<div class="gmail_quote">2011/2/2 crox <span dir="ltr"><<a href="mailto:linux.hacker@yandex.ru">linux.hacker@yandex.ru</a>></span><br><blockquote class="gmail_quote" style="margin: 0pt 0pt 0pt 0.8ex; border-left: 1px solid rgb(204, 204, 204); padding-left: 1ex;">
<div id=":1ge">02.02.2011, 19:13, "Juan Jose Garcia-Ripoll" <<a href="mailto:juanjose.garciaripoll@googlemail.com">juanjose.garciaripoll@googlemail.com</a>>:<br>
<div class="im">> ECL does have weak pointers and that could be used to implement weak hash tables, but it has not been done.<br>
</div>Sounds solvable... As far as I can see weak pointers are not documented yet?<br><div class="im"></div></div></blockquote><div><br>No, no documentation, but it is close to trivial.<br><br>(ext:make-weak-pointer o) => wp<br>
<br>(ext:weak-pointer-value wp) => o or NIL<br> <br></div><blockquote class="gmail_quote" style="margin: 0pt 0pt 0pt 0.8ex; border-left: 1px solid rgb(204, 204, 204); padding-left: 1ex;"><div id=":1ge"><div class="im">
> However, in many cases those dependencies can be worked around by library developers. In particular I doubt that trivial-garbage needs weak hash tables at all given that ECL has finalization.<br>
</div>Does it concerns weak hash tables? I think that it is better to implement them, rather than patching cl-gtk2. At least they will be helpful in porting some other packages, that depend on them.</div></blockquote></div>
<br>It would be easier to fix trivial-garbage instead, which is the one "supplying" weak hash tables to other systems.<br><br>Juanjo<br clear="all"><br>-- <br>Instituto de Física Fundamental, CSIC<br>c/ Serrano, 113b, Madrid 28006 (Spain) <br>
<a href="http://juanjose.garciaripoll.googlepages.com" target="_blank">http://juanjose.garciaripoll.googlepages.com</a><br>