You need not configure the printer from the HTML if you do not want to; I only wanted to show what is ultimately possible. I think it could make sense on corporate intranet. You can strip my example to what you consider essential if you like, but once we agree upon such a feature, it seems unwise to forbid such constructs.
In short, your idea was not that wide, but this is its natural scope. If this scope is too wide for you, you should reconsider whether your initial proposition is appropriate at all. Cheers Chris -----Original Message----- From: Sander [mailto:[EMAIL PROTECTED] Sent: Saturday, July 28, 2007 11:22 PM To: Křištof Želechovski; [EMAIL PROTECTED] Subject: Re: [Whatwg] Request for HTML-only print link Křištof Želechovski schreef: href="print://www.whatwg.org/specs/web-apps/current-work/" is no good; it asks the browser to find the resource using the print protocol. But the print protocol is for printing, not for finding resources; I imagine it could be used for finding out some printer configuration parameters (similar to the way printers with a network interface only can be configured) and to submit documents for printing, nothing more. How about <form action="print://host_name/printer_name/? href=&quo;http://www.whatwg.org/specs/web-apps/current-work/ <http://www.whatwg.org/specs/web-apps/current-work/&quo;&> &quo;& palette=mono& copies=3& mode=draft,booklet& stapled=top" method="post" ><input type="submit" value="Print me"></form >? It feels better to me. Of course, the arguments would be interpreted by the browser, not by the printer, contrary to what the syntax suggests, but I think this problem is much smaller and I can swallow it in spite of being a purist. The idea was not to dictate from HTML how the printer should behave (number of copies, color, etc.). This should be up to the visitor, who can manage that in the print prompt. The request was about an alternative to javascript:print() where there would be no need for client-side scripting. I probably gave this discussion a wrong turn by saying that a print-pseudo-protocol would perhaps be a good solution. I guess i should not have done so as my knowledge about such things is minimal.