RE: Occasional ssl23_read() failure

2003-12-31 Thread Francis.Vanhemmens
Well when you stay on F5 key, you force Internet Explorer to close the current connections and recreate new ones. ( It may even multiply the number of simultaneous connections as well. You can end up with 20 simultaneous connections for example, instead of the normal 2 to 4 ones ) Some maybe be

RE : OPenSSL Book ?

2003-12-17 Thread Francis.Vanhemmens
Can't you just rent it or something ? (Library, University ...) I dont know any pdf book on openssl around. <>

RE: Openssl API ?

2003-12-16 Thread Francis.Vanhemmens
I recommend this book, it's easy to read. And there is a explained sample doing what you ask in it. Network security with OpenSSL O'Reilly John Viega, Matt Messier & Pravir Chanda ISBN : 0-596-00270-X -Original Message- From: Jajang Kavita [mailto:[EMAIL PROTECTED] Sent: mardi 16 décembr

RE : SSL_accept error

2003-11-14 Thread Francis.Vanhemmens
Did you do something like : SSL_METHOD *meth = NULL; .// do some inits, maybe even mutlithread support . // load error strings, load algorithms . meth = SSLv3_server_method(); // operational implementation of accept is in there. if (SSL_CTX_new(meth)) if you are both server and

RE: PEM pass phrase hard coded

2003-11-13 Thread Francis.Vanhemmens
I had such a request to have such server start alone without human intervention. If you want a little more secure solution. You can for example, hardcode a 3DES key in your server, to encrypt/decrypt the PEM pass phrase, and eventually mess the result a little (but in a way you can reconstruct

RE: SSL_write() must be done twice to work properly

2003-11-13 Thread Francis.Vanhemmens
Did you try to adjust the size of buffers you write/read on both sides ? For example to 4096 bytes. I had such sort of problem in some of my C/S applications. If i remember correctly, for example, i was sending 5k, and reading by 4k, first read was done ok, then select say there are data to read

ASN1_sign, ASN1_verify

2003-11-04 Thread Francis.Vanhemmens
Hi everybody, i have a strange behavior ... (while investigationg it i noticed that i2d_ASN1_BIT_STRING shorten a bit string if there are null bytes at the end of the string which it was not doing in older versions.) But that's not why i write here : I upgraded a server application from an old