Thorsten Scherler a écrit :
On Tue, 2005-07-12 at 14:25 +0200, Cyriaque Dupoirieux wrote:
Ross Gardler a écrit :
Thorsten Scherler wrote:
On Tue, 2005-07-12 at 10:01 +0100, Ross Gardler wrote:
Thorsten Scherler wrote:
On Tue, 2005-07-12 at 10:08 +0200, Cyriaque Dupoirieux wrote:
Thorsten, you closed the FOR-320 and do not apply the patch on skins.
Does that mean that we will not improve skins anymore ?
I would not see any benefit from applying it to skins. ;-)
The patch already works for skins and there are many people already
using skins. Views are not ready to replace skins yet (we will need
pelt in views before that can happen). Please reopen the issue.
I will soon be on this improvment, but I don't understand what is the
impact of the project.skin property now that templates directly generate
XHTML.
Thorsten, how do you imagine to be multi-skins vith views ?
For now we can use a new forrest.properties property for that.
I added in http://svn.apache.org/viewcvs?rev=215900&view=rev
#views
project.view-defaultView=default.fv
Now you can add to you project specific forrest.properties
project.view-defaultView=pelt.fv
Ok, and where do I put my pelt.fv ?
I tried in :
*
forrest/build/plugins/org.apache.forrest.plugin.internal.view/resources/views,
* project/src/documentation/conf
it is not taken into account...
It is only taken into account in
project/src/documentation/content/xdocs, but I would like to have the
same behaviour of inheritance as for default.fv :
* Search for $path/myFile.fv
* If not found, try to take this -
project/src/documentation/conf/pelt.fv
* And - at last - if not found, take this -
forrest/build/plugins/org.apache.forrest.plugin.internal.view/resources/views/pelt.fv
Regards,
Cyriaque,
...and then follow
http://forrest.apache.org/docs_0_80/howto/howto-view-dsl.html
to develop pelt based on views.
Then you will end up with 1. css-file and 1. pelt.fv.
We will then add it to the view and we can choose 2 different skins
(default and pelt).
(That is how do we manage templates, how do we manage fv files
Maybe use of different directories to store templates ?
Maybe default.skinname.fv files ?
)
If we come to the point where we have views in the core and "view
plugins" that will be nearly the same. With a small extension that the
user can then as well define which views.xhtml plugin (s)he wants.
Like
project.views.xhtml.plugin=org.apache.forrest.plugin.output.viewHelper.xhtml
Only that we need to move views into the core to activate that.
Anyway if you can provide the css and view for pelt that would be more
then great.
:)
salu2
Regards,
Cyriaque,
You are as well admin and can reopen the issue (read on before you do).
Yes I know. But I don't want to go stamping on what you already did
without understanding your reasoning, you never know I might agree
with you when you explain it to me ;-)
Adding *new features* to skins *do not* help to make views ready to
replace skins. Did you thought about that?
I agree that spending the time adding new features is not a good
thing. However this patch already exists. as I said in my mail I agree
that it would need to be made configurable, I'm not sure anyone will
bother with that, I know I won't). Like I said in my original mail,
all I am suggesting is leaving it open so that if a user wants this
feature we can point them at the patch, or better still they will find
it without asking questions on the lists. If the issue is closed they
are unlikely to find it.
...and we *do not* need pelt in views, that is not true. Why do you mean
that?
If you want people to adopt views I strongly recommend that we have a
view that looks the same as pelt. Look at the number of sites on our
example sites that use it. Most of those sites will will not switch to
views unless at the point of switching their site looks the same as it
does not. Then some will start playing with views and customising
their sites look and feel because they will discover how easy it is.
Those who are not using pelt are, in most cases, using a customised
version of it. If you provide a view that looks like pelt it will be
much easier for them to recreate their own site in views. Therefore,
they are more likely to migrate to views.
If you do not provide this migration route most existing users will
stick with the deprecated skin for quite some time. There is no
motivation to move from it since they are perfectly happy with what
they have. The fact that it is cool, whizz bang technology and really
easy to customise is irrelevant if they are happy with what they have.
If we want to bring our existing user base to views we have to make it
very easy for them. This means, in my opinion, that we need a view
that looks exactly like pelt.
Anyway I am -1 to reopen this issue and apply it to skins because that
means we have to add yet another property to the skinconf,...
I said reopen, not necessarily apply it, I agree with David that it
would need to be configurable before we applied it and that involves
spending time on it, which I agree is not a good thing, our dev time
should go into views. However, the fact that we are deprecating the
skin does not mean we should put contributions in a place where they
will not be seen by existing users.
Ross
