On November 10, 2016 4:00:39 AM CST, Pierre Labastie <[email protected]> 
wrote:
>On 10/11/2016 08:24, DJ Lucas wrote:
>> Okay, so here is a little bit more complete example. I've moved back 
>> to using certdata.txt directly. I've replaced our external perl
>script 
>> with an unfortunately ugly function. I've populated our shared NSS DB
>
>> as a replacement DB, not a supplemental one. I've changed the logic 
>> for version to store the Version string in the ca-bundle.crt, and 
>> compare with the installed one.
>>
>> This doesn't actually install anything yet, it only throws the value 
>> of TEMPDIR to stdout and leaves the temporary directory in place so 
>> that the results of the conversion can be reviewed. I haven't messed 
>> with Java yet, or pulled in /etc/ssl/local but those are trivial 
>> additions at this point.
>>
>> Obviously outstanding is any explanatory text whatsoever. How to 
>> adjust NSS consumers to use the shared DB might be of some
>importance. 
>> (I'll try for that tomorrow). :-)
>>
>> As far as I can tell, the only remaining thing brought up in the 
>> previous thread was how to obtain and verify the file. I do like
>using 
>> the release branch as the default source (with version info as 
>> provided by Bruce's script on Anduin). Bruce, what do you think about
>
>> signing that file for verification? Or even automatically updating
>the 
>> date and md5sum of the file in the book -- changelog would need to be
>
>> skipped I think, but with that little concession, it should be 
>> reasonably easy to do from cron.
>>
>> P.S. I'd appreciate a second set of eyes on convert_pem(). It works 
>> well enough, but I could not get it to play nicely using only pipes. 
>> As a result, right now it uses a total of three temporary files per 
>> conversion. This is terribly inefficient and I'd like to fix it, but
>I 
>> just can't see it tonight.
>>
>> --DJ
>>
>>
>Quick look: why using "covert_pem" and not "convert_pem"?

Typo. Thank you. :-)

>
>Why would you need a loop in covert_pem for finding the "END" line? 
>Isn't the file examined for only one certificate?
>

It is, but it contains a lot more data than is necessary for the DER encoded 
file. We are only interested in that small part of it. There should be around 
10 lines that begin in ^END. I suppose we could limit to a range beginning at 
startline and just grab the first instance.

>I do not see where the octal/char conversion is done (missing format in
>
>printf?).

They are already formatted as \NNN.
From the manpage:
\<NNN>  Interprets <NNN> as octal number and prints the corresponding character 
from the character set.




-- 
Sent from my Android device with K-9 Mail. Please excuse my brevity.

-- 
http://lists.linuxfromscratch.org/listinfo/blfs-dev
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page

Reply via email to