<div dir="ltr"><div class="gmail_extra"><br><div class="gmail_quote">On Tue, Jun 11, 2013 at 10:16 PM, James M. Lawrence <span dir="ltr"><<a href="mailto:llmjjmll@gmail.com" target="_blank">llmjjmll@gmail.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div id=":24i" style="overflow:hidden">Threading in ECL still seems experimental because it fails some basic<br>
stress tests. I tend to classify thread problems as rather bad because<br>
the door is opened to corruption. Doesn't enabling it by default<br>
communicate the wrong message?<br></div></blockquote></div><br>Independently on whether the MP structures work or not, threading is needed to be able to embed ECL in multithreaded applications, even if only _one_ thread is needed. This applies to many of the present uses of ECL. Moreover, the multithreaded garbage collector should become the default, and Windows can only be built in multithreaded mode because of the way it handles exceptions.</div>
<div class="gmail_extra"><br></div><div class="gmail_extra">That said, it is not my intention to leave the MP structures broken (the former ones were not better, even if the POSIX brand seemed to give some assurance)<br>
<br clear="all">
<div><br></div>-- <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>
</div></div>