[Laszlo-dev] [JIRA] Resolved: (LPP-5816) simple test case for line numbers still off in script output

[Donald Anderson]

You're right, I misread the logic in Compiler.java.

On Apr 17, 2008, at 11:44 PM, P T Withington wrote:

> I thought nameFunctions is only on by default for debug and profile  
> builds?  I don't think it is on for a normal build.  So, I would  
> stick with nameFunctions implies trackLines.
>
> On 2008-04-17, at 18:15 EDT, Donald Anderson wrote:
>> Sanity check - it looks like nameFunctions is turned on by default.
>> Wouldn't tying them together increase the size of our DHTML runtime?
>>
>> Counterproposal - get rid of trackLines and have two options  
>> internally:
>>   trackFunctionLines
>>   trackSourceLines.
>>
>> By default only trackFunctionLines is on (and has the effect of  
>> today's nameFunction).
>> The 'nameFunction' option (no longer by default on), turns on both  
>> trackFunctionLines
>> trackSourceLines.  So by default you get line traces on functions  
>> only, but if you use
>> --option nameFunctions you get the whole shebang.
>>
>>
>> On Apr 17, 2008, at 4:44 PM, P T Withington wrote:
>>
>>> All good ideas.  For now, I think nameFunctions should imply  
>>> trackLines.
>>>
>>> On 2008-04-17, at 16:34 EDT, Donald Anderson wrote:
>>>>
>>>> On Apr 17, 2008, at 3:19 PM, P T Withington (JIRA) wrote:
>>>>
>>>>> Resolution: Fixed
>>>>>
>>>>> You're right, adding -DtrackLines results in:
>>>>>
>>>>> $lzc$class_hei1 = Class.make("$lzc$class_hei1", LzView, ["ae",
>>>>> /* -*- file: nexb.lzx#5.2 -*- */
>>>>> function ae () {
>>>>> with (this) {
>>>>> /* -*- file: #1 -*- */
>>>>> window.alert("hei")
>>>>> }}
>>>>>
>>>>> I wonder if some of the confusion is that this feature is  
>>>>> controlled by a 'compile-time constant' (-DtrackLines) and the  
>>>>> other is controlled by a compiler option (---option  
>>>>> nameFunctions).  I think they should both be options, since  
>>>>> trackLines should not affect code execution.
>>>>
>>>>
>>>> Maybe time to rethink how this is controlled.  Maybe --option  
>>>> nameFunctions should always enable
>>>> trackLines?  Is it helpful to break them apart?  If so, allow -- 
>>>> option trackLines ?
>>>> It seems haphazard - maybe that whole list in Compiler.java  
>>>> should be organized so
>>>> --option can be used with each one?  I could hook on a usage  
>>>> string for each as well.
>>>>
>>>>
>>>> --
>>>>
>>>> Don Anderson
>>>> Java/C/C++, Berk
>>>
>>
>>
>> --
>>
>> Don Anderson
>> Java/C/C++, Berkeley DB, systems consultant
>>
>> voice: 617-547-7881
>> email: dda at ddanderson.com
>> www: http://www.ddanderson.com
>>
>>
>>
>>
>


--

Don Anderson
Java/C/C++, Berkeley DB, systems consultant

voice: 617-547-7881
email: dda at ddanderson.com
www: http://www.ddanderson.com