to catch us any time (and use @ALL or @HELP to get our attention).

Full log below. Once we have regular logging I won't have to spam the list with this anymore ...

- Bert -

<matmo> hi all
<ritaf> hi
<bertf> hi folks
--> hilaire (~hilaire at bon74-1-88-184-136-97.fbx.proxad.net) has joined #etoys
<Guest85605> hi
<bertf> Randy: try a longer nick name, append soem characters
<jecel> hello
<hilaire> hello
<bertf> "randy" is probably registered to another person
<-> Guest85605 is now known as randycat
<randycat> that worked
<bertf> randycat: nice :) 
<bertf> cool, so let's start
<bertf> did anyone try the updates?
<randycat> The official notice of Richo's google award will be any minute - I hope nothing changed
<bertf> yes, that's awesome. I hear there was quite some discussion at ESUG which projects to support, right?
<randycat> it was tense
<bertf> but he's doing great, good choice
<kencausey> bertf: ah, good call
<bertf> does anyone have anything specific we need to talk about?
--> squeaker761 (~40379e74 at skylab.immuexa.com) has joined #etoys
<bertf> kencausey: you mean the web irc thing or what? "maintain"
<kencausey> yes, the web chat client
<-> squeaker761 is now known as bander
<bertf> ah, bander, hi
<bander> hi
<bertf> kencausey: well it;s working for now
<bander> just listening today :-)
<kencausey> true
<kencausey> bander: good luck with that. ;)
<bertf> kencausey: also, tim said he will continue to maintain the servers for now
<kencausey> sure
<kencausey> What does the @HELP prefix actually do?
<bertf> it makes my icon bounce ;)
<kencausey> For example, does the bot suggest alternate help sources?  Unfortunately people that use it rarely get answered.
<bertf> we all entered that as custom-highlight to our clients
<bertf> nothing to do with the bot
<kencausey> ah, I see, not bot related then
<bertf> bander: I did have a question for you
<bertf> I published a recategorization to the update stream
<bander> sure
<bertf> but comparing teh packages now to the trunk ones there are still a lot of differences
<bertf> like, where did flash go?
<bander> No clue. 3.9 road kill?
<bander> We should restore it
<bertf> do you think we should just make our own packages now? without ancestry?
<bertf> I thought about "adopting" some existing packages
<bander> How about adopting an older version?
<bertf> :)
<bertf> I thought about doing that with the earlies ones I can find
<bertf> in the 39a repo
<bander> The alternative is to move it into separate packages
<bertf> flash you mean
<bander> Yes
<kencausey> seems like the better plan to me
--> BackOrder (~4ad28298 at skylab.immuexa.com) has joined #etoys
<bander> I can look at that if it helps
<kencausey> it's time to transition to the core releases plus a basic built from the core with packages added
<kencausey> then etoys can start with the core and go from there
<bertf> yeah. well. I was so happy to have a reasonably categorized version today at 4 am that I published it
<kencausey> sure
<bertf> kencausey: yes that is the long-tem plan
<bertf> bander: that would be excellent
-*- saijanai_ unlurks and starts reading backlog
<bander> I'll see if I can find some time this week
<bertf> at least now we have some good tools to work with
-*- kencausey resists talking about non-etoys issues
<bertf> good boy ;)
<bertf> BackOrder: hi - do we know you?
<bander> IanT perhaps?
<bertf> no idea, that nick doesn't ring a bell
<bertf> in any case I want this to be done asap so we can start integrating the actual patches lingering on the tracker
<bander> Actually, I think we should keep the current version of the Flash package in the Etoys repo for the time being
<bertf> oh definitely
<bertf> I want all packages to at least have a copy in the etoys repo
<bander> Sounds reasonable
<bertf> so the etoys release will only contain packages from there
<bertf> which might have a master elsewhere but still
<ritaf> bertf: in Portugal, they are using an old Etoys version, 3.9 but they want move to Etoys 4.0. They say they did some work on the older version and the biggest problem is to do it all again for the new Etoys. What would you recommend for them to so? And especially, is the new update stream helping there?
<randycat> will any of the changes you are planning effect what Ricardo Moran creates for GSoC?
<bertf> ritaf: depends on what work they did
<ritaf> bertf: I don't know the details, I will ask.
<ritaf> bertf: maybe I shold just invite Luis for the next meeting
<bander> ritaf: the best thing to do would be to get them involved in the dev process
<bertf> randycat: shouldn't. we're not planning to do any big changes for this summer's release
<bertf> randycat: what we're talking about is just a different way to organize the code, not actually modifying teh code itself
<bertf> not in a big way anyhow
<bander> sorry folks, gotta run - eliot's dragging me out for lunch
--> karlram (~51aa9a7e at skylab.immuexa.com) has joined #etoys
<karlram> hello
<bertf> bander: thanks for stopping by
<ritaf> hi karlram
<bertf> karlram: hi. awesome :)
<karlram> hi
<bertf> you just missed andreas, we were talking about the new packages
<karlram> ah
<bertf> kencausey: do you have an idea for better loggin this channel?
<BackOrder> Hi guys! Yes, I am Ian Trudel. :)
<karlram> will there be logs ?
<karlram> hi
<bertf> karlram: sure
<karlram> cool
<kencausey> bertf: well, we might be able to invite clog as long as you want total logging and not something directly controllable.
<bertf> karlram: I just emailed you the backlog
<karlram> thanks
<kencausey> that woudl put logs at tunes.org/~nef/logs/
<bertf> kencausey: that would be awesome. just like for #squeak right?
<kencausey> BackOrder: Hi Ian
<kencausey> bertf: yes, trick is remembering who to contact
<BackOrder> I decided to take a peek in this chatroom after Bert's invitation on Squeak-Dev.
<kencausey> Brian Rice might know, shall I look into it?
<bertf> kencausey: please, yes. thanks
<bertf> I removed the broken "sm" log link from the topic today
<bertf> ok, logs, there was somethign else I wanted to ask ...
<bertf> ah, yes. karlram: I notiuced you added changesets to the tracker  but did not mark teh issues "resolved"
<bertf> are you happy with the way the tracker works now?
<bertf> or with the way we're using it?
<karlram> the process is a little unclear
<bertf> well we're abusing the system
<bertf> let me dig out the link
<karlram> i'm not sure what the diffrent thing stand for
<bertf> http://wiki.squeakland.org/display/sq/Process
<bertf> well it's unfortunate that they do not mean what they usually mean
<karlram> ok, i'll read that
<-- BackOrder (~4ad28298 at skylab.immuexa.com) has quit (Quit: CGI:IRC (Ping timeout))
<bertf> that's how we tried to operate, though it wasn't working too well I think
<karlram> but etoys 5 is our next release ?
<karlram> next target
<bertf> to me it feels more liek a minor release
--> BackOrder (~ian at 130-152.sh.cgocable.ca) has joined #etoys
<matmo> got to go soon. Copy of my notes re AV on Ubuntu Karmic (will be on swiki alter in the week): http://pastebin.com/J5E1jgK1
<bertf> matmo: thanks !
<karlram> it minor in some ways
<karlram> major if we get it up to squeak 4.1
<bertf> no chance
<karlram> :-)
<bertf> there are way too many issues to make that leap that soon :)
<karlram> ok
<saijanai_> question: is anyone working on how to make it easier to extend etoys? I'm just trying to nibble around the edges and it looks somewhat daunting
<bertf> so the development version we're working on now I'd like to call 4.1. 
<jecel> And I would say there is work to be done on the Squeak side as well, so it would be better to aim for Squeak 4.2 or something like that
<bertf> saijanai_: no one really. help welcome :)
<saijanai_> well, I'd like to, but I'm slow at digesting, so no promises
<karlram> name the issues you have with extending ?
<bertf> sure. but let us know where you feel you're hitting a wall
<saijanai_> getting started is the first issue. NOt sure where to start
<bertf> almost everythignis possible, but seldomly obvious and sometimes not trivial
<karlram> there should be some example project around....
<bertf> saijanai_: you wanted to add your own tiles IIRC right?
<saijanai_> right. I think you were the one who sent me info on the GSoC project? or was that someone else?
<bertf> not sure
<saijanai_> if youre not the one doing the work ont eh GSoC then it was someone else :-)
<bertf> the work is dan by Ricardo
<bertf> s/dan/done/
<saijanai_>  ah, OK, so Ricardo is the one who sent me the info
<randycat>  Ricardo is posting his progress on SqueakSource
<bertf> there is a lot of oral history around squeak ...
<bertf> some things seem to never have been  written up
<kencausey> many things
<saijanai_> and by the time someone tries, they've changed beyond recognition, or such is my impression
<bertf> which is both good and bad - good because the system is evolving, bad because it's harder to catch up
<matmo> how's the etoys book coming along?
<bertf> the ref manual?
<matmo> y
<bertf> ritaf would know
<bertf> last time I checked it looked pretty good
<karlram> bertf:true
<bertf> but it is a user's manual, not a developer's one
<ritaf> we have most of the text done, now one issue is formatting, so that it is better to read and then reviewing.
<bertf> kencausey: do you know which paper outlined the general principles of Etoys?
<bertf> like, the fat object model etc
<saijanai_> yeah, I've got many plans concerning etoys. The Scratch in Second Life could be extended in many ways (direct manipulation of prims via packet injection instead of LSL scripts, for example)
<bertf> saijanai_: basically, in etoys you only have a a single kind of object. no classes/inheritance etc.
<bertf> that object is called "Player"
<saijanai_> right. Based off of the Self design, with much kludging to make it work in Squeak
<bertf> everything you want to be scriptable, you must add to Player
<saijanai_> OK, sorta got that from what Ricardo told me
<bertf> no, Morphic comes from Self, but Etoys is yet another layer
<jecel> There is an indirection compared to Self - a Player may have several Costumes
<saijanai_> ah, OK. my bad
<bertf> right, what  jecel said
<bertf> so when you make an Etoys script it looks liek you are working with a morph, but in fact you are working with a Player
<randycat> bertf: I want to start a place where teachers and developers can download and discuss Ricardo's tools. What do you think about Google Groups?
<kencausey> bertf: no, jecel?
-*- kencausey checks new blue book
<kencausey> hmm, too old perhapos
<bertf> randycat: hmm. good question
<saijanai_> I'm willing to participate in such a discussion groiup.
<jecel> kencausey: I have never read a paper about the Etoys architecture, but did see one about Tweak which has some features in common
<karlram> I posted a simple changeset that adds camera tracking to Etoys on Etoys-dev
<bertf> randycat: you think the squeakland list would be inappropriate?
<kencausey> jecel: yeah, I'm not sure what paper bertf is refering too
<bertf> kencausey: maybe the one in my head ;)
<karlram> must be some paper from Alan
<jecel> is the squeakland list still linked to some web forums?
<bertf> we'd have to ask Scott
<karlram> he has some stuff on the vpri.org
<bertf> jecel: yes, though they are just spam repositories
<randycat> I though Google Groups might be more approachable by teachers and easier to archive.
<jecel> bertf: I was afraid of that
<bertf> randycat: then we have a problem with the squeakland list, which is supposed to be approachable by teachers
<kencausey> randycat: I like running our own tools whenever possible, but google groups seems reasonable for this case.
<saijanai_> http://www.pcs.cnu.edu/~rcaton/ESUG/ESUG.html   is the link Ricardo sent me
<karlram> saianai> ask question on etoys-dev mail list when you run into issus
<karlram> issues
<saijanai_> that's how I got the link :-)
<bertf> randycat: how about moving the whole squeakland list to google groups? (just speculating, not actually proposing it yet)
<bertf> ritaf: any comment? ^^^
<randycat> saijanai: I set u that link, but it is just a web page
<saijanai_> Ricardo?
<saijanai_> thanks
<bertf> saijanai_: Ricardo is doing the work, and Randy is mentoring him
<ritaf> bertf: let me think about that. I don't know yet what that would mean.
<saijanai_> ah, OK. Misread who sent what to whom then
<saijanai_> thanks anyways :-)
<bertf> ritaf: well for one it would mean we don't have to maintain the lists on our own
<randycat> bertf: can the squeakland list do all that Google Groups can in one environment?
<bertf> also, google is good about filtering spam
<bertf> randycat: I have no idea, I'm not too familiar with google groups
<saijanai_> google can require membership so you can just ban spammers
<saijanai_> I think
<kencausey> bertf: picking your battles seems like a good idea to me, running your own listserv is a bit of time sink
<karlram> bertf is there much spam on squeakland lists ?
<jecel> The Cobalt people use google groups and there is a new Smalltalk Research list there
<bertf> karlram: no, but on the forums there is
<karlram> ok, i never use the frums
<karlram> forums....
<bertf> does google groups have a forums-like view?
--> scottwal (~scottwal at 99-40-5-235.lightspeed.sntcca.sbcglobal.net) has joined #etoys
<bertf> oh, hi scottwal :)
<karlram> hi
<scottwal> hi!  [only here for a couple of moments -- have to head back out :-( ]
<bertf> we were wondering if there was a paper about the general architecture of Etoys?
<scottwal> Almost surely not...
<jecel> Yes, you can use google groups as a web forum (I do for over half of the list there in which I participate)
<bertf> or any documentation really
<randycat> bertf: Google groups has custom web pages, discussion threads, upload and download files - all in one package
<bertf> jecel: that sounds like a win
<kencausey> bertf: google groups can be used entirely through the web, so I think it supplants the idea of a web forum
<bertf> right. I'd love to have something I can use either as mailign list or as forum
<karlram> I have to read books to my daugther...  Bye
<ritaf> that sounds nice to me.
<bertf> karlram: thanks
<scottwal> and I have to pick up my daughter from the airport...  Bye
<-- karlram (~51aa9a7e at skylab.immuexa.com) has quit (Quit: CGI:IRC (EOF))
<jecel> I participate in an equal number of lists at yahoo groups and it is pretty reasonable too
<bertf> thanks to you too scott3
<ritaf> bye scott!
<-- scottwal (~scottwal at 99-40-5-235.lightspeed.sntcca.sbcglobal.net) has quit (Client Quit)
<bertf> jecel do you see any advantage to yahoo?
<bertf> (I've never seen it I guess)
<kencausey> Google seems like a better choice from a stability standpoint, at least this year
<randycat> bertf: ESUG uses Google Groups
<bertf> randycat: seems like a win to me
<jecel> bertf: I started the self-interest list on eGroups and that got bought by Yahoo. I think the two alternatives are roughly equivalent
<bertf> okay. Who volunteers to set this up?
<randycat> bertf: I will
<bertf> randycat: for the gsoc you mean
<randycat> yes
<matmo> bye all
<bertf> okay. I meant moving over the squeakland list and archives etc.
<randycat> bertf: we could try it for gsoc and see how we like it
<bertf> (assuming archives can be imported)
<bertf> randycat: good idea
<jecel> One great thing about the yahoo groups (google might have this too, but I am not an administrator for any list there) is that you can moderate just the first email from each new user, and once you see they are not spammers you let them post without moderation
<bertf> jecel: that's a feature we wanted to have for a long time
<jecel> Yeah, that reduces my workload to one email to look at every two weeks
<bertf> ritaf: can you ask marta how she would feel about moving the america-latina list?
<bertf> these are the lists we currently have http://lists.squeakland.org/mailman/listinfo
<ritaf> bertf: ok
<bertf> well, excellent. thanks guys - anything else we would want to discuss today?
<saijanai_> Just checked with John Dougan. Google has the same feature regarding moderation
<randycat> jecel: I just checked - there are management tools for postings
<bertf> saijanai_: thanks
<bertf> matmo: thanks!
<jecel> randycat: great!
<randycat> bertf: should I set up a Goggle Group for GSoC - once Ricardo's award is official?
<bertf> randycat: yes please. sounds like a good experiment
<bertf> I need someone to lead this discussion next week - I'll be in the air.
<randycat> OK - goodbye everyone and thanks
<bertf> thanks randycat. bye
<-- randycat (~cebcff48 at skylab.immuexa.com) has quit (Quit: CGI:IRC)
<bertf> or just come here next Monday, same place and time, and have a nice chat :)
<bertf> thanks everyone!
<saijanai_> cheers
<BackOrder> So long, folks. :)


More information about the Squeak-dev mailing list