[bt-devel] Next stable release

Gary Holmlund gary.holmlund at gmail.com
Sun Apr 15 16:53:50 MST 2012


On 4/15/2012 3:47 PM, Jaak Ristioja wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> On 16.04.2012 01:04, Gary Holmlund wrote:
>> On 4/15/2012 12:34 PM, Benjamin Misja wrote:
>>> Hi,
>>>
>>>
>>>
>>> Great to hear that you guys are preparing a new minor release.
>>>
>>>
>>>
>>> I am not a coder, so I lack the ability to check the code, but I
>>>   couldn't see anything on the tracker or in the commit comments
>>> about the recurring problem that BT doesn't accept unlock codes
>>> for locked modules. Here's the most recent report (I think):
>>>
>>> https://sourceforge.net/tracker/?func=detail&aid=3402987&group_id=954&atid=100954
>>>
>>>
>>>
>>>
>>>
>>> Blessings,
>>>
>>> Ben
>>>
>> I checked each of my BT installations to make sure the unlock was
>> still working. It worked for me everywhere but on my Windows
>> build. I debugged it and it looks like a sword bug. I got down to
>> BibleTime CSwordModuleInfo::unlockKeyIsValid. It does they
>> following: m_module->setPosition(sword::TOP);
>> m_module->getRawEntryBuf(); The getRawEntryBuf returns with an
>> empty string which causes the unlockKeyIsValid function to return
>> false.
>>
>> I tried to track down the issue in sword. The getRawEntryBuf
>> computes sizes, offfset, and check cache and eventually returns a
>> "". I did not know enough about how the sizes, offset, etc. work
>> to determine the exact cause.
>>
>> But, my point is that I don't believe this is a BibleTime problem
>> and I don't think we should hold 2.9.2 because of it.
> On the other hand, this bug is getting way too old and it is
> recurring. We should deal with it.
>
> Perhaps there exists some kind of workaround? What about something like:
>
>      m_module->setPosition(sword::TOP);
>      QString test;
>      while (m_module->error() == 0) {
>          test = /* bla bla */;
>          if (!test.isEmpty())
>              break;
>          m_module->increment();
>      };
>      // test "test" here...
>
> Btw, thanks Gary for debugging it this far! :)
>
>
> God bless!
> Jaak

Jaak,

Unfortunately it goes through the entire loop returning an empty 
string.    :(

Gary






More information about the bt-devel mailing list