Re: client-side only vs. libssh?
> I'm happy to correct all flaws and errors existing there. I've been open
> about that page since day 1 to get help to make it better.
I am happy we can work together on making that page better and more
>> -"not threadsafe": libssh is as much threadsafe as libssh2, in the
>> sense that all operations made on different sessions are threadsafe.
>> That is consistent with "Thread-safe: just don't share handles
>> simultaneously" on libssh2.org front page.
> I based that on comments in your mails/documentation. I'll update.
Hmm, I don't remember having written in the doc the thread-safety was
wrong. It's a closed case anyway.
>> -"Some limitations on >4gb files": Not any that I am aware of. Could
>> you explain what's actually lacking in libssh 0.4 ? There is a
>> sftp_seek64 call which takes a uint64_t parameter. That function
>> exists since 0.3 released one year ago.
> First, the page says it compares against libssh 0.3 so details in 0.4
> would of course not be covered. I based this on the API/docs/source and
> it claimed this. But with your feedback here I can update that to say it
> is a comparison against 0.4. Thanks!
Thanks for making the update. I was relating to 0.3, but of course now
the state of the art with libssh is 0.4.
>> -Speed claims: Please don't be ridiculous. No competent network
>> developer will take you seriously when you tell that libssh2 is 2.3
>> times faster that libssh.
> I did my tests and I presented what I did and I showed the results.
> Again, nobody else (including you) have argued or discussed them with
> me. I do not claim to have done any thorough and extensive tests but I
> did check the code and it looked like they were roughly comparable.
I saw the results several weeks after you wrote them. I began working
on real benchmarks before giving up. I'm sorry I waited so long before
writing this mail.
> I guess that makes me "no competent network developer".
You'd be no competent network developer if you seriously believed your
benchmark was representative of anything. By your explanation, I see
that was not your intent, but the way it was described on the page
could lead to confusion.
>> I am saying that it's ridiculous to make precise speed claims out of
>> that single biased test.
> They were not "precise" and all tests done by me will be biased.
2.3 is precise for me. I think it's all possible to do a test in an
unbiased manner. Thanks again for having removed the litigious parts.
>> I do not discuss the license of libssh and the blocking issues. These
>> are real and will hopefully be resolved in next major libssh release.
> Are you considering to change the license?
I was considering but other developers did not want. We do not see the
licensing as an issue.
>> I think a bit of competition in opensource is fair and leads to
> In this particular case, I'm far from convinced that this competition is
> good. We're making two separate and roughly comparable libs, both free
> and open with roughly the same purpose. I'd say we're dividing our
> efforts on two places instead of putting them all in one place...
What you say is true in a sense, but as been shown to be false for
other communities such as kde and gnome. We are at a point where we
have two codebase largely different and incompatible, having different
features, strengths, weaknesses and bugs. This makes the choice
possible for the final user, when a one-size-fits-all approach may not
be suited to the particular needs of that user.
I know you weren't at the origin of the libssh2 project. I just want to
add that the first release of libssh was made in 2003 and I never
refused any contribution.
> Thanks for all the corrections and clarifications!
I thank you for your quick intervention and cooperation. I want our two
projects to communicate and exchange in a constructive manner. Would
you want that we provide a comparison page between libssh and libssh2
on libssh.org ?
Received on 2010-01-23