New Editor
Locked
 
|
Tom suggested one of the features of PPD's canceled project 'jigsaw' (I think it was jigsaw, wasn't it?), wherein one image could be used as the IDE to drive another image running application code. This does not require that the application image have no IDE code in it; the 2 images may well contain identical code, so the application can leverage any IDE code you want. It's just that the image being used to debug has its own memory space and cannot be disrupted by an error in the application image. Processes in the application image can be displayed in a debugger UI in the IDE image, etc. Dave Stevenson[hidden email] From: Dennis Smith <[hidden email]> To: [hidden email] Sent: Tue, June 19, 2012 12:14:32 PM Subject: Re: [vwnc] New Editor Our applications are tightly bound to the IDE and extensions using our our business class browser, having the IDE as part of the application is criticial to us. I don't see it as a problem either. On 2012-06-19 1:10 PM, Terry Raymond wrote:
-- Dennis Smith +1 416.798.7948 Cherniak Software Development Corporation Fax: +1 416.798.0948 509-2001 Sheppard Avenue East [hidden email] Toronto, ON M2J 4Z8 [hidden email] Canada http://www.CherniakSoftware.com Entrance off Yorkland Blvd south of Sheppard Ave east of the DVP _______________________________________________ vwnc mailing list [hidden email] http://lists.cs.uiuc.edu/mailman/listinfo/vwnc |
|
|
Dave,
(I think it was jigsaw, wasn't it?) It was a Digitalk/Allen Wirfs-Brock project, originally presented at the Disneyland conference: Firewall. And So It Goes Sames ______________________________________________________________________ Samuel S. Shuster [|] VisualWorks Engineering, Store Project Smalltalk Enables Success -- What Are YOU Using? _______________________________________________ vwnc mailing list [hidden email] http://lists.cs.uiuc.edu/mailman/listinfo/vwnc |
|
|
Yes, Firewall, exactly. The details sometimes escape me.
Jigsaw was some kind of hybrid between the Digitalk and ParcPlace products. Seems ironic that jigsaw was canceled, but many years later a vaguely similar technology emerged with the ability to run ObjectStudio from VisualWorks. Dave Stevenson[hidden email] From: Samuel S. Shuster <[hidden email]> To: Dave Stevenson <[hidden email]> Cc: Dennis Smith <[hidden email]>; [hidden email] Sent: Tue, June 19, 2012 1:29:24 PM Subject: Re: [vwnc] New Editor Dave, (I think it was jigsaw, wasn't it?) It was a Digitalk/Allen Wirfs-Brock project, originally presented at the Disneyland conference: Firewall. And So It Goes
Sames
______________________________________________________________________
Samuel S. Shuster [|]
VisualWorks Engineering, Store Project
Smalltalk Enables Success -- What Are YOU Using?
_______________________________________________ vwnc mailing list [hidden email] http://lists.cs.uiuc.edu/mailman/listinfo/vwnc |
|
|
Dave, Seems ironic that jigsaw was canceled, but many years later a vaguely similar technology emerged with the ability to run ObjectStudio from VisualWorks. Jigsaw was the Java in VisualWorks work. And So It Goes Sames ______________________________________________________________________ Samuel S. Shuster [|] VisualWorks Engineering, Store Project Smalltalk Enables Success -- What Are YOU Using? _______________________________________________ vwnc mailing list [hidden email] http://lists.cs.uiuc.edu/mailman/listinfo/vwnc |
|
|
In reply to this post by Dave Stevenson-3
Isn't this dual image idea also one of the ideas in Craig Latta's Spoon project? He mentioned that he is trying to include VW as a target for Spoon. -Carl Gundel Liberty BASIC for Windows - http://www.libertybasic.com Run BASIC, easy web programming - http://www.runbasic.com
_______________________________________________ vwnc mailing list [hidden email] http://lists.cs.uiuc.edu/mailman/listinfo/vwnc |
|
|
In reply to this post by Samuel S. Shuster-2
I'm a little late on the discusion, but maybe you could check BlueMagic and MethodSniffer packages in the public repo. Two small extensions I made to the RefactoringBrowser, mainly in the text editor area. I think that everybody will agree, that many of those features should come in the basic RB, specially the RBBlueMagic package.
James Robertson did a screencast on it a few years ago: http://www.cincomsmalltalk.com/blog/blogView?showComments=true&printTitle=Smalltalk_Daily_12/30/08:_Enhanced_Browsing&entry=3408093267
Cheers, Emiliano. _______________________________________________ vwnc mailing list [hidden email] http://lists.cs.uiuc.edu/mailman/listinfo/vwnc |
Re: Platform Coherence (was: New Editor)
|
|
In reply to this post by andre
(oh good, you changed the subject)
On Tue, Jun 19, 2012 at 12:20 AM, andre <[hidden email]> wrote: > > On 19.06.2012, at 03:10, John Dougan wrote: > >> I'm not against platform coherence, I just don't think that it matters >> that much any more. > > I would like to strongly disagree! My experience is absolutely contrary. During the past years I have observed a significant relationship between platform coherence and sales. Making your desktop product look and feel like a web browser is the quickest way to lose money. > > However, it is important to distinguish between visual design and behavior. The term "Look & Feel" suggests a dependency where there actually isn't one. I agree with you that the /look/ is no longer that critical and a slick custom design is far superior to a half-baken native emulation. > > I can however not stress enough how critical platform coherent behavior, or "feel", is. I've put the following points together in the hope they may be helpful to others: I think we're closer on this that you might think. Feel matters, however users get their expectations on the feel from how it looks. In my experience there seems to be an "uncanny valley" that applies to UI…if it is close to something familiar to the user, but not close enough, the perceived user experience is worse than if the UI is more different. As an experiment, I gave users the VW default look in one release (with some adjustments)…and after a bit they were happier than they were with the emulated looks. They expected to see something that sorta looked and acted like a scrollbar in the position of a scrollbar, and there one was. But, it looked different enough that they didn't bring all their preconceptions to the table. There were enough capabilities missing from default look that I put the contract specified UI back later. I suspect that if I had been able to offer the users a superior set of capabilities in the context of the default look I might not have had to change it back. > (1) Terms and Layout: > (2) Interactivity: > For this sort of platform standards stuff VW is sufficient, if just barely, as long as you are willing to design the UI layouts and such once for each target platform. The various platform standards are different enough that it's probably too hard to automate enough to be worth the dev time. Maybe having VW startup be a touch more platform aware so it can reconfigure to the right application classes? The other thing that has happened is that the major platforms all have changed their platform standards over the last 15 years. Win95, WinNT, Win2000, WinXP, Vista, Win7, Win8, etc. each had significant changes to look and feel. This is part of what has made users (at least ones I've had to deal with) more flexible and makes VW's UI problem harder. > (3) Keyboard mapping: > Agreed. Not only does VW do the wrong thing, it is ridiculously hard to make it do the right thing. > (4) Selections: How widgets and editors behave with respect to making selections, inserting and replacing content, getting input focus, scrolling, etc. Fortunately there are only few differences, so this is less an issue. > Yup. I suspect a month of one developer could iron most of this out to where it would be good enough, which to me means that users mostly shouldn't have anything happen that contradicts their typing muscle memory….except where that target platform contradicts itself (make a text selection in various places in various Windows apps and hit the left arrow and see where the cursor ends up relative to the original selection). > (5) OS Integration: Your product needs to behave nicely in how it launches and quits, enters and leaves hibernation mode, uses native file and print dialogs, provides smooth copy/paste capabilities with other apps, supports scripting, handles "save", "save as", "revert to last saved", etc. > Agreed, to a point. I was assuming this level of integration discussion was outside the scope of the text editor discussion. Some of this (scripting, *full* clipboard support for non plain text items) can be very hard to provide, especially portably, and may not be worth Cincom's resources. If you need that kind of deep compatibility with the target platform you may be better off writing in whatever the platform systems language is (C, C++, Objective-C, etc.) or in an ST tuned more closely to the target platform (Smalltalk MT, Dolphin, ST/X, etc) and use the actual system facilities. Part of a dev's job is to pick the right tools. > At the end of the day, the equiation is very simple: The more alien an average end user feels with your product, the more money you lose. > Depending on how much value your app brings to the table. Users will overlook a lot if your app does something they really need they can't get easily elsewhere (eg. sales contact management programs circa 1999 or games today). This also get mitigated if the app has to run cross-platform in the user's environment. If they are actively using it on all of a Mac, a WinPC and a web interface, they will value platform coherence less (though not zero) and self coherence more. > It appears to me that Smalltalkers tend to center around their own views as developers a lot, while neglecting the demands of end users. IMO, this is the main reason why Smalltalk is stuck in a niche. I can live with a suboptimal and unconventional environment quite well. My end users can't. > Too broad a statement. There are Smalltalks that can do this stuff just fine. VW doesn't because of design directions that were taken back around 1992 or so to focus more on portability and generality (which made a lot of sense then). -- John Dougan [hidden email] _______________________________________________ vwnc mailing list [hidden email] http://lists.cs.uiuc.edu/mailman/listinfo/vwnc |
|
|
In reply to this post by Emiliano Pérez-3
On Tue, Jun 19, 2012 at 3:12 PM, Emiliano Pérez <[hidden email]> wrote:
I'm a little late on the discusion, but maybe you could check BlueMagic and MethodSniffer packages in the public repo. Two small extensions I made to the RefactoringBrowser, mainly in the text editor area. I think that everybody will agree, that many of those features should come in the basic RB, specially the RBBlueMagic package. +1 Emiliano, well done and thanks for sharing. Also, I'm in agreement that many of the features within your package should be baked into the current RB or there needs to be a better way to publish packages so that they can be easily found on the web and within the IDE. For example, I do quite a bit of Ruby development for work and the following site is great for finding new package(s):
_______________________________________________ Think different and code well, -Conrad _______________________________________________ vwnc mailing list [hidden email] http://lists.cs.uiuc.edu/mailman/listinfo/vwnc |
|
|
In reply to this post by Terry Raymond
On 20/06/2012, at 2:40 AM, Terry Raymond wrote: > One could use something like Opentalk to separate the IDE from the runtime. > But reusing an IDE that is not smalltalk would make it much harder to extend the IDE. This is a project I've been working on for several years - a Smalltalk IDE that is separated from the image by an asynchronous network connection. Imagine an X11 like widget toolkit where the widgets are high level objects like 'browser' and 'workspace'. The UI isn't written in ST, and is completely native, differing between platforms as far as L&F is concerned. This worked really well, although obviously it's for server-side applications unless you add different widgets to the UI server. OTOH, tools can be built in a generic fashion (Omnibrowser toolkit etc) that allow you to build new tools as long as they conform to existing UI paradigms. Image restore isn't difficult (i.e. windows come back to where they were etc). I was using it as a testbed for different UI models, such as document based view, where a document can be a class or a call hierarchy or a search result etc. I started on VW and then moved to Pharo, but I've not worked on it for a while. Not suggesting this is at all relevant to this thread, just responding to Terry's suggestion. Antony Blakey ------------- CTO, Linkuistics Pty Ltd Ph: 0438 840 787 One should respect public opinion insofar as is necessary to avoid starvation and keep out of prison, but anything that goes beyond this is voluntary submission to an unnecessary tyranny. -- Bertrand Russell _______________________________________________ vwnc mailing list [hidden email] http://lists.cs.uiuc.edu/mailman/listinfo/vwnc |
|
|
In reply to this post by Conrad Taylor
Thanks conrad. For the sake of this post here's a list of things that I think would be cool to integrate in the code editor:
- Open selected method/class with one click (hiperlink-like behavior).
- Most used refactorings by selecting the code and using a shortcut. - Automatic class/method/v.i. definition and spell corrector WITHOUT having to save the method. BlueMagic provided this three amongst a few more, but because at the moment, I didn't wanted to mess with the use of left and right click, all the functionality was accessed via the middle mouse button (wich was not all that intuitive).
Also a good navigation history and full support for undo/redo are a must. 2012/6/19 Conrad Taylor <[hidden email]>
Emiliano G. Pérez _______________________________________________ vwnc mailing list [hidden email] http://lists.cs.uiuc.edu/mailman/listinfo/vwnc |
|
Administrator
|
In reply to this post by Thomas, Arden
I would like to select and copy text anywhere they appear in the whole IDE: windows, editor, labels, menu, dialogs, combo box, etc.
Make 'explain' work anytime text is selected in the whole IDE: browsers, editors, changelist, parcel manager, etc. Allow exhaustive text search of entire image anytime text is selected. Classes, methods, definitions, comments, pragmas, primitives, etc. should be searchable. Thanks, Aik-Siong Koh |
|
Administrator
|
Make the code formatter in Debug behave the same as in Editor.
Aik-Siong Koh |
|
Administrator
|
In reply to this post by Thomas, Arden
Keep comments in place exactly where it is written. No extra tab, line or cr in front or behind it added or removed.
Aik-Siong Koh |
Re: New Editor
|
|
That's a formatter function, not an editor's though.
-Boris On 2012-08-16, at 8:01, "askoh" <[hidden email]> wrote: > Keep comments in place exactly where it is written. No extra tab, line or cr > in front or behind it added or removed. > > Aik-Siong Koh > > > > -- > View this message in context: http://forum.world.st/New-Editor-tp4635017p4644234.html > Sent from the VisualWorks mailing list archive at Nabble.com. > _______________________________________________ > vwnc mailing list > [hidden email] > http://lists.cs.uiuc.edu/mailman/listinfo/vwnc _______________________________________________ vwnc mailing list [hidden email] http://lists.cs.uiuc.edu/mailman/listinfo/vwnc |
| Free forum by Nabble | Edit this page |