[comp.protocols.appletalk] Pacer vs. Alisa

lui@CS.UCLA.EDU (04/12/88)

We are connecting our VAXes to our Macintoshes through the very popular
Kinetics FastPath gateway box. I'm comparing the two popular software
packages made by Alisa Systems and Pacer. We have a cluster of an 11/780
and an 8350, plus a LPS-40 Laser Printer.

Does anyone have any comments about either system? The LPS-40 is an
ethernet based printer; one of our main objectives is to make this
printer available to our Macintoshes.


	Stephen Lui
	UCLA Department of Computer Science and

	System Manager
	Hughes Aircraft Company
	Radar Systems Group

Physical Address:
	Centinela Ave. & Teale St.
	Culver City, CA.
	(213) 305-2085

Mailing Address:
	Hughes Aircraft Company
	M/S RC R49 2563
	P.O. Box 92426
	Los Angeles, CA
	90009-2426

	ARPA:  lui@cs.ucla.edu
	UUCP:  ...!{cepu,ihnp4,trwspp,ucbvax}!ucla-cs!lui
	Stephen Lui

	ARPA:  lui@cs.ucla.edu
	UUCP:  ...!{cepu,ihnp4,trwspp,ucbvax}!ucla-cs!lui

bmh@demon.siemens-rtl (Beatrice M Hwong) (04/12/88)

In article <11134@shemp.CS.UCLA.EDU> lui@CS.UCLA.EDU (Stephen Lui) writes:
>We are connecting our VAXes to our Macintoshes through the very popular
>Kinetics FastPath gateway box. I'm comparing the two popular software
>packages made by Alisa Systems and Pacer. We have a cluster of an 11/780
>and an 8350, plus a LPS-40 Laser Printer.
>
>Does anyone have any comments about either system? The LPS-40 is an
>ethernet based printer; one of our main objectives is to make this
>printer available to our Macintoshes.
>
>


Please post the results.  I think it is of general interest to all who
are concerned with Mac-VAX connectivity.

Beatrice M. Hwong
Siemens RTL
105 College Rd E.
Princeton, N.J. 08540
 609-734-3384

bmh%demon.siemens.com@princeton.edu

	

garrett@udel.EDU (Joel Garrett) (04/13/88)

In article <11134@shemp.CS.UCLA.EDU> lui@CS.UCLA.EDU (Stephen Lui) writes:
>We are connecting our VAXes to our Macintoshes through the very popular
>Kinetics FastPath gateway box. I'm comparing the two popular software
>packages made by Alisa Systems and Pacer.

I'm in the process of doing the same thing for my employer.

>Does anyone have any comments about either system?

I've only had a chance to dig into AlisaTalk so far, as I am still waiting for
a PacerShare beta tape (yes, it's an announced product, but it's still in
beta, I guess that's come to be the norm)

I ran into a few problems getting both packages to run at first (partially
due to incomplete installation instructions, partially due to voodoo
system problems)

One note to users of Wollongong WIN/TCP who want to get AlisaTalk or PacerLink:

If you are using the dedicated DEUNA driver for WIN/TCP, you're going to need
to call Wollongong and tell them you need to use the shared DEUNA/DEQNA
driver.  This isn't a big deal, all you need to do is get a new key from
TWG, which they'll read you over the phone, and then re-link the inet stuff
to include the shared DEUNA driver.  After this, just edit your startinet.com
file and the syconfig.com file and you're all set - For the dedicated deuna
driver, sysgen leaves out the XE driver while autoconfiguring the system at
boot time and both AlisaTalk and PacerLink need this.

Before I got this driver change from Wollongong, I tried installing AlisaTalk
with WIN/TCP not running.  For some reason the appletalk/vms bridge process
was unable to get the deuna's ethernet address and therefore was responding
to AppleTalk ARPs by saying "My ethernet address is FF.FF.FF.FF.FF.FF," which
is an ethernet broadcast address - not good.  This caused on the order of
about 10-20 ethernet broadcast packets per second to flood the net while I
was using the AlisaTalk file server.  Once I had the shared DEUNA WIN/TCP up
and running this problem mysteriously went away.

Both Alisa and Pacer provide printer spooling software that takes advantage
of VMS queuing services, AlisaPrint and PacerPrint.  Alisa also provides a
DEC printing package that will allow Mac users to access the LPS-40 and the
LN03-R.  I haven't dug much into that package, as we don't have any DEC
printers at our site.  I do know that the LPS-40 supports most of the
things that a LaserWriter can do (not including smoothing though, apple
handles that with 68000 code embedded in the postscript - or so I hear)

The one big difference I see between the two printing systems is how they
handle LaserWriters or other LocalTalk-connected printers.  AlisaPrint allows
you to leave your printers out on the LocalTalk net where they can still be
accessible to Mac users if the VAX should crash or have to be shut down for
PM, etc.  PacerPrint requires that the LaserWriters be connected directly
to the VAX with serial connections.  Granted, this solution probably yields
the best performance, but you lose a little flexibility in where you can
connect your printers.

As far as overall performance is concerned, Alisa's stuff seems to be a
little slower than Pacer's (all results aren't in, this test isn't complete)
which might be attributed to the way they implement their system, which
appears more flexible.

That's about all I've gotten so far - once AppleShare compatiblity is
available for both systems which should be in the near future (but how
near? :-)  the file servers for both systems will be greatly improved.

					Joel Garrett
					Research Associate
					CCM
					University of Delaware
					Newark, De  19716

					garrett@udel-ccm.arpa

bmh@demon.siemens-rtl (Beatrice M Hwong) (04/14/88)

In article <2044@louie.udel.EDU> garrett@udel.EDU (Joel Garrett) writes:
>In article <11134@shemp.CS.UCLA.EDU> lui@CS.UCLA.EDU (Stephen Lui) writes:
>>We are connecting our VAXes to our Macintoshes through the very popular
>>Kinetics FastPath gateway box. I'm comparing the two popular software
>>packages made by Alisa Systems and Pacer.
>
>
>>Does anyone have any comments about either system?
>
>
>As far as overall performance is concerned, Alisa's stuff seems to be a
>little slower than Pacer's (all results aren't in, this test isn't complete)
>which might be attributed to the way they implement their system, which
>appears more flexible.
There is a comparison of AlisaTalk, PacerShare, and CommUnity in Digital
Review (February 8, 1988) on p70.  The comparisons show that the Pacer is
either equal or faster than the other two using either EtherTalk or FastPath.
A disadvantage of Alisa is that it does not maintain HFS.
Experience here has been with PacerLink which needed a couple of minor revisions before it was working as promised.  Our PacerLink ties Ultrix on a Vax8600 to
Appletalk.
Beatrice M. Hwong
Siemens RTL
105 College Rd E
Princeton, N.J. 08540
609-734-3384
bmh%demon.siemens.com@princeton.edu

MANAGER@SKIDMORE.BITNET (Leo Geoffrion) (04/15/88)

bmh%demon.siemens.com@princeton.edu  writes:

>There is a comparison of AlisaTalk, PacerShare, and CommUnity in Digital
>Review (February 8, 1988) on p70.  The comparisons show that the Pacer is
>either equal or faster than the other two using either EtherTalk or FastPath.
>A disadvantage of Alisa is that it does not maintain HFS.
>Experience here has been with PacerLink which needed a couple of minor
>revisions
> before it was working as promised.  Our PacerLink ties Ultrix on a Vax8600 to
>Appletalk.

The edition of Alisa reviewed in that magazine is the old version which was not
HFS compatible.  At DEXPO, they were demonstrating their new version which is
fully AppleShare & HFS compatible.  This version is in Beta testing and is
supposted to be shipping in May or June according to the reps' at DEXPO.

We've been trying to decide between the two products for a while now, too.
Based on the literature that we've seen, we are leaning toward Alisa (the new
version...) for the following reasons:

   1.  It fully implements AppleShare for VMS, while Pacer uses proprietary
       code.  We fear that adopting Pacer may hurt us as other companies
       begin to implement Appleshare/VMS gateways.  Since AS for VMS is
       being promoted by Apple as a standard format, I'd prefer to go with
       standards.

   2.  Alisa uses a single VMS process to handle all file service requests,
       while Pacer creates a new process for each connection.  With a community
       of about 25 Mac's on our network, I'm not eager to see 25 or more
       VMS processes consuming memory and other resources -- particulary
       since our VAX computers are the older models that don't have lots
       of capacity.  Besides which, as any VMS system manager will attest,
       VMS does a poor job of creating and winding down processes.  I'm
       not eager to incur that overhead needlessly.

   3.  Pacer's pricing is out of sight.  Since they set their fees based
       on the number of simultaneous users, it gets expensive very very
       fast.  Back in the days, when Pacer was mainly a file transfer package,
       a per-user rate was tolerable since only a minority of users would
       be uploading simultaneously.  On the other hand, when doing file
       service, users are more likely to link up and stay attached all day.
       Then, I can easily have 30+ simultaneous users, Alisa's flat fee
       seems more realistic for file service where many users stay attached
       for long periods.  During those periods, most processing is local
       to the Mac, with occasional file requests to the VAX as documents
       are saved, etc.

Standard disclosure:
    I don't get paid by Alisa or Pacer to say these things.

    Our impressions are based on discussions with sales rep's and brief
    reviews at trade shows.  We have not installed wither package at our
    site.  We hope to do so later this summer, however.


===================================================================
Leo D. Geoffrion                  BITNET:  MANAGER@SKIDMORE.BITNET
Associate Director for             NYNEX:  (518) 584-5000 Ext. 2628
Academic Computing
Skidmore College
Saratoga Springs, NY  12866

krauskpf@uxe.cso.uiuc.edu (04/15/88)

If you have WIN/TCP on your VAX, then NCSA Telnet and CAP could
be your best alternatives.  Pacer and Alisa haven't caught up
in the terminal emulator department yet.

Tim Krauskopf
NCSA
timk@ncsa.uiuc.edu

garrett@udel.EDU (Joel Garrett) (04/18/88)

In article <66000018@uxe.cso.uiuc.edu> krauskpf@uxe.cso.uiuc.edu writes:
>
>
>If you have WIN/TCP on your VAX, then NCSA Telnet and CAP could
>be your best alternatives.  Pacer and Alisa haven't caught up
>in the terminal emulator department yet.
>
>Tim Krauskopf
>NCSA
>timk@ncsa.uiuc.edu

This is exactly what we are planning on doing - at least for terminal
services.  However, for file server/laserwriter support, you are going to
need more than WIN/TCP - you'll probably need Eunice BSD 4.3, and even that
might need work.  Might it be possible to build the CAP/Aufs/lwsrv under
Eunice and then run the result under the REX program that comes with WIN/TCP?
I looked into doing something similar for getting tn3270 under VMS but all
parties said that there were problems with this approach (I.e. the stuff would
look for Eunice-related gbl sections that aren't loaded with REX)

					Joel Garrett
					Research Associate
					CCM
					University of Delaware

					garrett@udel.edu

kenw@noah.arc.CDN (Ken Wallewein) (04/19/88)

>  We've been trying to decide between the two products for a while now, too.
>Based on the literature that we've seen, we are leaning toward Alisa (the new
>version...) for the following reasons:
>
>   1.  It fully implements AppleShare for VMS, while Pacer uses proprietary
>       code.  We fear that adopting Pacer may hurt us as other companies
>       begin to implement Appleshare/VMS gateways.  Since AS for VMS is
>       being promoted by Apple as a standard format, I'd prefer to go with
>       standards.

  We were in a similar quandry. We chose Pacer. Here are a few of the reasons.

  Pacer fully implements Appleshare for VMS, and as far as I can tell, does it
better than Alisa does. They use the Apple client software on the Mac, without
modification. Also, they use the NATIVE VMS file structure. This has two
results: speed, and transparency. You can use your Mac to look at your VMS
file structure. Note that the version Pacer is shipping now is BETA. It wass
announced BEFORE Alisa's, so I would suspect that there's is, too. Their
Virtual Disk server software, on the other hand, has been out for some time. 

>  2.  Alisa uses a single VMS process to handle all file service requests,
>      while Pacer creates a new process for each connection.  With a community
>      of about 25 Mac's on our network, I'm not eager to see 25 or more
>      VMS processes consuming memory and other resources -- particulary
>      since our VAX computers are the older models that don't have lots
>      of capacity.  Besides which, as any VMS system manager will attest,
>      VMS does a poor job of creating and winding down processes.  I'm
>      not eager to incur that overhead needlessly.

  On the other hand, this almost certainly translates to better speed of 
service. I guess you get what you pay for; you can't expect premium service on 
overloaded equipment.

>  3.  Pacer's pricing is out of sight.  Since they set their fees based
>      on the number of simultaneous users, it gets expensive very very
>      fast.  Back in the days, when Pacer was mainly a file transfer package,
>      a per-user rate was tolerable since only a minority of users would
>      be uploading simultaneously.  On the other hand, when doing file
>      service, users are more likely to link up and stay attached all day.
>      Then, I can easily have 30+ simultaneous users, Alisa's flat fee
>      seems more realistic for file service where many users stay attached
>      for long periods.  During those periods, most processing is local
>      to the Mac, with occasional file requests to the VAX as documents
>      are saved, etc.

  I definitely do _not_ agree. I think you will find that Tops (they've
announced a Mac file server due out later this year), Alisa, and Pacer all
have comparable prices for a comparable number of users, at least in the 30-
user ballbark. While you have a point regarding inactivity, a per-user license
allows sites to tailor the cost to the number of people who will _actually_
_use_ the facility, rather than than saying "your VAX will support 100 users,
therefore you must pay for 100 users". 

  Other considerations:

  a) Pacer runs on VAX/VMS, most Unix systems, Primes, Crays, and others. It
supports both Macintoshes and PCs. It runs over Ethernet, Appletalk, and
RS232. 

  b) TSSnet, which is now sold by Alisa (I believe it's a separate package,
and NOT cheap) supports VMS mail to Macs, among other DECnetish things. I
don't know the details. 

  One thing I have found; it's like walking on quicksand trying to get a
complete, accurate picture of things. They keep changing, but they're never
quite good enough! 

/kenw

cetron@utah-cs.UUCP (Edward J Cetron) (04/20/88)

In article <66000018@uxe.cso.uiuc.edu> krauskpf@uxe.cso.uiuc.edu writes:
>
>be your best alternatives.  Pacer and Alisa haven't caught up
>in the terminal emulator department yet.
>
>Tim Krauskopf

	The above is wrong for the Alisa package. It supports Alisaterminal
which is a resource which gives Versaterm, Versterm-pro, or MAC-240 the
capability to remote login from the mac to the VMS host.  It uses CTERM
protocols (VMS SET HOST XXX) and is MUCH more efficient then the NCSA telnet
program TO A VMS MACHINE.  It is also completely compatible with NCSA telnet
to TCP/IP machines.  We run Alisa for our VMS machines, and use NCSA telnet to
the unix boxes and for arpanet access.  I recommend BOTH NCSA telnet and Alisa.


-ed cetron
center for engineering design
university of utah
cetron@cs.utah.edu

derek@gucis.oz (Derek Austin) (04/23/88)

The article referenced above talks about improvements in a forthcoming
release of AlisaTalk.

I believe other disadvantages of AlisaTalk relative to PacerLink are:

	1. No support for standard VMS security

	2. Spooled LaserWriters must be on LocalTalk rather than on the VAX 

Are these things fixed in the new version?
Also, does anyone have any comments to make on PacerLink installation on
VMS. I have been having trouble getting the Kinetics side of things to work
on two different VAXes for reasons that the distributor here is unable to
explain. Anyone?      

	derek

garrett@udel.EDU (Joel Garrett) (04/25/88)

In article <607@gucis.oz> derek@gucis.oz (Derek Austin) writes:
>The article referenced above talks about improvements in a forthcoming
>release of AlisaTalk.
>
>I believe other disadvantages of AlisaTalk relative to PacerLink are:
>
>	1. No support for standard VMS security
>
AlisaShare, Alisa Systems' AppleShare-compatible system, will support
mappings into the standard VMS protection system much like PacerShare
currently does.  However, be aware, that there are many more permutations/
combinations of protections in VMS than are allowable in an AppleShare/VMS
mapping.
>
>	2. Spooled LaserWriters must be on LocalTalk rather than on the VAX 
>
Hmmm, maybe the DEC printing support package will also work for Serially-
connected LaserWriters... I'll have to ask the tech people at Alisa Systems...
>
>Also, does anyone have any comments to make on PacerLink installation on
>VMS. I have been having trouble getting the Kinetics side of things to work
>on two different VAXes for reasons that the distributor here is unable to
>explain. Anyone?      
>
Well, this realy depends upon how you want to configure the gateway box.
Are you, perchance, wanting to set up to run the combined gateway code,
(allows the box to route IP as well as AppleTalk-on-Ethernet) or just the
AppleTalk-Ethernet stuff?  Configuring for combined operation can be tricky.
Your description is pretty general, so one really can't make much of a
diagnosis on that.

					Joel

cetron@utah-cs.UUCP (Edward J Cetron) (04/29/88)

In article <607@gucis.oz> derek@gucis.oz (Derek Austin) writes:

>	1. No support for standard VMS security

		fixed (near as I can tell) in the appleshare version

>	2. Spooled LaserWriters must be on LocalTalk rather than on the VAX 


		I consider this a PLUS, on our microvaxen, the throughput
is MUCH greater through the ether then over a serial line.  And it allows the
macs to bypass the spoolers and go right to the printers if desired (which is
VERY useful when changing laserwriter names as equipment fails or is replaced)

-ed cetron
center for engineering design
university of utah.

ps. the biggest plus that I see with Alisa is that they support STANDARD MAC
APPLESHARE CLIENT SOFTWARE, and they use standard ethertalk throught the
kinetics.  As I understand it, Pacer uses some sort of proprietary client
software....

kenw@noah.arc.CDN (Ken Wallewein) (05/01/88)

>ps. the biggest plus that I see with Alisa is that they support STANDARD MAC
>APPLESHARE CLIENT SOFTWARE, and they use standard ethertalk throught the
>kinetics.  As I understand it, Pacer uses some sort of proprietary client
>software....

  Nope. Pacer uses the standard Apple client software, too, exactly as it
comes from Apple. What's more, Pacer's _server_ software on VMS uses the VMS
_NATIVE_ file structure, as opposed to those big virtual disks files that
Alias uses. This provides significant advantages in terms of performance, user
interface, and security.
 => You use standard Appleshare to manipulate your VMS directory structure. <=
I believe Alisa requires special (not yet released) software to do this. 

 Your impression is probably based on Pacer's _old_ server software, which,
having been around a while, supports BOTH Macs AND PCs on VMS, Unix, PRIMOS,
and Crays. It was in use long before Appleshare was even announced, let alone
delivered. It uses a special protocol which functions over real Ethernet, 
Appletalk, and RS232 serial links.

 /kenw
                                                                 A L B E R T A
Ken Wallewein                                                  R E S E A R C H
kenw@noah.arc.cdn                                                C O U N C I L

garrett@udel.EDU (Joel Garrett) (05/02/88)

In article <1396*kenw@noah.arc.cdn> kenw@noah.arc.CDN (Ken Wallewein) writes:
>>ps. the biggest plus that I see with Alisa is that they support STANDARD MAC
>>APPLESHARE CLIENT SOFTWARE, and they use standard ethertalk throught the
>>kinetics.  As I understand it, Pacer uses some sort of proprietary client
>>software....
>
>  Nope. Pacer uses the standard Apple client software, too, exactly as it
>comes from Apple. What's more, Pacer's _server_ software on VMS uses the VMS
>_NATIVE_ file structure, as opposed to those big virtual disks files that
>Alias uses. This provides significant advantages in terms of performance, user
>interface, and security.

Alisa's forthcoming AppleShare server, AlisaShare, also uses the Standard
Apple client software and a VMS native file structure.  I know, it isn't
available as a 1.0 product yet, but then again neither is PacerShare -
offically at least - it is still in Beta.

> => You use standard Appleshare to manipulate your VMS directory structure. <=
>I believe Alisa requires special (not yet released) software to do this. 

AlisaShare won't need anything like this either.  I kind of doubt if Alisa
will support their old way of doing things once this is released.

					Joel

cetron@utah-cs.UUCP (Edward J Cetron) (05/02/88)

In article <1396*kenw@noah.arc.cdn> kenw@noah.arc.CDN (Ken Wallewein) writes:
>
>  Nope. Pacer uses the standard Apple client software, too, exactly as it
>comes from Apple. What's more, Pacer's _server_ software on VMS uses the VMS
>_NATIVE_ file structure, as opposed to those big virtual disks files that
>Alias uses. This provides significant advantages in terms of performance, user
>interface, and security.

	This is very bizarre? If the above is true, then:

	1. why is Pacer trying to sell use SEPARATE client software. (at
		significant cost to us)

	2. Why does Pacer discuss tcp/ip style sockets when explaining why
		their software conflicts with Telnet.

	3. Why does my version of Alisashare work just fine with the NATIVE
		VMS file structure....

I always new Canada was different :-)

-ed
cetron@cs.utah.edu

kenw@noah.arc.CDN (Ken Wallewein) (05/06/88)

>In article <1396*kenw@noah.arc.cdn> kenw@noah.arc.CDN (Ken Wallewein)writes:
>>
>>  Nope. Pacer uses the standard Apple client software, too, exactly as it
>>comes from Apple. What's more, Pacer's _server_ software on VMS uses theVMS
>>_NATIVE_ file structure, as opposed to those big virtual disks files that
>>Alias uses. This provides significant advantages in terms of performance,user
>>interface, and security.
>
>        This is very bizarre? If the above is true, then:
>
>        1. why is Pacer trying to sell use SEPARATE client software. (at
>                significant cost to us)
>
>        2. Why does Pacer discuss tcp/ip style sockets when explaining why
>                their software conflicts with Telnet.
>
>        3. Why does my version of Alisashare work just fine with the NATIVE
>                VMS file structure....
>
>I always new Canada was different :-)
>
>-ed
>cetron@cs.utah.edu

  Oh, it is, it is :-)

  But let's get something cleared up here. This much is certain:

	a) Both Alisa and Pacer have well-established NON-Appleshare file/disk
service products. Neither of those used the native VAX/VMS directory structure
to represent the Macintosh directory structure. 

	b) Both Alisa and Pacer now have Appleshare file server HOST software
which runs under VAX/VMS, supporting _Apple's_ CLIENT software on the
Macintosh. 

  When we refer to these products, let's be clear about the category of server
to which we are referring. Now...

  1. Pacer's older product, pcLink, uses a single VAX file to represent a
Macintosh HFS volume. It requires special software on both the VAX and the
Mac. Alisa's older product is similar, except that, last I heard, it does not
properly support HFS. I doubt many people will care much longer. 

  2. pcLink runs over either Appletalk, Ethernet (your choice of protocol:
Pacer's custom PNP, or TCP/IP) or RS232 (not sure about the protocol there).
This provides support for both 'virtual disk' and (due to some device driver
magic) high-speed terminal connections. It also works, as I said before, with 
both Macs and (ugh) MS-DOS engines, served by VMX/VMS, Unix, Primos, etc.
PacerShare, on the other hand, uses EtherTalk, Macs to VMS only.

  3. What does 'just fine' mean? Do Mac files coexist in the same directories
as your VMS files, with AlisaShare? They do with PacerShare. Do they look any
different from your VMS files, aside from the icons? They don't with
PacerShare. The PacerShare SERVER uses the VMS _directory_ and file structures
_directly_. It maintains a desktop subdirectory in each VAX directory, to
contain Mac-specific information about the files in that directory/folder.
There's no other distinction between Mac and VMS files. 

  When I started to use the Mac to do VMS directory maintenance, the first
thing I did was make a lot of new directory/folders, and categorize things
better. It's a nice interface, but it's not QUITE transparent. All the VAX
files show up on the Mac as generic document icons. Blah. And I'd like to be
able to double-click on VAX files and have a terminal window open up running
the appropriate VAX/VMS image. As near as I can figure, it should be possible
to do exactly that, just by somehow (HyperCard?) invoking a Mac terminal
emulator and sending an appropriate command string to the VAX. Doesn't sound 
hard to me :-)

  I'm sure there's a difference between Alisa's and Pacer's Appleshare
servers, but to be honest, I'm no expert as to what they are. We are running
Pacer software. I tried unsuccessfully to get Alisa running here, but it
didn't like the load we were running on our Fastpath; at that time, there was
no Alisashare. Therefore, to some extent, I've been repeating what I've been
told about Alisa, albeit by people in a good position to know. Mea culpa. 

  I've yet to see _any_ significant comparison by _anyone_ with real
experiance with both products. 

  Nobody's said much so far about TSSnet. And I'm really looking forward to 
how Tops for VMS works. Tops is peer-to-peer; the idea of mounting a Mac 
volume from the VAX and backing it up to tape every night is intriguing.

/kenw