[prev in list] [next in list] [prev in thread] [next in thread] 

List:       quanta
Subject:    Re: [Quanta] wysiwyg
From:       Eric Laffoon <eric () kdewebdev ! org>
Date:       2008-02-18 21:26:08
Message-ID: 200802181326.08606.eric () kdewebdev ! org
[Download RAW message or body]

On Monday 18 February 2008 12:46:55 am Andrew wrote:
> Hi. Totally new to Quanta. Looking for a wysiwyg page editor to use on
> linux. Tried Nvu / Komposer but not too happy about the code it
> generates, and can't get on with Amaya at all. Quanta looks nice, but
> when I insert an image it totally loses the plot! The page looks ok on
> preview, but in vpl mode either the image does not show at all, or the
> rest of the page after the image does not show. Is there something
> simple I'm likely to be doing wrong?! I'm running ubuntu 7 with gnome.
>

First of all, the reason we set out to do VPL instead of WYSIWYG is because 
the reality of web design is what you see is never what somebody else gets. 
It's always different unless it's in fixed format CSS and they have the same 
fonts, screen resolution, etc... Visual Page Layout started out with another 
idea, that if you edit a paragraph that paragraph is edited and the rest of 
the document is not munched and trashed like loading it in a word processor 
that does HTML.

Nvu launched after VPL saying it was first, until someone must have pointed 
out the truth. They initially had virtually no DTD information, and later 
decided to declare a fixed DTD, which last I looked they did a poor job 
adhering to. Quanta's VPL adheres to your DTD as that is how it generated 
tags, just like in text mode. Try split mode if you want to learn to edit and 
see what it does.

Anyway as many people note VPL is not really ready for production use, and 
it's also not maintained currently. Here's why...

1) VPL is dependent on hooks in KTML. Initially it was coded building all this 
without those hooks and it was a nightmare that could never accomplish 
certain things. So we started over. This theoretical advantage tied VPL 
closely to your version of kdelibs.

2) The KHTML developers also decided to scrap their first attempt. This is 
insanely complex code. So we had to rework our code to a new API... again...

3) We had the Apple Webkit debacle where KHTML developers complained about how 
they didn't expose their CVS and the bulk code they exposed was impossible to 
integrate. Apple saw the problem and relented, but left us with a big 
question. They had just what we needed to enable a great VPL mode in there, 
but would those changed ever be integrated as they were so difficult to 
extract we might as well start over.

4) KDE4 development moved to first port and then enhance code, and as we share 
a development platform now with KDevelop we were not near ready to look at 
this, and we have no maintainer for VPL because...

5) Developer burn out has eaten two VPL maintainers. One did VPL as a project 
for his degree his last year of college. He still codes with our project, but 
is too burned out on all the headaches of VPL to start back.

It is not a matter of whether we want to support visual design done right. It 
is a matter of how we proceed, do we count on certain things getting into 
KHTML or not and how do we attract a developer. I'm resigned to this being 
such a tedious and demanding project that it requires a sponsored developer. 
I will NOT assign it to Andras and burn out our key developer for the whole 
project. Ironically it would take 10-20 users offering monthly what they 
would pay for shareware or 5-10 users offering yearly what they would pay for 
commercial packages. Outside of a few key large sponsors that is more than 
what we currently recieve for the whole project. I am entirely uncomfortable 
sponsoring someone one month, oweing them the next and letting them go after 
that and paying out of pocket while asking them to keep their calendar open 
in case some users actually want this later.

One thing to keep in mind here is that I have a track record of delivering 
what I say I will... although we could conjecture what I said about VPL, I 
don't recall saying it was going to be ready for production use yet, but as 
you can see our failures there were not for lack of trying.  It is going to 
take another effort. Sadly without a maintainer, and I don't have time or 
funds to do it myself, Quanta4 will debut without VPL. I am hoping to begin 
looking into this project in 6 months time as I am too busy and can't fund it 
personally now. 

I am moving ahead with other new design ideas including object templates, and 
this is where a visual mode makes sense to me... I want to be able to take a 
PHP include and see what it would look like in the page and edit text, CSS 
and loops in visual mode. This is where visual design is usable to me and my 
long term objective is to make a truly usuable visual design mode that can 
handle not just simple HTML but DTDs and can process linked parse trees of 
what goes to the server and the HTML it renders back. 

We have no interest in copying what I think has been lame, but we want to 
offer our users something original and productive. I just don't have time and 
resources (funds to sponsor or volunteer team members) to do it all.


-- 
Eric Laffoon
Project Lead - kdewebdev module
_______________________________________________
Quanta mailing list
Quanta@mail.kde.org
https://mail.kde.org/mailman/listinfo/quanta
[prev in list] [next in list] [prev in thread] [next in thread]