<div class="gmail_quote">On Wed, Jul 4, 2012 at 4:48 PM, David Moore <span dir="ltr"><<a href="mailto:dmoo1790@ihug.co.nz" target="_blank">dmoo1790@ihug.co.nz</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div class="im">On 04/07/12 13:30, Steve Hodge wrote:<br>
</div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="im">
From a design point of view it would be better if the client programs<br>
weren't connecting to the database directly - they should be getting all<br>
data via the backend process. Really they probably shouldn't even be<br>
aware that there is a database. There was some talk about going that<br>
way, not sure if anything has come of it (I don't really follow the main<br>
lists these days).<br>
<br>
I'm sure there will still be tools to access the database, it just won't<br>
be as convenient as it is now. I won't be happy if it happens either,<br>
but I do understand the reasoning.<br>
<br>
</div></blockquote>
<br>
I still don't get it. I would understand if there were good technical reasons for embedded vs external db. For example, faster or smaller or more reliable.</blockquote><div><br>Install complexity. Remote permissions for MySQL is one of the perennial issues that people have, it's come up on this list more than once. And note that the recent leap-second bug was triggered by MySQL, not MythTV itself. Versioning issues. As a Gentoo I was affected by the change in default character set that happened between versions awhile back and that was a big issue for some. Most users want mythtv to be an appliance. They don't want to have administer a database.<br>
<br>From memory, speed was actually one of the big reasons why they have stuck with MySQL, and if they do go embedded it'll probably be an embedded MySQL. Some of the scheduler queries are complex and slow. Also at the moment with database access spread through all the client apps and the system is closely tied to MySQL. There was some work done on a PostgreSQL port but the devs have flatly stated that the won't commit anything like that.<br>
</div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> But the argument that users might break it doesn't fly for me. So what if they do? It's their problem. If the solution means less functionality for me then I'm dead set against it.</blockquote>
<div><br>If you want your open source project to be successful you need to build a community around it. Telling users "it's your problem" is a good way to fail at that. And users having bad experiences put other users off. That said I think the motivation is not so much to protect users from themselves as to protect users from out of date third party tools.<br>
<br>Cheers,<br>Steve<br></div></div><br>