[hipl-dev] Re: [Merge] lp:~hipl-core/hipl/certificate-exchange into lp:hipl

  • From: Christof Mroz <christof.mroz@xxxxxxxxxxxxxx>
  • To: hipl-dev@xxxxxxxxxxxxx
  • Date: Tue, 13 Mar 2012 17:58:34 +0100

On 13.03.2012 17:31, Diego Biurrun wrote:
On Mon, Mar 12, 2012 at 09:29:29PM +0100, Diego Biurrun wrote:
On Thu, Mar 08, 2012 at 06:41:47PM +0100, Diego Biurrun wrote:
What is "param_cert + 1" supposed to be?  Are you trying to move the
pointer past the struct?  This is extremely brittle, see the cast
and the length calculation that are necessary.

I'm still mystified about this one.

Christoph confirms the intention.  Repairing this likely entails quite
a bit of fiddling with structures.

Related:
https://answers.launchpad.net/hipl/+question/151048

+    if (optc == 2&&  is_pos_natural_number(opt[1])) {

OK, what exactly is the problem here?  Why do you need the utility function?

Can the original author shed some light on this?

Christoph said scanf did not work for him.  I guess we'll trust his word.

I still consider explicit type checks an anti pattern which should be replaced by coercion, in this case scanf(), whenever easily possible.

As a fan of obscure facts I'd really like to know why scanf("%u",...) failed. Don't let my curiosity stall the merge, though.

Other related posts: