Hey guys,
Someone posted an interesting question on my blog
Hi Laurent,
I'm not sure if this is the correct place to post
this, but I was wondering why all the disciples don't try to contribute to the
same MVVM framework. Seems like Josh, Marlon, Karl, and Pete also have similar
tools (though of course unique in many ways). Seems like the MVVM Framework on
CodePlex is a good place to start. "Contrib" libraries with each
person's unique features could dissolve into the trunk over time as they
mature. What are your thoughts?
Thanks,
David Cuccia
I did my best to reply to David, the reply is at http://geekswithblogs.net/lbugnion/archive/2009/09/27/mvvm-light-toolkit-messenger-v2-beta.aspx#488823
If you want to chime in, feel free. If you prefer to answer on the group, it’s fine too. I have thought about that a few times myself too, and tried to summarize my feelings in my reply.
Cheers,
Laurent
--
Laurent Bugnion [Microsoft MVP, MCP]
Blog: http://blog.galasoft.ch
Support children in Calcutta: http://www.calcutta-espoir.ch
|
My business card as |
Agreeing with Josh and expanding a bit.
One really doesn’t need an MVVM framework.
MVVM is a fairly straightforward easy concept and not terribly hard to implement if you know C# and understand SL/WPF.
If you had to write your own MVVM framework it’ll take a SL/WPF expert about 1 day to do 80% of the work.
That makes MVVM frameworks extremely susceptible to Not Invented Here syndrome.
Additionally, one strictly doesn’t need a MVVM framework.
It’s definitely possible to build a micro app (ad) or small sized apps (twitter client) with ad-hoc MVVM architecture.
It’ll be darn stupid for someone to do that, but it’s more than possible
So couple ease of creation with lack of necessity and you get 20 MVVM frameworks.
There are some really great MVVM frameworks out there: Caliburn, MVVM Toolkit, Prism, WPF Toolkit, etc.
Each as josh said with its own unique style and flavour.
Personally, I believe once Microsoft ships something stable (and usable) in that area integrated into Visual Studio, it’s curtains down for most MVVM frameworks.
Which is roughly what happened in the Persistence Layer / ORM .net open source area when Microsoft shipped Linq2Sql and EF.
The stable and serious open source / commercial offerings stayed around (NH, IdeaBlade, etc) but the small ones just disappeared.
-- Justin
From: wpf-di...@googlegroups.com
[mailto:wpf-di...@googlegroups.com] On Behalf Of Josh Smith
Sent: September-29-09 2:19 PM
To: wpf-di...@googlegroups.com
Subject: [WPF Disciples] Re: A common framework, and why we don't do it
I think that having a bunch of strong-minded, vocal developers/architects try to agree on what's "best" will end in blood and tears. :) Seriously, we all have our own agendas and preferences. Why neuter our individual ambitions to create one framework of compromises that doesn't really meet any of our needs?
Agreeing with Josh and expanding a bit.
One really doesn’t need an MVVM framework.
MVVM is a fairly straightforward easy concept and not terribly hard to implement if you know C# and understand SL/WPF.
If you had to write your own MVVM framework it’ll take a SL/WPF expert about 1 day to do 80% of the work.
That makes MVVM frameworks extremely susceptible to Not Invented Here syndrome.
Additionally, one strictly doesn’t need a MVVM framework.
It’s definitely possible to build a micro app (ad) or small sized apps (twitter client) with ad-hoc MVVM architecture.
It’ll be darn stupid for someone to do that, but it’s more than possible
So couple ease of creation with lack of necessity and you get 20 MVVM frameworks.
There are some really great MVVM frameworks out there: Caliburn, MVVM Toolkit, Prism, WPF Toolkit, etc.
Each as josh said with its own unique style and flavour.
Personally, I believe once Microsoft ships something stable (and usable) in that area integrated into Visual Studio, it’s curtains down for most MVVM frameworks.
Which is roughly what happened in the Persistence Layer / ORM .net open source area when Microsoft shipped Linq2Sql and EF.
The stable and serious open source / commercial offerings stayed around (NH, IdeaBlade, etc) but the small ones just disappeared.
-- Justin
From: wpf-di...@googlegroups.com
[mailto:wpf-di...@googlegroups.com] On Behalf Of Josh Smith
Sent: September-29-09 2:19 PM
To: wpf-di...@googlegroups.com
Subject: [WPF Disciples] Re: A common framework, and why we don't do it
I think that having a bunch of strong-minded, vocal developers/architects try to agree on what's "best" will end in blood and tears. :) Seriously, we all have our own agendas and preferences. Why neuter our individual ambitions to create one framework of compromises that doesn't really meet any of our needs?
Exactly. I also mentioned in my reply on my blog that i would be more enclined to participate to a larger scale project if I didn't think that most of it will flow in the framework eventually, making most of my toolkit redundant and obsolete.
Cheers,
Laurent
--
Sent from mobile
-original message-
Subject: [WPF Disciples] Re: A common framework, and why we don't do it
From: Mike Brown <mbro...@gmail.com>
Date: 30.09.2009 04:03
I agree with what justin is saying WRT simplicity of implementing your own
MVVM solution. I think there is value in doing it yourself...it's like a
Jedi making his first lightsaber. Basically, it's a collection of patterns
(MVVM, DelegateCommand, Event Broker, Attached Behaviors) that are useful
for wpf development. There's a lot of nuances and variations in how to get
it all working and like Justin mentioned, any effort to make a single
framework will likely be trumped by an integrated solution from Redmond.
On Tue, Sep 29, 2009 at 6:20 PM, Justin Angel <J...@justinangel.net> wrote:
> Agreeing with Josh and expanding a bit.
>
>
>
> One really doesn’t need an MVVM framework.
>
> MVVM is a fairly straightforward easy concept and not terribly hard to
> implement if you know C# and understand SL/WPF.
>
>
>
> If you had to write your own MVVM framework it’ll take a SL/WPF expert
> about 1 day to do 80% of the work.
>
> That makes MVVM frameworks extremely susceptible to Not Invented Here
> syndrome.
>
>
>
> Additionally, one strictly doesn’t *need* a MVVM framework.
>
> It’s definitely possible to build a micro app (ad) or small sized apps
> (twitter client) with ad-hoc MVVM architecture.
>
> It’ll be darn stupid for someone to do that, but it’s more than possible
>
>
>
> So couple ease of creation with lack of necessity and you get 20 MVVM
> frameworks.
>
> There are some really great MVVM frameworks out there: Caliburn, MVVM
> Toolkit, Prism, WPF Toolkit, etc.
>
> Each as josh said with its own unique style and flavour.
>
>
>
> Personally, I believe once Microsoft ships something stable (and usable) in
> that area integrated into Visual Studio, it’s curtains down for most MVVM
> frameworks.
>
> Which is roughly what happened in the Persistence Layer / ORM .net open
> source area when Microsoft shipped Linq2Sql and EF.
>
> The stable and serious open source / commercial offerings stayed around
> (NH, IdeaBlade, etc) but the small ones just disappeared.
>
>
>
> -- Justin
>
>
>
> *From:* wpf-di...@googlegroups.com [mailto:
> wpf-di...@googlegroups.com] *On Behalf Of *Josh Smith
> *Sent:* September-29-09 2:19 PM
> *To:* wpf-di...@googlegroups.com
> *Subject:* [WPF Disciples] Re: A common framework, and why we don't do it
> Blog: http://blog.galasoft.ch <http://www.galasoft.ch/>
>
> Web: http://www.galasoft.ch
>
> Support children in Calcutta: http://www.calcutta-espoir.ch
>
>
>
> [image: cid:image0...@01C9C8AA.B722DA80]
>
> My
>
> business
>
> card
>
> as
>
> Microsoft Tag <http://www.microsoft.com/tag>
>
>
>
>
>
>
>
<<image001.png>>
hence the request for permission to reference peoples blogs sites etc...
It's coming soon.
The uber collection.
-----Original Message-----
From: wpf-di...@googlegroups.com [mailto:wpf-di...@googlegroups.com]
On Behalf Of Daniel Vaughan
Sent: 30 September 2009 08:57
To: WPF Disciples
Subject: [WPF Disciples] Re: A common framework, and why we don't do it