Squeak-dev June 2000

squeak-dev@lists.squeakfoundation.org

134 participants
686 discussions

Re: Unknown Protocol in ANSI draft ?
by R. A. Harmon 30 Jun '00

30 Jun '00

At 05:45 PM 6/30/00 +1200, Brent Pinkney <Brent(a)astarte.co.nz> wrote: >Hi, > >Does anones copy of the draft have a definition for the <abstractDictionary >factory protocol> > >I can only see a reference to it in section 5.7.22 where: > >5.7.22 Protocol: <IdentityDictionary factory> >Conforms To ><abstractDictionary factory> I couldn't find it defined either. I substituted <Dictionary factory> in my stuff that builds protocol objects, but I just noticed the graph of the protocols show it to be conforming to <collection factory>. Maybe an ANSI committee member will kindly set us straight. -- Richard A. Harmon "The only good zombie is a dead zombie" harmonra(a)webname.com E. G. McCarthy Spencer, Iowa

1 0

RE: ANSI draft in PDF format?
by Norton, Chris 30 Jun '00

30 Jun '00

Hi Stephan. >For me downloading the indexed version does *not* work with Netscape >Messenger, it seems to be the same problem as with the download of my >version (see below). I was able to download the indexed document yesterday with Netscape Communicator 4.7. Cheers, ---==> Chris

1 0

RE: [Q] Using FFI to call standard Windows dialogs
by Norton, Chris 30 Jun '00

30 Jun '00

Thanks for the quick reply Torsten & Bob J! I guess that my question has uncovered 2 problems: 1. I don't know a thing about printing 8-) 2. The FFI mechanism could use some better exception handling. I don't know if we can wrap an API call in such a way that we can recover from problems it causes, but IMHO it would be good to eliminate this crash/dump scenario. If / when I get the print dialog working, I'll post my work to y'all. Thanks for the FFI lessons! ---==> Chris

1 0

Re: New Web Browser
by Jon Hylands 30 Jun '00

30 Jun '00

On Fri, 30 Jun 2000 01:41:14 -0700, Alan Kay <Alan.Kay(a)disney.com> wrote: > I believe that it has been possible to embed morphs into text morphs > for some time now. I have used both of the below: Okay, thanks for the examples Alan... That's going to make my life a lot simpler. I did a bunch of work on it last night, and ported the parser and the layout engine. I now have web pages displaying on the Display in 2.7, although not using Morphs for layout yet. The layout engine has remained static for the past two years, and the web has not, so there are definitely some problems that need to get fixed. I'm going to spend some serious time this weekend working on this (moving the layout elements into Morphic), and hopefully will have something in a decent state by the end of the long weekend to release. No promises of course, but I'll do my best. I'm also planning on porting PIC/Smalltalk into Squeak (tonight), which should be fairly simple, and releasing it. Later, Jon -------------------------------------------------------------- Jon Hylands Jon(a)huv.com http://www.huv.com/jon Project: Micro Seeker (Micro Autonomous Underwater Vehicle) http://www.huv.com

1 1

[BUG][FIX] CCodeGenerator a minor detail
by Mats Nygren 30 Jun '00

30 Jun '00

Hi, while working with the CCodeGenerator I found a small bug. /Mats

1 0

Re: Belling the cat of complexity (was: Ship it with Squeak)
by Paul Fernhout 30 Jun '00

30 Jun '00

Stephan Rudlof wrote: > [Great ideas of using Squeak to make an "IndustrialStrength" code in a Squeak namespace using existing Squeak tools.] > "Göran Hultgren" wrote: > [Great elaborations on the concept.] These ideas are superb. This sounds like a good approach towards building something reliable using Squeak. The work being done with PocketSmalltalk for Squeak sounds somewhat similar to this approach. It seems like they are working out many of these issues because of their need to maintain a separate set of "PalmStrength" Smalltalk classes. Note that Smalltalk requires some resources to run which are not normally thought of as code. These include bitmaps, sounds, and fonts. They also include some things that could be generated by code but whose code origins may be lost -- like dictionaries of color mappings. In order to make this "IndustrialStrength" namespace complete, there needs to be some way to include such resources in the namespace (along with licensing information to avoid or flag for example the Microsoft font license issue.) There needs to be enough information about these resources that they can be at least isolated and understood as specific unbuildable objects, or ideally they need to be tied to code that regenerates them (for Dictionaries) or imports them from external files or hex representations (for bitmaps or sounds or fonts). All in all, your ideas sound like a good approach towards using the power of Squeak we have to bootstrap something even better. Thanks for the great ideas. -Paul Fernhout Kurtz-Fernhout Software ========================================================= Developers of custom software and educational simulations Creators of the Garden with Insight(TM) garden simulator http://www.kurtz-fernhout.com

1 0

RE: [Q] Using FFI to call standard Windows dialogs
by Jarvis, Robert P. (Contingent) 30 Jun '00

30 Jun '00

It looks like you're invoking the PrintDlg function, which takes a pointer to a PRINTDLG structure as its one parameter, not a window handle. That may explain the crash. The PRINTDLG structure looks like this: typedef struct tagPD { DWORD lStructSize; HWND hwndOwner; HGLOBAL hDevMode; HGLOBAL hDevNames; HDC hDC; DWORD Flags; WORD nFromPage; WORD nToPage; WORD nMinPage; WORD nMaxPage; WORD nCopies; HINSTANCE hInstance; LPARAM lCustData; LPPRINTHOOKPROC lpfnPrintHook; LPSETUPHOOKPROC lpfnSetupHook; LPCTSTR lpPrintTemplateName; LPCTSTR lpSetupTemplateName; HGLOBAL hPrintTemplate; HGLOBAL hSetupTemplate; } PRINTDLG, *LPPRINTDLG; How one would go about representing this in Squeak is beyond me. A few of the members of this structure are handles to other Win32 structures (e.g. hDevMode is a handle to a DEVMODE structure, hDevNames is a handle to a DEVNAMES structure, etc), which doubtless will increase the joy and fun of trying to get your print dialog up on the screen. As to "...how do I get a document to print..." - well, that's up to your application. PrintDlg just puts up the dialog, initialized and customized as per the structure above (and its substructures, such as DEVMODE and DEVNAMES), captures the user's input, and modifies the data in these structures appropriately. Printing is usually done by obtaining a DC (device context) on the appropriate printer and drawing away to your heart's content. An easy way to do this is to specify PD_RETURNDC in the Flags element of your PRINTDLG structure, which causes PrintDlg to fill in the hDC element of the PRINTDLG structure with a handle to an appropriate DC. Personally, I think printing is one of the most God-awful things an application ever has to do. It's messy, error-prone, varies from device to device and driver version to driver version, and is ecologically unsound to boot! <grumble> Let 'em read it on the screen if they want to know what's going on, that's what I say... :-) Bob Jarvis Compuware @ Timken > -----Original Message----- > From: Norton, Chris [mailto:chrisn@Kronos.com] > Sent: Thursday, June 29, 2000 10:26 PM > To: Squeak (E-mail) > Subject: [Q] Using FFI to call standard Windows dialogs > > > Hi Folks. > > I was experimenting with FFI on my Windows NT box and I kept > crashing my > image (virtual memory dumps). Can anybody help me with this call? > > ~~~~~~~~~~~~~~~~~~~~~~~~~~ > Win32Shell>> > shellPrintDialog: aHWND > "Win32Shell new shellPrintDialog: Win32Window getFocus" > > <apicall: bool 'PrintDlgA' (Win32Window) module: 'comdlg32.dll'> > ^self externalCallFailed > > ~~~~~~~~~~~~~~~~~~~~~~~~~~ > > I'm trying to launch the print dialog. > > Thanks for your thoughts! > > ---==> Chris > > PS> And my next question is, of course, how do I get a > document to print > once I've got the dialog launching from Squeak (e.g. how do I > provide the > dialog with a document to print?)... >

1 0

Re: Better solution [was: Re: [BUG][Linux] VM2.8 port eats processor cycles]
by Bert Freudenberg 30 Jun '00

30 Jun '00

On Fri, 30 Jun 2000, David T. Lewis wrote: > On Thu, Jun 29, 2000 at 10:15:28PM +0200, Bert Freudenberg wrote: > > On Thu, 29 Jun 2000, Tim Rowledge wrote: > > > > > > nanosleep(&sleepTime, &reminderTime); > > > Excellent - do you know if it actually portable to most unices, or is it > > > likely to be a hit and miss affair? > > > > It's a POSIX standard function. I checked (the manpages) on Irix and > > SunOS. For Suns, it seems you have to link with libposix4 though ... > > > > -- Bert > > > It will be somewhat hit and miss. > For example, BDSI does not seem to have it. > > bash-2.02$ uname > BSD/OS > bash-2.02$ man nanosleep > man: no entry for nanosleep in the manual. > Maybe in BSD there is some other short sleep than usleep? >From SunOS man page: "Interactions between usleep() and either alarm(2) or sleep(3C) are unspecified." Which seems bad to me. -- Bert

1 0

RE: [Q] Using FFI to call standard Windows dialogs
by Torsten.Bergmann＠phaidros.com 30 Jun '00

30 Jun '00

>It's easier because someone implemented WINAPI and PRINTDLG classes >before. You can do the same in Squeak. If you have too much time, that is. Yes - I would like to see a WINAPI keyword (and a MACAPI, UNIXAPI, ...) in Squeak. It is true that the PRINTDLG Structure is predefined. But MT has a structure generator for easy defining an appropriate structure class. The only thing you have to do is to copy and paste the C structure from the MSDN documentation and run the generator. Another thing I like is that the MT image wont crash if an API is called with a wrong parameter. (you get an Unhandled Exception then) And it is easier because MT knows imported lib's so you dont have to write down the lib if you call an API. (But it's posssible to specify one if there are naming conflicts) The only thing I disklike on MT is that I can rotate Win32 windows ;) bye Torsten

2 1

AW: Newbie helper -- Easier window resizing in Morphic
by FWr＠phaidros.com 30 Jun '00

30 Jun '00

Hi Jim, I think it doesn't make a real difference if you have 1 or 2 pixels to catch this stupid little yellow handle. And so I must say that I find Steffen's solution much more reliable. Anyhow I think it would be much easier to resize windows if there would be the well known property of a changing mouse cursor in a "catching area" around the window borders. Frank -----Ursprüngliche Nachricht----- Von: Jim Benson [mailto:jb@speed.net] Gesendet: Freitag, 30. Juni 2000 10:43 An: Squeak Betreff: Newbie helper -- Easier window resizing in Morphic Newbies, So, you find it hard to resize a window in Morphic because it is difficult to grab that one pixel on the window border? You can suck it up, be a man and learn to use a mouse like Engelbart intended, carpal tunnel syndrome and all. Or, a clever person like you can dig deep into the system and change the method: SystemWindow #panelRect to something like: panelRect "Answer the area below the title bar which is devoted to panes." ^ self innerBounds insetBy: (1 @ self labelHeight corner: 1 @ 1) which will give you a full 2 ( !!! ) pixels to grab. ( panelRect is the rectangle within a SystemWindow where the panes lie. Windows refers to this as the 'Client Rectangle' of a window, Macs refer to this as the 'Content Region'. Linux guys have an unpronounceable acronym, but I hear you can redirect it using stdio ;-) The smaller the panelRect, the more border there is to grab ). Now, it's not quite as pretty, but it sure makes it a lot easier to grab the side of the windows without the manual dexterity of an eye surgeon. It's especially useful to people who have higher resolution monitors where one pixel is a little bit harder to grab. Jim

1 0

Jump to page: