[Erp5-dev] [Erp5-report] r20493 - in /erp5/trunk/products/ERP5/bootstrap/erp5_xhtml_style: SkinTemplat...

[bartek]

Author: romain
Date: Mon Apr 14 16:22:36 2008
New Revision: 20493
URL: http://svn.erp5.org?rev=20493&view=rev
Revert http://svn.erp5.org/?view=rev&revision=20078, as it changes the DOM tree
and desactivate some CSS configuration.

Could you please specify WHICH css configuration does it break? So that
somebody can fix the css?

Bartek

This modification will be moved to experimental for now.
erp5/trunk/products/ERP5/bootstrap/erp5_xhtml_style/SkinTemplateItem/portal_skins/erp5_xhtml_style/field_render.xml
erp5/trunk/products/ERP5/bootstrap/erp5_xhtml_style/bt/revision
Modified: erp5/trunk/products/ERP5/bootstrap/erp5_xhtml_style/SkinTemplateItem/portal_skins/erp5_xhtml_style/field_render.xml
URL: http://svn.erp5.org/erp5/trunk/products/ERP5/bootstrap/erp5_xhtml_style/SkinTemplateItem/portal_skins/erp5_xhtml_style/field_render.xml?rev=20493&r1=20492&r2=20493&view=diff
==============================================================================
--- erp5/trunk/products/ERP5/bootstrap/erp5_xhtml_style/SkinTemplateItem/portal_skins/erp5_xhtml_style/field_render.xml (original)
+++ erp5/trunk/products/ERP5/bootstrap/erp5_xhtml_style/SkinTemplateItem/portal_skins/erp5_xhtml_style/field_render.xml Mon Apr 14 16:22:36 2008
@@ -68,7 +68,6 @@
class python: \' \'.join([x for x in [\'field\', field.is_required() and \'required\' or None, field_has_error and \'error\' or None, field.get_value(\'css_class\') or None] if x is not None])"\n
i18n:attributes="title" i18n:domain="ui">\n
<tal:block tal:repeat="html_tuple html_render">\n
- <div>\n
<label>\n
<tal:block tal:content="structure python: html_tuple[0]"\n
i18n:translate="" i18n:domain="ui" />\n
@@ -85,7 +84,6 @@
tal:content="python: field_errors[field_id].error_text"\n
i18n:translate="" i18n:domain="ui" />\n
<p class="clear" tal:condition="not:is_web_mode"></p>\n
- </div>\n
</tal:block>\n
</div>\n
</tal:block>\n
Modified: erp5/trunk/products/ERP5/bootstrap/erp5_xhtml_style/bt/revision
URL: http://svn.erp5.org/erp5/trunk/products/ERP5/bootstrap/erp5_xhtml_style/bt/revision?rev=20493&r1=20492&r2=20493&view=diff
==============================================================================
--- erp5/trunk/products/ERP5/bootstrap/erp5_xhtml_style/bt/revision (original)
+++ erp5/trunk/products/ERP5/bootstrap/erp5_xhtml_style/bt/revision Mon Apr 14 16:22:36 2008
@@ -1,1 +1,1 @@
-525
+526
_______________________________________________
Erp5-report mailing list
Erp5-report at erp5.org
http://mail.nexedi.com/mailman/listinfo/erp5-report

[Romain Courteaud]

Post by bartek

Author: romain
Date: Mon Apr 14 16:22:36 2008
New Revision: 20493
URL: http://svn.erp5.org?rev=20493&view=rev
Revert http://svn.erp5.org/?view=rev&revision=20078, as it changes the DOM tree
and desactivate some CSS configuration.

Could you please specify WHICH css configuration does it break? So that
somebody can fix the css?

Hello,

Functionnal tests were failing since 3 weeks (see for example:
http://mail.nexedi.com/pipermail/erp5-report/2008-April/021001.html).

As all functionalities of the CSS are not tested, and as your
modification is for now only used in experimental, I decided to move
your change to the experimental repo.

Regards,
Romain

[bartek]

Post by Romain Courteaud

Post by bartek

Author: romain
Date: Mon Apr 14 16:22:36 2008
New Revision: 20493
URL: http://svn.erp5.org?rev=20493&view=rev
Revert http://svn.erp5.org/?view=rev&revision=20078, as it changes the DOM tree
and desactivate some CSS configuration.

Could you please specify WHICH css configuration does it break? So that
somebody can fix the css?

Hello,
http://mail.nexedi.com/pipermail/erp5-report/2008-April/021001.html).

Would you mind me fixing it and putting it back?

I think the second failure can be fixed by adjusting the test itself,
and the first requires minor change to the css.

Post by Romain Courteaud
As all functionalities of the CSS are not tested, and as your
modification is for now only used in experimental, I decided to move
your change to the experimental repo.

I think it is better to keep the number of changes in experimental as
low as possible, otherwise we will run into compatibility problems. So,
like in this case: if there is a new feature in experimental which
requires only a minor change in the core, then it's better to make this
change in the core - this would reduce maintenance cost, and later
merging experimental features into the core will be much easier.

Bartek

Post by Romain Courteaud
Regards,
Romain
_______________________________________________
Erp5-dev mailing list
Erp5-dev at erp5.org
http://mail.nexedi.com/mailman/listinfo/erp5-dev

[Romain Courteaud]

Post by bartek

Post by Romain Courteaud
http://mail.nexedi.com/pipermail/erp5-report/2008-April/021001.html).

Would you mind me fixing it and putting it back?

I would be very happy if you could do it.

Post by bartek
I think the second failure can be fixed by adjusting the test itself,
and the first requires minor change to the css.

Those functionnal tests are only the top of the iceberg, and more fixes
are probably required.

Post by bartek

Post by Romain Courteaud
As all functionalities of the CSS are not tested, and as your
modification is for now only used in experimental, I decided to move
your change to the experimental repo.

I think it is better to keep the number of changes in experimental as
low as possible, otherwise we will run into compatibility problems. So,
like in this case: if there is a new feature in experimental which
requires only a minor change in the core, then it's better to make this
change in the core - this would reduce maintenance cost, and later
merging experimental features into the core will be much easier.

I completely agree with your point of point, except that I don't
consider the change done in 'field_render' as minor.

Currently, there is no automatic verification of the rendering of pages
by the CSS. It is only validated by developpers because we use the
system each day, but, some css classes are not checked usually,
because not used in most of the forms (like invisible, maybe figure, and
probably others).

In order to integrate your change, somebody has to check everything
manually and need to write tests for all css functionnalities.
With those tests, modifications will be easier to integrate later, as we
will all have certitude of CSS usability.

Personnaly, I don't know how to automatically guaranty the CSS. So, I
think you will have to check all the existing CSS manually.

As it was not done yet, I prefered to push the change to experimental,
where it is only used currently.

Regards,
Romain