<html>
<head>
<style><!--
.hmmessage P
{
margin:0px;
padding:0px
}
body.hmmessage
{
font-size: 12pt;
font-family:Calibri
}
--></style></head>
<body class='hmmessage'><div dir='ltr'><pre style="line-height: 21px; white-space: normal; color: rgb(68, 68, 68); font-size: 15px; "> > quality and performance of the new one though, just to avoid regression.<br></pre><div><br></div>Well..the main "quality" is portability.<div>Much is being potentially sacrificed for it.</div><div><br></div><div><br></div><div>There are potential other improvements, but also potential regressions.</div><div><br></div><div><br></div><div>Optimization might be better, i.e. by not using gcc, or worse, i.e. by going via C.</div><div><br></div><div><br></div><div>Compiler performance is likely worse, due to the extra text production and parsing.</div><div><div>It seems ok to me. I've been using a while. But not great. Skipping the .mc/m3cg file production helped a lot.</div></div><div><br></div><div><br></div><div>Debugging will be better on Darwin and NT -- where currently we have no type info, I hope to add significant typing.</div><div><br></div><div><br></div><div>"Ease of distribution" will take a step forward, but it isn't yet where I want it to be.</div><div>You know, you just want tar xf; configure; make; make install. No need for cross builds or "unusual" prerequesites, "just" a C compiler and Bourne shell and make.</div><div><br></div><div><br><div><div>Portability will increase.</div><div>Maintainability will increase.</div><div>We will no longer have a motivation for "chasing" newer gcc.</div><div> (nor patches e.g. for OpenBSD)</div><div><span style="font-size: 12pt; ">We will "automatically" port to "every" target. This is an exaggeration, but not by a lot.</span></div><div>The "porting" will work will decrease. </div><div>Porting work will further decrease when we have cooperative suspend.</div><div><br></div><div><br></div><div> - Jay<br><br><br><div><div id="SkyDrivePlaceholder"></div>> Date: Wed, 6 Feb 2013 09:34:09 +0100<br>> From: wagner@elegosoft.com<br>> To: jay.krell@cornell.edu<br>> CC: m3devel@elegosoft.com<br>> Subject: Re: [M3devel] and move to gcc 4.7?<br>> <br>> On Wed, 6 Feb 2013 08:29:03 +0000<br>> Jay K <jay.krell@cornell.edu> wrote:<br>> <br>> > Olaf, things working satisfactorily for you in Hudson now for AMD64_LINUX and AMD64_FREEBSD?<br>> Recent builds seem to have succeeded.<br>> <br>> > I'd like to bump m3cc/src/m3makefile up to 4.7 for AMD64_LINUX, FREEBSD. Or anything we have going decently in Hudson.Even though, granted, I'd like to remove that backend sooner rather than later anyway..AMD64_LINUX was at 4.7 but I backed it down to 4.5 today out of paranoia.<br>> <br>> Just try it.<br>> <br>> Before we remove the old backend, I'd like to have some validation of the<br>> quality and performance of the new one though, just to avoid regression.<br>> <br>> Olaf<br>> -- <br>> Olaf Wagner -- elego Software Solutions GmbH -- http://www.elegosoft.com<br>> Gustav-Meyer-Allee 25 / Gebäude 12, 13355 Berlin, Germany<br>> phone: +49 30 23 45 86 96 mobile: +49 177 2345 869 fax: +49 30 23 45 86 95<br>> Geschäftsführer: Michael Diers, Olaf Wagner | Sitz: Berlin<br>> Handelregister: Amtsgericht Charlottenburg HRB 77719 | USt-IdNr: DE163214194<br></div></div></div></div> </div></body>
</html>