From: Jamis Buck Date: 2004-07-30T04:17:49+09:00 Subject: Re: [ANN] Net::SSH 0.0.5 Carl Youngblood wrote: > Do you really mean "pure ruby"? I thought that ruby relied on OpenSSL > C code to do a lot of its stuff. I do mean pure Ruby. The Net::SSH stuff is all Ruby. Yes, it has dependencies on other libraries that are not pure Ruby, but Net::SSH itself consists of only Ruby code. I mean, if you get nit-picky enough, no Ruby library is "pure-Ruby", since Array, Hash, String, and the like are all implemented in C. Also, since "openssl" is one of the standard Ruby libraries, it's not as onerous a depenency as it could be. - Jamis > > On Thu, 29 Jul 2004 13:31:48 +0900, Jamis Buck wrote: > >>Net::SSH 0.0.5 is now available! (Yes, this comes hard on the heels of >>0.0.4, but I'm going out of town for a week and a half and figured I >>should release the work that has been done in the last few days, since >>it was significant.) >> >>Net::SSH is a pure-Ruby implementation of an SSH2 client. It may be >>obtained at: >> >> http://rubyforge.org/projects/net-ssh >> >>The complete (!) users manual and API documentation may be read at: >> >> http://net-ssh.rubyforge.org >> http://net-ssh.rubyforge.org/api >> >>Version 0.0.5 adds the much-anticipated SSH agent support, but (alas) >>only on Unix-ish systems. (The PuTTY "pageant" agent is sufficiently >>different from the OpenSSH ssh-agent utility that it will require a >>pretty bit of work to interface with... any takers?) >> >>This version also provides an "rb-keygen" command-line utility for >>generating keys. It mirrors a subset of the functionality of >>"ssh-keygen", so if you've got ssh-keygen available, you might as well >>use it for generating keys. If you don't, you can use rb-keygen instead. >> >>One particularly significant bug fix was suggested by Daniel Hobe: in >>0.0.4 and earlier, all of a user's private keys were loaded at startup >>time, which meant that if the keys were protected by passphrases, the >>user had to enter those passphrases for all keys up front, even if none >>of the keys were ever used. This was annoying. Now, key loading is >>delayed until the key is actually needed, so you never need to enter a >>passphrase for a private key unless it is actually going to be used. >> >>Various other improvements and fixes have been added as well--see the >>ChangeLog for the complete list. Also, don't forget to read the users >>manual--all planned chapters have been drafted, and I'm just awaiting >>comments. (Thanks to those of you that have submitted comments on the >>documentation already--they've helped a ton.) >> >>Special thanks for this release go to Daniel Hobe, without whose >>suggestions, comments, encouragement, and patches this release would not >>be nearly as slick. >> >>(I think we may be getting close to a 0.1 release!) >> >>-- >>Jamis Buck >>jgb3@email.byu.edu >>http://www.jamisbuck.org/jamis >> >>"I use octal until I get to 8, and then I switch to decimal." >> >> > > > . > -- Jamis Buck jgb3@email.byu.edu http://www.jamisbuck.org/jamis "I use octal until I get to 8, and then I switch to decimal."