kicad-developers team mailing list archive

[Dick Hollenbeck <dick@xxxxxxxxxxx>]

On 05/01/2013 03:26 PM, Dick Hollenbeck wrote:
> On 05/01/2013 12:18 PM, Lorenzo Marcantonio wrote:
>> I was testing stuff and well... in the kicad_brd file the sheet date is
>> not saved. I found this pearl in TITLE_BLOCK::Format:
>>
>>     /* version control users were complaining, see mailing list.
>>     if( !m_date.IsEmpty() )
>>         aFormatter->Print( aNestLevel+1, "(date %s)\n",
>>                             aFormatter->Quotew( m_date ).c_str() );
>>     */
>>
>> What is a field useful for if it's not saved? It always come up empty, in this
>> way.
>>
>> In many companies the date on a drawing is actually more important than the
>> drawing version number (in fact ISO/EN 7200:2004 requires the date of issue,
>> which is legally binding for a number of things). Other dates (approval,
>> modifications and so on) are optional but the "document date" is mandatory.
>>
>> I think the best solution would be putting the date as an editable field (with
>> the title, company and so on): this would eliminate the localization issue and
>> in the mean time allow a 'significant' date to be set (not necessarily that's
>> the last modified date, for a number of reasons). Also the vcs users wouldn't
>> have to complain (I don't see the problem if a date changes in a versioned
>> file... maybe they don't like it).
> 
> 
> OK, I remember now.
> 
> 
> Your suggestion of having it editable is OK for me.  But is not OK for others.  There are
> 4 schools of thought, and funny thing is you and I are in the same camp after all.
> 
> 
> 1) Make it editable, and fixed, not dynamically updated.  If I put in a date, it stays
> that way forever on disk too.
> 
> 
> 2) Make it dynamic (i.e. automatically updated), reflective of the last time of board
> *modification*, not last time of save.
> 
> 
> 3) Make it dynamic (i.e. automatically updated), reflective of the last time of board *save*.
> 
> 
> 4) Remove it, keep it out of the file, until the dispute is settled.
> 
> 
> We were at 4).  If we uncomment the code, we get 2), and then I get yelled at again by the
> VCS users.
> 
> 
> I don't want to keep getting yelled at.  :)   It gets hard to remember being yelled at.
> 
> Jean-Pierre is in camp 2).
> 
> I am in camp 1) or 4), just because the VCS users will eventually win, I recognize that,
> and I want the problem to go away.  I use VCS too, and reporting a new date for every
> change is not especially helpful to me, with my VCS user hat on, and advocating for those
> I know will complain later.  Been there, and had that conversation already.
> 
> 
> Help wanted.  :)
> 
> 
> Dick


Newsflash...

OK, JP is now OK with 1) also.

We have a solution if you can code it.

Thanks Lorenzo.