[vwnc] Does Cincom give up the Desktop ?
Locked
12
12
Re: [vwnc] Does Cincom give up the Desktop ?
 
|
Andre Schnoor wrote:
> > Alan, > >> - Fonts were mentioned a couple of times > > From an engineer's perspective, native fonts might seem like a minor > matter of design only. Real world end users however expect to be able to > use the fonts they have installed on their computer (and sometimes paid > for), at least for printing and editing. > that's especially true for "modern" Unixes that use Xft/Treetype to display antialiased fonts. Actually, Freetype has been around for some years now. Together with the Motif look, it looks outdated. _______________________________________________ vwnc mailing list [hidden email] http://lists.cs.uiuc.edu/mailman/listinfo/vwnc |
Re: [vwnc] Does Cincom give up the Desktop ?
|
|
I have been reading this thread, with a number of people weighing in with their opinions. I have been away from Smalltalk for a while, but I installed a recent 7.7 dev build on my Linux box. I spent half an hour going through the product briefly, just to see what's changed.
<rant> Apparently not much. It looks like yes, Cincom HAS given up the desktop. Here is what I found: - The Canvas Painter looks the same as it has for at least 15 years. It's never been intuitive, and it's still not. For instance, select the "Drop Target" tab and you get four fields labeled "Entry", "Over", "Exit", "Drop". Yes, I know what they're for. Does a novice? How much effort would it take to put a tooltip on each field, providing a little help to a new person? - The icons on the Palette haven't been updated in, uh, maybe ever. - I select "Launcher Help" from the Launcher, and I get an ugly grey box with a list, and at the bottom of the list it says "Move cursor over an item for help information". I move my cursor over everything I see, and nothing happens. - The AdHoc SQL window was awful 10 years ago, and it's awful now. When I enter a query, if any of the returned fields are a timestamp, I can't see the entire value. I have a field that contains "August 20, 2009 10:46:18 am" and all I can see in that field is "August 20, 2" and it gets cut off. Are you kidding? I remember complaining about this years ago. Nothing has happened. You cannot resize the column widths in the output. When I was at Morgan Stanley, I redid that app using the Aragon data set widget with tabs, and I could keep a different frequently-run query open in each tab, and I could rerun each query by hitting one key. Unfortunately I wrote this on Morgan Stanley's dime and they did not permit me to contribute it back to the product. But if I was able to do it, Cincom's engineers can do it easily enough. - I don't know the status of the DLLCC parser, but when I was using it frequently, it was years since I had a header file parse completely on the first attempt. I would always get errors that I had to fix by hand, since there were enhancements made to the world of header files that the parser didn't know about. I don't know if this has been fixed because I don't deal with those kinds of things any more, and I don't have a complicated header file lying around that I can use for testing. - Under the "Settings" tool, the selection dropdown for "Look and Feel" contains "Windows 95/NT" and "Windows 98/2000". Terrific. This just screams out "I am an up-to-date system." We all know the reason for this, of course. The UI received no attention for a period of six years due to the Pollock debacle. I don't want to mention names because it's not productive to do that, and everyone knows the history. The first Pollock parcels were written to the Cincom Store Repository in December 2001, and the effort was cancelled in November 2007. That's six years of work that wasn't spent redoing the Canvas Painter. Six years that wasn't spent rewriting the AdHoc SQL window. Or updating the DLLCC parser. Or writing new "Look and Feel" packages to emulate operating systems that came out in this century. Basically every issue with the product can be directly traced back to Pollock, and the opportunity cost involved in it. The comment for the "Pollock" bundle in the Cincom Store Repository says: "The goals of this new Framework are really quite simple. Make a GUI Framework that maintains all of the goals of the current VisualWorks GUI, that is flexible and capable enough to see us forward for at least the next decade." If it was "really quite simple" it would have been done by now. It would have been done years ago. There is nothing that's going to "see us forward for at least the next decade." More than half a decade was wasted, resulting in absolutely nothing. Maarten Mostert made a point earlier today which caused me to write this: > The ones who advocate Smalltalk most of the time underline its mallability and dynamic character. > If cincom is capable of underpinning a new modern UI's underneath 3000+ canvas type applications > then you Cincom can go to Gartner and tell them He guys this is what "we" have been doing recently > with user investements and dynamic languages. He is right, of course, but there is a corollary to what he wrote, which is this: We all talk about how productive Smalltalk is. But if an expert Smalltalk engineer can't redo the UI in six years, maybe it's not as productive as we have been led to believe. I would tend to think that there were people evaluating Smalltalk at the time who noticed this, and came to that conclusion. And the logical extension of that is: If we threw six years of work into the garbage trying to rework the UI, then it's difficult to see Cincom making a second attempt at it, so the UI we have now is the UI we are going to have for a long, long time(I had suggested in the past to look into redoing the UI using the wxWidgets package, but I was shot down for reasons I don't recall now). I hope I am wrong about this, but this is the way things look from here. </rant> As I wrote at the beginning, I have been away from Smalltalk for a while, so maybe there are things going on that I don't know about. I hope so. --Tom On Sat, Nov 28, 2009 at 2:46 PM, Holger Kleinsorgen <[hidden email]> wrote:
_______________________________________________ vwnc mailing list [hidden email] http://lists.cs.uiuc.edu/mailman/listinfo/vwnc |
Re: [vwnc] Does Cincom give up the Desktop ?
|
|
Thomas:
> That's six years of work that wasn't spent redoing the Canvas Painter. Six years that wasn't spent rewriting the AdHoc SQL window. Or updating the DLLCC parser. Or writing new "Look and Feel" packages to emulate operating systems that came out in this century. Basically every issue with the product can be directly traced back to Pollock, and the opportunity cost involved in it. I've been quiet on this thread, since GUI hasn't been my domain since September 2007 (September 10th to be exact, not November), but let's get some facts out there. In that "6 years", the Canvas Painter was indeed updated.: It added the Tree View of widgets, Layout of the tool windows was updated Added support for new widgets Tab Control Resizing Splitter Tree View Updated some of the the palette icons Totally changed the window layout mechanism and page Made it possible to almost directly make the Tab Control a direct substitute for the Notebook In that "6 years", the Looks and Feels were updated. Added the WinXP look Added the MacOSX 10.1 look In that "6 years" the Polling system was removed from the system In that "6 years" the Multi-Process UI was put into place. In that "6 years" the trigger event system was merged into the system. During those "6 years" I was ONLY full time on Widgetry for the last 3 1/2, and I was the ONLY one writing code for it. Even then, I spent at least 20% of my time working on Wrapper bug fixes and updates. For the first 2 1/2 years I spent only 20% of my time on Pollock/Widgetry, since during that time, it was considered a proof of concept only. None the less, "Wrapper" updates were done throughout. Widgetry did get to 1.0 in April of 2007, amazing given that I was all but alone in the 4 1/2 person years it took to work on it. Some things that Wrapper is missing are indeed because of those last 3+ years of working on Widgetry, because Widgetry "had" or these things were planned for Widgetrys future: A MacOSX 10.4 L &F. Totally configurable (by platform) keystroke / hot key support Direct native fonts support on all platforms Anti Aliased fonts on X11 platforms Native widgets on ALL platforms (OSX, Windows, X11, QT, GTK, Motif and more) NONE of the other things you mentioned: AdHoc SQL or DLLCC, are related in any way to the GUI work. No resources were diverted from those areas to GUI, be it Widgetry or Wrapper. Indeed, barely 1/6 of an additional resource on average was used to do actual coding work in the GUI area, almost exclusively in code review, throughout my tenure leading the GUI project. While the whole Widgetry story is still, and will probably forever be, a sore spot for me, I none the less still have great pride in the work I did while I was Lead of the GUI project. Significant work was done during that time. 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 |
Re: [vwnc] Does Cincom give up the Desktop ?
|
|
In reply to this post by Michael Lucas-Smith-2
Hi
Just a not fun point in this nice discussion. I always found amazing that during the last 6 years we have to remove the last character of the class we are looking for in the search form in the browser. I found that terribly sad that the tools that we used daily got this kind of obvious, annoying and laugthable "glitch". I cannot imagine somebody not seen it and as maarteen mentions if we have so cool and malable tools then it should take less than a couple of hours (min) to fix that. Bt apparently it takes years. I was always amazed that we could have hobbes running in 4 days but this bug not fixed for years. Stef On Nov 28, 2009, at 7:12 PM, Michael Lucas-Smith wrote: > >> >> For now I am still dissapointed with the pace and efforts you have >> been putting on the UI developement. We don't have a Pango editor nor >> a new Grid we don't even have a Win95 type date selector to mention a >> few. The podcast's bottom line confirmed that the centre of gravity of >> product deelopement was not on the screaming delai in the UI. > I tend to think of the UI as three parts: implementation, look and feel. > > Without a doubt, the 'look' part of the UI got next to no attention in > 7.7. The tools however did get a whole host of new icons which were only > possible because the implementation and feel got quite a lot of work. > Most of it you may not even notice because the changes are designed to > make our widgets behave more like native widgets. > > For example, when you click on a selected item in a list, it doesn't > deselect any more. I don't know how many years I've toiled with the > muscle memory of "the VW way" versus "every other application I have > way". Finally, they're one in the same. Many keyboard shortcuts in the > editor are now standard too - ctrl+f will actually do a find for example. > > Other cosmetic UI changes, such as the last line of a list showing even > if it's only partially visible. Icons can now have an alpha channel that > renders properly (although, not accelerated). > > There are numerous other changes too that all add up to a subtle > improvement in the quality of the UI experience. > > There are many many things that we can do to improve the look of the UI > - such as using the right colours on mac osx. As we refactor and improve > the underpinnings, the exterior will improve too. > > The debate about using native widgets vs emulated widgets is a long, hot > and bothered one. It's worth noting that while it may seem incredibly > difficult to emulate the Aero look and feel, the aero widgets first > appeared in Microsoft Word before aero made its first appearance in > Windows Vista - in other words, Word has (for a long time now) kept its > own widget set.. or in other terms, an "emulated" widget set. Ideally, > we'd be mixing and matching as if there's no tomorrow with the caveat on > portability of your application. > > Some of the work being done in ObjectStudio brings us closer to doing > that - but mainly for windows. For a real solution, we'd need to do it > cross platform. And there's the rub, you have to walk before you can > run. That's the real reason why Cairo is in preview for 7.7; we may know > how to use the cairo API but this is the first time we'd be trying to > support the cairo API. OpenGL is included in 7.7 as well as a > contributed package. This is the same story as cairo, except that we're > two steps behind on its maturity compared to cairo. > > I'm an eternal optimist, so I'll continue to see the positive side on > the progress that has been made and will to come, which probably comes > through in the podcasts. There is no direct correlation between optimism > and delivery. > > Pango, for instance, certainly does a good job of marrying modern text > layout techniques with platform fonts but in the process it abstracts > you far far away from how anything actually gets done and also abstracts > you away from the real fonts (you use their font matcher). Is that good > enough? I don't know. > > Straight from the horses mouth on the latest in font layout technology: > http://behdad.org/text/ -- even here it's unclear exactly what should be > done for high profile projects like Mozilla who have invested a great > deal of time and energy in to international text layout. > > When all is said and done, I'm still a huge supporter of more UI > improvements for 7.8 and hopefully our management will agree. The more > you guys and gals talk it up on the mailing list, the more confident > management has prioritising it - like Alan said, it's hard to sometimes > tell which of the bleeding super uber important priorities are really > the mosst bleeding super and uber. > > Michael > _______________________________________________ > 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 |
Re: [vwnc] Does Cincom give up the Desktop ?
|
|
In reply to this post by Steven Kelly
I second what Maarten, Andre, Thomas and Steven said.
Especially needed are antia-aliasing and alpha support, filters, gradients, modern widgets, modern font support. Additionaly VisualWorks is lacking Media support in general (not many image formats supported, jpg so slow that it is unusable; audio, video...). Easy to use (or automatic) animations are starting to get the norm in modern UIs (cf. MacOS X Core Animation, JavaFX). Judiciously used they can improve user experience a lot. Is there anything planned in this area? The default look on Mac OS X is awful. On the other hand the Linkuistics pundle looks great. I would like this to become part of VisualWorks proper. The Windows look needs an update as well. Best regards Thorsten Seitz |-------------------------------------> | "Steven Kelly" | | <[hidden email]> | | Gesendet von: | | [hidden email] | | | | | | 27.11.2009 00:20 | |-------------------------------------> >-------------------------------------------------------------------------------------------------------------------------------| | | | | | An| | "Arden Thomas" <[hidden email]>, "VWNC" <[hidden email]> | | Kopie| | | | Thema| | Re: [vwnc] Does Cincom give up the Desktop ? | | | | | | | | | | | >-------------------------------------------------------------------------------------------------------------------------------| GUIs are important to us, and I think to anybody who wants to sell a product. You can give something away for free that doesn't look quite right, but if you want money for it people expect a professional, modern-looking GUI. I second what Maarten, Andre and Thomas said. The main improvements I think we need are alpha and anti-aliasing - these being things that can't be added well in Smalltalk. The things I've added myself in Smalltalk, such as affine transforms, gradient fills, and dotted lines, would also be nice to get proper built-in support for. I think all these improvements need to end up in the base VW GraphicsContext, to be available to all subclasses. Otherwise you can have things that work on screen but that you can't print. Cheers, Steve > -----Original Message----- > From: [hidden email] [mailto:[hidden email]] On > Behalf Of Arden Thomas > Sent: 25 November 2009 18:52 > To: Maarten MOSTERT; VWNC > Subject: Re: [vwnc] Does Cincom give up the Desktop ? > > Hi Maarten; > > First, thank you for the feedback, I think we can appreciate the > perspective. > > I can assure you that plenty of UI work is still on the roadmap, and > that we too wish it was moving faster, but it is moving. > > With VW7.7 the Cairo work is in preview, and comes with the Cairo lib > dlls for use with VW. > Many significant and subtle ui and tool improvements are in 7.7. > > I had hoped for L&F work this release but we were unable to squeeze it > in. It is still on our roadmap. > > Also on the roadmap are things like significant improvement for the > font framework. > > Wrapper and other infrastructure improvements are still in the works. > Work and experiments have been done in this area, and some results can > be seen in the way the browsers and other tools are built and > assembled. I think it will be important to get these things right. > > Longer term we are likely to work on replacing the uibuilder and make > many improvements utilizing Cairo and font work improvements. > > Feedback like this helps keep the pressure on priority for this type > of work, so I invite others to weigh in on where their product > improvement priorities lie. > > Thanks! > > Arden Thomas > Cincom Smalltalk Product Manager > > > On Nov 25, 2009, at 8:36 AM, Maarten MOSTERT wrote: > > > Hi I just listened to industry misinterpretations 160 and I would > > like to express my disappointment. > > > > Previously the roadmap showed that users claimed a major and not > > minor overhaul of the GUI. This item disappeared > > later from the road map and now no longer seems to be a point? > > > > We can notice new icons which are nice for the IDE, but we have been > > doing that in our applications with Epigent > > and Cairo for ages now! > > > > I also remember very well MLS stating 2 years ago that wrapper was > > the most pure GUI api around and that it would > > be a breeze to modernise wrapper's interfaces? Since then MLS seemed > > to have been involved in any single package > > which comes with the product accept the one which every desktop > > users on the planet sees first when opening a VW > > application, its grand daddy and grand mum's look and feel packages. > > > > Do you guys understand that almost nothing came and will come with > > the product to address this problem for over a > > decade ? > > > > Why is it that new L&F's disappeared from the roadmap, does nobody > > at Cincom dares to tackle this? Is it to > > complicated for you guys to do so (even for MLS ?). Did you simply > > give up future development for the desktop? > > > > I started a new Vista UI myself as sometimes I really can't face it > > anymore. I agree that it is pain to do because > > of the endless graphics context copying and impressive levels in the > > class hierarchy. At the same time let's not > > forget that Cairo can make it a lot simpler as we can redraw without > > flickering allowing you to build simple > > straight forward good looking mvc L&F applications. > > > > Please don't take this personally as I really appreciate many of > > your personal efforts but who cares about a new > > launcher and new installers, when we are jeopardising so many end- > > users with grand mummy's look and feel? > > > > > > Regards, > > > > @+Maarten, > > > > > > > > _______________________________________________ > > 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 _______________________________________________ vwnc mailing list [hidden email] http://lists.cs.uiuc.edu/mailman/listinfo/vwnc --------- Diese E-Mail könnte vertrauliche und/oder rechtlich geschützte Informationen enthalten. Wenn Sie nicht der richtige Adressat sind oder diese E-Mail irrtümlich erhalten haben, informieren Sie bitte sofort den Absender und vernichten Sie diese Mail. Das unerlaubte Kopieren sowie die unbefugte Weitergabe dieser Mail sind nicht gestattet. This e-mail may contain confidential and/or privileged information. If you are not the intended recipient (or have received this e-mail in error) please notify the sender immediately and destroy this e-mail. Any unauthorised copying, disclosure or distribution of the material in this e-mail is strictly forbidden. ---------- _______________________________________________ vwnc mailing list [hidden email] http://lists.cs.uiuc.edu/mailman/listinfo/vwnc |
|
|
In reply to this post by Samuel S. Shuster-2
Hi,
What I think that's most importantly missing in VW right now is: 1) The possibility to use advanced graphics drawing features 2) Improved native window support Regarding the first point, VW should be extended with at least the possibility to use: - transparent paints/images - linear/radial gradients - anti aliasing support - dashed lines/curves - some more advanced 'path' drawing. ie. not just a polygon, but something that can consist out of bezier curves, lines, arc segments, and that can be filled and stroked (also with dashed lines) Cairo supports all of this, but it uses purely software based rendering - no hardware acceleration. I'd rather see the existing GraphicsContext extended with these features, relying on either the OS (which will use HW acceleration when available) or based on OpenGL. This will make the drawing much faster than is now the case with Cairo. For the 2nd topic, apart from the ideas of Andre Schnoor (Tool palettes, drawers, sheets, ...), you should be able to completely decorate a window yourself, ie. start of with a blank window with no title/border/close/min/max, and customize it yourself in VW. Also, a way to give a window a non-rectangular shape would be great. All of this in a platform independent way of course. Perhaps an idea here could be to move the window related code out of the VM into the image. As OS-es evolve, it will be easier to change and extend the code in Smalltalk than in the VM. Once you have this, you can improve the look&feel of your widgets and applications. On the subject of native widgets: I'm not sure whether going for native widgets is a good idea. I don't even know what a native widget is. My computer is running on Vista, but when I look at outlook, the entire user interface looks different than the one I see in the file explorer. Not even the scrollbars or the window borders are the same! The same goes for MS Word and Excel and if you're starting to use some of the Vista administration tools, the widgets you see there don't look at all like the ones you see in the aforementioned examples. If you open MSN messenger it looks completely different again. And these are all applications from the same manufacturer who also made the OS. Based on this, I don't think a native look exists (at least not on windows) and I wouldn't want Cincom to put effort into that direction. I think it's more important to offer the basic tools to be able to create your own user interface look. -Mark _______________________________________________ vwnc mailing list [hidden email] http://lists.cs.uiuc.edu/mailman/listinfo/vwnc |
Re: [vwnc] Does Cincom give up the Desktop ?
|
|
Am 30.11.2009 um 13:51 schrieb Mark Plas: > Cairo supports all of this, but it uses purely software based > rendering - no hardware acceleration. IIRC, it does so where this is possible. If the OS supports a certain operation, it just passes it over (e.g. to Quartz). Anyone please correct me if I'm wrong. The great advantage of Cairo is its Float screen coordinates. This eliminates many of the rounding errors and sub-pixel artifacts one currently has to deal with. > I'm not sure whether going for native widgets is a good idea. +1 I'd rather see a skinnable Uber L&F that can be easily configured and changed to individual needs. A new Look (on the same platform) often just means new widget images. This would also open up the possibilty for corporate design. Andre _______________________________________________ vwnc mailing list [hidden email] http://lists.cs.uiuc.edu/mailman/listinfo/vwnc |
Re: [vwnc] Does Cincom give up the Desktop ?
|
|
In reply to this post by Mark Plas
I cast my vote for improving the desktop as well. But like most realize
that functionality is king. For non-business use, I've recently been looking for a web site provider with nice web site development tools that others in my group can use to easily edit and update the site with. I found a provider that provides site templates that are totally based on Flash. While the initial Wow factor is high, I find the functionality for embedding other site apps (e.g. Google calendar), doing mashups, etc, is very low. It's great to have impressive looks, but not at the expense of functionality and features. I'd love to see a dataset/grid widget with a lot more functionality. The ability to embed a progress bar in column cells would be great. Alternating row colors seems to be fairly standard now. Not looking to recreate Excel, but built in support to linking Excel or OpenOffice would be very nice. Widget functionality hasn't seemed to change much in the last 15 years. For me the bottom line is Smalltalk needs to deliver solutions to my problems. And it's a real plus if those solutions look nice too. Thanks, Mark K. Petersen (Embedded image moved to Home Office: (319) Semiconductor Research file: pic16146.jpg) 406-4165 Cell: (845) and Development Center 235-4360 IBM Global Engineering Internet email: Solutions [hidden email] DMACS Wiki DMACS Forum _______________________________________________ vwnc mailing list [hidden email] http://lists.cs.uiuc.edu/mailman/listinfo/vwnc |
pic16146.jpg (9K)
Re: [vwnc] Does Cincom give up the Desktop ?
|
|
In reply to this post by Andre Schnoor
> I'd rather see a skinnable Uber L&F that can be easily configured and
> changed to individual needs. A new Look (on the same platform) often > just means new widget images. This would also open up the possibilty > for corporate design. With its easy to modify Artist classes this was exactly what Widgetry provided. @+Maarten, _______________________________________________ vwnc mailing list [hidden email] http://lists.cs.uiuc.edu/mailman/listinfo/vwnc |
Re: [vwnc] Does Cincom give up the Desktop ?
|
|
There are a few things that should be pointed out here:
-- The current release cycle (in release candidate mode now) was largely an "infrastructure" one. We did a lot of work on: international support with a fully Unicode VM and CLDR based locale support Re-hosted Store atop Glorp The former gives us a solid base for further internationalization; the latter makes it much easier to change the underlying Store schema <with minimal customer impact> in order to move Store forward. This was a large step, and something we've needed for awhile On the UI side of things, we've improved the "feel" a lot with more true to platform key bindings. As has been pointed out earlier, you'll be able to stop switching muscle memory when you fire VW up. Additionally, a lot of the tools work in the browser (for instance) was done partly in order to explore how best to enhance other UI areas. That work should start coming through soon. One of the large scale efforts that will be starting with ObjectStudio is an effort to pull the UI out of the VM, and into the image and associated DLLs. That effort will make it easier to replicate the same pattern on other platforms. When you take the Objective-C interface we added for the Mac alongside this, it lays the foundation for future work. So ultimately, we now have a better base on which to move a lot of these things forward. We like this level of feedback; it certainly helps keep us grounded in reality :) James Robertson Cincom Smalltalk Product Evangelist http://www.cincomsmalltalk.com/blog/blogView Talk Small and Carry a Big Class Library On Nov 30, 2009, at 9:26 AM, Maarten MOSTERT wrote: >> I'd rather see a skinnable Uber L&F that can be easily configured and >> changed to individual needs. A new Look (on the same platform) often >> just means new widget images. This would also open up the possibilty >> for corporate design. > > With its easy to modify Artist classes this was exactly what Widgetry provided. > > @+Maarten, > > _______________________________________________ > 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 |
Re: [vwnc] Does Cincom give up the Desktop ?
|
|
In reply to this post by Andre Schnoor
++2 on nixing native widgets.
VisualAge had a good multi-platform native widget implementation based on Motif. Start forking application processes and then using semaphores to synchronize; on first wait from UI action you discover the limitations. Semaphores didn't behave well because of UIProcess event synchronization from OS callbacks. "Invalid operation during callback" errors when in OS callback and semaphore attempts to wait. There was no fix without VM work or a lot of low-level event re-queuing. Much as I'd love to see native widgets for VisualWorks, I'd be concerned that it wouldn't happen and that the end result could be as limited as VisualAge was. There is too much work and cost involved in adding native widgets to VW. Widgetry is an example of what happens when too much is attempted without resources or will to carry through. Pollock/Widgetry work had been an enduring excuse for decisions to defer fixes and improvements to Wrapper. Many good contributions were simply deferred and lost over several years, and ultimately Widgetry itself died with them. It was an anticipated outcome on a path reluctant to change; after a point, one decides to sit back and watch the wreck or be amazed in the event of survival. VW needs to evolve through practical and incremental changes where no single step is beyond what customers will easily migrate to. Consider how most software now evolves through the ability to discover and update itself. Get a process of incremental change in place before stepping toward radical change. Most of the fixes currently in VW releases should be through incremental updates; releases are for changes that affect behavior of existing interfaces or add marketable functionality. Development environments should evolve. Public and private protocols should be easily identified. Paul Baumann > -----Original Message----- > From: [hidden email] [mailto:[hidden email]] On Behalf Of Andre Schnoor > Sent: Monday, November 30, 2009 8:50 AM > Am 30.11.2009 um 13:51 schrieb Mark Plas: > > I'm not sure whether going for native widgets is a good idea. > +1 This message may contain confidential information and is intended for specific recipients unless explicitly noted otherwise. If you have reason to believe you are not an intended recipient of this message, please delete it and notify the sender. This message may not represent the opinion of IntercontinentalExchange, Inc. (ICE), its subsidiaries or affiliates, and does not constitute a contract or guarantee. Unencrypted electronic mail is not secure and the recipient of this message is expected to provide safeguards from viruses and pursue alternate means of communication where privacy or a binding message is desired. _______________________________________________ vwnc mailing list [hidden email] http://lists.cs.uiuc.edu/mailman/listinfo/vwnc |
Re: [vwnc] Does Cincom give up the Desktop ?
|
|
In reply to this post by Mark Plas
Hi,
1. I second that the most important is to first provide good transparent paint, anti aliasing and linear/radial gradients. Then, at least, one has the opportunity to fix the problems with the bad look oneself. 2. I think that the integration with the surroundings, as C, Java and ObjectiveC must be improved and more "off the shelf" not requiring a lot of tuning before getting it work. 3. I hope that the irritating blocking of VisualWorks processes while moving a window around on the screen will be removed. 4. I would like to see (better) documentation of Glorp. 5. easier deploying mechanisms, better support for "team development" and Continous Integration. 6. improvement of the GUI FW. Björn On 2009-11-30 13:51, Mark Plas wrote: > Hi, > > What I think that's most importantly missing in VW right now is: > > 1) The possibility to use advanced graphics drawing features > 2) Improved native window support > > Regarding the first point, VW should be extended with at least the possibility to use: > - transparent paints/images > - linear/radial gradients > - anti aliasing support > - dashed lines/curves > - some more advanced 'path' drawing. ie. not just a polygon, but something that can consist out of bezier curves, lines, arc segments, and that can be filled and stroked (also with dashed lines) > > Cairo supports all of this, but it uses purely software based rendering - no hardware acceleration. I'd rather see the existing GraphicsContext extended with these features, relying on either the OS (which will use HW acceleration when available) or based on OpenGL. This will make the drawing much faster than is now the case with Cairo. > > For the 2nd topic, apart from the ideas of Andre Schnoor (Tool palettes, drawers, sheets, ...), you should be able to completely decorate a window yourself, ie. start of with a blank window with no title/border/close/min/max, and customize it yourself in VW. Also, a way to give a window a non-rectangular shape would be great. All of this in a platform independent way of course. > > Perhaps an idea here could be to move the window related code out of the VM into the image. As OS-es evolve, it will be easier to change and extend the code in Smalltalk than in the VM. > > Once you have this, you can improve the look&feel of your widgets and applications. > > On the subject of native widgets: I'm not sure whether going for native widgets is a good idea. I don't even know what a native widget is. My computer is running on Vista, but when I look at outlook, the entire user interface looks different than the one I see in the file explorer. Not even the scrollbars or the window borders are the same! The same goes for MS Word and Excel and if you're starting to use some of the Vista administration tools, the widgets you see there don't look at all like the ones you see in the aforementioned examples. If you open MSN messenger it looks completely different again. And these are all applications from the same manufacturer who also made the OS. > > Based on this, I don't think a native look exists (at least not on windows) and I wouldn't want Cincom to put effort into that direction. I think it's more important to offer the basic tools to be able to create your own user interface look. > > -Mark > > > _______________________________________________ > 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 |
Re: [vwnc] Does Cincom give up the Desktop ?
|
|
In reply to this post by Andre Schnoor
On Nov 28, 2009, at 11:27 AM, Andre Schnoor wrote: > Travis, > >>> Just a thought: If Pango could be enhanced to support native fonts, >>> contributing this feature to the Pango project would probably be a >>> good investment. >> >> I'm not sure what you mean Andre. Pango supports multiple backends, >> including OSX. In what way do you want it to be "native"? > > IIRC, last time we talked about Pango it was impossible to use, for > example, the Lucida Grande font (at least from within Smalltalk). > Looking at the Pango site now, I see it claims to support native > fonts through ATSUI. That sounds like a great progress. > > Sorry, obviously I missed that. Yes, must have been a misunderstanding. It's always been possible to use Lucida Grande and compose with it's fractional font widths, at least to the best of my understanding. Sorry for any confusion I may have caused historically. -- Travis Griggs Objologist Light travels faster than sound. This is why some people appear bright until you hear them speak... _______________________________________________ vwnc mailing list [hidden email] http://lists.cs.uiuc.edu/mailman/listinfo/vwnc |
Re: [vwnc] [Bulk] Re: Does Cincom give up the Desktop ?
|
|
In reply to this post by Madhu-21
I'm afraid that explanation doesn't help me to understand what it is that
you are having a problem with. See below.
My intention was not to say that ObjectStudio 8 doesn't support ActiveX controls. My intention in that matter was the following: It is? As far as I know, ObjectStudio 8 has supported ActiveX controls all along. VisualWorks has not, but the 7.7 release will. Can new users get the latest ObjectStudio 8 and try out ActiveX controls? The I don't understand this. For ObjectStudio Classic, the answer for both questions above is Yes. So my Yes, we're well aware of that, and we have a project underway to do a major reworking of the ObjectStudio widgets to bring that up to speed. The issue there is primarily that the widgets were written with Microsoft's preferred technology of the time (MFC) which they have since deprecated, and the widgets really need to be pulled out of it. So that work is in progress. Modelling tool: Using Glorp would not be a problem with the Modelling tool. In fact, the revised Mapping tool will use Glorp, and it works with the Modelling tool. Designer: Could you be more specific? IDE: I believe that item is the same thing you're asking for 3 points earlier, updating the native widgets. Regards, --
Alan Knight [|], Engineering Manager, Cincom Smalltalk
_______________________________________________ vwnc mailing list [hidden email] http://lists.cs.uiuc.edu/mailman/listinfo/vwnc |
|
|
>>>>I'm afraid that explanation
doesn't help me to understand what it is that you are having a problem
with.....
I completely agree ObjectStudio 8 supports ActiveX controls, its that some ActiveX controls which were working in ObjectStudio Classic do not work anymore with ObjectStudio 8. I had reported some of the problems to the ObjectStudio mailing list. >>Designer:
The designer had never been friendly for users. But over the time I got used to it. Now with newer ObjectStudio versions, it has become even more unfriendly. >>>>Could you be more specific?
Here are few specific examples: For my tests, I used ObjectStudio 8.1 on Windows Vista Ultimate.
1. Start ObjectStudio, create new controller, add a new static text to the
form, double click on it. You would observe that the current focus is in the
Name field and the name is not shown completely. Press 'Home' key, then it would
be shown.
2. Start ObjectStudio, create new controller, add four buttons to the form, holding the control key, click on each button to select all buttons. You would observe that its not visible which buttons are selected and which are not. 3. Start ObjectStudio, create new controller, add four buttons to the form, place them horizontally next to each other, now with the mouse try to select the two buttons on to the right, you would observe that instead of selecting the two buttons on the right, the two buttons on the left would be selected (sometimes only the left most button is selected). These problems might be small ones but they are quite annoying. Problems
like these started to appear from ObjectStudio 7.1.2 or 7.1.3 onwards.
>>IDE: In http://www.cincomsmalltalk.com/userblogs/cincom/blogView?content=roadmap this was in the priority list for ObjectStudio 8: Update the look of the IDE with modern widgets. I would have preferred the feel of IDE to be on the priority list rather than the look of the IDE. From time to time, I work with .net IDE, I still prefer ObjectStudio classic class browser over the nice looking .net IDE. >>>>I believe that item is the same thing you're asking for 3
points earlier, updating the native widgets.
Not really. 3 points earlier I was talking about the end user application look and this one is regarding the IDE. For me, in the case of IDE, the first priority would be the feel and then the look. Here's an example of what I mean: In the refactoring browser, copying some text to cliboard and then using Ctrl+V to paste it in the browser would select the pasted text. One day I saw James posted in one of his screencasts "Platform Faithful Pasting in VW", then I asked if we can make the cursor to blink, Michael replied saying that its there in public store but not integrated. In my opinion, these things should get priority so that these are made available by default in ObjectStudio and for existing ObjectStudio users, it would be closer to how they are used to work with Classic versions. Regards,
Madhu. From: [hidden email]
Sent: Monday, November 30, 2009 11:30 PM
To: [hidden email] ; [hidden email]
Subject: Re: [Bulk] Re: [vwnc] Does Cincom give up the Desktop
? My intention was not to say that ObjectStudio 8 doesn't support ActiveX controls. My intention in that matter was the following: It is? As far as I know, ObjectStudio 8 has supported ActiveX controls all along. VisualWorks has not, but the 7.7 release will. Can new users get the latest ObjectStudio 8 and try out ActiveX controls? The I don't understand this. For ObjectStudio Classic, the answer for both questions above is Yes. So my Yes, we're well aware of that, and we have a project underway to do a major reworking of the ObjectStudio widgets to bring that up to speed. The issue there is primarily that the widgets were written with Microsoft's preferred technology of the time (MFC) which they have since deprecated, and the widgets really need to be pulled out of it. So that work is in progress. Modelling tool: Using Glorp would not be a problem with the Modelling tool. In fact, the revised Mapping tool will use Glorp, and it works with the Modelling tool. Designer: Could you be more specific? IDE: I believe that item is the same thing you're asking for 3 points earlier, updating the native widgets. Regards, --
Alan Knight [|], Engineering Manager, Cincom Smalltalk
_______________________________________________ vwnc mailing list [hidden email] http://lists.cs.uiuc.edu/mailman/listinfo/vwnc |
«
Return to VisualWorks
|
1 view|%1 views
| Free forum by Nabble | Edit this page |