<?xml version="1.0" encoding="ISO-8859-1"?>

<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/">
	<channel>
		<title>Data Access Worldwide Forums - DataFlex Technology Stack Testing</title>
		<link>https://support.dataaccess.com/Forums/</link>
		<description>A forum dedicated to discussing, testing, and shaping the future of the Technology Stack project.</description>
		<language>en</language>
		<lastBuildDate>Wed, 15 Jul 2026 11:03:33 GMT</lastBuildDate>
		<generator>vBulletin</generator>
		<ttl>60</ttl>
		<image>
			<url>https://support.dataaccess.com/Forums/images/misc/rss.png</url>
			<title>Data Access Worldwide Forums - DataFlex Technology Stack Testing</title>
			<link>https://support.dataaccess.com/Forums/</link>
		</image>
		<item>
			<title>Various string operations compared</title>
			<link>https://support.dataaccess.com/Forums/showthread.php?73379-Various-string-operations-compared&amp;goto=newpost</link>
			<pubDate>Tue, 07 Jul 2026 10:49:39 GMT</pubDate>
			<description><![CDATA[So, inspired by Samuel and Marco's thoughts about strings...]]></description>
			<content:encoded><![CDATA[<div>So, inspired by Samuel and Marco's <a href="https://support.dataaccess.com/Forums/showthread.php?73355-TechStack-WishList-better-array-assignement-syntax" target="_blank">thoughts about strings</a>, I built some tests and ran them under the various runtimes we (where &quot;<i>we</i>&quot; refers to the chosen few from the TechStack training at FlexRoads ;)) have access to in DF27 (pre-pre-alpha ;)).<br />
<br />
I've summarised the results (only one run each - yes, I know I should do it several times and average the results :o) in a Google Sheet <a href="https://docs.google.com/spreadsheets/d/1ryo_NbMEGVBqZrSUoJY1LkCCocIiNoplPtTy8Bbhv5Q/edit?usp=sharing" target="_blank">here</a>, but have also attached them as a PDF here <a href="https://support.dataaccess.com/Forums/attachment.php?attachmentid=18226&amp;d=1783419955"  title="Name:  DF 27 (pre-alpha) string tests - Sheet1.pdf
Views: 26
Size:  51.4 KB">DF 27 (pre-alpha) string tests - Sheet1.pdf</a> (the forum doesn't like .xlsx files it seems).  All results are in milliseconds (for which I had to write my own arithmetic routines as the TimeSpan stuff isn't working yet).<br />
<br />
The basis of the tests is that I load a string with A-Z, a-z comma-separated values (so 52 characters, with commas between then) 5,000 times (so the resulting string is 519,999 characters long).  I do that by first loading them into a UChar array and then converting that to a string (<i>very</i> quick under all implementations).  For comparison I also load a string array with the same characters (but without the commas, so just 260,000 of them).<br />
<br />
I then do a bunch of string-things on those: split to an array (the new runtimes do this in native code, rather than DF code, so are <i>very</i> much faster), finding and replacing, replaces, joining, inserting, removing, finding and overstriking, etc.<br />
<br />
It was a bit of a pain to get the same essential code to run under all the different toolchains, but it is done now.  I'd really like it if there was an easy way to check which toolchain you are using, but I guess that will come in time.<br />
<br />
Mike<br />
<br />
PS - Can't seem to spell &quot;operations&quot; today!  :o<br />
<br />
PPS - The bold column headings are what (SysConf(SYSCONF_OS_NAME)) returns for each of the five extant toolchains, plus, in the Windows cases, &quot;- 64 bit&quot; or &quot;- 32 bit&quot; depending on the compile-time value of Is$Win64.</div>


	<div style="padding:10px">

	

	

	

	
		<fieldset class="fieldset">
			<legend>Attached Files</legend>
			<ul>
			<li>
	<img class="inlineimg" src="https://support.dataaccess.com/Forums/images/attach/pdf.gif" alt="File Type: pdf" />
	<a href="https://support.dataaccess.com/Forums/attachment.php?attachmentid=18226&amp;d=1783419955">DF 27 (pre-alpha) string tests - Sheet1.pdf</a> 
(51.4 KB)
</li>
			</ul>
		</fieldset>
	

	</div>
]]></content:encoded>
			<category domain="https://support.dataaccess.com/Forums/forumdisplay.php?101-DataFlex-Technology-Stack-Testing">DataFlex Technology Stack Testing</category>
			<dc:creator>Mike Peat</dc:creator>
			<guid isPermaLink="true">https://support.dataaccess.com/Forums/showthread.php?73379-Various-string-operations-compared</guid>
		</item>
		<item>
			<title>A funny one... ;)</title>
			<link>https://support.dataaccess.com/Forums/showthread.php?73378-A-funny-one-)&amp;goto=newpost</link>
			<pubDate>Tue, 07 Jul 2026 07:11:34 GMT</pubDate>
			<description><![CDATA[Wrote "Object oXYZ" but forgot to add "is a cABC".  I was using the 0.1.5+wasm-32 toolchain at the time, so was amused to see the error: "*NO_IMAGE,...]]></description>
			<content:encoded><![CDATA[<div>Wrote &quot;Object oXYZ&quot; but forgot to add &quot;is a cABC&quot;.  I was using the 0.1.5+wasm-32 toolchain at the time, so was amused to see the error: &quot;<b>NO_IMAGE, INVALID LITERAL&quot;</b>.  ;)<br />
<br />
Mike</div>

]]></content:encoded>
			<category domain="https://support.dataaccess.com/Forums/forumdisplay.php?101-DataFlex-Technology-Stack-Testing">DataFlex Technology Stack Testing</category>
			<dc:creator>Mike Peat</dc:creator>
			<guid isPermaLink="true">https://support.dataaccess.com/Forums/showthread.php?73378-A-funny-one-)</guid>
		</item>
		<item>
			<title>Oh no!  :(</title>
			<link>https://support.dataaccess.com/Forums/showthread.php?73374-Oh-no!-(&amp;goto=newpost</link>
			<pubDate>Fri, 03 Jul 2026 14:55:56 GMT</pubDate>
			<description><![CDATA[It seems that my favourite nasty hack: 
 
Code: 
--------- 
While (num_arguments >= iArg) 
     Move iArg& to sArg 
           : 
           : 
    ...]]></description>
			<content:encoded><![CDATA[<div>It seems that my favourite nasty hack:<br />
<div class="bbcode_container">
	<div class="bbcode_description">Code:</div>
	<hr /><code class="bbcode_code">While (num_arguments &gt;= iArg)<br />
&nbsp; &nbsp;  Move iArg&amp; to sArg<br />
&nbsp; &nbsp; &nbsp; &nbsp; &nbsp;  :<br />
&nbsp; &nbsp; &nbsp; &nbsp; &nbsp;  :<br />
&nbsp; &nbsp;  Increment iArg<br />
Loop</code><hr />
</div>Is no longer allowed: &quot;<i>Error 4391 ... Indexed and split fields are no longer allowed, use the reflection API.</i>&quot;  :(<br />
<br />
Mike</div>

]]></content:encoded>
			<category domain="https://support.dataaccess.com/Forums/forumdisplay.php?101-DataFlex-Technology-Stack-Testing">DataFlex Technology Stack Testing</category>
			<dc:creator>Mike Peat</dc:creator>
			<guid isPermaLink="true">https://support.dataaccess.com/Forums/showthread.php?73374-Oh-no!-(</guid>
		</item>
		<item>
			<title><![CDATA[No "LINK ERROR" on new compilers]]></title>
			<link>https://support.dataaccess.com/Forums/showthread.php?73371-No-quot-LINK-ERROR-quot-on-new-compilers&amp;goto=newpost</link>
			<pubDate>Thu, 02 Jul 2026 12:18:45 GMT</pubDate>
			<description><![CDATA[If you compile a program while a previous version of it is running, under the "old" toolchains (27.0.0-windows32/64) you get a "- LINK ERROR: UNABLE...]]></description>
			<content:encoded><![CDATA[<div>If you compile a program while a previous version of it is running, under the &quot;old&quot; toolchains (27.0.0-windows32/64) you get a &quot;<font color="#0000cd">- LINK ERROR: UNABLE TO OPEN FILE...</font>&quot; with a red circled exclamation-point making it clear it didn't work.<br />
<br />
However under the &quot;new&quot; toolchains this doesn't happen.  Instead the compiler reports &quot;<font color="#0000cd">----Compile Finished in <i>n.n</i> seconds ---- ... ----Executable written: ...</font>&quot; , which last statement <u>really</u> isn't true.  To compound the confusion, if you then realise your error, stop the running program, go back to the Studio and recompile, you <i>then</i> get &quot;<font color="#0000CD">- No changes detected, skipping compile ... ----Compile Finished in 0 seconds ---- ... ----Executable written: ...</font>&quot; (again untrue) - you have to make a change to an involved source file in order to force the compiler into action.  'Tis confusing for a poor fuddled developer.  :confused:<br />
<br />
Mike</div>

]]></content:encoded>
			<category domain="https://support.dataaccess.com/Forums/forumdisplay.php?101-DataFlex-Technology-Stack-Testing">DataFlex Technology Stack Testing</category>
			<dc:creator>Mike Peat</dc:creator>
			<guid isPermaLink="true">https://support.dataaccess.com/Forums/showthread.php?73371-No-quot-LINK-ERROR-quot-on-new-compilers</guid>
		</item>
		<item>
			<title>Debugging throwing a COM error</title>
			<link>https://support.dataaccess.com/Forums/showthread.php?73367-Debugging-throwing-a-COM-error&amp;goto=newpost</link>
			<pubDate>Tue, 30 Jun 2026 10:25:51 GMT</pubDate>
			<description>Doing a bit of debugging, using (in this case) the 27.0.0+windows-64 toolchain, I am getting a COM error: 
 
 
Code: 
--------- 
C:\Program...</description>
			<content:encoded><![CDATA[<div>Doing a bit of debugging, using (in this case) the 27.0.0+windows-64 toolchain, I am getting a COM error:<br />
<br />
<div class="bbcode_container">
	<div class="bbcode_description">Code:</div>
	<hr /><code class="bbcode_code">C:\Program Files\DataFlex 27.0\Bin64\Studio.exe<br />
COM object method invocation error<br />
<br />
<br />
Internal Error: boost::any_cast() failed unexpectedly. Contained type, if any, is incorrect<br />
<br />
<br />
Error: 4399<br />
<br />
<br />
Please report this error to Data Access Corporation<br />
<br />
<br />
GET_COMEVAL (14038) - oDebuggerEngine (882) - at address 109466<br />
MSG_ONCOMEVALUATEQUICKINFO (13473) - oCodeComplete (5296) - at address 146976<br />
[start] - at address 292193</code><hr />
</div>When hovering over a DateTime variable (dtStart), just after doing: Move (CurrentDateTime()) to it - the breakpoint is on the next line: I'm just trying to check what is in it.<br />
<br />
Edit: The &quot;Locals&quot; pane shows for dtStart &quot;Internal Error: boost::any_cast() failed unexpectedly. Contained type, if any, is incorrect&quot;. <br />
<br />
Mike</div>

]]></content:encoded>
			<category domain="https://support.dataaccess.com/Forums/forumdisplay.php?101-DataFlex-Technology-Stack-Testing">DataFlex Technology Stack Testing</category>
			<dc:creator>Mike Peat</dc:creator>
			<guid isPermaLink="true">https://support.dataaccess.com/Forums/showthread.php?73367-Debugging-throwing-a-COM-error</guid>
		</item>
		<item>
			<title>DateTime arithmetic no (yet) working</title>
			<link>https://support.dataaccess.com/Forums/showthread.php?73365-DateTime-arithmetic-no-(yet)-working&amp;goto=newpost</link>
			<pubDate>Mon, 29 Jun 2026 15:00:37 GMT</pubDate>
			<description><![CDATA[Just saying ('coz I'm sure it will get fixed in due course), but in messing around with measuring how fast things were working I noticed that my...]]></description>
			<content:encoded><![CDATA[<div>Just saying ('coz I'm sure it will get fixed in due course), but in messing around with measuring how fast things were working I noticed that my usual:<br />
<div class="bbcode_container">
	<div class="bbcode_description">Code:</div>
	<hr /><code class="bbcode_code">&nbsp; &nbsp; DateTime stStart dtEnd<br />
&nbsp; &nbsp; TimeSpan tsTook<br />
&nbsp; &nbsp; &nbsp; &nbsp; :<br />
&nbsp; &nbsp; &nbsp; &nbsp; :<br />
<br />
&nbsp; &nbsp; Move (CurrentDateTime()) to dtStart<br />
<br />
&nbsp; &nbsp; <font color="#006400">// Thing that I'm measuring...</font><br />
<br />
&nbsp; &nbsp; Move (CurrentDateTime()) to dtEnd<br />
&nbsp; &nbsp; Move (dtEnd - dtStart) to tsTook</code><hr />
</div>Doesn't (yet) work in the 0.1.5+native-64/32 toolchain versions.  tsTook seems (in the debugger) to hold a value similar to 4:14:25:51.912 (4+ days! ;)) while the actual time was well under 1 second, while Showln (String(tsTook)) displays 02:25:51 (no milliseconds), from which I concluded that TimeSpan handling code is a bit broken ATM.<br />
<br />
So I've had to hand-roll my own code to work out and display the difference between two DateTimes.<br />
<br />
Mike</div>

]]></content:encoded>
			<category domain="https://support.dataaccess.com/Forums/forumdisplay.php?101-DataFlex-Technology-Stack-Testing">DataFlex Technology Stack Testing</category>
			<dc:creator>Mike Peat</dc:creator>
			<guid isPermaLink="true">https://support.dataaccess.com/Forums/showthread.php?73365-DateTime-arithmetic-no-(yet)-working</guid>
		</item>
		<item>
			<title>Tell what toochain is used</title>
			<link>https://support.dataaccess.com/Forums/showthread.php?73362-Tell-what-toochain-is-used&amp;goto=newpost</link>
			<pubDate>Fri, 26 Jun 2026 15:34:12 GMT</pubDate>
			<description>Is there a mechanism for checking which toolchain a program has been compiled under?  Like IS$WIN64. 
 
Mike</description>
			<content:encoded><![CDATA[<div>Is there a mechanism for checking which toolchain a program has been compiled under?  Like IS$WIN64.<br />
<br />
Mike</div>

]]></content:encoded>
			<category domain="https://support.dataaccess.com/Forums/forumdisplay.php?101-DataFlex-Technology-Stack-Testing">DataFlex Technology Stack Testing</category>
			<dc:creator>Mike Peat</dc:creator>
			<guid isPermaLink="true">https://support.dataaccess.com/Forums/showthread.php?73362-Tell-what-toochain-is-used</guid>
		</item>
		<item>
			<title>Timeline please</title>
			<link>https://support.dataaccess.com/Forums/showthread.php?73361-Timeline-please&amp;goto=newpost</link>
			<pubDate>Fri, 26 Jun 2026 15:24:32 GMT</pubDate>
			<description>Yes, I am whining.  
 
I know I know, you said the preview build is coming soon, and I know the people that attended that Amsterdam conference got...</description>
			<content:encoded><![CDATA[<div>Yes, I am whining. <br />
<br />
I know I know, you said the preview build is coming soon, and I know the people that attended that Amsterdam conference got the preview build. It's just annoying to see people talking and using the preview already and I am left with nothing to play with. <br />
<br />
Frank Cheng</div>

]]></content:encoded>
			<category domain="https://support.dataaccess.com/Forums/forumdisplay.php?101-DataFlex-Technology-Stack-Testing">DataFlex Technology Stack Testing</category>
			<dc:creator>Frank Cheng</dc:creator>
			<guid isPermaLink="true">https://support.dataaccess.com/Forums/showthread.php?73361-Timeline-please</guid>
		</item>
		<item>
			<title>Welcome to the Technology Stack testing forum</title>
			<link>https://support.dataaccess.com/Forums/showthread.php?73358-Welcome-to-the-Technology-Stack-testing-forum&amp;goto=newpost</link>
			<pubDate>Thu, 25 Jun 2026 19:21:26 GMT</pubDate>
			<description>Welcome to the Technology Stack testing forum. This new section is dedicated to discussing, testing, and shaping the future of the Technology Stack...</description>
			<content:encoded><![CDATA[<div>Welcome to the Technology Stack testing forum. This new section is dedicated to discussing, testing, and shaping the future of the Technology Stack project: the development of a new DataFlex runtime that, among many other improvements, will bring back multi-platform support.<br />
<br />
The first major target platform is WebAssembly, enabling WebApps with offline capabilities that can be packaged as real mobile applications. This will be followed by server-side WebApps and FlexTron desktop applications on Windows. Looking further ahead, we also plan to support additional platforms such as Linux and macOS through a new Web Server.<br />
<br />
The first DataFlex release to include this new runtime will be DataFlex 2027. A Technology Preview is coming soon, and trainees from FlexRoads have already received an early build. We expect an extended first release cycle and plan to provide regular monthly updates as we work toward a stable first release.<br />
<br />
In the coming preview, you can expect the first hands-on look at the new runtime in action, including the foundations for WebAssembly support and the path toward offline-capable, cross-platform DataFlex applications. We look forward to your feedback, questions, and real-world testing as this work progresses.</div>

]]></content:encoded>
			<category domain="https://support.dataaccess.com/Forums/forumdisplay.php?101-DataFlex-Technology-Stack-Testing">DataFlex Technology Stack Testing</category>
			<dc:creator>Harm Wibier</dc:creator>
			<guid isPermaLink="true">https://support.dataaccess.com/Forums/showthread.php?73358-Welcome-to-the-Technology-Stack-testing-forum</guid>
		</item>
		<item>
			<title>TechStack - WishList  - better array assignement syntax</title>
			<link>https://support.dataaccess.com/Forums/showthread.php?73355-TechStack-WishList-better-array-assignement-syntax&amp;goto=newpost</link>
			<pubDate>Wed, 24 Jun 2026 17:12:34 GMT</pubDate>
			<description>Please, please, please...  
 
provide support to have a easier, and more direct way of assigning multiple items to an array...   
 
basically,...</description>
			<content:encoded><![CDATA[<div>Please, please, please... <br />
<br />
provide support to have a easier, and more direct way of assigning multiple items to an array...  <br />
<br />
basically, instead of <br />
<div class="bbcode_container">
	<div class="bbcode_description">Code:</div>
	<hr /><code class="bbcode_code">Move &quot;A&quot; to sArray[0]<br />
Move &quot;B&quot; to sArray[1]<br />
Move &quot;C&quot; to sArray[2] <br />
...</code><hr />
</div>Something like <br />
<br />
<div class="bbcode_container">
	<div class="bbcode_description">Code:</div>
	<hr /><code class="bbcode_code">Move [&quot;A&quot;,&quot;B&quot;,&quot;C&quot;] to sArray</code><hr />
</div>It does not need to be necessarily like this.  but you got the idea.. <br />
<br />
I would appreciate it a lot!</div>

]]></content:encoded>
			<category domain="https://support.dataaccess.com/Forums/forumdisplay.php?101-DataFlex-Technology-Stack-Testing">DataFlex Technology Stack Testing</category>
			<dc:creator>Samuel Pizarro</dc:creator>
			<guid isPermaLink="true">https://support.dataaccess.com/Forums/showthread.php?73355-TechStack-WishList-better-array-assignement-syntax</guid>
		</item>
	</channel>
</rss>
