On Wed, Mar 31, 2010 at 8:11 PM, Faré <span dir="ltr"><<a href="mailto:fahree@gmail.com" target="_blank">fahree@gmail.com</a>></span> wrote:<br><div class="gmail_quote"><blockquote class="gmail_quote" style="margin: 0pt 0pt 0pt 0.8ex; border-left: 1px solid rgb(204, 204, 204); padding-left: 1ex;">
There are many things to clarify about logical pathnames and the use<br>
thereof or not, but I don't think this should block a release of ASDF<br>
2. Should it? And so I'd like to declare it an ASDF 2.1 or ASDF 3<br>
issue, and invite you to focus on blocking issues for ASDF 2. (Of<br>
course, if someone has a working, portable, solution that makes<br>
everyone happy, I'll gladly merge and commit it.)<br></blockquote><div><br>I have uploaded a newer version of the feature that allows two different syntaxes<br><br>:logical-host "MY-HOST"<br>:logical-pathname "MY-HOST:MY-BASE"<br>
<br>The three pathnames that are defined are "xxx;SRC;*.*.*", "xxx;DATA;*.*.*" and "xxx;OUTPUT-FILES;*.*.*" where the first two map to the source directory of the system and the second one to the translated directory for FASL and other generated files. The "xxx" here denotes the argument provided by the user. Note that the code will complain if there are wildcards.<br>
</div><blockquote class="gmail_quote" style="margin: 0pt 0pt 0pt 0.8ex; border-left: 1px solid rgb(204, 204, 204); padding-left: 1ex;">
I think the asdf-bundle extension is more important, because<br>
underlying it is the structure of how we deliver extensions, and how<br>
we play well with ECL, especially as it might entail some slight<br>
backwards incompatibility for ECL that I'd rather not happen *after*<br>
we release ASDF 2.<br>
</blockquote></div><br>Actually if you plan an immediate release I would leave that for 2.1 because I would rather sort out that thing of the incompatibility later, and also because I believe there is room for improvement in the patch, including support for other implementations.<br>
<br>For the record, even if the ASDF crew decided to leave the bundle out of the core and thus expel ECL's extensions from it as well, I would continue to ship everything packed together with ECL, so that a single REQUIRE is needed.<br clear="all">
<br>Juanjo<br><br>-- <br>Instituto de Física Fundamental, CSIC<br>c/ Serrano, 113b, Madrid 28006 (Spain) <br><a href="http://tream.dreamhosters.com" target="_blank">http://tream.dreamhosters.com</a><br>