Re: [phpGroupWare-developers] status of trunk in svn

[Maât]

Sigurd Nes a écrit :
> Dave Hall wrote:
>   
>> On Wed, 2008-04-23 at 20:51 +0200, Sigurd Nes wrote:
>>     
>>> Dave Hall wrote:
>>>       
>>>> On Wed, 2008-04-23 at 08:08 +0200, Dave Hall wrote:
>>>>         
>>>>> On Wed, 2008-04-23 at 09:58 +0200, Sigurd Nes wrote:
>>>>>           
>>>>>> Dave Hall wrote:
>>>>>>             
>>>>>>> On Mon, 2008-04-21 at 17:55 +0200, Sigurd Nes wrote:
>>>>>>>               
>>>>>>>> Dave Hall wrote:
>>>>>>>>                 
>>>>>>>>> On Sat, 2008-04-19 at 11:09 +0200, Sigurd Nes wrote:
>>>>>>>>>                   
>>>>>>>>>> Dave Hall wrote:
>>>>>>>>>>                     
>>>>>>>>>>> On Sat, 2008-04-19 at 00:26 +0200, Sigurd Nes wrote:
>>>>>>>>>>>                       
>>>>>>>>>>>> Dave Hall wrote:
>>>>>>>>>>>>                         
>>>>>>>>>>>>> On Thu, 2008-04-17 at 18:53 +0200, Olivier Berger wrote:
>>>>>>>>>>>>>                           
>>>>>>>>>>>>>> Le jeudi 17 avril 2008 à 11:03 +0200, christian bac a écrit :
>>>>>>>>>>>>>>                             
>>>>>>>>>>>>>>> I can propose diffs from my version and the trunk, but last 
>>>>>>>>>>>>>>> time I did
>>>>>>>>>>>>>>> it, skwashd talled that there were new versions in the 
>>>>>>>>>>>>>>> <<resight tree>>,
>>>>>>>>>>>>>>> that already had the same fixes.
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> When will the resight tree changes be merged to trunk.
>>>>>>>>>>>>>>>                               
>>>>>>>>>>>>> This is a question for Sigurd to answer.
>>>>>>>>>>>>>
>>>>>>>>>>>>>                           
>>>>>>>>>>>> I'm glad to see that the activity is picking up, there has been 
>>>>>>>>>>>> very little of
>>>>>>>>>>>> it on savanna (except for ged) lately - and the sync has mostly 
>>>>>>>>>>>> moved code one
>>>>>>>>>>>> way - and the code needs manual tweaking before sync:
>>>>>>>>>>>>
>>>>>>>>>>>> 1) The resight project really need a method for quickly switch from
>>>>>>>>>>>> developer-mode to end-user mode - suspending the notices to be 
>>>>>>>>>>>> printed due to
>>>>>>>>>>>> frequently testing and demonstration of the system for 
>>>>>>>>>>>> non-developers.
>>>>>>>>>>>>                         
>>>>>>>>>> When: allow 1)
>>>>>>>>>>                     
>>>>>>>>> Someone just needs to write the code which was proposed last time this
>>>>>>>>> was discussed.  I am sure the community would be happy to accept the
>>>>>>>>> code when provided.
>>>>>>>>>
>>>>>>>>>                   
>>>>>>>> Why not just allow the extra line of code in functions.inc.php:
>>>>>>>>                                
>>>>>>>> if(isset($GLOBALS['phpgw_info']['server']['log_levels']['global_level'])
>>>>>>>>  &&
>>>>>>>> $GLOBALS['phpgw_info']['server']['log_levels']['global_level'] == 'N')
>>>>>>>> {
>>>>>>>>  echo "\n<br>" . lang('ERROR Notice: %1 in %2 at line %3', $error_msg,
>>>>>>>> $error_file, $error_line) . "<br>\n";
>>>>>>>> }
>>>>>>>>                 
>>>>>>> This has been discussed in the past.  The current code means developers
>>>>>>> have to fix their code.  It is in the trunk - not a stable branch.  A
>>>>>>> sustainable fix was proposed last time this was discussed, I am sure the
>>>>>>> community would be happy to accept that code any time.
>>>>>>>
>>>>>>>               
>>>>>> Has anyone looked at it?
>>>>>> The point is that it has to be simple to switch to and from debugging 
>>>>>> mode.
>>>>>>             
>>>>> It was explored in depth back in January - here is the conclusion
>>>>>
>>>>> http://thread.gmane.org/gmane.comp.web.phpgroupware.devel/6395/focus=6398
>>>>>           
>>>> Sorry the correct link is
>>>> http://thread.gmane.org/gmane.comp.web.phpgroupware.devel/6395/focus=6400
>>>>
>>>>         
>>> So display_errors set to off will suppress the notices from being displayed?
>>> That's really in effect what I have been asking for.
>>>       
>> That is part of what was proposed back in January - keep in mind that it
>> is a complete solution to be implemented.
>>
>>     
> Could we live with the check on ['global_level'] == 'N' in the meanwhile?
> ('N' is the default value)
>
> Regards
>
> Sigurd
>   
honestly i can live with it and without it... i can even change it 
myself if something goes wrong.

what i cannot do myself is merging the missing code into trunk so that
we can go on with phpgroupware

sorry to insist but compared to the whole project being locked down this
issue of having a var being set to whatever value you want seems incredible.

Let's merge this code with all the rest that's missing into trunk... if
something need to be adapted or optimized that can be adressed later ?

We are in trunk... so if that little part of the code brings drawbacks
we can have the very same debate in a few days with the code inside
trunk and the trunk ready for .18 testing and stabilization

my 2cts

Maât