Article on wine development strategy

Reece Dunn msclrhd at
Sat Apr 18 09:22:34 CDT 2009

2009/4/18 Susan Cragin <susancragin at>:
> I would like to put in my two cents for making Dragon NaturallySpeaking 10 work at about "50%."
> DNS 10 is a terrible program in a lot of ways. It's interface is over-engineered, clumsy, unintuitive, and packed with features that the average user never even looks at.
> People who don't know DNS say, "I'd love to dictate into word, format by voice command, etc."
> People who know DNS well, say "just give me accuracy and speed, and forget everything else." Those people usually tend to be the natural DNS constituency: either disabled or want to churn out a tremendous amount of text without typing.
> I think my point here is that happiness depends not only on % but on the type of program. Not all programs are created equal. Not all features are a good idea.

That also brings up a good point as to why focusing on applications -
even those used by a large number of people - is only part of the
equation: every user is different.

If you take a sample of Office, PhotoShop or iTunes users and see what
features they use, there are bound to be differences. You use iTunes
just as a media player? Want to download songs from the iTunes store?
Want to sync with your iPod? All of these hit different areas.

Can't hear? Fixing sound issues isn't going to help this user.

Can't see? This user isn't going to be fussed about graphical issues.
They will be more interested in accessibility support (MSAA,
UI/Automation) and would use something like DNS to interact with the
applications. Or perhaps get an MSAA/UIAutomation to Gnome/KDE
accessibility APIs so they can use the system's screen reader.

So here, the user-by-user approach is more useful.

- Reece

More information about the wine-devel mailing list