Conference irocz::terminal_servers

	again...

	I'm using ds700,ds900 with ACB callback server software
	(that provide slip/ppp service for our internal group).
	I desire to add an additional server, ds90m in the system
	however i've losten NAS V1.5 ECO2 binary kit.
	v1.5 eco4 and up are unsuitable version for me (required
	function i need removed(or corrected) from v1.5 eco4 and up).
	so, i need v1.5 eco2 that currently working on ds700/900).
	i need your help.
	Thanks in advance.
			shinozaki/digital tokyo

RE: .1

>	v1.5 eco4 and up are unsuitable version for me (required
>	function i need removed(or corrected) from v1.5 eco4 and up).

What "feature" are you reliant on, that's unique to ECO 2 and ECO3?  Are
you planning to run ECO2 forever?

We normally consider ECO releases to be cumulative, and threfore do not
keep the older ones online.

Regards,

Dave

	re:.2

	Hi, Dave
	Thank you for information.

	>>> What "feature" are you reliant on, that's unique to
	>>> ECO 2 and ECO3?
	I execute steps to call back,
	  1.define the port characteristics of DECserver for reverse LAT
	    and enable modem control. default protocol is PPP.
	  2.dial thru the port.
	  3.after modems are connected, send service messages thru port.
	  4.set the port of DECserver to local/PPP to begin PPP session.
	    TSM> set port x modem control disable
	    TSM> def port x access local
	    TSM> def port x modem control enable
	    TSM> log port x
	    TSM> def port x access remote
	  On V1.5 ECO4 and up, When execute above procedure, the line
	  dropped and attached process of LTA terminal is logged out
	  even if modem control disabled.
	  Is that behaviour result of security enhancement or bug?
	  V1.5 ECO2 and earlier never drop the line. 

	>>> Are you planning to run ECO2 forever?
	  It is only too true (however, if future DNAS provide callback
	  feature for VMS/NT/95 stations, I will replace my custom
	  software to dnas/asm).

	>>> We normally consider ECO releases to be cumulative,
	>>> and threfore do not keep the older ones online.
	I'm waiting anyone who presents the kit NAECO2015.A for my ds90m.

		regards,
			f.shinozaki

RE: .3

>	I execute steps to call back,
>	  1.define the port characteristics of DECserver for reverse LAT
>	    and enable modem control. default protocol is PPP.
>	  2.dial thru the port.
>	  3.after modems are connected, send service messages thru port.
>	  4.set the port of DECserver to local/PPP to begin PPP session.
>	    TSM> set port x modem control disable
>	    TSM> def port x access local
>	    TSM> def port x modem control enable
>	    TSM> log port x
>	    TSM> def port x access remote

Ugh.  What a "hack"...  but I guess if it works for you...

>	  On V1.5 ECO4 and up, When execute above procedure, the line
>	  dropped and attached process of LTA terminal is logged out
>	  even if modem control disabled.
>	  Is that behaviour result of security enhancement or bug?

Well, I don't know, actually.  My guess is that it's a side effect.  I'm
pretty sure we don't routinely do any testing of the scenario you are using.
I'd be hard pressed to say that a modem should _ever_ have remained attached
when you perform a LOGOUT PORT x command.  And certainly you shouldn't expect
the LAT service to remain connected when you logout the port.  Maybe you were
taking advantage of a "bug" that we fixed?

>	>>> Are you planning to run ECO2 forever?
>	  It is only too true (however, if future DNAS provide callback
>	  feature for VMS/NT/95 stations, I will replace my custom
>	  software to dnas/asm).

DNAS V2.2 with CBCP to support Microsoft clients will be out this summer.

Regards,

Dave

  Hi,

  I think the image your looking for can be copied from NPSS::PUBLIC$DSK:
  [INTERCHANGE.DECSERVERS.DNAS]MNENG2_V15_Bl95B.SYS.

  As Dave pointed out it's no longer supported so we won't be able to make
  any corrections if you experience problems with it. Also, it will be deleted
  from the area tomorrow (March 27) so please copy it as soon as possible.

  Could you please post the exact steps and sequence of commands your using?
  I'm still a little confused as to what your doing and I'd like to make sure
  we didn't inadvertently break something.  I'm not sure what you mean by
  "dial thru the port", "send service messages thru port".  In step 4 are you
  logging out the port that your currently logged into?   It sounds like the
  you were taking advantage of a bug in earlier code but I'd like to know what
  the bug was.

          2.dial thru the port.
          3.after modems are connected, send service messages thru port.
          4.set the port of DECserver to local/PPP to begin PPP session.
            TSM> set port x modem control disable
            TSM> def port x access local
            TSM> def port x modem control enable
            TSM> log port x
            TSM> def port x access remote


  bob richard

  NPSS::PUBLIC$DSK:[INTERCHANGE.DECSERVERS.DNAS] also contains a DS700/DS900
  version of the code. It's called WWENG2_V15_BL95B.SYS.

    re:-1
    >>> I think the image your looking for can be copied from
    Thx. I obtained the firmware image from your pointer.
    	/shin

	re:.4

	>>> Ugh.  What a "hack"...  but I guess if it works for you...
	It is not "hack".
	Terminal server say;

	Local> help set port modem
	MODEM CONTROL specifies whether the server manipulates modem signals.
	Disable MODEM CONTROL when a port is connected to a device that does not
	support modem signals.

	So, V1.5 ECO2 is correct and V2.0 is incorrect which should be
	corrected.
		/shin

RE: .8

	> It is not "hack".

	Oh, I beg to differ.  Just because it works for you, and does something
	that you desparately need, doesn't mean it's not a hack.  I just means
	that it's a hack that's near and dear to your heart.

	> Local> help set port modem
	> MODEM CONTROL specifies whether the server manipulates modem signals.
	> Disable MODEM CONTROL when a port is connected to a device that does
	> not support modem signals.

	Yeah, so?  Nothing in the foregoing implies that you should be able to
	manipulate the MODEM CONTROL status of a port to avoid the modem	
	dropping carrier when you logout the port.  It says that you use this
	port characteristic to match the port to the device attached.  The
	modem always _supports_ modem control signals.  Using this port 
	characteristic to sometimes suppress the effect of port logouts is
	a "hack".  From your perspective, a clever and useful one.

	I suggest that in the long term, callback could better be accomplished
	using legitimate callback protocols (e.g. RADIUS, PPP LCP Callback PPP
	Callback Control Protocol).  You see, I happen to think that all of 
	ACB is a "hack"!  But that's just my viewpoint.

	> So, V1.5 ECO2 is correct and V2.0 is incorrect which should be
	> corrected.
	
	I respectfully disagree.  Maybe Bob Richard will find your arguments
	compelling; I do not.  But then Bob is the one who issues the ECOs,
	not me.

	Regards,

	Dave

  Logging out a port is always expected to cause DTR to be momentarily 
  de-asserted  regardless of the setting of Modem Control.  The difference is
  that with Modem Control enabled it's de-asserted for 5 seconds and with 
  Modem Control disabled it's de-asserted for 100 ms.  We'll look into why
  different versions behave differently.