trinity-devel@lists.pearsoncomputing.net

Message: previous - next
Month: July 2011

Re: [trinity-devel] A users perspective on Trinity

From: Robert Xu <robxu9@...>
Date: Wed, 6 Jul 2011 09:00:07 +0800
On 2011-07-06, Darrell Anderson <humanreadable@...> wrote:
> Rebranding: To what extent? I don't think renaming all files is necessary. I
> think most of the branding issues are resolved. If the user is addressing
> app names, I am content with the old "K" naming scheme. I also like the idea
> that with the "K" naming convention people can see and remember the KDE3
> roots, which I think is a selling point.

You make a good point here.

>
> KOffice: I never piped in on that debate, but I see no reason to maintain
> most of KOffice. LibreOffice is more than acceptable as a replacement. I do
> think the wiki needs to be updated with instructions how to build
> LibreOffice to ensure native KDE3/Trinity file picker support (using the
> --enable-kde build option). I think a handful of apps from KOffice should be
> maintained, such as Kivio, Krita, etc. I don't see a need to maintain the
> main apps.

When you put it that way... Is it possible to maintain a subset of
KOffice? Like Darell said, only a handful of apps? Or is it all or
nothing?

>
> Digikam: Are all the new features provided in the QT4 version necessary? I
> am no camera junkie, but I depend upon Digikam in KDE3 to interface with my
> digital camera. I'd hate to see that app disappear in Trinity. I'm happy
> with the older version of Digikam. The appeal of Trinity is that of being
> light weight compared to KDE4 and GNOME 3. I see no reason not to maintain
> light weight versions of various apps.
>
> Browser: I don't see a viable solution to providing a built-in web browser
> for Trinity. Konqueror in KDE3 never was as extensible or usable as Firefox.
> That is not going to change soon in KDE4. I think web browsers should not be
> a concern or component of Trinity. I still think Konqueror in KDE3/Trinity
> is the best file manager available. I hope that does not change.
>
> Light weight desktop: I don't use KDE 4 enough to add to the perception of
> being bloated. I agree that KDE4 from upstream is configured for bleeding
> edge hardware with all the various desktop effects enabled by default.
> Likewise with GNOME 3, which works only on hardware with 3D video
> accelerators. I have noticed that since the advent of GNOME 3 and Unity,
> many people are now offering "light weight" distros. The new Porteus
> portable system, a successor replacement for Slax, uses Trinity 3.5.12.
> Other distros are now offering Xfce and LXDE as choices over KDE4 and GNOME
> 3. Many people are unhappy with the direction of KDE4, GNOME 3, Unity. I
> foresee no reconciliation because the people controlling those environments
> are on a different plane of existence than people who want light weight but
> flexible desktops.
>
> In other words, there is a healthy market for Trinity. Stay focused on the
> desktop and existing apps in the source tree.
>
> Will users compare Trinity to KDE4? I think that is inevitable and
> unavoidable. Reviewers are likely to notice what "features" are available in
> KDE4 and unavailable in Trinity. With that said, the same can be done with
> comparing Xfce and LXDE to KDE4. There is a significant difference and
> always will be --- and should be. Any reviewer who argues otherwise is
> missing the point.
>
> I vote for not worrying about trying to maintain Trinity as feature rich as
> KDE4 or backporting features. There is too much a price to pay. The user who
> is attracted to Xfce, LXDE, and Trinity are not looking for every single
> feature imaginable. They want a flexible but stable desktop. They want a
> sense of continuity and cohesiveness with apps. Trinity provides that.
>
> The competition for Trinity is Xfce and LXDE, not KDE4.
>
> Complete the cmake conversion. Resolve a few dozen critical and paper cut
> bugs. Update the wiki and web sites to emphasize the light weight appeal. I
> expect to read many positive reviews when Trinity 3.5.13 is announced.
>
>

Well put in many ways.

-- 
later, Robert Xu + rxu AT lincomlinux DOT org