kde-www module

Christoph Cullmann cullmann at babylon2k.de
Thu Oct 17 08:55:48 UTC 2002


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Thursday 17 October 2002 10:46, Neil Stevens wrote:
> On Thursday October 17, 2002 01:35, Christoph Cullmann wrote:
> > What do my commits break ????
> > - your newdesign page
>
> It removes features and reduces flexiblity, all without discussing first.
>
> I put the code into kde-www because I thought we were going to cooperate.
>
> > What does they archive:
> >
> > 1) it is now possible to simply put kde-www on one server for mirrors,
> > on different servers for the "real-world" *.kde.org stuff as it is atm.
>
> That won't work.  You're assuming, once again, that in the server we were
> going to put every KDE site.  sysadmins told you no, and everyone here
> told you no.  That's an inefficient use of the server resources.
? Read my sentences again, I said:
on different servers for the "real-world" *.kde.org stuff as it is atm
Now what does you mean with one server ? Perhaps my english is bad, but 
doesn't that sentences shows clearly that I am aware that the pages need to 
be on seperate servers for the real life ? The first part only says: it is 
POSSIBLE to mirror kde-www on one server. Now, what is wrong with that ? 
Don't want mirrors ?

>
> > 2) the kde-www stuff now only needs on server side var, the php include
> > path to be adjusted and per page only the $site_root given as relative
> > path and that /images, /inc and /styles is located in the serverroot.
>
> My design only required one var in 99% of the files!  You had to point to
> one site.inc file.  The other vars go into that one measly site.inc file.
>
> This is only a problem if we agreed to your mirroring scheme, which we
> didn't.
My design don't need a site.inc nor does it need more than one var per page, 
why is that worse, even without any mirror in thought ? To build a page needs 
to be as simple as possible, why it that bad ?

>
> > Now my question: What makes that worse ? K, perhaps styles per page can
> > be enabled again, but that is a matter of 2 lines. If we reconstruct
> > everything from scratch, we should do it right and not only because the
> > newdesign on usabilty created by you and jasone does work as it does. I
> > am simply not willed just to take your broken stuff only because it
> > works for usabilty.kde.org/newdesign.
>
> The code does work.  It just supports features you don't care about, and
> doesn't support the feature that only you want.
>
> > If you only want to revert because it breaks your newdesign study, than
> > that is no point, it will break kate.kde.org, too and all other pages
> > will be need to redone in each case.
>
> What will be broken?  I thought we started from scratch?
Than where is the problem with a more flexible design ?

>
> > But all that may be disscussed again tomorrow, 19 MEST, know.
>
> What's the point?  Unless we can make decisions, this whole project will
> fail.  And I thought it was clear that the decision was made *not* to put
> everything on the same server.  Your plan requires that.

???????????????????????????????????

PLEASE READ MY MAIL !

Where did my design require one server ???
Could you point me out where it is "less flexible" and where does it hardcode 
"one server" for all pages ? Where does it remove working features ? My 
design allows that:

a) everthing is one one server (testing, mirrors, ...)
b) every area/app page is located on different servers (as long as you have 
the inc/, images/, styles/ dir there. (if you need it, we can reintroduce 
$baseurl to be able to have /images and /styles located on a different 
server)

Btw. Is constructive work to simply say: your stuff is borked, mine is better 
? I have not rewritten your stuff, only changed some vars and implementation 
details, why is that bad ? And don't tell me anything about communication or 
"work together", the stuff is in cvs for testing/working/improving, you even 
didn'T tell root66 that you used his design, changed it a lot and simply put 
it on the net, which now may cause that he won't work further with us on that 
page.

cu
Christoph



- -- 
Christoph Cullmann
Kate/KDE developer
cullmann at kde.org
http://kate.kde.org
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.0 (GNU/Linux)

iD8DBQE9rnshyPjDGePm9UIRAvTXAKCB4J/TeRxkehzcSu99MbXKOC4ypACfZZrQ
FWKsCrExljS7fRdlWeLhpN0=
=o2pA
-----END PGP SIGNATURE-----




More information about the kde-www mailing list