Re: [Ada] made me hate programming

["Yannick Duchêne (Hibou57)" <yannick_duchene@yahoo.fr>]

Le Mon, 23 Aug 2010 17:51:34 +0200, Warren <ve3wwg@gmail.com> a écrit:
>> You got it (about the compile-time bug-box, especially when you do not
>> have any other choice for the compiler)
>
> It can be a fun exercise to see what is provoking the
> bug box.
Except when this ends to understand there will be no good workaround and  
you will have to change many thing (especially do not like it if what has  
to be changed.... is at the interface level)

> While it is best to report those, I do find that
> the present reporting "process" is too time consuming.
> I find it is quicker and eaiser to find a work-around
> and hope that the bug has already been reported.
>
> Only as a last resort, I'll report it.  They need to
> make that process easier somehow. For example, allow
> the user to just send all the *.adb and *.ads files
> used.
I see a good reasons for that: this avoid local configuration or project  
layout to interfere. This is just like when an application hang for some  
user special configuration reason which are not of the application  
responsibility and you have to be sure of that.

Also, just think bug repository may be actively feed: just try to imagine  
yourself at the other side of this stream. “Ouch!” you will say.

I agree this is tedious (I have multiple bug reports pending because I am  
not sure of some submission requirement details), but just think of it as  
a little fee you give back (no one enjoy to b a slave). I suppose  
commercial support may be more relaxing regarding how bug reports are  
received (this probably may even be the source of a whole application with  
ugly make files and etc).

> Instead they insist on *ada files and don't want
> anything that is not involved.  Gnat tries to name
> the involved sources, but I have found that this isn't
> reliable - resulting in more futzing around.
Yes, that is why I prefer this: “remove a component, see if the bug still  
express, remove a component, see if the bug ...”, until I can get  
something small enough and as much as possible free of any external  
dependencies or external libraries (if needed, you may also create an  
local extract of library dependencies). This is good not just in the  
eventuality of a bug report, but also for yourself if ever you like to  
archive sources exposing bugs for you own regression tests or just to not  
forget what you already learned about the bugs of a given application.



-- 
* 3 lines of concise statements is readable, 10 pages of concise  
statements is unreadable ;
* 3 lines of verbose statements may looks unuseful, 10 pages of verbose  
statements will never looks too much pedantic