Adding features indiscriminately has a very real price, even for 'free' software. Every feature that we add - particularly those that are compiled into the web server (like Zend Optimizer) - has the potential to cause administrative problems, particularly with regards to stability. This means that it needs to be tested, which takes time - time that is taken away from other things.
Furthermore, even if a given piece of software is stable and doesn't cause problems, it's still another variable that we have to keep in mind and check should problems develop.
Also, we can't just add stuff to our servers as it is requested. If we add a given feature, customers expect it to be supported to some degree. In other words, if it does cause problems there will be people who are unhappy if we just say, "oops!" and get rid of it. As such, while we can make some changes (ie. the installation of Perl modules) pretty routinely, for any major changes that could impact a lot of people we have to be a bit more cautious and consider whether it's worth it.
Really, maintaining several hundred servers that tens of thousands of customers depend on is much different than maintaining software installed on a local PC. Changes of this magnitude are not something that is done lightly, as any repercussions are magnified across tens of thousands of web sites.
Finally, there is another factor - we are a business. It takes time to do the things that we do, and that equates to money spent paying administrators to do all of the above. That means that we have finite resources to do all of the things that customers want us to do. Spending that time working on features that a very small number of customers want (ie. Zend Optimizer) takes time away from our implementing things that a large number of customers want (ie. the afore-mentioned PHP 5/MySQL 4.1 upgrades) - which translates into lost sign-ups and unhappy customers. That's not something you can afford to do as a business if you want to stay competitive.
So, we manage this by doing our best to gauge what our customers want and deliver to the extent that we can. We will occasionally take a chance and do something nobody else is doing or that is somewhat unproven, but only when we think the technology is cool and has potential (Jabber comes to mind here). Supporting a 'feature' that makes it so that software developers can charge for code intentionally obfuscated so that their customers can't touch it doesn't fall into that category, so really comes down to customer demand.
Probably. I believe it is, actually. I'm pretty sure Zend Optimizer is free too. Price has never really been the limiting factor for our lack of support, though.
I will be honest with you - if your web site's needs are not being met by DreamHost due to a lack of support for a given piece of software and you don't want to wait for an upgrade that may never come, you should consider another host (at least for that specific site). We're here to serve you, and if we are unable to do so we completely understand if you look elsewhere. However, we do have to support a lot of other people as well, and when each of them has different needs and priorities it can be a major juggling act. Rather than be everything to everyone we sometimes have to make decisions as to where we will allocate our resources/risk.
- Jeff @ DreamHost
- DH Discussion Forum Admin