Mr. Sundman will look into it.

Mr. Withington keeps adding more and more funcitonallity to the 
debugger and sometimes Mr. Sundman has trouble keeping up.

I do plan to revisit the debugger stuff this month.

jrs

On May 15, 2006, at 2:11 PM, P T Withington wrote:

> [Re-directing to laszlo-dev, as I think this should be of general 
> interest to the community.]
>
> Probably I'm just not a very good documenter.  The stuff I wrote after 
> E.g., was supposed to be an example.
>
> Try typing: `Debug.warn('This is a hotlink: %w', {a: 1, b: 2})` into 
> the eval pane of the Debugger and I think it will be obvious?
>
> There is stuff written up at:
>
> http://www.laszlosystems.com/lps-3.2/docs/reference/debug.html
>
> see Debug.format and Debug.formatToString.  And 
> Debug.{debug,info,warn,error} all have the same capabilities.
>
> Any Object that is displayed using %w (in legal's %s too), will 
> automatically be hot-linked.  For backwards compatibility, if you just 
> give any of the Debug.format-like methods a bunch of arguments instead 
> of using the format string, it will behave like old Debug.write and 
> just output the arguments separated by spaces, but again, any Objects 
> will automatically be hot linked (hm, that might be only in legal's 
> too).
>
> Here's an example from the LFC:
>
>    Debug.error("%w.makeChild(%w, %w) when %w.__LZdeleted", this, e, 
> async, p);
>
> All the objects referred to with %w will be hot linked.
>
> Maybe Mr. Sundman can offer some advice on how to make the 
> documentation more user friendly?
>
> On 2006-05-15, at 13:09 EDT, Sarah Allen wrote:
>
>> maybe I'm not very smart, but I looked at the LZX reference (and the 
>> Dguide which seems to just include reference text with different 
>> formatting) and I couldn't quite guess what formatt options would 
>> give me Debugger output with hotlinks... any chance for an example or 
>> two?
>>
>>
>> On Thu, May 4, 2006 at  3:45 PM, P T Withington wrote:
>>
>>> FYI, rather than concating the string yourself, use the printf 
>>> capability of the debugger, then the objects that you 'present' 
>>> become clickable/inspectable.  Also, should this be a warning 
>>> instead of just a write?  (Note that Debug.warn will capture a 
>>> backtrace, if available, whereas Debut.write will not.  If you have 
>>> any ideas on how I can encourage people to use the printf style in 
>>> messages (like when we are rewriting the LFC) I would appreciate it. 
>>>  I think it makes debug messages much more useful.)
>>>
>>> E.g.:
>>>
>>> lzx> Debug.warn('foo: %s', LzBrowser.getLoadURLAsLzURL())
>>> WARNING: foo: http://tin-woodsman.local:8080/lps-legals/test/smoke/ 
>>> simple.lzx?lzt=swf&debug=true
>>> lzx> Debug.inspect(«¡LzURL!#71| http://tin-woodsman.local:80...»)
>>> «¡LzURL!#71| http://tin-woodsman.local:8080/lps-legals/test/smoke/ 
>>> simple.lzx?lzt=swf&debug=true» {
>>> file: simple.lzx
>>> host: tin-woodsman.local
>>> path: /lps-legals/test/smoke/
>>> port: 8080
>>> protocol: http
>>> query: lzt=swf&debug=true
>>> }«¡LzURL!#71| http://tin-woodsman.local:8080/lps-legals/test/smoke/ 
>>> simple.lzx?lzt=swf&debug=true»
>>> lzx> On 2006-05-04, at 18:29 EDT, jgrandy wrote:
>>>
>>> [Note to myself:  Fix %s formatting so that %#s means 'print 
>>> readably' which means quote strings.]
>>>
>>>> Change 42031 by [EMAIL PROTECTED] on 2006/05/04 15:26:14
>>>>
>>>>    Summary:Integrate patch from Oliver fixing quoting in error message
>>>>    
>>>>    Bugs Fixed:LPP-1597
>>>>    
>>>>    Technical Reviewer:  (pending)
>>>>    QA Reviewer:  (pending)
>>>>    Doc Reviewer:  (pending)
>>>>
>>>> Affected files ...
>>>>
>>>> ... //depot/lps-dev/WEB-INF/lps/lfc/services/LzLoadQueue.as#35 edit
>>>>
>>>
>>
>


_______________________________________________
Laszlo-dev mailing list
[email protected]
http://www.openlaszlo.org/mailman/listinfo/laszlo-dev

Reply via email to