MySQL Forums
Forum List  »  Connector/ODBC

Re: Last stable/fast version?
Posted by: Bill Williams
Date: April 11, 2005 09:44AM

3.51.06 and 3.51.09 seem to work well (fast and stable), but I'd caution you that these connection problems/delays may be partially the fault of the 4.1.x Server series. Both 3.51.10 and 3.51.11 have had serious problems for some people, although there's a developement patch (3.51.11-2) that appears to fix some of the serious issues. You can find that at Peter Harvey's site under the downloads section:
http://www.peterharvey.org/

At this point, we're using 4.0.23 (linux) connecting with 3.51.06 via windows. This is extremely quick and stable, with none of the "date retuned as text" or "catastrophic connection failure" nonsense you'll see on these forums. We won't be moving either until we see some stability back in the driver.

It's beginning to look as if MyODBC is being treated as an afterthought, considering the major issues that people are having with the driver. This is surprising and saddening, considering that a solid ODBC driver is vital to tens of thousands of exisitng applications. The current bugs afflicting the driver have a HUGE impact on these applications, forcing many to remain with old drivers and the 4.0.x server tree. Moreover, these problems have been in existence for months, since 3.51.10 was released. Granted, 3.51.11 eventually was patched, but considering the major issue of DATE values being returned as TEXT, the patch process was inexcusably long. Finally, it appears as if 4.1.x server has at least some blame for this, as the .NET driver is experiencing similar speed issues.

Options: ReplyQuote


Subject
Written By
Posted
April 05, 2005 08:09AM
April 11, 2005 08:51AM
Re: Last stable/fast version?
April 11, 2005 09:44AM
April 13, 2005 09:46AM
April 13, 2005 09:36AM


Sorry, you can't reply to this topic. It has been closed.

Content reproduced on this site is the property of the respective copyright holders. It is not reviewed in advance by Oracle and does not necessarily represent the opinion of Oracle or any other party.