[Mysql-cocoa-commits] CVS: Documentation/website archives.html,1.3,1.4 index.html,1.19,1.20
Brought to you by:
sergecohen
From: Serge C. <ser...@us...> - 2004-08-09 16:07:59
|
Update of /cvsroot/mysql-cocoa/Documentation/website In directory sc8-pr-cvs1.sourceforge.net:/tmp/cvs-serv28494/website Modified Files: archives.html index.html Log Message: Updated the website to include the release of version 2.3.0 and 2.3.1. Updated also the API documentation. 2004-08-09; Serge Cohen. Index: archives.html =================================================================== RCS file: /cvsroot/mysql-cocoa/Documentation/website/archives.html,v retrieving revision 1.3 retrieving revision 1.4 diff -C2 -d -r1.3 -r1.4 *** archives.html 30 Aug 2003 13:43:22 -0000 1.3 --- archives.html 9 Aug 2004 16:07:49 -0000 1.4 *************** *** 18,21 **** --- 18,25 ---- <a name="home"><p>Home page of the MySQL Cocoa project can be found <a href="index.html">here</a></p></a> + <p><em>25 August 2003:</em> Today I finally had the time to update the web site. The information you find here should now be up to date (especially the <a href="#files">Files</a>, <a href="#docs">Documentation</a> were completely outdated – I had no time to update them since I first wrote them in January 2002 –).</p> + + <p><em>20 August 2003:</em> Releases the version 2.2.1 of the framework: minor bug correction. The previous versions of the framework were not able to handle properly the integer with more than 32bits (MySQL BIGINT). This is now corrected (indeed the other branches of the framework are also corrected in CVS, so if you are still using the version 1, I encourage you to get the fix from CVS). Unfortunately I have not found a solution to know from the MySQL server if a retrieved integer is signed or not, hence the framework assumes always signed int: this can cause a trouble if you have a unsigned column (either for BIGINT or INT) and you are using number bigger than the maximum SIGNED number. Thanks to Lorenz Textor for pointing the bug to me.</p> + <p class="withlist"><em>18 August 2003:</em> Released the version 2.2.0 of the framework. The API is kept unchanged, the main difference is in the project organisation (in Project Builder). I also have tried to make the aim of each target more clear, and cleaned up the build-styles, in that respect I have defined 3 different names for each of the flavours (corresponding to 3 different targets) of the framework (which BTW have also changed):</p> <ol class="intext"> Index: index.html =================================================================== RCS file: /cvsroot/mysql-cocoa/Documentation/website/index.html,v retrieving revision 1.19 retrieving revision 1.20 diff -C2 -d -r1.19 -r1.20 *** index.html 4 Mar 2004 22:21:14 -0000 1.19 --- index.html 9 Aug 2004 16:07:49 -0000 1.20 *************** *** 29,32 **** --- 29,41 ---- <a name="archive"><p>Archives of the updates can be found <a href="archives.html">here</a></p></a> + <p><em>09 August 2004:</em> Corrected a small bug in the new method -quoteObject:, it was not taking care of nil or NSNull object, now it will return the string NULL (without surrounding quotes) for these cases. This is version 2.3.1. Normally the API documentation should be updated to the latest version of the framework as of today.</p> + + <p class="withlist"><em>05 August 2004:</em> Aaron Jacobs pointed me to the solution of the signed/unsigned integer (which indeed was already in a mail from Lorenz Textor). This also gave me the way to discriminate between TEXT and BLOB columns. After implementing the bug fixes, the version 2.3.0 of the framework is now released, the new features/bug fixes are:</p> + <ul class="intext"> + <li class="intext"><em>Integers</em> : All integers should now be correctly be converted in NSNumber. Indeed I had to implement a NSNumber subclass to handle properly the unsigned numbers as the Foundation NSNumber lass does not retain if the stored number is signed or not (if one create a NSNumber with a large unsigned int, and outputs it later, the output is negative). Now even large UNIGNED BIGINT should be output correctly positive.</li> + <li class="intext"><em>TEXT/BLOB</em> : Now the framework is able to discriminate between these two types (and related SMALL... and LONG...). TEXT will be converted to NSString (using the proper encoding, if the encoding of the MySQL server is supported by Cocoa), whilst BLOB will be converted to NSData.</li> + <li class="intext"><em>quoteObject</em> : This is a new method added to MCPConnection to quote any object, using the object class to know how it should be quoted.</li> + </l> + <p><em>04 March 2004:</em> With the comments of Dave Winter, I have corrected the examlpe in the HTML doc, should be Ok now (except maybe a \312 character used for tabbing...). Documentation should be updated to 2.2.3 by now.</p> *************** *** 39,46 **** <p><em>27 August 2003:</em> Added a documentation on how to build MySQL from sources. You will find a link to these inscription in the <a href="#docs">Documentation</a> section. Following these instructions you should be able to build MySQL server and client sides, with prebound dynamic libraries (as well as static libraries), openssl support for both server and client... and finally getting the <code>libmysqld.a</code> library which is basically a embedded MySQL server running in its own thread, so your program can contains its own server (not yet supported by SMySQL framework... but might come some time soon).</p> - <p><em>25 August 2003:</em> Today I finally had the time to update the web site. The information you find here should now be up to date (especially the <a href="#files">Files</a>, <a href="#docs">Documentation</a> were completely outdated – I had no time to update them since I first wrote them in January 2002 –).</p> - - <p><em>20 August 2003:</em> Releases the version 2.2.1 of the framework: minor bug correction. The previous versions of the framework were not able to handle properly the integer with more than 32bits (MySQL BIGINT). This is now corrected (indeed the other branches of the framework are also corrected in CVS, so if you are still using the version 1, I encourage you to get the fix from CVS). Unfortunately I have not found a solution to know from the MySQL server if an retrieve integer is signed or not, hence the framework assumes always signed int: this can cause a trouble if you have a unsigned column (either for BIGINT or INT) and you are using number bigger than the maximum SIGNED number. Thanks to Lorenz Textor for pointing the bug to me.</p> - <!--/Updates--> --- 48,51 ---- |