One of the most challenging parts of development and support of SugarCRM is the fact that we are support matrix is quite a wide net ( this is something I’ve talked about at conferences a few times now ). As a product, it’s something that sets us apart in the marketplace, making us an attractive application and platform for thousands of organizations all around the world looking to take control of their CRM and business processes, so we don’t want to take that freedom away. However as time moves on so does software, and we want to make sure we can continue to deliver the most modern, stable, customizable, and flexible solution possible. With that said, in order to keep our support matrix sane we do need to look at the support of older upstream components and see if it makes practical sense to continue support of them. Here’s a some of the ones that have reached or are quickly approaching “legacy” status, and in turn we are looking to discontinue support for:
- PHP 5.2 since it has been officially discontinued by PHP.net
- Windows Server 2003 and Windows XP as server platforms which are 8 and 10 years old right now respectively and currently outside of Mainstream Support from Microsoft for well over a year now. There is also not a native URL Rewrite module for IIS 5.1 ( part of Windows XP ) or IIS 6 ( part of Windows Server 2003 ) which creates potential security problems for any systems hosted on this platform.
- MySQL 5.0 since it has been in Extended Support since late 2009 and be fully retired in 2012.
- Internet Explorer 7 since it has been superceded by Internet Explorer 8 on all supported platforms.
- Safari 4 since it has been superceded by Safari 5 on all current Apple OSes ( OS X 10.5, 10.6, 10.7 )
- FreeTDS and php_mssql extensions for connecting to SQL Server, both with are not compatible with PHP 5.3
- php_mysql extension, since it has been long been supplanted by the current php_mysqli extension and has several known security holes.
- Office 2003 since mainstream support ended 14 Apr 2009
- Lotus Notes 7.0 since support ended 30 Apr 2011
By doing this it will help us to concentrate support on the most popular upstream component versions out there currently, be able to support newer versions of these components sooner, and develop features for the product which can take advantage of the advanced capabilities not present in these older versions to ensure greater performance and stability of the product. What does everyone think about this? Are we moving too fast with these discontinuations, or does it seem right on to what we need to do? Let us know in the comments or on Twitter your thoughts.