STATUS

   1
   2   OpenSSL STATUS                           Last modified at
   3   ______________                           $Date: 1999/03/14 01:16:42 $
   4
   5   DEVELOPMENT STATE
   6
   7     o  OpenSSL 0.9.2: Freezed!
   8        - No more feature commits, please.
   9        - Commit bugfixes and cleanups only, please.
  10        - Already successfully tested platform:
  11          o  FreeBSD 2.2.7    Ben Laurie            OK
  12          o  FreeBSD 3.1      Ralf S. Engelschall   OK
  13          o  Solaris 2.6      Ralf S. Engelschall   OK except for ctype warns
  14          o  Linux RH 5.1     Steve Henson          OK
  15          o  Linux 2.0 RH 5.2 Ralf S. Engelschall   OK
  16          o  Linux 2.2 RH 5.2 Peter 'Luna' Altberg  OK
  17          o  WinNT,95 VC++ 5  Steve Henson          OK
  18
  19        Proposed release time: Monday, March 15th 1999
  20
  21     o  OpenSSL 0.9.1c: Released on December 23th, 1998
  22
  23   RELEASE SHOWSTOPPERS
  24
  25     o Compilation warnings: ctype-related int vs. char
  26     o Compilation error: "unsigned long*" and "int*" under AIX
  27     o Why are NULL ciphers in front of the cipher spec?
  28       (see my posting "Bug?!: Cipher Suite and NULL Ciphers" in openssl-dev)
  29     o A Linux alpha user reported: "alpha.s does not have bn_div_word()"
  30       (see posting "ALPHA error" in openssl-dev)
  31     o Undefined BN symbols in assembler stuff on Solaris
  32       (see posting "openssl-SNAP-19990308-2130 on Solaris-2.6 SC4" in openssl-dev)
  33
  34   AVAILABLE PATCHES
  35
  36     o Solaris AS error (kenji@miyake.org)
  37     o getenv in ca.c and x509_def.c (jaltman@watsun.cc.columbia.edu)
  38     o s3_lib.c - export clients (levitte@stacken.kth.se)
  39     o linux dynamic libs (colin@field.medicine.adelaide.edu.au)
  40     o MingW support (niklas@canit.se)
  41     o crypto/des/enc_read.c bugfix (mike@cs.mun.ca)
  42
  43   IN PROGRESS
  44
  45     o Steve is currently working on:
  46         X509 V3 extension code including:
  47         1. Support for the more common PKIX extensions.
  48         2. Proper (or at least usable) certificate chain verification.
  49         3. Support in standard applications (req, x509, ca).
  50         4. Documentation on how all the above works.
  51         Next on the list is probably PKCS#12 integration.
  52
  53     o Mark is currently working on:
  54         Folding in any changes that are in the C2Net code base that were
  55         not in the original SSLeay-0.9.1.b release.  Plus other minor
  56         tidying.
  57
  58     o Ralf is currently working on:
  59         1. Support for SSL_set_default_verify_paths(),
  60            SSL_load_verify_locations(), SSL_get_cert_store() and
  61            SSL_set_cert_store() functions which work like their existing
  62            SSL_CTX_xxx() variants but on a per connection basis. That's needed
  63            to let us provide full-featured per-URL client verification in
  64            mod_ssl or Apache-SSL.
  65            => It still dumps core, so I suspend this and investigate
  66               again for OpenSSL 0.9.3.
  67         2. The perl/ stuff to make it really work the first time ;-)
  68            => I'll investigate a few more hours for OpenSSL 0.9.2
  69         3. The new documentation set in POD format under doc/
  70            => I'll investigate a few more hours for OpenSSL 0.9.2
  71         4. More cleanups to get rid of obsolete/old/ugly files in the
  72            source tree which are not really needed.
  73            => Done all which were possible with my personal knowledge
  74
  75     o Ben is currently working on:
  76         1. Function Prototype Thought Police issues.
  77         2. Integrated documentation.
  78         3. New TLS Ciphersuites.
  79         4. Anything else that takes his fancy.
  80
  81   NEEDS PATCH
  82
  83   OPEN ISSUES
  84
  85     o  The Makefile hierarchy and build mechanism is still not a round thing:
  86
  87        1. The config vs. Configure scripts
  88           It's the same nasty situation as for Apache with APACI vs.
  89           src/Configure. It confuses.
  90           Suggestion: Merge Configure and config into a single configure
  91                       script with a Autoconf style interface ;-) and remove
  92                       Configure and config. Or even let us use GNU Autoconf
  93                       itself. Then we can avoid a lot of those platform checks
  94                       which are currently in Configure.
  95
  96        2. The xxx.org -> xxx.h generation:
  97           It's not obvious for which file xxx.org is the source.
  98           Suggestion: Rename xxx.org to xxx.h.in (Autoconf style), this way
  99                       one sees that xxx.h.in is the input for xxx.h
 100
 101           Status: Mark +1
 102
 103     o  The installation under "make install" produces a very
 104        installation layout: $prefix/certs and $prefix/private dirs.  That's
 105        not nice. Ralf suggests to move the two certs and private dirs either
 106        to $prefix/etc/, $prefix/lib/ or $prefix/share. Alternatively
 107        we could also not install the certs at all.
 108
 109        Status: Ralf +1 for both not installing the certs at all and
 110                        moving it to $prefix/etc/. +0 for $prefix/lib/
 111                        and $prefix/share.
 112                Paul: why is it not nice?
 113                Ralf: because it messes up the install dir when
 114                      $prefix is not a dedicated area like /usr/local/ssl.
 115                      When we move them to a standard subdir like
 116                      etc/ lib/ or share/ we don't mess up things
 117                      when $prefix is /usr or /usr/local, etc.
 118                      Additionally it makes package vendors life
 119                      easier....
 120
 121     o  Support for Shared Libraries has to be added at least
 122        for the major Unix platforms. The details we can rip from the stuff
 123        Ralf has done for the Apache src/Configure script. Ben wants the
 124        solution to be really simple.
 125
 126        Status: Ralf will look how we can easily incorporate the
 127                compiler PIC and linker DSO flags from Apache
 128                into the OpenSSL Configure script.
 129
 130     o  The perl/ stuff needs a major overhaul. Currently it's
 131        totally obsolete. Either we clean it up and enhance it to be up-to-date
 132        with the C code or we also could replace it with the really nice
 133        Net::SSLeay package we can find under
 134        http://www.neuronio.pt/SSLeay.pm.html.  Ralf uses this package for a
 135        longer time and it works fine and is a nice Perl module. Best would be
 136        to convince the author to work for the OpenSSL project and create a
 137        Net::OpenSSL or Crypt::OpenSSL package out of it and maintains it for
 138        us.
 139
 140        Status: Ralf thinks we should both contact the author of Net::SSLeay
 141                and look how much effort it is to bring Eric's perl/ stuff up
 142                to date.
 143                Paul +1
 144
 145     o The EVP and ASN1 stuff is a mess. Currently you have one EVP_CIPHER
 146       structure for each cipher. This may make sense for things like DES but
 147       for variable length ciphers like RC2 and RC4 it is NBG. Need a way to
 148       use the EVP interface and set up the cipher parameters. The ASN1 stuff
 149       is also foo wrt ciphers whose AlgorithmIdentifier has more than just
 150       an IV in it (e.g. RC2, RC5). This also means that EVP_Seal and EVP_Open
 151       don't work unless the key length matches the fixed value (some vendors
 152       use a key length decided by the size of the RSA encrypted key and expect
 153       RC2 to adapt).
 154
 155     o Properly initialize the PRNG in the absence of /dev/random.
 156
 157     o > NO_RSA (ejs@bfd.com)
 158       > ./Configure -DNO_IDEA -DNO_RC5 -DNO_RC4 -DNO_RC2 -DNO_RSA -DNO_ERR linux-elf
 159       > I tried for a whole day to do this and could not get it to work.  Linux
 160       > machine, kernel 2.0.36 and 2.2.1, redhat 5.2 latest, gcc and egcs , no
 161       > go.  I also noticed the even with -DNO_IDEA, _DNO_RC2, etc. the make
 162       > still goes into those subdirectories and 'makes'.
 163
 164   WISHES
 165
 166     o  Damien Miller:
 167        "How about making the each of the locations compile-time defined. I
 168        would like to (for example) put binaries in /usr/bin, configuration
 169        data, certs and keys in /etc/openssl/certs and /etc/openssl/keys, etc.
 170        This would also be a great boon to binary package makers.  The
 171        SSLeay-0.9.1b RPM already includes some patches which do some of this.
 172        I can forward them if you wish."
 173
 174     o  Mats Nilsson <mats.nilsson@xware.se>:
 175        "Add reference counting to all substructures of X509 etc. For instance,
 176        X509_NAME lacks a reference counter, while EVP_PKEY has one.  I'm
 177        making COM-wrappers for selected parts of SSLeay for a project of ours,
 178        and has found this inconsistency in copy semantics annoying."
 179