Another scary bug??

[Monte Goulding]

Thanks Bob

Lots of parts of the engine that deal with strings have two code paths. One for native encoding and one for unicode. This is an attempt to maintain as close to pre LC 7 performance as possible. I suspect based on your description that the bug only impacts the native encoding side of things. I can’t recall anyone working on anything like this for 8.1 but I could be wrong there.

You never know this could be bug of the week on `This week in LiveCode` ;-)

Cheers

Monte

> On 8 Sep 2016, at 8:09 AM, Bob Sneidar <bobsneidar at iotecdigital.com> wrote:
> 
> And now, even scarier, I copied the values out of the field I was testign with, and pasted the value into the code which uses hard coded values and THAT WORKS NOW when a second ago before I pasted the value, it did NOT work!!!
> 
> This is definitely an issue where the formatting of the text is coming into play. 
> 
> I submitted a bug report 18343. I attached a sample stack. I'd be interested if this is happening under 8.1 RC2. They may have fixed this and I don't know it yet. I'd also be interested if this is happening with Windows 8.0.1, but I have a real job here and have spent a LOT of time chasing this one. 
> 
> Bob S
> 
> 
>> On Sep 7, 2016, at 14:45 , Bob Sneidar <bobsneidar at iotecdigital.com> wrote:
>> 
>> Here's a nice little twist! This works if the arguements for itemOffset come from fields. it only breaks if you hard code them (or in my case pull the values from a database. I am willing to bet a nice steak dinner this is being caused by uniCode. The engine is probably expecting to search for unicode values and instead is finding UTF8 (or whatever it is). 
>> 
>> Bob S
>> 
>> 
>>> On Sep 7, 2016, at 13:59 , Bob Sneidar <bobsneidar at iotecdigital.com> wrote:
>>> 
>>> Thanks Panagiotis. I will submit a bug report to that effect. Unfortunately I need wholeMatches because contacttypes can be IT or SITES among other things and replacing IT in contacttypes will result in SE instead of SITE. I am working around it but this can bite someone in the arse hard if they are not carefule. 
>>> 
>>> Bob S
>>> 
>>> 
>>>> On Sep 7, 2016, at 11:10 , panagiotis merakos <merakosp at gmail.com> wrote:
>>>> 
>>>> It looks like "set the wholeMatches to true" is the culprit. Removing this
>>>> line results in the expected behavior.
>>>> 
>>>> On Wed, Sep 7, 2016 at 8:36 PM, Richard Gaskin <ambassador at fourthworld.com>
>>>> wrote:
>>>> 
>>> 
>>> 
>>> _______________________________________________
>>> use-livecode mailing list
>>> use-livecode at lists.runrev.com
>>> Please visit this url to subscribe, unsubscribe and manage your subscription preferences:
>>> http://lists.runrev.com/mailman/listinfo/use-livecode
>> 
>> 
>> _______________________________________________
>> use-livecode mailing list
>> use-livecode at lists.runrev.com
>> Please visit this url to subscribe, unsubscribe and manage your subscription preferences:
>> http://lists.runrev.com/mailman/listinfo/use-livecode
> 
> 
> _______________________________________________
> use-livecode mailing list
> use-livecode at lists.runrev.com
> Please visit this url to subscribe, unsubscribe and manage your subscription preferences:
> http://lists.runrev.com/mailman/listinfo/use-livecode