XAML is inherently object-oriented since its elements represent CLR classes.
This means that an element derived from another XAML element
inherits attributes of its parents. For example, a System.Windows.Controls.Button
derives from System.Windows.Controls.ButtonBase, which derives from System.Windows.
Controls.ContentControl, which derives from System.Windows.FrameworkElement,
which derives from System.Windows.UIElement.
Therefore the Button element ( <Button..../> in good old XAML jargon...) has very few attributes of its own
but still boasts a lenghty list of attributes that it has inherieted from
classes above it in the hierarchy, such as Width and Height.
XAML in a nutshell ... Page 4 Chapter 1: Introduction to XAML
About renaming Blend... Why not Blend _ER (for those who like cooking...)
or even Blend_edict XVI... (for those who like the Pope...) Come on be serious...
This is a wish list so I'll answer any questions about my understanding via email - if anyone wants a "debate" with me.
I look forward to all of the cool features that will come out in 5.
Hopefully, in parallel, I'll get my Blend_O.
Software life cycle:
Requirements: fix this; Design: break it down; Devel: get it working then fix it; Testing: if it ain't broke don't fix it; Help: it's broke, here's how to fix it; Sales: it will be fixed in the next version.
The MS Program manager need to watch some experienced developers new to SL try to develop a silverlight app. It's painful. (I'll put down how many hours I couldn't bill due to these)
Use the browsers' CultureInfo/locale info automatically. No code should have to be written. (0.5 hour)
LOB controls (DataGrid and DataForm) in SL Toolkit should include a BusyIndicator by default, and should be bound automatically. (0.5 hour)
Provide just ONE standard LOB blueprint, for the simplest Departmental Database. You'll find enough surprises to fix a lot of issues. The following lists the bugs one developer found over several days.
NavigationPage and security requires a lot of tweaks. e.g. "Keep me logged in" is async, so we cannot do security checks until user state is known.(5 hours)
Provide a fix for DomainDataSource with ComboBoxes, Kyle McClellan's fix should be rolled into the toolkit. See [url]http://blogs.msdn.com/b/kylemc/archive/2010/06/18/combobox-sample-for-ria-services.aspx[/url] (3 hours)
I don't know why, but DataForm ComboBoxes requiring datasources have to be loaded from a DataSource inside a DataForm control. (3 hours)
Package with a SDK with a deployment troubleshooting tool. I've got a troubleshooter for basic WCF, but still need some help pulling out information from the IIS metabase. (5 hours). The WCF stack needs to provide better logging than just 404, and
it should log in a hosting scenario where users do not have access to a lot of tools. See [url]http://www.redmountainsw.com/wordpress/archives/asp-net-4-0-silverlight-hosting-tweaks[/url]
Figuring out SQL server issues with my hoster set up. The thousands of linux hosters have a basic system for accessing the database. There is no consistent way of doing this with ASP.net. You need to provide a standard hosting setup that the hosters
can copy into their environment. One-Click is good, but you need to make this as easy as possible for the hosters. (2 hours)
SqlWebAdmin needs to be hosting friendly.(1 hour)
The toolkit controls need to support a "print" mode. It is so easy in HTML, and we miss being able to print things for free. Specifically, it will strip away scrollviews, replace textboxes which overflow in screen with an expandable textbox. If we
can mark UIElements as non-printable, then logos etc can be stripped off. It doesn't matter if pages don't break properly. We would like to see SL being back to parity with HTML.
If you are a MS Program Manager, feel free to contact me and I'll give you a list of basic issues that SL needs to address before working on fancy features.
Re: The MS Program manager need to watch some experienced developers new to SL try
to develop a silverlight app. It's painful. The MS Program manager need to watch
some experienced developers new to SL try to develop a silverlight app.
It's painful. (I'll put down how many hours I couldn't bill due to these)
Holly Cooooooow... 19 hours spent stocked on those issues... Are you kidding me?
Silverlight might not be a 100% perfect but in your case I'm sorry to say that
MS Program Manager aren't maybe the ones who should bare the blame...
You believe you have a list of brilliant ideas and solutions to issues in mind... Well then be our guest...
MS Program Managers like you call them... have put on line a Web site especially for that... http://silverlight.mswish.net
You propose a brilliant idea and if other "experienced developpers" feel that it's worth it...
they'll vote them up...it's just as easy and simple as that...
Thanks, MawashiKid. I note that you spend a lot of time on the forum helping people out. So I have a measure of respect for your abilities. Don't you think there are some aspects of Silverlight that should be made simpler, e.g. troubleshooting deployment issues?
The MS devs spend a lot of time on these forums asking people to check a number of settings when these could have been automated.
I've programmed in lots of environments with lots of tools over my 35+ years as a professional programmer. As I was reading your list of points, the thing I was thinking was, "It takes a LOT LONGER than he is saying to solve those problems and ones like
them!" I think you have raised really excellent points.
For instance, I have just spent about 3 DAYS trying to figure out how to get validation to work in my current application. First was trying to understand the new INotifyDataErrorInfo mechanism, which, as it turns out is relatively simple. But then I tried
a bunch of tests that revealed that the Validator.ValidateObject, which I was already using, doesn't work well. I have a UI with several tabs and bound fields on each one. When I tell the Validator to validate the object, I expect it to validate the object.
However, if a validation error condition exists and it is on a tab that isn't currently visible, it doesn't throw an error.
It is these kinds of things that waste unbelievable amounts of time. The whole attribute validation mechanism seems half-baked for real life applications.
I find myself spending DAYS trying to figure out why some quirky things don't work as expected.
Like Theming. Why is it that legends in charts don't display correctly in half the themes? Why can't someone LOOK at the SL samples that SHOW that the charts don't theme right? Why don't all the controls, like DatePicker have ValidationStates? Why is
it that using the new implicit theme gives a different result than an explicit theme? etc. etc. etc.
ANYTHING that can be done with Silverlight to decrease programmer time figuring out these sorts of things is worth doing for SL 5 imho.
... time passes ...
I don't know that it makes a difference, but I should point out that the tabs mentioned above each have their own View and are bound to the same datacontext.
Thanks mtiede, for a moment I felt really stupid and alone. :) Goodness, elements not validating when hidden. This is getting pretty bad don't you think? It's like back in the early days of a technology there's 50% of functionality one should avoid, but the
problem is not knowing which.
Hey guys... have you seen Adobe is announcing 3D support for Flash ?
Since 1st version of SL I'm saying that we must have 3D in SL. Xaml did it since WPF, and now, you now what ? Flash will do it before SL !
I can't imagine worst situation.
Flash has good sound management, will have 3D and have the popularity.
SL Team is losing its time on OOB that nobody wants (for a customer, a web app is not a semi desktop app that needs to be installed).
Pleaaaase : SL is a very good product, even with some little problems, but we LOVE IT, we are SUPPORTING IT since first version, but we NEED TO BE LISTEN : stop wasting time on OOB and supply a plugin that can do, AT LEAST, what FLash is doing well since
a long time or what it will support soon.
Our customers doesn't understand one word about RIA services (something marvelous), nor Linq, nor the beauty of Xaml. They just speak this way "Flash can do this and that, can you make a Web app with Silverlight that is doing the same ?". No less, no more.
Please ear what we know about customers. You speak with other developers, cool, but we speak with customers... They do not want any OOB, they want good sound, (with loop), good appealing look & feel (so 3D), printing (invoice, online orders) with preview,
fast loading and no more.
All the rest are technical points for technicians. But we do not sell anything to technician, we have customers that know very well what Flash did, is doing, and will do.
Let us offering at least the same things. Sincerely, other goals are just wasted time, only one thing is important : what our customers want to buy. Other debates have no meaning.
please....
Olivier Dahan (MVP Silverlight 2011 , MVP CAD 2010, MVP C# 2009)
www.e-naxos.com
Blog: www.e-naxos.com/blog
English is not my native language so be indulgent and do not hesitate to ask for clarification! Thanks.
Ziggo
Member
4 Points
5 Posts
Re: Re: Re: Silverlight 5 Wishlist
Jul 05, 2010 03:18 PM | LINK
MawashiKid
Contributor
4165 Points
806 Posts
Re: Re: Silverlight 5 Wishlist
Jul 05, 2010 03:50 PM | LINK
RE: Make Silverlight more Object Oriented Friendly....
Maybe you should spend a bit of time reading this...
http://books.google.ca/books?id=6t8FE32tuTgC&pg=PA4&lpg=PA4&dq=XAML+object+oriented&source=bl&ots=1A2uoci7EX&sig=FkjPgyHPcHkUBeKrfvInE54smvk&hl=en&ei=5PIxTMKtNoOC8gad4J2QAw&sa=X&oi=book_result&ct=result&resnum=3&ved=0CCEQ6AEwAg#v=onepage&q=XAML%20object%20oriented&f=false
Paragraph starting with...
XAML is inherently object-oriented since its elements represent CLR classes.
This means that an element derived from another XAML element
inherits attributes of its parents. For example, a System.Windows.Controls.Button
derives from System.Windows.Controls.ButtonBase, which derives from System.Windows.
Controls.ContentControl, which derives from System.Windows.FrameworkElement,
which derives from System.Windows.UIElement.
Therefore the Button element ( <Button..../> in good old XAML jargon...) has very few attributes of its own
but still boasts a lenghty list of attributes that it has inherieted from
classes above it in the hierarchy, such as Width and Height.
XAML in a nutshell ... Page 4 Chapter 1: Introduction to XAML
About renaming Blend... Why not Blend _ER (for those who like cooking...)
or even Blend_edict XVI... (for those who like the Pope...) Come on be serious...
K2P2
Participant
1134 Points
401 Posts
Re: Re: Silverlight 5 Wishlist
Jul 05, 2010 04:36 PM | LINK
Thanks guys for responding.
This is a wish list so I'll answer any questions about my understanding via email - if anyone wants a "debate" with me.
I look forward to all of the cool features that will come out in 5.
Hopefully, in parallel, I'll get my Blend_O.
Requirements: fix this; Design: break it down; Devel: get it working then fix it; Testing: if it ain't broke don't fix it; Help: it's broke, here's how to fix it; Sales: it will be fixed in the next version.
teyc
Contributor
2401 Points
606 Posts
Re: Silverlight 5 Wishlist
Jul 07, 2010 11:35 PM | LINK
The MS Program manager need to watch some experienced developers new to SL try to develop a silverlight app. It's painful. (I'll put down how many hours I couldn't bill due to these)
If you are a MS Program Manager, feel free to contact me and I'll give you a list of basic issues that SL needs to address before working on fancy features.
MawashiKid
Contributor
4165 Points
806 Posts
Re: Silverlight 5 Wishlist
Jul 08, 2010 12:54 AM | LINK
Re: The MS Program manager need to watch some experienced developers new to SL try
to develop a silverlight app. It's painful. The MS Program manager need to watch
some experienced developers new to SL try to develop a silverlight app.
It's painful. (I'll put down how many hours I couldn't bill due to these)
Holly Cooooooow... 19 hours spent stocked on those issues... Are you kidding me?
Silverlight might not be a 100% perfect but in your case I'm sorry to say that
MS Program Manager aren't maybe the ones who should bare the blame...
You believe you have a list of brilliant ideas and solutions to issues in mind... Well then be our guest...
MS Program Managers like you call them... have put on line a Web site especially for that...
http://silverlight.mswish.net
You propose a brilliant idea and if other "experienced developpers" feel that it's worth it...
they'll vote them up...it's just as easy and simple as that...
teyc
Contributor
2401 Points
606 Posts
Re: Re: Silverlight 5 Wishlist
Jul 08, 2010 07:01 AM | LINK
mtiede@swtec...
Contributor
5273 Points
2200 Posts
Re: Re: Re: Silverlight 5 Wishlist
Jul 08, 2010 11:40 AM | LINK
teyc,
I've programmed in lots of environments with lots of tools over my 35+ years as a professional programmer. As I was reading your list of points, the thing I was thinking was, "It takes a LOT LONGER than he is saying to solve those problems and ones like them!" I think you have raised really excellent points.
For instance, I have just spent about 3 DAYS trying to figure out how to get validation to work in my current application. First was trying to understand the new INotifyDataErrorInfo mechanism, which, as it turns out is relatively simple. But then I tried a bunch of tests that revealed that the Validator.ValidateObject, which I was already using, doesn't work well. I have a UI with several tabs and bound fields on each one. When I tell the Validator to validate the object, I expect it to validate the object. However, if a validation error condition exists and it is on a tab that isn't currently visible, it doesn't throw an error.
It is these kinds of things that waste unbelievable amounts of time. The whole attribute validation mechanism seems half-baked for real life applications.
I find myself spending DAYS trying to figure out why some quirky things don't work as expected.
Like Theming. Why is it that legends in charts don't display correctly in half the themes? Why can't someone LOOK at the SL samples that SHOW that the charts don't theme right? Why don't all the controls, like DatePicker have ValidationStates? Why is it that using the new implicit theme gives a different result than an explicit theme? etc. etc. etc.
ANYTHING that can be done with Silverlight to decrease programmer time figuring out these sorts of things is worth doing for SL 5 imho.
... time passes ...
I don't know that it makes a difference, but I should point out that the tabs mentioned above each have their own View and are bound to the same datacontext.
teyc
Contributor
2401 Points
606 Posts
Re: Re: Re: Re: Silverlight 5 Wishlist
Jul 08, 2010 01:54 PM | LINK
alexblank
Member
4 Points
2 Posts
Re: Silverlight 5 Wishlist
Jul 09, 2010 02:23 PM | LINK
Android support for Silverlight. This way, I can develop for the web / PC, and still leverage my codebase for Win Phone 7, Android, etc.
odahan
Star
8828 Points
1496 Posts
Re: Re: Silverlight 5 Wishlist
Jul 10, 2010 05:50 AM | LINK
Hey guys... have you seen Adobe is announcing 3D support for Flash ?
Since 1st version of SL I'm saying that we must have 3D in SL. Xaml did it since WPF, and now, you now what ? Flash will do it before SL !
I can't imagine worst situation.
Flash has good sound management, will have 3D and have the popularity.
SL Team is losing its time on OOB that nobody wants (for a customer, a web app is not a semi desktop app that needs to be installed).
Pleaaaase : SL is a very good product, even with some little problems, but we LOVE IT, we are SUPPORTING IT since first version, but we NEED TO BE LISTEN : stop wasting time on OOB and supply a plugin that can do, AT LEAST, what FLash is doing well since a long time or what it will support soon.
Our customers doesn't understand one word about RIA services (something marvelous), nor Linq, nor the beauty of Xaml. They just speak this way "Flash can do this and that, can you make a Web app with Silverlight that is doing the same ?". No less, no more.
Please ear what we know about customers. You speak with other developers, cool, but we speak with customers... They do not want any OOB, they want good sound, (with loop), good appealing look & feel (so 3D), printing (invoice, online orders) with preview, fast loading and no more.
All the rest are technical points for technicians. But we do not sell anything to technician, we have customers that know very well what Flash did, is doing, and will do.
Let us offering at least the same things. Sincerely, other goals are just wasted time, only one thing is important : what our customers want to buy. Other debates have no meaning.
please....
www.e-naxos.com
Blog: www.e-naxos.com/blog
English is not my native language so be indulgent and do not hesitate to ask for clarification! Thanks.