Hi Carlos,

On Tue, Nov 19, 2024 at 08:14:48AM -0500, Carlos O'Donell wrote:

On 11/17/24 5:46 AM, Helge Kreutzmann wrote:

quoted

Without further ado, the following was found:

Issue:    No "in succession" here on purpose?

"When no I<key> is provided, use B<sethostent>(3), B<gethostent>(3), and "
"B<endhostent>(3)  to enumerate the hosts database.  When one or more I<key> "
"arguments are provided, pass each I<key> to B<gethostbyaddr>(3)  or "
"B<gethostbyname2>(3), depending on whether a call to B<inet_pton>(3)  "
"indicates that the I<key> is an IPv6 or IPv4 address or not, and display the "
"result."

This text is for the "hosts" database.

quoted

… and in many other paragraphs (I can provide you the full list). Last
time you said, that this is under review with glibc - are there any
results on this?

As an upstream glibc maintainer I can comment on this.

The getent program today iterates over the keys passing them in succession to the calls
as listed for the "hosts" database i.e. nss/getent.c:hosts_keys.

I think the text could be improved by making it consistent for "hosts."

Thanks!

Each one of these instances is different so they each should be enumerated and reviewed
separately.

Would you mind proposing some patches?

Have a lovely day!
Alex

-- 
<https://www.alejandro-colomar.es/>