It's strange. I loaded last version of XMLParser and all work very well (at 3 times faster)<br><br><div class="gmail_quote">2010/6/25 Igor Stasenko <span dir="ltr"><<a href="mailto:siguctua@gmail.com" target="_blank">siguctua@gmail.com</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;">2010/6/22 Denis Kudriashov <<a href="mailto:dionisiydk@gmail.com" target="_blank">dionisiydk@gmail.com</a>>:<br>
<div><div></div><div>> Hello.<br>
><br>
> Cog performance is incredible. My application improved 3 times speedup. And<br>
> I now does not need special caching logic in my program.<br>
><br>
> But I found very strange behavior in xml dom parsing of big text with many<br>
> russian text (by XML-Parser XMLDOMParser class).<br>
><br>
> russionXml := String streamContents: [:s | s nextPutAll: '<root>'.<br>
> 1000 timesRepeat: [<br>
> s nextPutAll: '<item>'.<br>
> s nextPutAll: 'ĘŁ×ĮŠŅĻĢÄ'.<br>
> s nextPutAll: '</item>'].<br>
> s nextPutAll: '</root>'<br>
> ].<br>
><br>
> [XMLDOMParser parseDocumentFrom: russionXml] timeToRun .<br>
><br>
> In pharo 1.1 I have<br>
><br>
> 567 ms with old VM.<br>
> 6673 ms with Cog VM.<br>
><br>
> But all other stuff that I tested work faster with Cog (at 3-4 times<br>
> faster).<br>
><br>
<br>
</div></div>Looks like the code using different paths.<br>
I suspect that in one case, some operations are done primitively,<br>
while in another one - interpreted.<br>
<div><div></div><div><br>
><br>
><br>
><br>
><br>
><br>
><br>
<br>
<br>
<br>
--<br>
Best regards,<br>
Igor Stasenko AKA sig.<br>
<br>
</div></div></blockquote></div><br>