New browser for OpenVMS in field test, Itanium only so far - VMS

This is a discussion on New browser for OpenVMS in field test, Itanium only so far - VMS ; Richard Maher wrote (excerpted): > Unfortunately the license fees for these moth-balled products are not cheap > and all of the new-feature-enabling revenue goes to DECforms (a product > which you don't want to touch with a barge-pole) > As ...

+ Reply to Thread
Page 4 of 4 FirstFirst ... 2 3 4
Results 61 to 62 of 62

Thread: New browser for OpenVMS in field test, Itanium only so far

  1. Re: Elvis is dead - get over it!

    Richard Maher wrote (excerpted):

    > Unfortunately the license fees for these moth-balled products are not cheap
    > and all of the new-feature-enabling revenue goes to DECforms (a product
    > which you don't want to touch with a barge-pole)
    >


    As one of the original developers of DECforms, I wonder why you warn
    people away from it?
    John Sauter (John_Sauter@systemeyescomputerstore.com)

  2. Re: Elvis is dead - get over it!

    On Oct 11, 9:07*pm, "Richard Maher"
    wrote:
    > Hi Paul,
    >
    > Congatulations with the software port and especially the *new* VMS
    > customers!
    >


    Thanks! It's a lot of hard work, but the reward is worth it. I'll put
    in a good plug for the DSPP as well.
    Not a lot of one on one support from them, but they do make the
    products available. Software products
    at least.

    > > And VMS really needs some kind of block mode terminal
    > > emulation...

    >
    > I am at a loss to know why anyone would want to do this. Having said that,
    > you may wish to look at the VMS Forms products: -
    >

    Actually, the reason is pretty simple; character by character telnet
    is not efficient, and never really has been. VT* terminals have always
    used it, and often to really good effect, but, especially over slow
    links, char by char transmissions are awful. You see things like half
    a screen refresh, pause, and then complete. Or a user who is
    typing can easily get ahead of the transmissions. That happens even
    on a well constructed LAN, as there are always moments where the
    network peaks.

    As you mentioned, I did look at those products, but they were both too
    expensive (for the customers) and also rather complex, and they still
    occasionally exhibit issues with screens pausing in the middle of a
    redraw.

    What I did was write a little server that accepts TN3270
    transmissions, negotiates with the terminals, and then converts a
    simple "screen buffer" into TN3270 data streams. Does the same for
    reading the strings.

    The program then manages just this little buffer (1924 plus a few
    overhead bytes for a 24X80 screen, a little bit more for larger screen
    sizes of course...), which is then sent and received via a normal TCP
    socket. All at one time.

    3270 will not redraw until it gets the entire transmission, so to the
    user's perception, there may be a slight delay before the screen
    refreshes, but when it refreshes, it does so at lightening speed.
    Psychological perhaps, as the total amount of time to refresh the
    screen is about the same, but they like it better.

    Okay, I cheated a little, it is connected into PortMapper, but the end
    result is that an application program is started when a connection is
    established and during it's lifetime, it communicates to the screen
    via this little old buffer. By specifying that certain hex values
    indicate things like modify, highlight, underline, color, and so
    forth, I can do pretty nice 3270 screens. And there are plenty of free
    TN3270 emulators out there.

    This is not deployed yet, I currently use a modified Putty telnet
    client, and yep, I am getting some complaints about screen refresh
    issues. I have to modify a bit of Cobol, some smaller amount of
    Fortran to make it all work. Fortunately, the screen handling code is
    all segregated and easy to change. I just keep finding new little
    "gotcha's" to pin down and fix.

    To move to the web is a great idea, and I have been recoding the
    interface into an Ajax driven set of screens. That is going pretty
    well too, and some customers want that pretty badly. (I'd like it in
    terms of running a service bureau...)
    But some customers just want the "green screen" bit, so- give 'em what
    they want.

    One advantage of the bit I described above is that no user ever logs
    into the native VMS system; security is handled at the application
    level, which is pretty good in a lot of ways. More coding to setup of
    course, but once working, it works nicely.

    This is also trivial to do under Linux.


    > . TDMS (which has *just* been ported to Alpha, and presumably Itanium, in
    > another amazing example VMS Management vacillations and ineptitude) TDMS was
    > sort of block-mode for ACMS (our CICS-like TP monitor)
    >
    > . FMS which has some form of IBM3270 emulation mode
    >
    > Unfortunately the license fees for these moth-balled products are not cheap
    > and all of the new-feature-enabling revenue goes to DECforms (a product
    > which you don't want to touch with a barge-pole)
    >
    > What does the product do? What language? Java or have you looked at the SMG$
    > run-time library?
    >
    > Why not use a browser for display?
    >
    > Cheers Richard Maher
    >
    > "PR" wrote in message
    >
    > news:899f1c14-d2f2-4ece-8c08-1748188ccf48@m32g2000hsf.googlegroups.com...
    > On Oct 10, 9:28 pm, David J Dachtera
    > wrote:
    >
    >
    >
    > > PR wrote:

    >
    > > > On Oct 5, 4:53 am, "Richard Maher"
    > > > wrote:
    > > > > Hi Michael,

    >
    > > > > > Then why do you bother with VMS at all?

    >
    > > > > 'cos it's the best server on the planet you ****ing idiot!

    >
    > > > > If only the incompetent, self-serving, vms-appologists at HP would

    > open the
    > > > > flood-gates and let the users *INTEGRATE* their VMS apps with this
    > > > > feature-rich, cheap, ubiquitous, full-function client platform(s) then

    > maybe
    > > > > we might get somewhere?

    >
    > > > > Regards Richard Maher

    >
    > > > > "Michael Kraemer" wrote in message

    >
    > > > >news:gc9spk$1lc$00$1@news.t-online.com...

    >
    > > > > > Richard Maher schrieb:

    >
    > > > > > > Personally, I love FireFox and Firebug! I love Flex (and

    > FlexBuilder)! I
    > > > > > > love .NET (a bit less)! Silverlight is getting better! I love Java
    > > > > Applets!
    > > > > > > (and the new 1.6_10 jnlp deployment options) I love Chrome's

    > Application
    > > > > > > Launch shortcuts! I love HTML/DOM/Javascript! I love the price of
    > > > > laptops,
    > > > > > > PCs and Macs! I have no problem with Windows! *And so do the

    > users!*
    >
    > > > > > Then why do you bother with VMS at all?

    >
    > > > Amen brother 0 thou do preach to the choir!

    >
    > > > I've said it before and I will say it yet again, from a dead zero
    > > > start, I've put in five VMS installations in a little less than a
    > > > year, including the time to convert the darn software, spec out RX26xx
    > > > servers, install em, and even - get this -
    > > > *get paid!*

    >
    > > O.k. You've told us THAT you did it. Well Done! Congrats!

    >
    > > Now, tell us HOW you did it.

    >
    > The how is simple, I already have software in the field, though it is
    > running on other platforms. One of my clients wanted to upgrade and
    > the platform they were on was just too expensive to stay on and
    > upgrade. *So they asked for an alternative, and yes indeed, they
    > suggested Windows.
    >
    > This is where it got really easy - I just put the facts, mostly
    > numbers, down on paper for them.
    >
    > The VMS based solution sold itself based purely on the numbers. This
    > particular customer has 110 terminals scattered around 4 locations,
    > and replacing all those with Windows based machines was a big cost
    > factor. I did replace them by the way, but with a very basic PC
    > running Linux. Total cost for the PC's and the network gear and
    > install was less than the cost of buying Windows and a Windows
    > terminal emulator. Also, no real problem with virii. Just everything
    > came together. (Now, I tore my hair out getting the software
    > converted. And VMS really needs some kind of block mode terminal
    > emulation... in fact I have that about 3/4 of the way done. It makes
    > the server responses seem lightening fast, even over a pretty slow
    > remote link.)
    >
    > And boy are they happy. My software looks and acts just about the same
    > under VMS and it does under z/OS or on an iSeries machine. Sweet. Much
    > lower cost to deploy.
    >
    > And when the owner of that particular company was out golfing one
    > afternoon, his golfing partner was complaining about his software
    > costs. Of course, he turned his partner on to me... and the first new
    > sale happened. It was smaller, about 50 PC's on an existing network. A
    > modified Putty took care of the terminal problems there. Another happy
    > customer.
    >
    > And so it went - 4 more times. And a couple existing customers are
    > wanting to move to the HP solution, as soon as I have time to do it.
    >
    > The how is easy.
    >
    > > > If 100 more people would get out there and evangelize VMS, a whole
    > > > hell of a lot of GOOD things would happen.
    > > > It is a GREAT server, and affordable.

    >
    > > Again, tell us HOW you did/do it.

    >
    > You talk to people, and you prove to them that VMS is not something
    > mysterious or arcane or unsupported or whatever. Mostly you TALK to
    > them. And there is no need to complain about how something or another
    > seems to have poor support - after all - YOU are going to support it
    > aren't you?
    >
    > Talk to them. Just tell them the truth - VMS is a GREAT OS and we can
    > use it to cost effectively put a solution in place for you. The only
    > catch is you have to believe in the OS and in your software.
    >
    > > Success leaves clues. Find the clues, document the process and make it
    > > REPEATABLE!

    >
    > It is more engineering than science, and actually, more art than
    > engineering. You just have to believe in what you are trying to sell
    > them.
    > If you don't believe in it, what they heck are you doing selling it in
    > the first place? It isn't ethical to do that to begin with.
    >
    > > > Now if they would just hurry up with 8.4 so I could run the thing in
    > > > an Integrity Virtual Machine, boy would I ever be a happy duck!

    >
    > > Why not run it on I64 "bare metal" and be even happier? (No PH-UX to
    > > deal with!)

    >
    > I do of course, but the machine I was thinking of is one of my
    > development machines, and like any small business, I have to squeeze
    > every last nickel, make everything as efficient as possible, and
    > control every single cost. It isn't any fun, but it keeps the business
    > from going bankrupt - or my credit card from maxing out!
    >
    > I use a single Itanium machine to run HP-UX, Linux, and Windows.
    > Running VMS as a VM would be more cost effective and efficient than
    > having to constantly boot into and out of VMS.
    >
    > -Paul
    >
    > > D.J.D.



+ Reply to Thread
Page 4 of 4 FirstFirst ... 2 3 4