Re: [gentoo-user] Re: Has semantic-desktop really become compulsatory for kmail?

[Zeerak Waseem]

On Fri, 12 Feb 2010 15:46:57 +0100, BRM <bm_witness@×××××.com> wrote:

> ----- Original Message ----
>
>> From: Zeerak Waseem <zeerak.w@×××××.com>
>> On Fri, 12 Feb 2010 10:53:04 +0100, Neil Bothwick wrote:
>> > On Fri, 12 Feb 2010 05:19:43 +0100, Zeerak Waseem wrote:
>> >> But I do find it silly, that the various applications that aren't
>> >> dependent of the DE, to require a dependency of the DE. It just seems
>> >> a bit backwards to me :-) I simply don't understand.
>> > That just shows that they are still partially dependent on the DE,
>> KMail
>> > also needs various KDE libraries. KDE was designed as a cohesive DE,
>> not
>> > just a bunch of applications with a common look and feel. KDE apps are
>> > intended to be run on a KDE desktop, anything else is a nice bonus.
>> Indeed, and it is a noble pursuit.
>> But from a marketing aspect, it would make more sense to have things
>> that aren't
>> -vital- for the app, unlike kde-libs in this case, to be soft (is this
>> the
>> correct term?) dependencies.
>> Both aspects could be satisfied by having symantic-desktop as an
>> optional dep.
>> It's not a vital function for kmail to be able to tag and index all the
>> files on
>> the computer (which is what the symantic-desktop does if I understand
>> correctly), it's a nifty thing for KDE users, and soon probably Gnome
>> users as
>> well, but for anyone else, it's a nifty thing -if- they feel the need
>> for it.
>> Much like most other bits of software :-)
>
> Obviously you don't understand the reason for the dependency.
> It does not exist so that Kmail can index all the files on the system
> but for the opposite -
> so that Kmail can participate in the search by allowing the system to be
> able to search _its_ data.
>
> And, btw, you're not turning it off within Kmail, but at the system - DE
> - level.
> The application itself will still check to see if it could participate,
> only to have nothing turned on to support so then it doesn't do anything.
>

Right, but then when the DE isn't a DE, but a window manager, a minimal  
one, then it's kind of a strange for a function to be forced outside of  
the specific DE. Well it seems strange to me anyway.

>> In the end there isn't a right or wrong, but just a standpoint.
>
> Question: are you a software developer?
>
> Kmail probably has the dependency the way they do b/c it is far easier
> to make it one and let the system determine not to support the
> functionality
> than it is to litter the codebase with "if (symanticDesktopEnabled)..."
> code.
>

An aspiring one, yes. And it probably easier to just make it a hard dep,  
however if the quality of their application for anyone -not- using kde.  
It's sensible for it being set if you have kde, but if you don't have kde  
it just seems very out of place.

>> Some don't mind
>> the bloat (we can agree that it's bloat if you're just going to disable
>> the
>> function as soon as it's been installed, right?) and don't consider it
>> to be the
>> slightest bit akin to bloat, whilst to others it's an unnecessary
>> feature forced
>> on them (mainly thinking of the people not using kde, but also those
>> kde-users
>> that just disable it) and thus becomes bloat.
>
> No more than it is bloat for gcc to support mmx/sse/sse2/sse3/sse4 when
> your processor cannot.
>
> Ben
>
>
>

I hadn't considered that particular thing, but yes, in a sense you're  
right. I mean there is difference for a compiler and a mail app, with gcc  
you can compile for another system so the it supports things your  
processor doesn't support doesn't necessarily mean that you won't need the  
support, with a mail app you can... But essentially, if you know you'll  
never need to compile for another processor, then yes I'd consider it  
bloat.

-- 
Zeerak