On Tue, Aug 16, 2011 at 23:05, Dan Allen <span dir="ltr"><<a href="mailto:dan.j.allen@gmail.com">dan.j.allen@gmail.com</a>></span> wrote:<br><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
<div class="im">On Tue, Aug 16, 2011 at 22:57, Shane Bryzak <span dir="ltr"><<a href="mailto:sbryzak@redhat.com" target="_blank">sbryzak@redhat.com</a>></span> wrote:<br></div><div class="gmail_quote"><div class="im">
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div text="#000000" bgcolor="#FFFFFF">
Of course, but we break that rule. Solder is one example, there's
multiple utility classes in the implementation that are required to
compile other modules.</div></blockquote><div><br></div></div><div>I consider that a bug (or a work in progress, depending on how you look at it).</div></div></blockquote><div><br></div><div>An API is an API. If it's an "internal API", then perhaps it should be an SPI. Under no circumstances should we be depending on implementation classes between modules. Otherwise, we are changing the definition of an implementation.</div>
<div><br></div><div>-Dan</div><div><br></div></div>-- <br><div>Dan Allen</div>Principal Software Engineer, Red Hat | Author of Seam in Action<br>Registered Linux User #231597<br><br><div><a href="http://www.google.com/profiles/dan.j.allen#about" target="_blank">http://www.google.com/profiles/dan.j.allen#about</a><br>
<a href="http://mojavelinux.com" target="_blank">http://mojavelinux.com</a><br><a href="http://mojavelinux.com/seaminaction" target="_blank">http://mojavelinux.com/seaminaction</a><br></div><br>