Hi Troy,
Using verse compression doesn't seem to change things in any way.
Thanks,
Brian
On 12/04/2009 10:34 PM, Troy A. Griffitts wrote:
> Brian, if individual verses can grow this large, could you try changing
> the compression to be verse level and see if this fixes the problem. I
> didn't
Sorry, I guess I *thought* that field was being used for the key
text, rather than the value. It should be sufficient.
Matthew
___
sword-devel mailing list: sword-devel@crosswire.org
http://www.crosswire.org/mailman/listinfo/sword-devel
Instructions to
On Fri, Dec 4, 2009 at 10:54 PM, Troy A. Griffitts wrote:
> I believe 2k should be quite enough for the key value (e.g., "Mat.1.1").
> Do you still believe we have a case that warrants upping this limit?
Sorry, I guess I that field was being used for the key text, rather
than the value. It should
I believe 2k should be quite enough for the key value (e.g., "Mat.1.1").
Do you still believe we have a case that warrants upping this limit?
Troy
On Thu, 2009-12-03 at 19:31 -0500, Matthew Talbert wrote:
> Sorry to keep sending things...
>
> Doesn't swmodule.cpp:501 need to be this:
>
Brian, if individual verses can grow this large, could you try changing
the compression to be verse level and see if this fixes the problem. I
didn't write the z drivers, so I'm unaware where the limits are, but
this might give us a clue if you get different results. Then we can
take our clues to
On Thu, Dec 3, 2009 at 6:33 PM, Jonathan Marsden wrote:
> Troy A. Griffitts wrote:
>
>> All warnings which cause -Werror to break (for the systems I've tried)
>> are removed and some reported by others which I do not see are also
>> fixed. There are still some warning patches I have not applied,
Sorry to keep sending things...
Doesn't swmodule.cpp:501 need to be this:
const unsigned short int MAX_CONV_SIZE = 1024 * 1024;
as well?
Matthew
___
sword-devel mailing list: sword-devel@crosswire.org
http://www.crosswire.org/mailman/l
Also, I noticed that you changed the configuration to always use
HAVE_ICUSWORD if ICU is present. While this is desirable on Windows,
it will mean scrapping all the custom transliterators in sword/icu.
Basically if _ICUSWORD_ is defined, it is assumed all transliterators
will be loaded directly fro
Hey Troy,
Could you explain what you did with need_local_regex and mingw? The
test never actually worked, which was a real pain, but I don't want to
build with sword's regex either, as it appears I'll be forced to do
now (assuming the test for mingw actually works now). I don't really
understand a
Troy A. Griffitts wrote:
All warnings which cause -Werror to break (for the systems I've tried)
are removed and some reported by others which I do not see are also
fixed. There are still some warning patches I have not applied, as I do
not see any problems on my end with a new F12 g++ 4.4.2 too
In preparation for a SWORD 1.6.1 release, I've just committed what I
hope are all the outstanding patches which have been submitted.
I've updated the autoconf configuration to actually work when different
features are turned off or missing.
FTPLib support should be brought up to the same level as
11 matches
Mail list logo
