KDE Cryptography Module

Alon Bar-Lev alon.barlev at gmail.com
Wed May 23 21:59:59 BST 2007


Hello All,

I wish to end my contribution to this thread. It seems that I offended
too many people during this discussion.

First of all I must say I regret if anyone of you was offended. This
was not my intention. I tried to discuss technical details only and I
regret people find this offending.

I also regret that my English skill are not to great, which not making
it easier for us to communicate correctly.

Aaron approach of "Show me the code" offer may be seemed great at
first glance, but I must say that this is the worse cooperation I
expected to have.

A better cooperation would have, no... we don't wish to add his
feature, we go side by side with Qt features. I would have accepted
this approach gladly.

Since I am not core KDE developer, reimplementing core component alone
without the maintainer cooperation is a waste of resources, both for
the implementer and the maintainer.

Especially when we know that the KSSL interface is going to be
modified. And I even don't have a good example of what you thinking
about.

Working in incremental mode, without clear destination of where and
how we should be in the final stage, is also a waste of resources, as
much work will need to be redone later on.

The best approach is cooperation, we, the QCA developers, provide our
insights and together with current maintainer reach to a state that
all components are cryptographic friendly. We can divide the
components, produce a roadmap together and start implementation.

So please understand Aaron, your proposal is not practical, and it
likely cause me to fail in my efforts.

I know cryptography is not simple to implement, this is why we offer
our experience. But without planning and commitment the resources
cannot be justified.

I know that I am not complete crazy since at least one but me, Oswald,
fully understand the implications.

If you ever have a clear roadmap and have smartcard milestone within
with commitment of all maintainers, I will be happy to help
implementing this.

I reached to similar situation in GnuPG, OpenSSH and maybe soon
GnuTLS. I end up maintaining external patch since upstream don't
provide what users require. Who knows... Maybe I will do the same
mistake and provide an external patch for KDE 4 as well.

Thank you for this discussion,
Alon Bar-Lev.




More information about the kde-core-devel mailing list