[sword-devel] 1.7.2 release

[Костя Маслюк]

2014-03-17 18:04 GMT+04:00 Troy A. Griffitts <scribe �� crosswire.org>:
> Dear Костя,
>
> Is there a way to include a link to the diff in the bug instead of simply to
> you entire repository. I'd lime to review and comment on your fix (as I'm
> using my phone to check my mail) and have spent 5 minutes or so trying to
> get the website to just switch to the branch you mention and it doesn't
> respond, at least on Chrome for Android. I've seen other people post links
> to diffs, in the past. Is their a way to create a link which shows the work
> done in your branch?

http://pastebin.com/EXC9WRBx
(using gitorious i didn't found a way to get diff for several commits)

I m not familiar with module making pipeline so i anyway need
someone's guidelines to make correct changes.

>
> Regarding the bug, there are 2 separate issues in my mind.
>
> First, the use of distinct sorting algorithms between module creation and
> module consumption could cause this problem. I don't know if this is the
> problem, but I've always worried about it. We have 3 build scenarios for
> SWORD which worry me for this bug, i11n supported by: ICU, Qt, neither. If
> the module build tools configuration differs from the client configuration,
> I am worried our binary sort will loop, though it's only a suspicion we use
> i11n code in the sort algo (toupper).

No, sorting works fine for me, this issue is about strongs padding.

I assume i m is not clear enough on tracker.


1. Looping occurs when system lookups for entry that look like Strong
number, but it is not in format of Strong number e.g "0001" should be
"00001" if we talking of Strong's lexicon. Because in module, entry
"0001" will be placed after (or before no matter) "00010" because of
sorting algorithm and incrementing from "0001" will goto "00002" each
time. Is it necessary to explain it more detailed?

2. There are some modules that should have StrongsPadding=false in
theirs conf file but they haven't (like ZhEnglish). By default
StrongsPadding=true.

3. There are some modules that are Strong lexicons but have corrupted
entries like "0001" and we have no check procedure during module
creation.

>
> The second issue is user-displayed order of keys. The binary sort is an
> index for fast lookup. We've talked about keeping another index which
> preserves the 'book order' of the module. The LD driver is meant only for
> fast lookups, not for showing a module sequentially.
>
> Troy
>
> On March 17, 2014 5:23:18 AM MST, "Костя Маслюк" <kostyamaslyuk �� gmail.com>
> wrote:
>>
>> Please, consider this fix:
>>
>> http://www.crosswire.org/tracker/browse/MOD-61 and
>> http://www.crosswire.org/tracker/browse/API-150 is similar.
>>
>>
>> Actually i got weird things as things are not threaded in time. Blessings.
>>
>> 2014-01-14 14:02 GMT+04:00 Chris Little <chrislit �� crosswire.org>:
>>>
>>>  On 1/14/2014 12:51 AM, Костя Маслюк wrote:
>>>>
>>>>
>>>>  What about the "inconsequent lexicon index" fix i have posted here a
>>>>  month ago? I realize that it would not be threated as Critical, but it
>>>>  is still bad behaivior that affects every Sword frontend.
>>>>
>>>>  I also realize that fix would not be complete, but no one even asked me
>>>>  for deeper investigation, if people have no time.
>>>
>>>
>>>
>>>  The patch you submitted just covers up the underlying problem without
>>>  solving anything (or finding the root of the problem), so I would say it
>>> is
>>>  not a candidate for inclusion in Sword.
>>>
>>>  Some brief discussion of the bug is here:
>>>  http://www.crosswire.org/tracker/browse/MOD-61
>>>
>>>  If you can find the cause of these modules being built incorrectly (or
>>> being
>>>  interpreted incorrectly), I'd love to see a fix be committed.
>>>
>>>  --Chris
>>>
>>>
>>>
>>> ________________________________
>>>
>>>  sword-devel mailing list: sword-devel �� crosswire.org
>>>  http://www.crosswire.org/mailman/listinfo/sword-devel
>>>  Instructions to unsubscribe/change your settings at above page
>>
>>
>> ________________________________
>>
>> sword-devel mailing list: sword-devel �� crosswire.org
>> http://www.crosswire.org/mailman/listinfo/sword-devel
>> Instructions to unsubscribe/change your settings at above page
>
>
> --
> Sent from my Android device with K-9 Mail. Please excuse my brevity.
>
> _______________________________________________
> sword-devel mailing list: sword-devel �� crosswire.org
> http://www.crosswire.org/mailman/listinfo/sword-devel
> Instructions to unsubscribe/change your settings at above page