Are ASP.NET user interfaces essentially dead now?

[Rob R. Ainscough]

VS 2005 I have:
ClickOnce deployment
User's that hate and or don't want to use an IE Client (don't blame them)

I don't see how ASPX web pages are going to survive? With .NET 2.0 and
clickonce deployment my app is 427KB (even with modem dialup speed it
doesn't take long to download) -- the user gets a very friendly secure
WindowsForm app (most of them don't even notice they're not under IE
anymore) that performs considerably faster than any ASP.NET app.

The road map as I see it:
IE client -- back to static just clickonce links that open up
WindowsForms apps
Vista -- .NET 2.0 built in (no need to download)
WebServices -- called from WindowsForms apps (keeps it secure and
firewall friendly)
WindowsForms are a HELL of a lot more secure (no IE attached activex
components, no data miners, click monitors, etc. etc.)

The way I see it -- user interaction is going to move back to WindowsForms
since the IE client by definition is just NOT user friendly, NOT programmer
friendly, and has a ton of other issues surrounding it in terms of security
and performance and flexibility.

Don't get me wrong, web development will still exist (web services and basic
static content), but I believe anyone doing serious business applications
using the web will migrate to this approach -- it really is a win win.

 

[Peter Rilling]

And what about Linux and Mac and other platforms. It is arrogant to assume
that all people visiting ASPX pages are only Windows users.

All technology has its place.

Not sure what you mean by "IE client by definition is not user friendly".
The browser has nothing to do with whether a site is friendly or not. Now
being a website does limit its potential since it can only stream HTML and
other browser technology, but if you program it right, then it works fine.
I have seen some rich clients that would make Frank Lloyd Wright spin in his
grave.

 

[Bruce Barker]

thats MS's hope. there are a couple issues

1) no MS o/s comes with .net installed. its currently an addon (22mb
download)
2) click once is a great way to spead viruses (as the code does not run in a
sandbox), and MS has not tackled the author validation issues.
3) requires ms o/s (apple is still > 6% share)
4) many company firewalls will block click once (maybe even smart clients)
5) click once only supported by IE (<90% share).
6) more personal devices with browsers

with these issues, and ajax getting popular again, browser apps will be
around for a while.

-- bruce (sqlwork.com)

 

[Marina]

However, using winforms requires that the client has the .NET framework
installed. At this point that is a big leap to take.

Where as an ASP.NET app can work just fine on other browsers/platforms with
proper care by the developer, there is no way a developer can make winforms
work on a platform that does not have an implementation of the .NET
framework installed.

And yes, while the application in winforms will be secure, it may be to
secure. If the sandbox is too tight, the application may not be able to
make a web service call back to the web server for example. Now the user has
to go configure security on their machine - what average user is going to
know how to do that?

Another issue is someone checking their email or the news from a kiosk.
Those machines are locked down pretty tight - the odds of them being open
enough to run your winforms application? I don't know.

Personally, I don't see winforms as a viable option for web development for
the forseeable future. But that's me, I could be way off.

 

[Rob R. Ainscough]

As I understand it .NET 2.0 is going to be available on both those
platforms. Not being arrogant, being realistic -- unless your business
targets that specific small group of people but it really is just a matter
of where the money is at.

The browser (IE or whatever you choice) is a NOT a user friendly
environment -- some users don't even know where the "Back button" is, nor
where the settings are in IE to enable a site to work, nor the settings to
enable pictures to display or not. Then there is the issue of the 100's of
activex components that get loaded and install. If you hadn't checked
around 1 in 5 people use the web, that is pretty pathetic. The number #1
reason they don't use it is because their affraid of spyware and viruses, #2
is because they don't understand how to use it -- and I don't blame them.

Sure I can code a great site but I can't manage the client's IE. I've seen
rich clients that take minutes to load and many that don't even render
because of client (IE) issues (either not updated, missing patches,
corrupted activex components, etc. etc.). It maybe too late to bring back
user confidence, but continuing with HTML rendered intefaces is definitely
NOT the future -- it'll be a combination of HTML web site for static links
to the real interface (.NET 2.0 windows forms apps) to do the work and do it
faster, more secure, and with a hell of a lot less development time to
implement.

 

[Rob R. Ainscough]

As I said,

1) Vista includes .NET 2.0
2) click once is a managed process and will be more secure WSE 3.0
3) .NET 2.0 will be available for mac/linux
4) not relevant -- no more blocking than using web services
5) see 3)
6) .NET 2.0 compact framework for smart devices

No doubt browsers will continue but they will be doing less and less of the
interactive interface work -- they are fundamentally not suited to get the
job done well along with the myraid of other issues surrounding IE client
(or whatever client you choose). Sure I can spend 6 months working on web
based app interface, or I can spend 2 months for the same windowsforms based
interface and get better functionality with equal exposure to all types of
platforms/users.

Continuing using IE for all the interface work is NOT gonna help bring in
more customers/users -- most users that see the "Your going to an unsecure
web site...blah blah" click on No and that's that -- potential customer
gone. Why, because they have there security setup to do that and they
didn't even know it -- could have happened because they installed Norton
Internet Security. But think about it -- what do you see lining the store
shelves -- it is at the point where there are move Antivirus, spyware,
internet security tools than another other combine form of
applications/games.

Be a user for a day and take yourself out of the developer role -- walk into
your local Fry's Electornics or BestBuy or Compusa or whatever
software/hardware superstore and look around -- 50% of the software has
something to do with keeping you system "clean" -- this is a pretty sad
state of affairs. Sticking with an IE client to do the interface ALL the
interface work is a dead end regardless of your platform.

Rob.

 

[Rob R. Ainscough]

This seems to be a recurring theme in all the responses:

"...browsers/platforms with proper care by the developer..." there is
proper care and then there is jumping thru hoops, you want to avoid the hoop
jumping. But beyond that, no matter how much care you put into the ASP.NET
app it has several flaws:

1. it relies on a browser that can be modified by other malicious web sites
2. takes longer to develop ASP.NET interface than a windowsform app
3. slow and inefficient even if you do the work to cache everything just
right, the rendering will aways be slower
4. there more things you can do in a WindowsForm interface that you either
can't do in a web page or requires consider work and effort to do it

If the sandbox is that tight, they probably can't do much with the PC
regardless. And communicating with a Web Service is not required the
developer can choose how they want to communicate -- direct to SQL servers,
web services, or even local services that might be on that PC.

I see Winforms doing the major amount of interface work and leaving the web
pages for mostly static work -- just a way to get to the clickonce link.
Ultimately the folks paying the development bill want the fastest solution
possible with the features they need. ASP.NET is not that solution.

Rob.

 

[Peter Rilling]

How does "clickonce" solve the problem up version updates? One nice thing
about a webserver, is the site can be updated without the user having to do
anything.

 

[Marina]

ASP.NET doesn't rely on anything more then any other web technology does.

As far as the sandbox, it will be tight enough that you can't call SQL
server, web service, or anything local. I gave web services as an example -
but it won't be able to do plenty of other things. That is just how it will
be, to prevent the user from running some program that is then going to run
some malicious code. So, you are not going to gain anything here, as you are
either going to have to have a user that is sophisticated enough to set
security policy or is an admin - both of which are highly unlikely for most
users.

This is if the user is going to have the framework installed to begin. This
is a big if. So what if Vista comes with it. That is meaningless. What about
everyone on xp, 2000, 98 (yes, that happens), linux, mac, etc? That is a lot
of other people who dont' have the framework by default, and in some cases
can't, unless somehow it gets implemented on other platforms. I have not
heard anything about this being a definite - and even if it happens, again,
people have to actually install it on their machine!

I think what you are saying sounds all nice in concept. But in practice, it
is just not feasible in many scenarios at this point.

I don't understand the point of this post, really? Is it to convince us of
something? I mean, you can say that everyone on this newsgroup is wrong and
do what you feel is best, and that's that. Do you feel a need to convert us
all?

How about in 2 years, we revisit this issue. My guess is, HTML will still be
around, ASP.NET will still be around. WinForms will be making headway in
very specific narrow types of situations where they are a good fit - but
they certainly will not replace the web.

 

[Marina]

Not that I agree with Rob. But, clickonce solves it in terms of the user
receiving the latest copy next time the app is run. Because in theory the
first thing that should happen, is the applications checks to see if there
is a new version on the web server, and downloads it if there is. Until the
app is run again, it doesn't really matter if the old version is still
sitting there since it isn't being run anyway.

 

[Sylvain Lafontaine]

The same thing has been told 3 years ago with Flash: with the capability of
displaying fully animated web pages with Flash, how could HTML survive?

For some data or graphical applications, using ClickOnce might be a good
idea (isn't it for this that it has been developed?) but there is a lot more
about HTML and ASP.NET (PHP, ColdFusion, etc.) than just barebone data
editing.

 

[Wiktor Zychla [C# MVP]]

As I understand it .NET 2.0 is going to be available on both those

platforms. Not being arrogant, being realistic -- unless your business
targets that specific small group of people but it really is just a matter
of where the money is at.

I doubt .NET will ever be available on Unix platform to the extent where you
could think of stable WinForms applications. As for target groups - if you
target the Win-based intranets then everything is beautiful. If you target a
global web you cannot just say that Unix-users do not exists or you just do
not want to see them. Do you imagine google as a click-once Windows desktop
application?

Wiktor Zychla

 

[Rob R. Ainscough]

At some point a port has to be open if the end user wants contact with the
outside world. If the user is VPN then no issues at all.

I think my statement is about the future, not Today -- I'm more interested
in the future as today has got user acceptance problems which I assume you
don't believe -- however the statistic are available on the web if you want
to see why people aren't using IE (of any flavor) and jumping on the web.
Something needs to be done and keep a concept/technology like and IE client
as the only way to do it is not good for anyone. The future will be a
hybrid of the two IMHO.

It is feasible today and being used today -- I'm not sure why you think
people are OK with having a IE Client but NOT ok with have .NET framework?
Dialup download might be a little painful, but not unreasonable at 22 MB --
certainly no more unreasonable than SP2 or the numerous other software
patches (on any platform).

I agree in 2 years HTML will still be around so will ASP.NET, but those
technologies will become redundant and as those spending the money on
development start to see this other approach that is better, costs less,
then they will start to migrate also. HTML will be around for a long time,
but it is clearly a very worky and dependant technology.

Put a check mark in your calendar and wait, watch, and see.

 

[Rob R. Ainscough]

If you want to do any serious graphics processing, IE client and the
numerous component downloads (like Flash) is not going to approach the power
of DX being used in a WindowsForm -- and when Vista is released the
graphical potential via a clickonce WindowsForm app using read 3D
accleration via DX will simply put Flash to shame. AND this is where the
advertisers will want to go, not some 2D flash animations and mouse hovers &
clicks.

I agree that HTML isn't going away overnight, but it's days are numbered and
it's roll in the future will be a much smaller one (as it should be). Lets
be real, we've tried to expand and extend HTML to do what it was never
intended to do -- it has been tweaked to the point of bloated exhaustion and
rendering nightmares, it is burdened by the drag of being platform
independant. And is anything truely platform independant, NO -- but we seem
to strive endlessly to try and make it so at untold cost & productivity
where the needs of the few seem to out weight the needs of the many
(different topic for a different day).

Rob.

 

[Edwin Knoppert]

I would thank the day MS or other invents something much better for
browsers.
Like a windows system in browser with hyperlink navigating and still
considered thin client.
Better develop support.
Not the nonsense html stuff we have now.
I guess java goes a bit in that direction, well i don't know much about that
though.

 

[Bruce Barker]

number 4 is relevent - a firewall can detect applet downloads and block
(typically you strip the tag). as click once installs the app on the hard
drive and it runs in the open sandbox its as dangerous as active/x.

vista will allow companies block click once, except from trusted (internal)
sites.

microsoft has never suggested .net support for any o/s other than windows.
there is an educational version for another o/s, but you are not allowed to
write commerical software for it.




-- bruce (sqlwork.com)

 

[Kevin Spencer]

It is interesting to me that, when Microsoft adds another technology to the
toolbox, so many people seem to think that it is somehow supplanting or
replacing some other technology, or making it obsolete. I don't see how
anyone can assume that ClickOnce is going to replace ASP.Net, or make web
applications obsolete. It is simply another tool in the toolbox. No doubt it
will be found useful for many situations, and that it will replace ASP.Net
solutions in a number of instances. But from what I've read, even from
Microsoft, it is not expected to supplant ASP.Net, or in any way make web
applications go the way of the dinosaur. In fact, ClickOnce is just one of
many emerging Microsoft technologies that, rather than limit us, provide us
with more choices. It provides us with the ability to create solutions that
are more appropriate to the given requirements. Among these solutions, some
will be ASP.Net, some will be ClickOnce, some will be Web Services, some
with be Indigo, some will be XAML, etc., etc.

Think of it this way. There is a peg board out there of requirements. Some
of the holes are square. Some are round. Some are trapezoid. Some are oval.
Now, ASP.Net is like a round peg. We have had to shave our ASP.Net pegs a
bit to get them to fit into the oval holes. Now, Microsoft has created an
oval peg. I don't think it would fit well in the round holes. But it is
certainly easier to fit it into the oval ones.

--
HTH,

Kevin Spencer
Microsoft MVP
..Net Developer
You can lead a fish to a bicycle,
but you can't make it stink.

 

[Patrice]

Not exactly but an interesting move is that Vista allows to descibe a
Windows UI using XML markup. As such you can have a web site serving this
markup and the user see a Windows UI.
Also it's likely that the web will progress one day (for example through
XForms). I'm optimist by nature as you see...

Really depends what we are doing but for maintream modules having a web
application is IMO still cleaner than having a buch of local applications,
it would be likely overkill. We do have also local applications when the
requirements are heavy regarding the UI...

 

[Rob R. Ainscough]

Google doesn't really have much of an interface? Even if .NET 2.0 never
appears for Unix -- which I believe it will give the resource tossed into
this by Microsoft -- I thought Unix had Windows emulators?

But like I said, the more complex interface the less the fit to a IE type
client -- the Global case is no different, Windows is everywhere and is
dominant, that's why uncle Bill has a global size wallet.

 

[Rob R. Ainscough]

I didn't say that, it is removing some of the load from one dated and over
extended technology into another that is more suited at the task.

And Microsoft soft originally said VB6 would migrate to .NET 100% -- and
they changed their minds later on.

The goal is to use IE as little as possible and be more secure -- bring IE
back inline with its roots and original intent is a good thing.

But think about the interface technology and the end user -- I can't see
ASP.NET, HTML, etc. etc. surviving -- they invented Cache management, place
more things for the developer to think about that don't directly apply to
working complex solutions.

Besides why use up server bandwidth and processor/memory resources when you
have a client with the power and resources to do the interface processing?
Sending a ton of the same rendering information back and forth -- this is a
good thing? No it's poor way to accomplish a task and uses up precious
bandwidth. The entire concept is bad.

I just finish discussing clickonce technology with several large and small
clients and they instantly WANT IT NOW! I don't think you realize the
frustrate end users go thru with their client browsers -- they hate it and
IT departs hate the support they have to perform on it -- give IT a choice
of a firewall setting vs. supporting a client browser, they opt for the
firewall setting every time -- life would be so much simpler.

But you point the exact flaw, XAML, AJAX, more and more and more of the
same -- lets just load up the server with as many services as possible til
64GB of memory and 4 64bit processors just isn't enough for the server to
even boot.

ClickOnce is a lot more than pushing out the developer's favorite tool of
the day -- hey I'm in a Java mood today lets manage mgmt to convert over to
Java, oh now I'm in a ASP.NET mood, I'm bored now lets try Indigo, how about
XAML, everything has to be SOAP...it is getting silly. End result is a
business stuck with maintaining something that is now "out of fashion".