Squeak-dev January 2004

squeak-dev@lists.squeakfoundation.org

156 participants
396 discussions

BitBlt>>pixelAt: does 0-based indexing?
by Martin Kuball 07 Jan '04

07 Jan '04

Hi! I'm trying to do some image processing in Squeak. I load an image into a Form and than use a BitBlt peeker to examine pixel values. After some hacking I noticed that the indexing is not 1-based as in other Smalltalk methods. Is this on purpose or a bug? Martin

12 12

[ENH][3.7a] ChangeListEnh-nk
by ned＠squeakland.org 07 Jan '04

07 Jan '04

from preamble: "Change Set: ChangeListEnh-nk Date: 7 January 2004 Author: Ned Konz Some enhancements for the ChangeList and VersionsBrowser: - compare (one or several) change(s) in the ChangeList with the current and all the historical versions in the image. - ignore LF characters when diffing method text. May want this to be a preference, I'm not sure. - ignore whitespace changes when looking for unchanged methods. May want this to be a preference, I'm not sure. "!

1 0

Re: [Squeakland] Q: Leverage Overlapping Player to Etoys?
by Ned Konz 07 Jan '04

07 Jan '04

On Wednesday 07 January 2004 4:19 am, Scott Wallace wrote: > (1) The beginnings of just such a thing are already present in the > Squeakland image. Look in the 'collections' section of a viewer of > any playfield-like object, and you'll see the following tile: Ah, I'd not seen this. > (2) With a few small changes, one can arrange matters such that a > player-reference tile (the kind you tear off via the orange > halo-handle) will forward any putative message dispatched to it via > #tellAllContents: on to the tile's referent. This allows > player-reference tiles to function as "aliases" for the purpose of > scripting, as Alan suggests in the excerpt quoted below. > > With this change, one can use a Holder (or any other PasteUpMorph) as > a scripting iterator. Yes, this makes sense to me. I'd like to go a bit further with it and let player-valued phrases be used (i.e. 'Fred's playerAtCursor'), as well. After the 10th I'll send out a CS that lets us deal with player-valued phrases as first-class objects. > This is just an existence proof. To become useful, I think we will > want to extend the etoy vocabulary to allow programmatic creation and > deletion of the player-reference tiles, etc. Creation, anyway. Wouldn't we want to provide some more general way to delete morphs (including tiles)? Right now we have "removeAllContents" in Holder; or "doMenuItem: #delete". Would it make sense to add a 'delete' verb to some vocabulary that could be used for *any* morph? Or was this avoided for a good reason? Another possibility, BTW, might be to give the TrashCan a Holder-like vocabulary. > CAVEAT: The #tellAllContents: tile is only available in the > *Squeakland* image, not in any Squeak.org image at present. Seems to work in the 3.7a Squeak image if you load 0170textInEToys-sw.cs but I haven't tested fully. Given these ideas, would it make sense to have a new Morph/Player type that could be used for the return value from collection-valued slots (like, for instance, #incomingConnections when using Connectors, or #submorphs for a Morph)? It would be a specialized Holder that would poll the value of its expression periodically and update its contents with player tiles. The polling could be omitted if it weren't visible. Modifying its contents would cause the associated collection setter to be called (if any). Dropping a non-Tile Morph into it would actually drop a Tile and would return the morph to its original position. -- Ned Konz MetaMagix Stanwood WA (360) 629-1091

1 0

About SM2
by ducasse 07 Jan '04

07 Jan '04

Hi goran Do you want to have SM2 for 3.7? What should be done from an harvesting process point of view? I'm trying to have a clearer view of what we have to do. Stef

4 5

RE: About harvesting
by Oca Emilio 07 Jan '04

07 Jan '04

Hi, May be it can be established a partnership program. Who submits something (ENH,BUG,FIX,etc) to the list privately asks, begs or even pays a friend/partner/co-author/co-worker to review/comment his submission. Next time that favor could be back. This way reviewing could become a bit more personal and submitters get more involved in the "made-it-to-the-image" process. I hope this helps. Regards. Emilio > -----Mensaje original----- > De: Cees de Groot [mailto:cg@tric.nl] > Enviado el: Miércoles, 07 de Enero de 2004 10:03 > Para: squeak-dev(a)lists.squeakfoundation.org > Asunto: Re: About harvesting > > > Markus Gaelli <squeak-dev(a)lists.squeakfoundation.org> said: > >how about sorting the submitted fixes/enhancements according to the > >number of "made-it-to-the-image"-_reviews_ the submitter did in the > >last time? > > > Heh. Cool idea. And then a button to sort according to SqueakPeople > rating, maybe :-) > > Stephane&MarkusD are right, of course. I bow my head in shame and will > really really make a serious attempt to work with BFAV today. For me, > that has been the biggest hurdle so far - get comfortable > with BFAV so I > can indeed spend 5 or 10 minutes on something productive with it. > > -- > Cees de Groot http://www.tric.nl <cg(a)tric.nl> > tric, the new way helpdesk/ticketing software, VoIP/CTI, > web applications, custom development > >

1 0

Re: [Squeakland] Q: Leverage Overlapping Player to Etoys?
by Scott Wallace 07 Jan '04

07 Jan '04

(1) The beginnings of just such a thing are already present in the Squeakland image. Look in the 'collections' section of a viewer of any playfield-like object, and you'll see the following tile: This line of script, when executed, runs the designated script in each object within the holder. To a separate email (to hold down size) I shall attach a project, IteratorExample1, which shows a simple use of this feature. But this is not using aliases yet, this is having to have the actual objects in the holder. But it's a point of departure. (2) With a few small changes, one can arrange matters such that a player-reference tile (the kind you tear off via the orange halo-handle) will forward any putative message dispatched to it via #tellAllContents: on to the tile's referent. This allows player-reference tiles to function as "aliases" for the purpose of scripting, as Alan suggests in the excerpt quoted below. With this change, one can use a Holder (or any other PasteUpMorph) as a scripting iterator. I illustrate this style in a separate project, "IteratorWithTiles", which I also attach to a separate memo, rather than here, to keep individual memo size down. The IteratorWithTiles project contains the code changes that allow this style of iteration-over-aliases, and also contains a new version of the IteratorExample1 project in which the iteration is now done by way of tiles in a Holder. In short, a Holder can (after these changes) be set up to hold Reference tiles: and a Holder's scripts can iterate across the objects referred to by the reference tiles within it: This is just an existence proof. To become useful, I think we will want to extend the etoy vocabulary to allow programmatic creation and deletion of the player-reference tiles, etc. Awaiting your comments... Cheers, -- Scott CAVEAT: The #tellAllContents: tile is only available in the *Squeakland* image, not in any Squeak.org image at present. At 2:37 PM -0800 1/6/04, Alan Kay wrote: >Hi Ned -- > >Yes, we should have aliases (shortcuts) that we can use in holders >the way we do in variables. The name tile does do this job for >putting an alias of a player into a player-valued variable, but I >don't think we can generally use it in other useful places at >present.

1 1

Iterator with tiles (was Re: [Squeakland] Q: Leverage Overlapping Player to Etoys?)
by Scott Wallace 07 Jan '04

07 Jan '04

>(2) With a few small changes, one can arrange matters such that a >player-reference tile (the kind you tear off via the orange >halo-handle) will forward any putative message dispatched to it via >#tellAllContents: on to the tile's referent. This allows >player-reference tiles to function as "aliases" for the purpose of >scripting, as Alan suggests in the excerpt quoted below. > >With this change, one can use a Holder (or any other PasteUpMorph) >as a scripting iterator. I illustrate this style in a separate >project, "IteratorWithTiles", which I shall attach to a separate >memo, rather than here, to keep memo size down... I see that even after having separated the two, my previous email still got bounced, or at least delayed, by the Squeak list. Anyway, attached is the "IteratorWithTiles" project. Again, note that this must be loaded into a Squeakland image. Cheers, -- Scott

1 0

[BUG] Vars missing in debugger...
by goran.krampe＠bluefish.se 07 Jan '04

07 Jan '04

Hi! I have no idea why I haven't seen this before but try this do-it in a workspace: 'asd' do: [:cll | self halt] ...now - where is the cll var in the debugger pane? It should be down in the right pane - it is in my old email image (3.2) at least. Haven't tried tracking or looking any further - perhaps someone else knows this MUCH better. :) regards, Göran

6 6

file locking primitive
by Julian Fitzell 07 Jan '04

07 Jan '04

This has come up before but we need the file locking primitive in Squeak to support our work in OmniBase. There has been enough interest in this in the past that I think it's worthwhile coming up with a primitive interface that is implementable in all the common VMs. I don't know anywhere near enough about file locking on the various platforms to know what this is. To keep the number of primitives down I'm guessing we want one that takes a boolean indicating whether to lock or unlock the file. fctl() on *nix seems to allow you to speficy a range using a length and a start position relative to either the beginning of the file, the current position, or the end of the file. Do we want to maintain the same interface? Or is it going to be easier to have the primitive always take positions relative to the start? or to the current position? These questions are based mainly on ignorance about how Squeak deals with the file descriptors, etc. Any pointers on how to get the FD from the primitive or comments on the correct interface for the primitive would be much appreciated. We're going to look at implementing something for the unix VM on Monday, but if we can come up with something that makes others happy then hopefully others can jump in with implementations for the other VMs. Julian

7 19

building an up to date VM for unix
by Martin Kuball 07 Jan '04

07 Jan '04

Hi! I'm trying to build an up to date VM. I searched the internet and found a reference to the anonymous cvs repository on sourceforge. I checked out the ned-branch, installed VMMaker in my 3.7 image and generated everything. Than I noticed that there is no platform/unix/config/configure script and even no configure.in. So what now? Martin

8 10

← Newer
1
...
32
33
34
35
36
37
38
39
40
Older →

Jump to page: