From: PJ6 on
OK, fair enough. You make good points.

I'm in the middle of writing an article on factors that affect a developer's
productivity in computer languages and architecture, with a particular focus
on end product stability (especially under change), development time
required, and estimation reliability.

Someone else made a start along these lines talking about what's wrong with
C++:
http://www.ittybittycomputers.com/IttyBitty/CppHarm.htm

Part of the article points out flaws in web development - namely: browsers
need to accept lower-level UI instructions (HTML is too high level, too
crippled), and complex UIs are where you need a compiler validation the
most - referential integrity checks, strong typing, etc.

That looks like you're real email address.

If you're so inclined, I'd like to send you the article and get your
criticism when I'm done before I actually roll it all into a lecture. My
hatred of web development probably clouds my judgment, and I could use some
feedback from someone who has an opposite inclination.

Regards,
Paul

"Mr. Arnold" <Arnold(a)Arnold.com> wrote in message
news:eOEFAgNLLHA.4796(a)TK2MSFTNGP02.phx.gbl...
<...>


From: Registered User on
I believe you both over-value your anecdotal view of developers. My
experience indicates the individual's OO skillset has much to do with
their ability to develop both web or desktop apps. The skills and
practices are eminently transferable between the two.

Generally developers should not be responsible for producing a design
as all too often the design follows the code. This stream of
consciousness methodology can lead to work-arounds on top of
work-arounds when design flaws appear. The ersatz logic for not
correcting the design is always 'too much of an investment' in the
code that has already been written. The first and only path explored
is the one that gets taken, often with less than optimum results.

A proper design is paramount to both web and desktop apps. The problem
with using the term 'web designer' is its ambiguous meaning. There are
major differences between designing a web site and designing a web
application. The former is more about 'look & feel' (brochure-ware)
than anything else.

When pointing fingers at designers or developers, one should always
consider management's role in the entire process. The worst kind of
manager is the one who assumes because they are in charge they
automatically know all the right answers and solutions before they
understand the problem.

regards
A.G.

On Mon, 26 Jul 2010 11:33:52 -0400, "Mr. Arnold" <Arnold(a)Arnold.com>
wrote:

>On 7/26/2010 10:16 AM, PJ6 wrote:
>> "Mr. Arnold"<MR. Arnold(a)Arnold.com> wrote in message
>> news:upAX4JFJLHA.5720(a)TK2MSFTNGP02.phx.gbl...
>>> So?
>>> Compared to real applications -- please?
>>> I work in both environments, and what the desktop flyboy jockeys need to
>>> learn is n-tier, object oriented programming, design patterns, TDD and
>>> DDD in enterprise level development, which can be used at the desktop
>>> just as they are being used in Web based solution. The desktop flyboy
>>> jockeys are no where in the ballpark.
>>
>> Are you serious?
>>
>> When I came out of thick client development I was shocked at the lack of
>> skills the people that called themselves web application "developers" had.
>
>And how many years ago was that? Those days are long gone as .Net
>Architects for Web applications are starting to use the various .NET
>technologies and they are coming into play more and more.
>
>> All they knew was markup, and they spoke of "the code behind" and "scripts"
>> with shades of fear.
>
>Not anymore and particularly so with the usage of WPF, Silverlight and
>WCF RIA.
>
>> None of them had any inkling at all about what
>> object-oriented programming is. I've been all over the industry for well
>> over a decade and I can say that while there's been change, web developers
>> as a whole (not you) remain the sorriest lot of them all.
>
>So, when they have this kind of learning material available and they
>know what they are doing after they use the material, because they can
>see it in action and see the source code on how it's done, then what?
>
>http://www.dofactory.com/Patterns/Patterns.aspx
>http://www.lhotka.net/cslanet/
>
>And I have been on the MS platform since 1996 or so and I have seen
>sorry Windows forms developers to this day -- don't kid yourself now
>about the expertise level of a Windows forms developer over and Windows
>Web developer in today's, because I know better.
>
>> You may personally
>> work with good people, but really - you can't be serious when you say web
>> developers generally have a clue.
>
>I am dead serious about what is happening in today's environment. The
>days you're talking about are disappearing fast.
>
> I have reworked so many *devastatingly*
>> bad designs wrought by web "developers", designs that would make you cry.
>
>I have done the same thing on Windows forms applications, and prior to
>that, on the mainframe platform too back in the 80's and early part of
>the 90's, which was looked at code that was badly written and horror.
>
>> No
>> thick client I have seen has come *close* to the horror, the abject
>> failures, I've seen these people create.
>
>Bad programming and application design is not limited to any development
>environment. And I have seen some nightmares on the Windows desktop for
>applications.
>>
>> And web UI development in general... things that are trivial, take seconds
>> to do in a thick client, things that just work the first time, can take
>> forever to do for a web interface, or are simply impossible.
>
>I and a whole lot of others are not experiencing those issues.
>
>About the only thing a Web UI can't do well is keep state, but that's
>changing too with the HTML5, and there are other way to keep state as well.
>
>> The very idea
>> of using markup and script for an application's UI - where you actually use
>> it to WORK - is flawed at its very core and I can point out in a thousand
>> examples by comparison as to why this is true. Square peg, round hole.
>
>It's not going away, and the Web application's footprint is minuscule,
>as compared to a Windows forms based solution (thin client or not) where
>lots of things can go wrong and the attack vector is great. I have been
>there and done that.
From: "Mr. Arnold" MR. on

"PJ6" <noone(a)nowhere.net> wrote in message
news:OPV8ewOLLHA.4636(a)TK2MSFTNGP06.phx.gbl...
> OK, fair enough. You make good points.
>
> I'm in the middle of writing an article on factors that affect a
> developer's productivity in computer languages and architecture, with a
> particular focus on end product stability (especially under change),
> development time required, and estimation reliability.
>
> Someone else made a start along these lines talking about what's wrong
> with C++:
> http://www.ittybittycomputers.com/IttyBitty/CppHarm.htm
>
> Part of the article points out flaws in web development - namely: browsers
> need to accept lower-level UI instructions (HTML is too high level, too
> crippled), and complex UIs are where you need a compiler validation the
> most - referential integrity checks, strong typing, etc.
>
> That looks like you're real email address.
>
> If you're so inclined, I'd like to send you the article and get your
> criticism when I'm done before I actually roll it all into a lecture. My
> hatred of web development probably clouds my judgment, and I could use
> some feedback from someone who has an opposite inclination.
>

You can post a link to your article if you like. That's not my real email
address. Like I side, I have worked on both sides of the fence developing
Win desktop, Console and Windows service applications. I have also developed
application in the Web domain, ASP.NET UI, Web services and n-tier. It's not
as bad as you make it out to be on the Web side, and they are paying big
$$$$$ for the expertise.