hipl-core team mailing list archive

[Diego Biurrun <diego@xxxxxxxxxx>]

On Fri, Aug 20, 2010 at 11:53:45PM +0200, Christof Mroz wrote:
> On Fri, 20 Aug 2010 21:44:57 +0200, Diego Biurrun <diego@xxxxxxxxxx> wrote:
>
>>> revno: 4908
>>> committer: Christof Mroz <christof.mroz@xxxxxxxxxxxxxx>
>>> branch nick: trunk
>>> timestamp: Fri 2010-08-20 16:59:41 +0200
>>> message:
>>>   Clear ctx->error to 0 prior to processing each socket.
>>>
>>>   Maybe that's the wrong place to patch so here's some background info.
>>
>> Then why did you commit instead of sending a patch and discussing on the
>> ml first?
>
> Thanks for pointing this out, I'll be more cautious next time I alter  
> program logic.
> In the future, filing a bug and attaching the fix in launchpad would be  
> the cleanest way in my opinion (for bookkeping purposes), and the list is 
> gonna receive that anyway; or would it be best to send it to this list as 
> an old-fashioned patch file?

I myself am fine with patches, I guess it depends on whether you are
afraid it might get lost otherwise.

> What about "cosmetic" (const-correctness etc.) and documentation (docs,  
> more verbose debug messages,...) changes? May I commit these directly?

Yes, sure.

> Concerning this commit in particular, I think it's quite safe resetting 
> an error code at a place where no processing has taken place yet (it's 
> just one stack frame below hipd_main).

It just struck me as odd that you committed something with the log
message - let me paraphrase for a sec - "Change X.  Gee, I have no
idea if this is correct, but I committed anyway...".  To me this
sounds like a sign that another pair or eyeballs would help you.

There is no hurry, just send a patch or whatever and wait for somebody
to comment.  If nobody comments you can still commit in a few days, no
harm done.  But possibly somebody more experienced can give you some
valuable insights before that...

Diego