Patch

[Dirk Hohndel]

First. Do not top post. It breaks the flow of conversation.

On Sun, 2014-03-09 at 00:22 +0800, Rishubh Jain wrote:
> I am sorry for the silly mistakes...i am willing to correct my
> mistakes...I understand the indent problem and the commit message I
> will definitely improvise that, and will surely remove the qDebug()
> calls....

Good.

> but i still didnt get the idea of not using what i did..isnt it the
> same thing saving on the messge box or saving where you were
> editing?..please explain me if possible..and if you consider my trying
> to understand worthy of explanation..If not then it is ok

I don't really know what else to tell you. The check before the message
box is shown tells us that the user is in the middle of editing a dive.
Because of the way we store the information that is being entered, that
dive could be in an inconsistent state. Data entered will get lost -
which is NOT the result that the user expects when hitting 'save'.

So the user first needs to decide if they want to keep the changes to
the dive that they made. That's why we are telling them to go back
there.

Yes, you can cancel and just quit. As I said in my previous response
(way below, thanks to your top-reply) that would be acceptable.

But simply calling file_save() there is flat out wrong.

/D


> 
> On Saturday, 8 March 2014 9:29 PM, Dirk Hohndel <dirk at hohndel.org>
> wrote:
> 
> On Sat, 2014-03-08 at 23:00 +0800, Rishubh Jain wrote:
> > Hey Dirk
> > 
> > 
> > Please review my patch once more ...I think its usefull and i guess
> > even few people like thiago thinks it will be usefull, 
> > 
> > It makes the UI a little more friendly
> 
> You have not read my response. You have not reacted to the explanation
> in my response. Instead you keep sending messages urging me to
> reconsider and apply your patch.
> 
> You have not addressed the issues that I raised with the formatting of
> the patch (commit message), the formatting of the code (indentation),
> or
> the issue that at the point where you inserted the code, calling
> file_save() means that you will try to save an inconsistent state
> since
> you are IN THE MIDDLE OF EDITING A DIVE.
> 
> No, the patch is wrong. You CANNOT simply save there.
> 
> Yes, I might be willing to consider a patch that tells the user
> "you're
> in the middle of editing, do you really just want to throw away
> everything?" (not those words, the idea) and offers "quit anyway" "go
> back".
> 
> I will not accept a patch that calls file_save() at this spot.
> 
> Also, when redoing the patch, remove the qDebug() calls, please
> 
> 
> /D
> 
> 
> 
> 
> 
>