Re: MFC & Guis
- From: BobF <rNfOrSePeAzMe@xxxxxxxxxxx>
- Date: Fri, 27 Jan 2006 08:23:12 -0600
Daniel,
I think we're in agreement on all but one side point. I believe it *is* M$
responsibility to provide the purchasers of their development products with
anything and everything to make development effort on the target platforms
as productive as possible.
<IMO>
The fact that so many GUI wheels are reinvented by so many devs is an
indication that the vendor should focus more effort on building a greater
level of support in this area. That could be anything from better
documentation to better dev software support.
</IMO>
Thanks - this has been a good discussion. I'm always interested other's
perspectives on this kinda' stuff.
On Fri, 27 Jan 2006 10:58:37 GMT, Daniel James wrote:
> In article news:<1vhsi0ag1x56x$.8wtztiiscqfj.dlg@xxxxxxxxxx>, BobF wrote:
>> IMO, application development is still very much an art. We all see GUI's
>> that deviate from the "standard" - some we really like, some we hate.
>
> At the end of the day what matters most to the user is being able to use the
> software to get the job done. Users of Windows software form expectations
> about how an application will behave; they recognize the standard Windows
> controls and GUI metaphors like the menus and toolbars, and they learn that
> these are normally work in the same way in every application. This "if you've
> learnt one Windows app you can use them all" feature is very powerful -- and
> is one of the things that has contributed to Windows's widespread adoption.
>
> GUIs should be designed to take advantage of the existing familiarity users
> will have with the platform. Don't deviate from the platform's native "look
> and feel" or you'll make life harder for your users, make everything work the
> way they will expect. This is sometimes known as the "principle of least
> surprise".
>
> If you design an app with a GUI that deviates from the Windows norm for no
> good reason you will give your users an unnecessarily hard time when they try
> to learn to use your app. Worse: you will erode their grasp of the implicit
> "rules" of Windows GUI design that they have been learning from all other
> apps, and so /increase/ the effort they will have to expend to pick up other
> apps.
>
> This *is* evil, it *should* be argued against.
>
> Note that what I'm talking about here is essentially messing with the
> non-client areas of the windows -- the parts that Windows will draw for you
> in a standard way. Trying to change standard look and feel of those things is
> not only unnecessary expenditure of programmer effort, it violates the
> "principle of least surprise".
>
> None of this is to say that one should not be innovative in the way that
> information is presented and manipulated within the client areas. The
> Audacity example you linked elsewhere is a good case in point -- the
> non-client areas look like a Windows app (and, indeed, the nonclient areas of
> Audacity on linux or MacOS X have the proper appearance for those platforms)
> while the client areas exhibit behaviour specific to the application domain.
>
>> The bottom line is that "non-standard" GUI development isn't supported
>> directly by M$ tools or documentation, even though M$ developers obviously
>> put much effort into customizing M$ product GUI's.
>
> There are two issues here:
>
> - MS don't provide much in the way of tools or support for changing the
> non-client characteristics of application windows because it isn't desirable
> to encourage that this be done (I'm surprised they do it themselves, and that
> they even make it possible - it violates all good design guidelines).
>
> - MS don't provide much in the way of tools or support for implementing
> customized window clients because it's not their job. The client area is
> *your* responsibility, and only you can know what data you want to represent
> there, what representation of those data is going to be meaningful to the
> user, and what manipulations of those data your program is going to support.
>
>> This forces people to reach out for help.
>
> People are not being forced to ask for help to do things that most people
> will tell them they shouldn't be thinking of doing at all (or, at least, not
> without a very good and unusual reason).
>
>> In this particular case, the OP not only didn't [get] help, but was
>> also chastised for even thinking about trying to dress up an application.
>
> I may have stated my point a little too bluntly in my earlier posting here,
> and I'm sorry if any offence was caused. The OP's opening gambit "As we know
> MFC makes ugly GUIs" was a bit of a red rag, though ...
>
> MFC doesn't "make GUIs" at all (Windows makes the nonclient areas, and the
> application makes the client areas); these GUIs are neither more nor less
> "ugly" because of MFC (MFC just makes the programming easier - blame the
> Windows API not MFC if you don't like how it looks); "ugliness" is not really
> the isse (what matters is usability); and I don't consider the appearance of
> an MFC/Win32 app to be ugly anyway (unless running on XP in non-Classic
> mode).
>
>> Anyway, I'm not here to pick a fight. I just hate seeing people post
>> perfectly valid questions here, only to receive unjustified criticism in
>> return.
>
> I have no desire to fight, either. I may have overreatced to that question,
> but it did seem to me to be a cry for more of all that is worst in GUI
> programming.
>
> Cheers,
> Daniel.
.
- Follow-Ups:
- Re: MFC & Guis
- From: Daniel James
- Re: MFC & Guis
- References:
- MFC & Guis
- From: Jessica Weiner
- Re: MFC & Guis
- From: Daniel James
- Re: MFC & Guis
- From: M
- Re: MFC & Guis
- From: BobF
- Re: MFC & Guis
- From: Pete Delgado
- Re: MFC & Guis
- From: BobF
- Re: MFC & Guis
- From: M
- Re: MFC & Guis
- From: BobF
- Re: MFC & Guis
- From: Daniel James
- MFC & Guis
- Prev by Date: Re: populate a combobox with a returned array from an external dll
- Next by Date: Re: "Modal like" disabling window.
- Previous by thread: Re: MFC & Guis
- Next by thread: Re: MFC & Guis
- Index(es):
Relevant Pages
|