Re: removing w3c logos from distributed versions of validators/link checker

On 23-Mar-09, at 3:41 PM, Ville Skyttä wrote:
> 1) docs/checklink.html still has a reference
> to ../images/valid_icons/valid-xhtml10-blue.png.  Maybe get rid of  
> the whole
> validity badge there?

Yes, I think so.

> 2) images/w3c.png still exists in CVS, and bin/checklink and
> docs/checklink.html still refer to it instead of no_w3c.png.  The  
> easiest
> solution would seem to be to overwrite images/w3c.png with the current
> images/no_w3c.png (and revert the related changes to MANIFEST and  
> SIGNATURE).

That was the idea indeed. I was thinking the makefile/tarball making  
script could basically do a `mv images/no_w3c.png images/w3c.png` but  
that might not be right.

> On the other hand this will lead to local installations displaying  
> the online
> w3c.png which is the "real" W3C logo by default - IIUC this was not
> desirable.

How would they end up with the image, if it is not distributed? Or do  
you mean that installations done from CVS would display the w3c logo?  
That would indeed by a little problematic, though acceptable (the main  
issue being, after all, tarballs going to distribs like debian and  
fedora).

>  Another solution would be to change bin/checklink and
> docs/checklink.html in CVS to refer to no_w3c.png, and customize the  
> version
> running on official validator servers to refer to w3c.png instead  
> (or to make
> sure no_w3c.png is actually the real w3c.png there).  This way  
> images/w3c.png
> could stay in CVS, it'd just not end up in the dist tarball.

I suppose that works, indeed. Probably simpler, and although it would  
be extra work whenever updating the instance on validator.w3.org, that  
seems like something the w3c team could live with.

> 3) images/no_w3c.png is missing from MANIFEST and SIGNATURE, but  
> fixing this
> depends on how 2) above is solved (and SIGNATURE will be autogenerated
> anyway, no need to modify it except when finalizing a release).

Gotcha, I wasn't certain about SIGNATURE. I vote we follow your  
(second) solution.

-- 
olivier

Received on Monday, 23 March 2009 20:09:21 UTC