OpenBSD Journal

Mozilla

Contributed by jose on from the modern-browsers dept.

Peter Strömberg writes: "The lizard runs dynamically! :-)

http://marc.theaimsgroup.com/?l=openbsd-ports-cvs&m=105294901802254&w=2 " Almost. As noted in the commmit message some work needs to be done, but overall this is a huge step forward. It builds and runs dynamically. Thanks, Peter, and thanks to everyone who worked hard to make this work.

(Comments are closed)


Comments
  1. By Schubert () on http://schubert.cx/

    cd /usr;
    export or set CVSROOT=anoncvs@favoritecvsserver:/cvs
    cvs update -P ports/www/mozilla
    or
    cvs co -P ports/www/mozilla
    comment out the BROKEN line in the Makefile.

    Oh and http://ftp.mozilla.org/ is probably faster than ftp://ftp.mozilla.org/

    Comments
    1. By mirabile () mirabile@bsdcow.net on http://MirBSD.BSDadvocacy.org/

      Hm. I don't like mozilla still, but congrats.
      Btw, the umlauts are fubar'd (i.e. not escaped,
      lynx displays "Strmberg")

      Comments
      1. By Bruno Rohée () bruno@rohee.com on mailto:bruno@rohee.com

        Your term must be FUBAR, just checked with lynx in a PuTTY terminal, umlauts displayed.

        Comments
        1. By mirabile () mirabile@bsdcow.net on http://MirBSD.BSDadvocacy.org/

          This is because your lynx (and lynx' defaults)are broken.

          You must set
          assume unknown charset = utf-8
          assume local charset = iso-8859-1

          on OpenBSD (on systems that support utf-8 natively,
          such as Windows 2000, you can use utf-8 for the
          latter as well).

          Or try it with an old Mac browser, they don't use
          latin1 (iso-8859-1 or windows-1252) either.

          And check the fucking source. If you don't see
          an ö [if that renders, ö] the source
          is broken.

          Comments
          1. By Bruno Rohée () bruno@rohee.com on mailto:bruno@rohee.com

            I think the default charset for a browser unless specified otherwise in the HTTP header is supposed to be iso-8859-1. Should lookup the relevants RFCs...

          2. By Bruno Rohée () bruno@rohee.com on mailto:bruno@rohee.com

            When no explicit charset parameter is provided by the sender, media subtypes of the "text" type are defined to have a default charset value of "ISO-8859-1" when received via HTTP. Data in character sets other than "ISO-8859-1" or its subsets MUST be labeled with an appropriate charset value. See section 3.4.1 for compatibility problems.
            Your settings broke lynx, but the default configuration is right.

            Comments
            1. By mirabile () mirabile@bsdcow.net on http://MirBSD.BSDadvocacy.org/

              The XHTML standard specifies that, if no characterset is names in either the

              or the

              headers, the user agent must assume utf-8 (or
              us-ascii).

              The RFC is for servers. If they don't know it's
              latin1 or something else, _they_ must send a
              HTTP header with the right content-type to me.

            2. By mirabile () mirabile@bsdcow.net on http://MirBSD.BSDadvocacy.org/

              The XHTML standard specifies that, if no characterset is names in either the <br> <br> or the <br> <br> headers, the user agent must assume utf-8 (or <br> us-ascii). <br> <br> The RFC is for servers. If they don't know it's <br> latin1 or something else, _they_ must send a <br> HTTP header with the right content-type to me. <br>

            3. By mirabile () mirabile@bsdcow.net on http://MirBSD.BSDadvocacy.org/

              The XHTML standard specifies that, if no characterset is names in either the
              <?xml version="1.0" encoding="foo"?>
              or the
              <meta http-equiv="content-type"
              content="text/html; charset=foo" />
              headers, the user agent must assume utf-8 (or
              us-ascii).

              The RFC is for servers. If they don't know it's
              latin1 or something else, _they_ must send a
              HTTP header with the right content-type to me.

              Comments
              1. By Bruno Rohée () bruno@rohee.com on mailto:bruno@rohee.com

                This applies on in case of reading a document from a file, not from HTTP, since HTTP explicitely states that specifying no charset in the HTTP header implies that it is in iso-8859-1. I don't see which logic could lead to think otherwise. A charset is always specified in HTTP, be it explicit or implied.

                Comments
                1. By mirabile () mirabile@bsdcow.net on http://MirBSD.BSDadvocacy.org/

                  This does however not apply to XHTML parsers, sincethe XHTML standard specifies it differently, and
                  since the standard is newer - you get it.

                  --

                  Let's draw a line here. You're right, but I'm
                  correct as well. We have two standards which
                  conflict in one position, and to choose one is
                  a task one has to do for himself.
                  I still believe that, if the charset is not
                  explicitly given, it should be stripped even
                  to 7bit ASCII (no EBCDIC :), and configured my
                  sendmail this way too (voluntarily breaking
                  RFC2821, not 821 tho).

                  Let's focus on more important things, okay? :)

    2. By Anonymous Coward () on

  2. By Anonymous Coward () on

    3.3 or -current? Also, does this mean that Firebird (Phoenix) will run dynamically too?

    Comments
    1. By Anonymous Coward () on

      I compiled it on 3.3 today. No problems so far.
      But I wish https support would be included in the port...

    2. By francois briere () francois@dynasty.ath.cx on mailto:francois@dynasty.ath.cx

      Yep it work on -current ELF binary dynamically linked.
      And, compilation is now really fast, 41 minutes
      without mail/news '--disable-mailnews' on a duron 1ghz.

      http://dynasty.ath.cx:81/~francois/current_moz.jpg

  3. By schubert () on http://schubert.cx/

    http://schubert.cx/tmp/obsd_moz.jpg

    Make sure you get all the patches (the cvs mirror I normally use doesn't have all the latest checkins because some of them are only hours old)

  4. By schubert () on http://schubert.cx/

    The port seems to be working very well now. It is no longer marked as broken, and is now built with --enable-crypto and tests are disabled.

    One caveat it seems that --enable-extensions is not used and that is neccessary for the Personal Security Manager (PSM) to be built which is neccesary for accessing SSL sites and imap servers. You can easily add this to the Makefile.

    Comments
    1. By Peter Strömberg () on

      If you don't use --enable-extensions or --disable-extensions you'll get the default extensions.

      All --enable-extensions does is to add 'xmlterm access-builtin datetime finger cview' to the extension list.

      Comments
      1. By schubert () on http://schubert.cx/

        I've bashed my head against a wall a few times wondering why after a build with --enable-crypto and you go to use something with SSL and mozilla gives you the error "You must have the Personal Security Manager (PSM) installed to..." message and then gone back and compiled with --enable-extensions and poof it works.

        But I'll take your word on it.

        It seems odd if you don't use --enable-extensions you get the default ones but if you add it.. you get the default ones PLUS the ones you listed above... you'd think --enable-extensions alone would just give you the default extensions and nothing more but then again this is mozilla we're talking about here so logic can probably be discarded.

  5. By Anonymous Coward () on

    Anyone tried the port on 3.3?
    Is there a phoenix flavor?

    Comments
    1. By schubert () on http://schubert.cx/

      Yes, it does work on 3.3. And as of 05/18 or so, there is a port of the new firebird 0.6. Whether or not that works I'm not sure yet.

      Comments
      1. By Anonymous Coward () on

        Ok I installed the mozilla-firebird port but that is just mozilla 1.4b. I really like Phoenix a whole lot better :-(

        Is there a way to get that running?

        Comments
        1. By schubert () on http://schubert.cx/

          Doh! and I went to sleep with my build still going and you're right it is. What needs to be done is MOZ_PHOENIX="1" needs to be set in the Makefile at CONFIGURE_ENV. Configure args probably also need:
          --enable-extensions=default,-inspector,-irc,-venkman,-content-packs,-help
          and --disable-mailnews and --enable-plaintext-editor-only (these are all from the extremely hard to find phoenix compile howto on mozilla.org)

          Comments
          1. By Anonymous Coward () on

            No go. Just makes mozilla again.

            Dang a FLAVOR would be awesome here....

        2. By Peter Strömberg () on

          Fuckup by me, now the real firebird is built and installed

        3. By Anonymous Coward () on

          I applied the patches and build Firebird 0.6 (dynamically), but it currently segfaults (silently) on startup -- at least on my machine.

          Comments
          1. By Anonymous Coward () on

            It worked for me but there was no run script. I had to use ldconfig for all libs and run the binary directory.

            Comments
            1. By Peter Strömberg () wilfried@spamcop.net on mailto:wilfried@spamcop.net

              The script is called MozillaFirebird (the default name)

              And you don't need to ldconfig, you can use LD_LIBRARY_PATH=. ./MozillaFirebird-bin instead.

              Comments
              1. By Anonymous Coward () on

                Yep, found it :-)

                Maybe an idea to change it to firebird or mozillafirebird instead. I was simply not expecting the caps that's why I created my own script (and then found the one provided!).

Latest Articles

Credits

Copyright © - Daniel Hartmeier. All rights reserved. Articles and comments are copyright their respective authors, submission implies license to publish on this web site. Contents of the archive prior to as well as images and HTML templates were copied from the fabulous original deadly.org with Jose's and Jim's kind permission. This journal runs as CGI with httpd(8) on OpenBSD, the source code is BSD licensed. undeadly \Un*dead"ly\, a. Not subject to death; immortal. [Obs.]