Stefan,

Another way to include b2evo content on your page would be to use an RSS aggregator and insert that into your page.

I've used Feedburner widgets for that. This is a nice easy trick. From what you're describing, this would solve your issue of putting content on the main page.

Once people click on the actual permalink, then that would bring them to your b2evo skin.

I looked into that, but that won´t probably work: I do not want a a part of the blog in my page an then redirect for the complete to the normal skin, I want everything integrated in my page (not administration stuff). The feeds do not allow calender, extended posts, feedback form etc.
This does not seem to be an option.

Are there other solutions?

I got another questions, what get-variables does b2evo (blog.php?a=1&b=2 etc.) use? They should not interfere with my variables.

Regards,
Stefan

I want the Admin panel etc. only be available using SSL

Search for blueyed and https ;)

I want b2evo to use a special class (class="something") for all local URLs, I need this because my site can use Ajax to load new pages, how can I do that?

If you require js already the just use js to add the class to local links?

Unfortunately b2evo does not always comply to XHTML strict, although I enabled $use_strict somewhere. I already edited a lot of files to ensure it complies to strict. Have I missed something to enforce strict?

Can you point me to pages that fails validation, I'll happily fix in 2.4.x / cvs ... I like green ticks ;)

I noticed that there are lot of messages in english, but by default it should use german. Is it a problem of my installation or is it not fully translated?

Sorry, I only speak/read english, blueyed is german though, so I would have thought german was 100% translated/implemented?

The functions used in a_noskin.php are not very well documented in the file, is there somekind of documentation what parameters exist that I can push into this arrays?

Urm, there is a way to get to them via the "docs" link in the menu at the top of the screen ... I normally read the files though :p

The PHP code is often mixed with the output, I do not want the user to click somewhere what I haven´t seen and then everything breaks because something inserts a lot of stuff into my page. Is there an option to find out all possible outputs by these functions?

Most of the output is overridable with the params

I noticed it sometimes includes some javascript stuff, which doesn´t work becuase I haven´t included any javascript files from b2evo in my page. Is there some documentation about that?

Probably not, but what js fails ( urls would be helpful )

¥

Search for blueyed and https

It does not work perfectly, because the rsc-folder is not loaded encrypted. In case I do the trick in the _advanced.php the rsc folder is always loaded encrypted. Can I redefine $rsc_url at the beginning of admin.php without any problems?

If you require js already the just use js to add the class to local links?

JS uses this class-attribute to identify the links to modify, because only 99% of the local links have to be modfied.
Next problem, b2evo does not use local links.

I tried to make it use local links by setting $baseurl to './' and all paths in _advanced.php to $baseurl.'blogs/. I also modified two preg_replace uses, one in _advanced.php (the setting port, host etc.) and another in _blog.class.php. Also the blog itself uses local links now, it´s very often redirecting to itself in an endless loop (categories, comments and opening single entries), search and archive are working fine:

Request:
GET ./blog.php?cat=5
Answer:
301 Moved Permanently
Location: ./blog.php?cat=5

Any idea why it gets redirected?

Can you point me to pages that fails validation, I'll happily fix in 2.4.x / cvs ... I like green ticks

There is some img code which has none-css fallback, in one function there is
if (!$use_strict)..., but in another there is always non-css fallback.
Other problems (only with XHTML 1.1) are: There is no lang-attribute, it´s now called "xml:lang" and a table is required to have a tbody if there is thead and tfoot (in the calender).

Sorry, I only speak/read english, blueyed is german though, so I would have thought german was 100% translated/implemented?

At least for me it´s not. poEdit tells me: 2824 strings in total, 544 fuzzy (translation does not really match/incomplete) and 364 not translated. Unfortunately in this 364 strings there are several important ones.

Probably not, but what js fails ( urls would be helpful )

It was something with comments, but because it´s broken I can´t find out exactly at the moment.

Regards,
Stefan

Edit:
I found out that there are three (?) positions in the code where the uri is checked if it´s canonical (also called canoncical in _skin.funcs.php).
Because the canonical uri is always local now, but the request uri is absolute it failed and redirects to itself. I modified the code there in a way only the end is compared after the last slash. It seems to work now.

Edit 2:
the javascript code:

<script type="text/javascript">
				//<![CDATA[
				if( typeof addEvent == "function" && typeof init_dynamicSelect == "function" )
				{
					addEvent( window, "load", init_dynamicSelect, false );
					
				}
				
				//]]>
			</script>

and

<script type="text/javascript">var b2evoCanvas = document.getElementById( "p" );</script>

You should be able to alter the $rsc_url; in admin.php just after the config files are called :

/**
 * Do the MAIN initializations:
 */
require_once dirname(__FILE__).'/conf/_config.php';
$rsc_url = 'https://domain.com/rsc/';

I think you'll end up with fun and games if you start trying to use relative links in evo, it's far easier to just amend your js to treat all http://domain.com as local links ( which is what I do on my blog )

There is some img code which has none-css fallback, in one function there is
if (!$use_strict)..., but in another there is always non-css fallback.

There's 2,800 files in evocore, any chance of narrowing it down to the files that you've noticed it in?

I find that the best thing to do with the calendar is to throw it in the bin, for me it's not worth the server overheads for what is ( essentially ) a bauble.

To get rid of the js error just add a fake function ;)

<script type="text/javascript">function init_dynamicSelect(){}</script>

¥