Thread: RE: [JSch-users] Getting "Auth fail" after 7 successful connections.
Status: Alpha
Brought to you by:
ymnk
From: Oberhuber, M. <Mar...@wi...> - 2006-05-22 09:15:11
|
Hi Eric / Atsuhiko, =20 I'd very much like to see the the char[] change for the password/passphrase in Jsch, since i think that security is a really big issue. =20 I think there are ways of making this change backward compatible, so that not every clent code needs to be changed. For instance, interface UserInfo2 could be defined holding the getPasspharse() method return a char[]. Then, Jsch could internally work with char[] and do the client interface call ...if userinfo instanceof UserInfo2... Cheers, Martin -- Martin Oberhuber - WindRiver, Austria +43(662)457915-85 =20 ________________________________ From: jsc...@li... [mailto:jsc...@li...] On Behalf Of Eric Meek Sent: Friday, May 19, 2006 4:33 PM To: Andrei Tchijov Cc: Atsuhiko Yamanaka; jsc...@li... Subject: Re: [JSch-users] Getting "Auth fail" after 7 successful connections. =09 =09 Andrei,=20 I've submitted these changes with several other changes and most have been accepted. For some reason, however, the equals methods were not accepted. I have been keeping a separate source tree in my CVS with changes that were not accepted but I need in my project. Two of the biggest changes are a prototype implementation of a java "ssh-agent" for JSch and the internal storing passwords and passphrases in JSch as a char array and not immutable Strings. The password/passphrase change is absolutely necessary in our project to ensure the passwords are erased when usage is finished and will not hang around in memory after program termination (basically the reason getText() was deprecated in JPasswordField).=20 I can understand why these two changes haven't made it into the main source since they are non-trivial changes. Non-trivial in the sense that the Agent change requires many change to JSch and the char change requires a change to the UserInfo interface forcing every JSch user to change their code.=20 Eric "If it takes three days to predict tomorrow's weather, what good is it?" -Peter Kastner On May 19, 2006, at 9:02 AM, Andrei Tchijov wrote: Eric,=20 Have you ever tried to make this update part of "official" source tree? What was the answer from JCraft folks (if any)? I can not see any reason for them to object, but than, I was really surprised to find such and obvious problem in a first place. =20 Andrei On May 19, 2006, at 8:53 AM, Eric Meek wrote: Andrei,=20 If you want a little more specific version, here is an equals method I have added to IdentityFile.java in each JSch release since 0.1.21 to ensure proper behavior when using Identities in a Collections. I also have implemented equals methods in Channel and Session classes if you would like them since both classes are also used in collections. Eric=20 public boolean equals(Object o) { boolean b =3D false; if (o instanceof IdentityFile) { IdentityFile tmp =3D (IdentityFile) o; b =3D tmp.encrypted =3D=3D encrypted; b =3D b && tmp.identity.equals(identity); b =3D b && (tmp.type =3D=3D type); b =3D b && (tmp.keytype =3D=3D keytype); for (int i =3D 0; (key !=3D null) && (i < key.length) && b; i++) { b =3D b && (key[i] =3D=3D tmp.key[i]); } } else { b =3D super.equals(o); } return b; } "To be prepared for war is one of the most effectual means of preserving peace." - George Washington, in First Annual Address to both houses of Congress, January 8, 1790 On May 19, 2006, at 7:03 AM, Andrei Tchijov wrote: Excellent!!! This is exactly what is happening. May I make a small suggestion? It looks like JSch.addIdentity attempts to prevent same identity file been added multiple times :=20 public void addIdentity(String foo, String bar) throws JSchException { Identity identity =3D new IdentityFile(foo, this); if (bar !=3D null) { identity.setPassphrase(bar); } synchronized (identities) { if (!identities.contains(identity)) { // This will not work, unless IdentityFile implements equals() identities.addElement(identity); } } } But unfortunately IdentityFile class lucks "equals( Object o )" method which makes "contains" method return false every time. As soon as I added this=20 =09 =09 public boolean equals( Object o ) { if( o instanceof IdentityFile ) { return identity.equals( ((IdentityFile)o).identity); } else { return super.equals( o ); } } =09 =09 to com.jcraft.jsch.IdenityFile class, my problem went away. =09 =09 =09 =09 Thank you very much for quick response. Andrei Tchijov On May 18, 2006, at 11:53 PM, Atsuhiko Yamanaka wrote: Hi, +-From: Andrei Tchijov <an...@tc...> -- |_Date: Thu, 18 May 2006 23:01:23 -0400 ______ | |I am experiencing very strange problem. In my application I connect =20 |to remote server and setting up local tunnel. I can do it 7 times in =20 |a row. On 8th attempt I am getting |... |Caused by: com.jcraft.jsch.JSchException: Auth fail | at com.jcraft.jsch.Session.connect(Session.java:390) | at com.jcraft.jsch.Session.connect(Session.java:143) |... |Once I get into this state, all consecutive attempts end up the same =20 |way. ... |PS I am using JSch 1.28 on Mac OS X (both client and server side). =20 |Authentication is password base. It seems an exception has occurred in a trial for public-key authentication. Have you invoked "jsch.addIdentity" method? If you don't invoke it, public-key authentication will not be tried. I guess that you have invoked "jsch.addIdentity" 8 times. =20 Is this guess correct? By invoking that method, 8 key will be cached=20 internally and the "UserAuthPublicKey.start" method has tried public-key=20 authentication 8 times with wrong keys, and sshd has rejected the connection because of so many failures. 7 times failure may be acceptable for that sshd, but 8 times failure not. Such a behavior must be depended on its=20 implementation, I guess. Sincerely, -- Atsuhiko Yamanaka JCraft,Inc. 1-14-20 HONCHO AOBA-KU, SENDAI, MIYAGI 980-0014 Japan. Tel +81-22-723-2150 +1-415-578-3454 Fax +81-22-224-8773 Skype callto://jcraft/ =09 ------------------------------------------------------- Using Tomcat but need to do more? Need to support web services, security? Get stuff done quickly with pre-integrated technology to make your job easier Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo =09 http://sel.as-us.falkag.net/sel?cmd=3Dlnk&kid=3D120709&bid=3D263057&dat=3D= 121642 =09 _______________________________________________ JSch-users mailing list JSc...@li... =09 https://lists.sourceforge.net/lists/listinfo/jsch-users |
From: <ym...@jc...> - 2006-05-22 09:43:59
|
Hi, +-From: "Oberhuber, Martin" <Mar...@wi...> -- |_Date: Mon, 22 May 2006 11:14:17 +0200 _______________________ | |I'd very much like to see the the char[] change for the |password/passphrase in Jsch, since i think that security is a really big |issue. I have modified the code[1] to add * Session.setPassword(byte[]) and * JSch.addIdentity(String identity, byte[] passphrase), are they not enogh? [1] http://www.jcraft.com/jsch/jsch-0.1.29-rc4.zip Sincerely, -- Atsuhiko Yamanaka JCraft,Inc. 1-14-20 HONCHO AOBA-KU, SENDAI, MIYAGI 980-0014 Japan. Tel +81-22-723-2150 +1-415-578-3454 Fax +81-22-224-8773 Skype callto://jcraft/ |
From: Andrei T. <an...@tc...> - 2006-05-22 10:53:18
|
Atsuhiko, What about adding "equals()" method to IdentityFile? Andrei On May 22, 2006, at 5:43 AM, Atsuhiko Yamanaka wrote: > Hi, > > +-From: "Oberhuber, Martin" <Mar...@wi...> -- > |_Date: Mon, 22 May 2006 11:14:17 +0200 _______________________ > | > |I'd very much like to see the the char[] change for the > |password/passphrase in Jsch, since i think that security is a > really big > |issue. > > I have modified the code[1] to add > * Session.setPassword(byte[]) and > * JSch.addIdentity(String identity, byte[] passphrase), > are they not enogh? > > [1] http://www.jcraft.com/jsch/jsch-0.1.29-rc4.zip > > > Sincerely, > -- > Atsuhiko Yamanaka > JCraft,Inc. > 1-14-20 HONCHO AOBA-KU, > SENDAI, MIYAGI 980-0014 Japan. > Tel +81-22-723-2150 > +1-415-578-3454 > Fax +81-22-224-8773 > Skype callto://jcraft/ |
From: Eric M. <me...@cs...> - 2006-05-22 11:53:33
|
Atsuhiko, I've implemented the "equals()" method in Channel.java, Session.java and IdentityFile.java if you would like to use or bounce of that code for adding to JSch. If you still don't have the code I sent you let me know and I'll send it. Eric "Only those who have the patience to do simple things perfectly will acquire the skill to do difficult things easily." - Johann Von Schiller On May 22, 2006, at 6:52 AM, Andrei Tchijov wrote: > Atsuhiko, > What about adding "equals()" method to IdentityFile? > Andrei > > On May 22, 2006, at 5:43 AM, Atsuhiko Yamanaka wrote: > >> Hi, >> >> +-From: "Oberhuber, Martin" <Mar...@wi...> -- >> |_Date: Mon, 22 May 2006 11:14:17 +0200 _______________________ >> | >> |I'd very much like to see the the char[] change for the >> |password/passphrase in Jsch, since i think that security is a >> really big >> |issue. >> >> I have modified the code[1] to add >> * Session.setPassword(byte[]) and >> * JSch.addIdentity(String identity, byte[] passphrase), >> are they not enogh? >> >> [1] http://www.jcraft.com/jsch/jsch-0.1.29-rc4.zip >> >> >> Sincerely, >> -- >> Atsuhiko Yamanaka >> JCraft,Inc. >> 1-14-20 HONCHO AOBA-KU, >> SENDAI, MIYAGI 980-0014 Japan. >> Tel +81-22-723-2150 >> +1-415-578-3454 >> Fax +81-22-224-8773 >> Skype callto://jcraft/ > |
From: Eric M. <me...@cs...> - 2006-05-22 13:14:10
|
Hi, It's a start but doesn't really address the concern. If these are the only changes to JSch, the UserInfo interface still requires the passwords and passphrases to be returned as a Strings, hence the security concerns. I really like Martin's idea of a separate interface and think that is the way to go. Split out all methods that won't change from UserInfo into a new User interface. Extend the User interface by UserInfo and UserInfo2 and then deprecate the UserInfo interface. Here is a list of what I would like to see changed in JSch in regard to the Strings * all internal storage of the passwords or passphrases from Strings to byte[] * all methods requiring password or passphrases as Strings be overridden to also accept char[] * add a UseInfo2 getPassword, getPassphrase returning char[] * add a UIKeyboardInteractive2 promptKeyboardInteractive returning char[][] instead of String[] On another note, I've implemented most of these changes my local copy of JSch (working on the separate interfaces now) but merging them back would take some work since, as you know, I've formatted the source with Jalopy http://sourceforge.net/projects/jalopy In the interests of making the JSch source more hacker friendly, you should use some set of formatting rules and publish them. If you were to define a coding standard for JSch, I would conform to it and then considering these changes would be much simpler. I would highly recommend Jalopy because you could post your formatting rules making it much easier to conform. Eric "The past is a guidepost, not a hitching post." - L. Thomas Holcroft On May 22, 2006, at 5:43 AM, Atsuhiko Yamanaka wrote: > Hi, > > +-From: "Oberhuber, Martin" <Mar...@wi...> -- > |_Date: Mon, 22 May 2006 11:14:17 +0200 _______________________ > | > |I'd very much like to see the the char[] change for the > |password/passphrase in Jsch, since i think that security is a > really big > |issue. > > I have modified the code[1] to add > * Session.setPassword(byte[]) and > * JSch.addIdentity(String identity, byte[] passphrase), > are they not enogh? > > [1] http://www.jcraft.com/jsch/jsch-0.1.29-rc4.zip > > > Sincerely, > -- > Atsuhiko Yamanaka > JCraft,Inc. > 1-14-20 HONCHO AOBA-KU, > SENDAI, MIYAGI 980-0014 Japan. > Tel +81-22-723-2150 > +1-415-578-3454 > Fax +81-22-224-8773 > Skype callto://jcraft/ > > > ------------------------------------------------------- > Using Tomcat but need to do more? Need to support web services, > security? > Get stuff done quickly with pre-integrated technology to make your > job easier > Download IBM WebSphere Application Server v.1.0.1 based on Apache > Geronimo > http://sel.as-us.falkag.net/sel? > cmd=lnk&kid=120709&bid=263057&dat=121642 > _______________________________________________ > JSch-users mailing list > JSc...@li... > https://lists.sourceforge.net/lists/listinfo/jsch-users |
From: Andrei T. <an...@tc...> - 2006-05-22 13:27:09
|
I strongly support Eric comment about coding standards. Frankly, I am amazed that such a complicated product could be some-what successfully maintained (and developed) with current state of a source code. Just simple fixing of indents will go a long way. I am using Eclipse and every time I need to look into JSch source code, first thing I do is applying 'Source/Format' to the file in question. Instantaneously it becomes 1000 times more readable. Please to not take any offense to this comment. JSch is wonderful product and it is crucial part of many projects (including mine). All we wont is just to make it a little bit more "hacker friendly". Andrei On May 22, 2006, at 9:14 AM, Eric Meek wrote: > Hi, > > It's a start but doesn't really address the concern. If these are > the only changes to JSch, the UserInfo interface still requires the > passwords and passphrases to be returned as a Strings, hence the > security concerns. > > I really like Martin's idea of a separate interface and think that > is the way to go. Split out all methods that won't change from > UserInfo into a new User interface. Extend the User interface by > UserInfo and UserInfo2 and then deprecate the UserInfo interface. > > Here is a list of what I would like to see changed in JSch in > regard to the Strings > * all internal storage of the passwords or passphrases from > Strings to byte[] > * all methods requiring password or passphrases as Strings be > overridden to also accept char[] > * add a UseInfo2 getPassword, getPassphrase returning char[] > * add a UIKeyboardInteractive2 promptKeyboardInteractive > returning char[][] instead of String[] > > > On another note, I've implemented most of these changes my local > copy of JSch (working on the separate interfaces now) but merging > them back would take some work since, as you know, I've formatted > the source with Jalopy http://sourceforge.net/projects/jalopy > > In the interests of making the JSch source more hacker friendly, > you should use some set of formatting rules and publish them. If > you were to define a coding standard for JSch, I would conform to > it and then considering these changes would be much simpler. I > would highly recommend Jalopy because you could post your > formatting rules making it much easier to conform. > > Eric > > > "The past is a guidepost, not a hitching post." - L. Thomas Holcroft > > On May 22, 2006, at 5:43 AM, Atsuhiko Yamanaka wrote: > >> Hi, >> >> +-From: "Oberhuber, Martin" <Mar...@wi...> -- >> |_Date: Mon, 22 May 2006 11:14:17 +0200 _______________________ >> | >> |I'd very much like to see the the char[] change for the >> |password/passphrase in Jsch, since i think that security is a >> really big >> |issue. >> >> I have modified the code[1] to add >> * Session.setPassword(byte[]) and >> * JSch.addIdentity(String identity, byte[] passphrase), >> are they not enogh? >> >> [1] http://www.jcraft.com/jsch/jsch-0.1.29-rc4.zip >> >> >> Sincerely, >> -- >> Atsuhiko Yamanaka >> JCraft,Inc. >> 1-14-20 HONCHO AOBA-KU, >> SENDAI, MIYAGI 980-0014 Japan. >> Tel +81-22-723-2150 >> +1-415-578-3454 >> Fax +81-22-224-8773 >> Skype callto://jcraft/ >> >> >> ------------------------------------------------------- >> Using Tomcat but need to do more? Need to support web services, >> security? >> Get stuff done quickly with pre-integrated technology to make your >> job easier >> Download IBM WebSphere Application Server v.1.0.1 based on Apache >> Geronimo >> http://sel.as-us.falkag.net/sel? >> cmd=lnk&kid=120709&bid=263057&dat=121642 >> _______________________________________________ >> JSch-users mailing list >> JSc...@li... >> https://lists.sourceforge.net/lists/listinfo/jsch-users > |
From: Eric <eri...@ul...> - 2006-05-22 15:03:18
|
Le Lundi 22 Mai 2006 15:14, Eric Meek a =E9crit=A0: > Hi, > It's a start but doesn't really address the concern. [...] > > On another note, I've implemented most of these changes my local copy > of JSch (working on the separate interfaces now) but merging them > back would take some work since, as you know, I've formatted the > source with Jalopy http://sourceforge.net/projects/jalopy > > In the interests of making the JSch source more hacker friendly, you > should use some set of formatting rules and publish them. If you > were to define a coding standard for JSch, I would conform to it and > then considering these changes would be much simpler. I would highly > recommend Jalopy because you could post your formatting rules making > it much easier to conform. > Eric I maintain my local changes with a subversion vendor branch, and if Atsuhik= o=20 has his own repository, It would be easier for thoose who want to keep in=20 sync to have some read-access to maintain their local branch, thus branchin= g=20 and merging would be easier. Anyway I'm glad to see others are harvesting the source to spot bugs and fi= nd=20 enhancements in jsch. So It would be very nice that thoose who like the two Eric's have a certain= =20 level of understanding of Atsuhiko's code (many thanks to him) give a littl= e=20 bit of their time java-documenting some essential parts of the code. IMHO it would greatly increase our abilities to quickly find bugs, and=20 hopefully help ymk fix them (even if it's useless as he releases bugfixes a= t=20 lightspeed compared to other projets...). But it needs Coding Guidelines from ymnk. [out-of-topic]: Like Andrei, I use eclipse, and source formatting is since = no=20 longer a headache (I specialy appreciate automatic import lines and=20 refactoring tools). =2D-=20 Eric D=C9CORNOD Ing=E9nieur en d=E9veloppement d'applications ULP Multim=E9dia Universit=E9 Louis Pasteur courriel: eri...@ul... |
From: <ym...@jc...> - 2006-05-24 00:58:38
|
Hi, +-From: eri...@ul... -- |_Date: Mon, 22 May 2006 17:03:53 +0200 ___ | |IMHO it would greatly increase our abilities to quickly find bugs, and |hopefully help ymk fix them (even if it's useless as he releases bugfixes |at lightspeed compared to other projets...). But it needs Coding |Guidelines from ymnk. What is a "Coding Guidelines"? I don't waist my time to write such a thing. If you need it, write it by yourself and post it here. We will review it. Here is a Java Code Style formatter generated Eclipse 3.2RC4, http://www.jcraft.com/jsch/misc/eclipse-java-code-formatter-20060522.xml For your conveniences, I'll reformat the code with it at formal 0.1.29 release timing, but please keep in your mind that, that format definition may be changed and I'll not promise to reformat the code at every release. I'm working a committer of Eclipse.org, but I'm not an Eclipse user. I can not write the code without Emacs. Sincerely, -- Atsuhiko Yamanaka JCraft,Inc. 1-14-20 HONCHO AOBA-KU, SENDAI, MIYAGI 980-0014 Japan. Tel +81-22-723-2150 +1-415-578-3454 Fax +81-22-224-8773 Skype callto://jcraft/ |
From: Andrei T. <an...@tc...> - 2006-05-24 01:10:03
|
Atsuhiko, Thank you very much. This is excellent news. Andrei PS Emacs is good, but you really should try Eclipse. I was much like you (only I used to do it in Vi), but Eclipse's re-factoring capabilities are highly addictive (along with bunch of other features which really make you work faster - no gimmick). And unlike many other IDE's overhead of bringing your projects under Eclipse "control" is minimal (especially if you are using Ant). PPS Sorry to be annoying, but are you going to make me completely ecstatic and add equals() method to IdentityFile class as well? On May 23, 2006, at 8:58 PM, Atsuhiko Yamanaka wrote: > Hi, > > +-From: eri...@ul... -- > |_Date: Mon, 22 May 2006 17:03:53 +0200 ___ > | > |IMHO it would greatly increase our abilities to quickly find > bugs, and > |hopefully help ymk fix them (even if it's useless as he > releases bugfixes > |at lightspeed compared to other projets...). But it needs Coding > |Guidelines from ymnk. > > What is a "Coding Guidelines"? I don't waist my time to write such > a thing. > If you need it, write it by yourself and post it here. We will > review it. > > Here is a Java Code Style formatter generated Eclipse 3.2RC4, > http://www.jcraft.com/jsch/misc/eclipse-java-code- > formatter-20060522.xml > > For your conveniences, I'll reformat the code with it at formal 0.1.29 > release timing, but please keep in your mind that, that format > definition may > be changed and I'll not promise to reformat the code at every release. > I'm working a committer of Eclipse.org, but I'm not an Eclipse user. > I can not write the code without Emacs. > > > Sincerely, > -- > Atsuhiko Yamanaka > JCraft,Inc. > 1-14-20 HONCHO AOBA-KU, > SENDAI, MIYAGI 980-0014 Japan. > Tel +81-22-723-2150 > +1-415-578-3454 > Fax +81-22-224-8773 > Skype callto://jcraft/ > > > ------------------------------------------------------- > All the advantages of Linux Managed Hosting--Without the Cost and > Risk! > Fully trained technicians. The highest number of Red Hat > certifications in > the hosting industry. Fanatical Support. Click to learn more > http://sel.as-us.falkag.net/sel? > cmd=lnk&kid=107521&bid=248729&dat=121642 > _______________________________________________ > JSch-users mailing list > JSc...@li... > https://lists.sourceforge.net/lists/listinfo/jsch-users |
From: <ym...@jc...> - 2006-05-24 02:03:56
|
Hi, +-From: Andrei Tchijov <an...@tc...> -- |_Date: Tue, 23 May 2006 21:09:52 -0400 ______ | |PS Emacs is good, but you really should try Eclipse. I was much like |you (only I used to do it in Vi), but Eclipse's re-factoring |capabilities are highly addictive (along with bunch of other features |which really make you work faster - no gimmick). And unlike many |other IDE's overhead of bringing your projects under Eclipse |"control" is minimal (especially if you are using Ant). Don't worry about it :-) When in Rome, do as the Romans do. As a committer for Eclipse.org project, I'll use Eclipse in contributing plug-ins to Eclipse.org's CVS repository and I have understood what a wonderful/excellent IDE Eclipse is, and also I understand how useful Apache Ant is, because I have been contributing code to that project. FYI, next Apache Ant formal release 1.7 will include my sftp task as an optional task. |PPS Sorry to be annoying, but are you going to make me completely |ecstatic and add equals() method to IdentityFile class as well? The latest IdentityFile has included following method, public boolean equals(Object o){ if(!(o instanceof IdentityFile)) return super.equals(o); IdentityFile foo=(IdentityFile)o; return identity.equals(foo.identity); } It is not so simple to check the equallity correctly, becuase its internal state(ecrypted or decrypted) will be changed. So, I have simply checked it according to where it is from. It must be enough for almost of all case. As other solution, I may restrict how many identity files are cached. In fact, OpenSSH's ssh client allow you to hadle only two keys(id_rsa and id_dsa). Sincerely, -- Atsuhiko Yamanaka JCraft,Inc. 1-14-20 HONCHO AOBA-KU, SENDAI, MIYAGI 980-0014 Japan. Tel +81-22-723-2150 +1-415-578-3454 Fax +81-22-224-8773 Skype callto://jcraft/ |
From: Andrei T. <an...@tc...> - 2006-05-24 02:12:11
|
Wonderful. Can not wait to get my hands on 0.1.29. Thanks again. Andrei On May 23, 2006, at 10:03 PM, Atsuhiko Yamanaka wrote: > Hi, > > +-From: Andrei Tchijov <an...@tc...> -- > |_Date: Tue, 23 May 2006 21:09:52 -0400 ______ > | > |PS Emacs is good, but you really should try Eclipse. I was > much like > |you (only I used to do it in Vi), but Eclipse's re-factoring > |capabilities are highly addictive (along with bunch of other > features > |which really make you work faster - no gimmick). And unlike many > |other IDE's overhead of bringing your projects under Eclipse > |"control" is minimal (especially if you are using Ant). > > Don't worry about it :-) When in Rome, do as the Romans do. > As a committer for Eclipse.org project, I'll use Eclipse in > contributing > plug-ins to Eclipse.org's CVS repository and I have understood what a > wonderful/excellent IDE Eclipse is, and also I understand how useful > Apache Ant is, because I have been contributing code to that project. > FYI, next Apache Ant formal release 1.7 will include my sftp task > as an optional task. > > |PPS Sorry to be annoying, but are you going to make me completely > |ecstatic and add equals() method to IdentityFile class as well? > > The latest IdentityFile has included following method, > > public boolean equals(Object o){ > if(!(o instanceof IdentityFile)) return super.equals(o); > IdentityFile foo=(IdentityFile)o; > return identity.equals(foo.identity); > } > > It is not so simple to check the equallity correctly, becuase its > internal > state(ecrypted or decrypted) will be changed. So, I have simply > checked it > according to where it is from. It must be enough for almost of all > case. > > As other solution, I may restrict how many identity files are cached. > In fact, OpenSSH's ssh client allow you to hadle only two keys > (id_rsa and > id_dsa). > > > Sincerely, > -- > Atsuhiko Yamanaka > JCraft,Inc. > 1-14-20 HONCHO AOBA-KU, > SENDAI, MIYAGI 980-0014 Japan. > Tel +81-22-723-2150 > +1-415-578-3454 > Fax +81-22-224-8773 > Skype callto://jcraft/ |
From: B. S. S. <sc...@sm...> - 2006-05-24 02:41:28
|
Hi ymnk, Atsuhiko Yamanaka wrote: > For your conveniences, I'll reformat the code with it at formal 0.1.29 > release timing, but please keep in your mind that, that format definition may > be changed and I'll not promise to reformat the code at every release. > I'm working a committer of Eclipse.org, but I'm not an Eclipse user. > I can not write the code without Emacs. > > At this point, I'd like to throw my 2 cents in as well. I also like to use basic text editors, although I am growing very fond of the NetBeans IDE. As you know, I also contribute patches and enhancements to JSch. In order to make the code more legible when I am working with it, I use jalopy. It was very easy for me to add to the existing build.xml, and the code comes out looking great in any editor. I simply made the following changes to the build.xml file: diff -r1.3 build.xml 14a15 > <property name="jalopy" location="jalopy"/> 35a37,57 > > <target name="jalopy" depends="compile"> > <echo>${jalopy}</echo> > <taskdef > name="jalopy" > classname="de.hunsicker.jalopy.plugin.ant.AntPlugin" > > > <classpath> > <fileset dir="${jalopy}" includes="**/*.jar" /> > </classpath> > </taskdef> > > <jalopy > convention="${jalopy}/jalopy.xml" > loglevel="warn" > threads="2" > > > <fileset dir="${src}" includes="**/*.java" /> > </jalopy> > </target> > 39c61 < <target name="dist" depends="compile" --- > <target name="dist" depends="jalopy" I created a "jalopy" folder that holds the following files: jakarta-oro.jar jalopy-ant.jar jalopy.jar jalopy.xml jdom.jar log4j.jar If you are interested, I would be happy to email these files to you directly, or send you the links for where to get these files directly. Thanks. |
From: <ym...@jc...> - 2006-05-24 03:09:37
|
Hi, +-From: "B. Scott Smith" <sc...@sm...> -- |_Date: Tue, 23 May 2006 22:40:53 -0400 ___________ | |At this point, I'd like to throw my 2 cents in as well. I also like to |use basic text editors, although I am growing very fond of the NetBeans |IDE. As you know, I also contribute patches and enhancements to JSch. In |order to make the code more legible when I am working with it, I use |jalopy. It was very easy for me to add to the existing build.xml, and |the code comes out looking great in any editor. I simply made the |following changes to the build.xml file: ... |If you are interested, I would be happy to email these files to you |directly, or send you the links for where to get these files directly. Thank you for your suggetions, but I'll not continue this issue anymore. It was tribial for me to occur such a flame war. I'll respect for your style, but every people has their own styles. It is immposible to make all of them to satisfy. Should I also include .emacs ? How about configurations for vi, IntelliJ, JBuilder and JEditor? If this issue continues on this list, I'll stop to reformat the code by Eclipse, to be fair for every other users. Sincerely, -- Atsuhiko Yamanaka JCraft,Inc. 1-14-20 HONCHO AOBA-KU, SENDAI, MIYAGI 980-0014 Japan. Tel +81-22-723-2150 +1-415-578-3454 Fax +81-22-224-8773 Skype callto://jcraft/ |
From: Andrei T. <an...@tc...> - 2006-05-24 03:19:17
|
End result will be pretty much the same regardless which tool was used to do reformatting. I say that Atsuhiko should use what ever is more convenient to him (take less of his time). I am sure that Jalopy is wonderful tool, but in this situation the important part is to get readable code. It does not really matter how. Let stop talk about it and let Atsuhiko finish 0.1.29. Andrei On May 23, 2006, at 11:09 PM, Atsuhiko Yamanaka wrote: > Hi, > > +-From: "B. Scott Smith" <sc...@sm...> -- > |_Date: Tue, 23 May 2006 22:40:53 -0400 ___________ > | > |At this point, I'd like to throw my 2 cents in as well. I also > like to > |use basic text editors, although I am growing very fond of the > NetBeans > |IDE. As you know, I also contribute patches and enhancements to > JSch. In > |order to make the code more legible when I am working with it, > I use > |jalopy. It was very easy for me to add to the existing > build.xml, and > |the code comes out looking great in any editor. I simply made the > |following changes to the build.xml file: > ... > |If you are interested, I would be happy to email these files to > you > |directly, or send you the links for where to get these files > directly. > > Thank you for your suggetions, but I'll not continue this issue > anymore. > > It was tribial for me to occur such a flame war. I'll respect for > your style, > but every people has their own styles. It is immposible to make > all of them > to satisfy. Should I also include .emacs ? How about > configurations for > vi, IntelliJ, JBuilder and JEditor? If this issue continues on > this list, > I'll stop to reformat the code by Eclipse, to be fair for every > other users. > > > Sincerely, > -- > Atsuhiko Yamanaka > JCraft,Inc. > 1-14-20 HONCHO AOBA-KU, > SENDAI, MIYAGI 980-0014 Japan. > Tel +81-22-723-2150 > +1-415-578-3454 > Fax +81-22-224-8773 > Skype callto://jcraft/ > > > ------------------------------------------------------- > All the advantages of Linux Managed Hosting--Without the Cost and > Risk! > Fully trained technicians. The highest number of Red Hat > certifications in > the hosting industry. Fanatical Support. Click to learn more > http://sel.as-us.falkag.net/sel? > cmd=lnk&kid=107521&bid=248729&dat=121642 > _______________________________________________ > JSch-users mailing list > JSc...@li... > https://lists.sourceforge.net/lists/listinfo/jsch-users |