Building libxml2 on OpenVMS/VAX - VMS

This is a discussion on Building libxml2 on OpenVMS/VAX - VMS ; Hi everybody, first off, happy new year to all comp.os.cms members. To my problem: Since 2 days I try to build the libxml2 library on my Vax (a VaxStation 4000/96 with OpenVMS 7.3 with Compaq C V6.4-005). I encounter some ...

+ Reply to Thread
Results 1 to 7 of 7

Thread: Building libxml2 on OpenVMS/VAX

  1. Building libxml2 on OpenVMS/VAX

    Hi everybody,

    first off, happy new year to all comp.os.cms members.

    To my problem: Since 2 days I try to build the libxml2 library on my
    Vax (a VaxStation 4000/96 with
    OpenVMS 7.3 with Compaq C V6.4-005).

    I encounter some problems that seem to be significant and I really do
    not know any solution to that.
    Compiling the sources gives me an informational message on each
    compiled object saying

    CC/NAMES=(SHORTENED)/FLOAT=D_FLOAT/object=DKA300:
    [AW.LIBXML2-2_6_30.DEBUG]HTMLTR
    EE.OBJ; HTMLTREE.C
    typedef long double trio_long_double_t;
    ........^
    %CC-I-LONGDOUBLENYI, In this declaration, type long double has the
    same
    representation as type double on this platform.
    At line number 156 in DKA300:
    [AW.LIBXML2-2_6_30]TRIODEF.H;1.

    A FAQ I found on the HP site just says: user action: be aware of that.
    Well ...

    The compilation procedure gets thru and results in the libxml.olb
    library. Another problem is, that I get one undefined symbol
    (fp_class). Overall, these problems seem to be resolvable but I have
    the string feeling, that
    I won't be able to build that library on OpenVMS /VAX. As far as I
    know it can be build on an Alpha. With the
    libxml2 distribution there's also an vms build procedure (which I
    use).

    I had to modify the compiler command, since my cc/decc version has no
    such compiler options, originally
    it was

    cc_opts = "/NAMES=(SHORTENED)/FLOAT=IEEE/IEEE_MODE=DENORM_RESULTS"

    and changed that to

    /NAMES=(SHORTENED)/FLOAT=D_FLOAT

    I still get those "%CC-I-LONGDOUBLENYI" messages and I really have no
    clue what to do/change to get
    a "clean" libxml.olb.

    My basic questions is: Did anybody manage to build the libxml2 library
    on OpenVMS/VAX? Are there
    any modifications I can do to get that library done?

    One more: It's been a little while I've programmed on OpenVMS, just
    switched on my VAX'es like 2 weeks ago and want to make them ready for
    further programming / porting software.

    I need the libxml2 library to install the perl libxml2 parser to
    implement a perl-driven xml/xslt homepage with the WASD http server.
    It will be another private hobbyist VMS homepage. Originally I wanted
    to build the
    Sablotron Parser, but I do not have an C++ compiler. Tried to install
    the c++ compiler but get an vmsinstal
    installtion error that reports some undefined logicals. So for now I
    leave the c++ installation for later.

    Any hints to get me to an result are greatly appreciated. If I forgot
    to give any important info, please let me know.

    Once again, thanks for you help!

    Andreas W Wylach

  2. Re: Building libxml2 on OpenVMS/VAX

    In article <28f0bbf8-f6c7-4417-90f2-0ea8aa489ece@i72g2000hsd.googlegroups.com>, "Andreas W. Wylach" writes:
    > %CC-I-LONGDOUBLENYI, In this declaration, type long double has the


    Note the form: "%fac-I-name, text". That "-I-" indicates
    informational message, not error. Since someone bothered to enter
    "long double" in the source, they might have thought it was different
    than "double". On some other platoform, maybe it is.

    >
    > The compilation procedure gets thru and results in the libxml.olb
    > library. Another problem is, that I get one undefined symbol
    > (fp_class). Overall, these problems seem to be resolvable but I have
    > the string feeling, that
    > I won't be able to build that library on OpenVMS /VAX. As far as I
    > know it can be build on an Alpha. With the
    > libxml2 distribution there's also an vms build procedure (which I
    > use).


    Either fp_class is there, or it isn't. Make sure you fix this.

    > I had to modify the compiler command, since my cc/decc version has no
    > such compiler options, originally
    > it was
    >
    > cc_opts = "/NAMES=(SHORTENED)/FLOAT=IEEE/IEEE_MODE=DENORM_RESULTS"
    >
    > and changed that to
    >
    > /NAMES=(SHORTENED)/FLOAT=D_FLOAT


    I know of no version of the DEC C compiler for Alpha which didn't
    accept /float=ieee. I'd pick IEEE or G_FLOAT over D_FLOAT. Or did
    I misread what your writing? If this is from an attempt to build
    on DEC C on VAX then there is not IEEE, and I'd use D_FLOAT.

    > I still get those "%CC-I-LONGDOUBLENYI" messages and I really have no
    > clue what to do/change to get
    > a "clean" libxml.olb.


    Since the message is just for your information, and not an error,
    the build is clean. If youreally want to, you can use "#ifdef __vms"
    to change the liine of code from "long double" to "double" for this
    platform.

    > One more: It's been a little while I've programmed on OpenVMS, just
    > switched on my VAX'es like 2 weeks ago and want to make them ready for
    > further programming / porting software.
    >
    > I need the libxml2 library to install the perl libxml2 parser to
    > implement a perl-driven xml/xslt homepage with the WASD http server.
    > It will be another private hobbyist VMS homepage. Originally I wanted
    > to build the
    > Sablotron Parser, but I do not have an C++ compiler. Tried to install
    > the c++ compiler but get an vmsinstal
    > installtion error that reports some undefined logicals. So for now I
    > leave the c++ installation for later.


    The C++ compiler installation should not depend on any unusual
    logical names. Make sure the compiler version you are installing
    supports the VMS version you are running.



  3. Re: Building libxml2 on OpenVMS/VAX

    On 7 Jan., 15:33, koeh...@eisner.nospam.encompasserve.org (Bob
    Koehler) wrote:
    > In article <28f0bbf8-f6c7-4417-90f2-0ea8aa489...@i72g2000hsd.googlegroups.com>, "Andreas W. Wylach" writes:
    >
    > > %CC-I-LONGDOUBLENYI, In this declaration, type long double has the

    >
    > Note the form: "%fac-I-name, text". That "-I-" indicates
    > informational message, not error. Since someone bothered to enter
    > "long double" in the source, they might have thought it was different
    > than "double". On some other platoform, maybe it is.


    I know ... i changed that to "double" only, erasing the "long".

    >
    >
    >
    > > The compilation procedure gets thru and results in the libxml.olb
    > > library. Another problem is, that I get one undefined symbol
    > > (fp_class). Overall, these problems seem to be resolvable but I have
    > > the string feeling, that
    > > I won't be able to build that library on OpenVMS /VAX. As far as I
    > > know it can be build on an Alpha. With the
    > > libxml2 distribution there's also an vms build procedure (which I
    > > use).

    >
    > Eitherfp_classis there, or it isn't. Make sure you fix this.


    I just try to get it compiled using the fallback solution in the code,
    means not having fp_class
    Taking a look in the fp_classh header file, it is only defined for
    Alpha ...

    >
    > > I had to modify the compiler command, since my cc/decc version has no
    > > such compiler options, originally
    > > it was

    >
    > > cc_opts = "/NAMES=(SHORTENED)/FLOAT=IEEE/IEEE_MODE=DENORM_RESULTS"

    >
    > > and changed that to

    >
    > > /NAMES=(SHORTENED)/FLOAT=D_FLOAT

    >
    > I know of no version of the DEC C compiler for Alpha which didn't
    > accept /float=ieee. I'd pick IEEE or G_FLOAT over D_FLOAT. Or did
    > I misread what your writing? If this is from an attempt to build
    > on DEC C on VAX then there is not IEEE, and I'd use D_FLOAT.



    I BUILD ON VAX .. like stated above several times, not Alpha. Also the
    headline of my posting says OPENVMS / VAX, does it?


    >
    > > I still get those "%CC-I-LONGDOUBLENYI" messages and I really have no
    > > clue what to do/change to get
    > > a "clean" libxml.olb.

    >
    > Since the message is just for your information, and not an error,
    > the build is clean. If youreally want to, you can use "#ifdef __vms"
    > to change the liine of code from "long double" to "double" for this
    > platform.
    >


    I know that ... it is not bothering me

    > > One more: It's been a little while I've programmed on OpenVMS, just
    > > switched on my VAX'es like 2 weeks ago and want to make them ready for
    > > further programming / porting software.

    >
    > > I need the libxml2 library to install the perl libxml2 parser to
    > > implement a perl-driven xml/xslt homepage with the WASD http server.
    > > It will be another private hobbyist VMS homepage. Originally I wanted
    > > to build the
    > > Sablotron Parser, but I do not have an C++ compiler. Tried to install
    > > the c++ compiler but get an vmsinstal
    > > installtion error that reports some undefined logicals. So for now I
    > > leave the c++ installation for later.

    >
    > The C++ compiler installation should not depend on any unusual
    > logical names. Make sure the compiler version you are installing
    > supports the VMS version you are running.


    I'll check that, that might be the reason .... I thought it was for my
    OpenVMS/VAX 7.3 but
    I'll double check that later on. Thanks for the hint.


  4. Re: Building libxml2 on OpenVMS/VAX

    In article <0fb01ceb-0031-4c01-83a4-1490c2fe8d2d@q39g2000hsf.googlegroups.com>, yyyc186 writes:
    >
    > You need to start with a known base and work forward. AMD 64-bit
    > chips are now the industry standard. Intel, to paraphrase Vin Diesel,
    > is the KG-used-to-Be of chip makers. Ubuntu is the direction the
    > Linux market is going. Red Hat is tainted by Oracle and SuSE is
    > suffering from Novel's continued fornication with MS.
    >
    > Moving forward from the current Ubuntu base, you add on Logicals,
    > ACL's, and clustering.


    VMS is VMS down to the kernel design. If you start with Ubuntu
    you may be able to get a close API, but lots of us will be able
    to distinguish it from VMS, no matter what you pile on top.


  5. Re: Building libxml2 on OpenVMS/VAX

    yyyc186 wrote:
    > Moving forward from the current Ubuntu base, you add on Logicals,
    > ACL's, and clustering. This will require some kernel gutting and a
    > break from the shared kernel of other distros. You will still be able
    > to use the Debian distribution packaging, but the installer will need
    > to be modified.


    It is by definition impossible to create VMS based on Ubuntu. The
    best you can hope for is a VMS look-a-like. And I doubt that is
    interesting for many.

    BTW, Ubuntu would be a very bad choice for starting distro - it
    is created for maximum functionality to compete with Windows.

    > Once a stable kernel with Logicals, ACL's, and clustering is in place,
    > you then set about developing a DCL shell with Lexical functions.


    Peoples definition of kernel varies, but I find it hard to
    come up with one that includes ACL's.

    Arne

  6. Re: Building libxml2 on OpenVMS/VAX

    yyyc186 wrote:
    > You need to start with a known base and work forward. AMD 64-bit
    > chips are now the industry standard. Intel, to paraphrase Vin Diesel,
    > is the KG-used-to-Be of chip makers.


    ????

    Try look at the calendar - it says 2008 not 2005.

    Intel is completely dominating the PC market at this time and
    AMD has huge problems.

    Arne

  7. Re: Building libxml2 on OpenVMS/VAX

    On 10 Jan., 02:21, Arne Vajh°j wrote:
    > yyyc186 wrote:
    > > Moving forward from the current Ubuntu base, you add on Logicals,
    > > ACL's, and clustering. This will require some kernel gutting and a
    > > break from the shared kernel of other distros. You will still be able
    > > to use the Debian distribution packaging, but the installer will need
    > > to be modified.

    >
    > It is by definition impossible to create VMS based on Ubuntu. The
    > best you can hope for is a VMS look-a-like. And I doubt that is
    > interesting for many.
    >
    > BTW, Ubuntu would be a very bad choice for starting distro - it
    > is created for maximum functionality to compete with Windows.
    >
    > > Once a stable kernel with Logicals, ACL's, and clustering is in place,
    > > you then set about developing a DCL shell with Lexical functions.

    >
    > Peoples definition of kernel varies, but I find it hard to
    > come up with one that includes ACL's.
    >
    > Arne


    Well, the Ubuntu distro is not the badest tho. Some monthes ago
    I tried to install Ubuntu on a USB stick, but didn't succeed, cause
    there was problems with swapping and some other stuff that didn't
    work out.
    So I ended up installing the regular debian on the 4GB stick, since I
    had
    not really the need with x-windows. It's still rocking and I have it
    with me all the time with some of my software.

    The main thing I like about ubuntu is the fact that it is based on
    debian. And when it comes to linux, debian is my Nr. 1
    choice. 3 of my development servers and one internet server
    in my company is debian. Never had any major problems!

    Andreas W. Wylach


+ Reply to Thread