[Opendnssec-develop] RE: Postponement of 2.0.0a4 release

Jerry Lundström jerry at opendnssec.org
Tue Apr 29 19:52:31 UTC 2014

On 29 apr 2014, at 20:05, Yuri Schaeffer <yuri at nlnetlabs.nl> wrote:

Hash: SHA1

It is unclear why we didn't notice it before, but once observed it
took me quite a while to reproduce reliably. Probably because there
isn't even a usecase where one would have to call 'enforce'. Still it
indicates a problem with the current implementation.

This isnt really that strange.

We currently just test "the happy path", meaning that we have a path of
execution that we run and expect a specific result in each step which is
more or less success.

For example:
1. Run list
2. Grep in list
3. Add to list
4. Grep in list for new entry

What happened when you tested is that you started another path within a
path and it failed badly :(

Now with the error that emerged I saw no point to dig deeper into it. It is
actually really easy to abuse transactions. I can explain why in more
detail if anyone wants to know over TS or something similar.

Then we will implement MySQL, and pick up where we left converting tests.

MySQL is not a biggy, its so similar to SQLite that you can mostly say that
if it works in one it will work to 98% on the other. And the implementation
time is very short.

The current state of affairs can be found at

We will merge the PR when we have it on par of the current code to not
break functionality.

Jerry Lundström - OpenDNSSEC Developer
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opendnssec.org/pipermail/opendnssec-develop/attachments/20140429/eab3f076/attachment.htm>

More information about the Opendnssec-develop mailing list