Re: [Re: [Re: [Re: [Re: [libxml++] UTF8 support]]]]



Murray Cumming wrote:

I am not being GNOME-centric. Glib::ustring will be available for general
non-GTK, non-GNOME use.

Isn't Glib a C++ wrapper around the C library 'glib' ? I don't see any
reason why any C++ project should need to use that. I can see why C
people try to reinvent OO techniques instead of learning C++. But
anyways, let's not get more off-topic.

My point is really always the same: I don't want to be tied to any
particular unicode library. I outlined the two major reasons earlier:
Getting undesired dependencies, having to deal with two string APIs
when one should suffice, and finally the performance cost of having to
copy each string twice.

I don't know of any other C++ unicode string class
that is not tied to a GUI toolkit.

What does it matter whether the unicode string class is tied to a
particular GUI toolkit or not, as long as I can choose it ? Your point
is valid only under the assumption that libxml++ has to fix the choice.
Then it would indeed be important to be as independent as possible. But
that has nothing to do with my proposal.

Please try to read my responses instead of
just repeating your vague accusation that my involvement with GNOME is somehow
a disadvantage to you.

I'm not accusing you of anything. I'm just suggesting that your
preference for using Glib::ustring for libxml++ is due to your bias. I
don't think you can refute that that easily.

And I called you 'GNOME centric' because you have been telling me again
and again that you can't imagine anybody wanting to use any unicode
library but Glib. Hell, I didn't even know about Glib::ustring's
existence untill you mentioned it to me !

Yes, performance is the only issue that I can think of. Luckily I can sidestep
that with my usual "profile-before-optimising" argument.

I mentioned (repeatedly) two other (to me obvious) arguments. What about
them ?


glibmm requires only glib. That's really not very much, particularly as every
single linux and solaris installation will have glib installed in the near
future, and there's a high chance that the same will be true of glibmm.

*sigh*. I'm developing things *right now* on linux and irix (and there
are lots of other platforms I want it to run, too, such as *bsd,
windows, etc.). And our quality management protocols require me to make
the dependencies explicit, meaning I even have to package libxml2 myself
and ship it with the distribution. I just don't *want* to be dependant
on anything that's not strictly necessary !

Given
your slightly irrational fear of glib I'm surprised that you don't object to
our use of pkg-config too as that was also produced by people involved with
GNOME.

Talk about irrational arguments. Now you are putting words into my
mouth. I don't have a problem with 'people involved with GNOME'. I don't
care. Please stop with these childish arguments.

Anyways, this thread has already taken way too much time. Time I should
be spending on more productive work. I have made my point, now you as
the self-acclaimed 'lead developer' have to choose the direction for
libxml++. Good luck !

Regards,
		Stefan





[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]