<div dir="ltr"><div dir="ltr"><br></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">ср, 10 февр. 2021 г. в 15:17, Bryan Frimin <<a href="mailto:bryan@frimin.fr">bryan@frimin.fr</a>>:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">Hello Sergey,<br>
<br>
> Is there some difference from this implementation?<br>
> <a href="https://github.com/basho/erlang-pbkdf2" rel="noreferrer" target="_blank">https://github.com/basho/erlang-pbkdf2</a><br>
The two libraries are very similar. The main differences are:<br>
- Released under ISC license;<br>
- Uses the new Erlang crypto API;<br>
- Strictly follows the RFC implementation instead of being compliant<br>
with the Apache CouchDB.<br>
<br></blockquote><div>Ah, nice! Thanks!</div><div> </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
> We have a ticket in `epgsql` to replace our "homemade" implementation of<br>
> this algorithm <a href="https://github.com/epgsql/epgsql/issues/172" rel="noreferrer" target="_blank">https://github.com/epgsql/epgsql/issues/172</a>, but never<br>
> actually had time to look into it.<br>
Not sure I understand the motivation you can have to use external<br>
dependency instead of your implementation.<br>
<br></blockquote><div><br></div><div>I guess, mainly the rule of thumb "try to not implement your own crypto, when there are already available well-tested libraries for that".</div><div>Also, as the ticket says, Elixir's postgres driver uses an external library for pbkdf2 as well, doubling this opinion.</div><div> </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
Regards,<br>
<br>
-- <br>
Bryan Frimin<br>
<a href="https://www.frimin.fr" rel="noreferrer" target="_blank">https://www.frimin.fr</a><br>
<a href="mailto:bryan@frimin.fr" target="_blank">bryan@frimin.fr</a><br>
</blockquote></div></div>