alexis@ccnysci.UUCP (Alexis Rosen) (05/12/89)
Finally, a Peer-to-Peer non-dedicated distributed AppleShare!
For a long time now, people have been wishing for an AppleShare server
which didn't require a dedicated Mac. For just as long, people have been
buying TOPS instead. But now we are beginning to see just how bad TOPS
really is. It's totally incompatible with AppleShare, and Sun has no
interest in or desire to implement AFP compatability in the near future,
despite markting claims to the contrary. There are also a host of bugs
that make using DBMSs with TOPS difficult to impossible.
Finally, a solution is at hand. IPT will shortly be releasing
Peer-to-Peer AppleShare, which does NOT require a dedicated machine for
file service. You can read all about it on page 1 of next week's
MacWeek, but here is a summary:
1) Peer-to-Peer AppleShare runs on Macs, PCs, and Unix machines.
2) Price for the Mac software is *** $150 *** per server!
3) Claimed speed is "slightly faster than AppleShare."
4) Fully compliant with AFP.
5) Does NOT send out garbage serial-number packets on the net.
6) To be released by June 1.
Now, all this comes from talking to their marketing VP. She is a very
capable person technically, so the chances of her being wrong because of
a lack of understanding the subject are pretty much nil. As to whether
or not they are overestimating their speed or compatability, I will know
soon, as we are receiving beta copies early next week.
Note that the price for setting up a network initially looks like it's
about 1/2 the price of a TOPS network. After all, TOPS is $289 per Mac
while this is $150 per Mac. (I am using list prices here for
convenience, but this should scale down to street prices pretty evenly.)
In fact, IPT's server is MUCH cheaper even than that. For example, one
of my clients has a network of about 100 Macs and a few dozen PCs.
Forgetting the PCs for now, it would cost $28,900 to network them on
TOPS, and probably only ten percent of the machines (or less) would
actually be file servers. So they could actually buy AppleShare on eight
dedicated Mac SEs to serve the net, at the same price. In fact, they did
just that, before I started working with them.
On the other hand, consider the cost of setting up this same net with
Peer-to-Peer Appleshare. If you want ten machines as servers, your total
cost is *$1500* since you only need to buy copies of the software for
the servers. All the other machines on the net use the AppleShare client
server software that comes with all Macintoshes.
For smaller networks, when less money is available, the difference is
even more impressive. For a ten Mac network, with one server, TOPS would
cost $2890, and AppleShare would cost $2600 (or more, depending on what
kind of Mac you use for a server). IPT's AppleShare would cost $150.
Also note that Tops consumes almost 70K more memory than the AppleShare
Client software. That means that every user who doesn't need to serve up
his disk gets back 70K. That's enough room for QuickMail, QuickKeys, or
a bunch of smaller INITs, for example. On a 1MB Mac, it could instead
mean the difference between being able to run MultiFinder or not.
Administrators of large networks will be glad to know that IPT's server
does not send out serial-number packets every few seconds, unlike TOPS.
These packets can really clog a large network. They are especially
antisocial when you are putting Macs on an EtherNet cable along with
lots of other machines, a practice which is becoming more and more
common at universities and large companies.
When I get the Beta copies next week, there are a number of interesting
questions that I will then be able to answer. For starters, how much
memory does a server use? How much does this slow down the CPU? How does
is respond under heavy load? Is it *really* compliant with AFP,
including byte-range locking and the Desktop Manager calls?
Will this software really performs according to IPT's claims? I'll let
you all know, but for now it seems likely that they've done it right.
IPT has been in the business since the Mac first came out, so they've
got the experience to do it.
Assuming it works, this will have an incredible impact on the Mac
market. It will demolish TOPS virtually overnight, which might well be
what Sun really wants anyway (after all, Sun's game plan calls for NFS
to rule the world, not TOPS, and they've done nothing to integrate the
two in almost two years). It will also consolidate the Mac networking
market around AFP, and put even greater pressure on 3Com to finish up
it's AFP services sooner.
It is interesting to watch history repeat itself. In early 1986
InfoSphere, publisher of MacServe, virtually owned the networking
market. 3Com was only a bit player. There was nothing else, except this
upstart called TOPS. The great thing about TOPS was that it allowed
people to use the same folders on the same disks at the same time,
whereas MacServe only let one person have write access to a given disk
(or volume, actually) at the same time. InfoSphere refused to upgrade
their product and as a result their market share went from upwards of
90% to about 0% today. (Amazingly, from the ashes of MacServe rose
Liason, a truly wonderful product which bridges multiple AppleTalk
nets.)
Today, the same thing is about to happen. TOPS dominates the market.
They are starting to lose market share to AppleShare because they are
incompatible with AFP (and thus certain important programs, such as
FoxBase). But they still hold their position on the basis of TOPS's
ability to run in the background and serve folders from every Mac on the
net. Now IPT is introducing a product that does all this, and is also
fully compatible with AFP and AppleShare, at a fraction of the price.
For every thing there is a season, and TOPS' season is just about over.
It's about time.
I have absolutely no affiliation with either IPT or TOPS, except as an
unsatisfied customer of TOPS' and as a Beta site for both companies.
Copyright 1989 by Alexis M. Rosen.
Please do not reprint this (distribution on the internet is OK) because
it will serve as the basis of the review I am writing.
---
Alexis Rosen
alexis@ccnysci.{uucp,bitnet}
alexis@rascal.ics.utexas.edu (last resort)
You can also try alexis@sci.ccny.cuny.edu, but it may not work yet.alexis@ccnysci.UUCP (Alexis Rosen) (05/13/89)
In article 1968@ccnysci.UUCP> I wrote: > >Finally, a Peer-to-Peer non-dedicated distributed AppleShare! > [et-very-long-cetera.] Perhaps it would be wise to wait until I post my evaluation early next week before you call them. If you want to talk to them right away, though, their address is: Information Presentation Technologies (IPT) 23801 Calabasas Road Suite 2008 Calabasas, CA 91302 Their telephone number is (818) 347-7791. I spoke to Olivia Fazela, but probably anyone there can answer questions. Mention that you saw my posting on the net- It won't get you a deal, but the more respect companies have for the net as an influence on their sales, the better it is for all of us... --- Alexis Rosen alexis@ccnysci.{uucp,bitnet} alexis@rascal.ics.utexas.edu (last resort)
desnoyer@Apple.COM (Peter Desnoyers) (05/15/89)
In article <1968@ccnysci.UUCP> alexis@ccnysci.UUCP (Alexis Rosen) writes: > >buying TOPS instead. But now we are beginning to see just how bad TOPS >really is. You know what one of the "best" TOPS features is? Mount a volume published on a PC with TOPS. Use option-copy to copy only the text contents of a file from the PC to your Mac. Your mileage may vary, but in my situation (PC TOPS -> localtalk -> bridge -> thin ethernet -> mac) there is about a 50% chance that the file will be corrupted, with garbage characters and random stuff from the Mac's memory inserted somewhere in the middle. (I.e. TOPS warning messages, pieces of previously read files that must have been sitting in the disk cache, etc. It's truly amazing what shows up.) Peter Desnoyers
tim@hoptoad.uucp (Tim Maroney) (05/16/89)
In article <1980@ccnysci.UUCP> alexis@ccnysci.UUCP (Alexis Rosen) writes: >Mention that you saw my posting on the net- It won't get you a deal, but >the more respect companies have for the net as an influence on their sales, >the better it is for all of us... Wrong as usual, Alexis. Those of us who share negative information in our possession are more than accustomed to overly sensitive companies threatening us with legal action and economic sanctions. As far as I'm concerned, we'd all be better off if the marketing bozos didn't even know about the network, leaving it as a valuable information resource for those of us on the other sides of things. An information resource that contains only positive, smiley face things and can't warn people about traps and obstacles they may face is a useless resource. Increased marketing awareness of the networks can only lead to increased pressure to suppress critical messages. -- Tim Maroney, Consultant, Eclectic Software, sun!hoptoad!tim "God must be a Boogie Man." -- Joni Mitchell
amanda@intercon.UUCP (Amanda Walker) (05/17/89)
In article <7319@hoptoad.uucp>, tim@hoptoad.uucp (Tim Maroney) writes: > Wrong as usual, Alexis. Polite as usual, Tim :-(. > Increased marketing awareness of the networks can only lead to > increased pressure to suppress critical messages. Oh, I don't know... our company finds networks like Usenet very useful in keeping our marketing and technical support on track. Of course, we *are* in the networking business, so we may be biased :-). Apple seems to think pretty highly of Usenet, at least lately with the System 7.0 stuff... -- Amanda Walker <amanda@intercon.UUCP> InterCon Systems Corporation -- "You don't have to take my word for it--I'll convince you!" --Gurshuran Sidhu
werner@molokai.sw.mcc.com (Werner Uhrig) (05/17/89)
In article <7319@hoptoad.uucp>, tim@hoptoad.uucp (Tim Maroney) writes: > In article <1980@ccnysci.UUCP> alexis@ccnysci.UUCP (Alexis Rosen) writes: > >Mention that you saw my posting on the net- It won't get you a deal, but > >the more respect companies have for the net as an influence on their sales, > >the better it is for all of us... > > Wrong as usual, Alexis. Those of us who share negative information in ^^^^^^^^^^^^^^ RUDE > our possession are more than accustomed to overly sensitive companies > threatening us with legal action and economic sanctions. As far as I'm ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ALARMIST ... > Increased marketing awareness of the networks can only lead to ^^^^^^^^^^^^^^^^ WRONG > increased pressure to suppress critical messages. where is your data to proof these allegations, Tim? just one person that a) appreciates Alexis, b) has not heard of any threats to be taken seriously c) appreciates the support companies like Apple, SuperMac, Symantec, etc have begun to provide to us here on the net - not least because they have not only learned to fear our negative articles, but also appreciate our positive articles and business support. ---Werner (fed up with articles without meaningful contents!) -- --------------------------> please send REPLIES to <------------------------ INTERNET: uhrig@mcc.com (if unavailable: werner@rascal.ics.utexas.edu) UUCP: ...<well-connected-site>!milano!werner ALTERNATIVE: werner@astro.as.utexas.edu OR werner@utastro.UUCP
dee@XAIT.Xerox.COM (Donald Eastlake) (05/19/89)
In article <2360@molokai.sw.mcc.com> werner@molokai.sw.mcc.com (Werner Uhrig) writes: >In article <7319@hoptoad.uucp>, tim@hoptoad.uucp (Tim Maroney) writes: >> our possession are more than accustomed to overly sensitive companies >> threatening us with legal action and economic sanctions. As far as I'm > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > ALARMIST >> Increased marketing awareness of the networks can only lead to > ^^^^^^^^^^^^^^^^ > WRONG >> increased pressure to suppress critical messages. > where is your data to proof these allegations, Tim? Seems to me that their used to be useful comments to the net from Joel West until something related to threats of legal action concerning a message of his caused him to stop posting altogether. Although not exactly the same thing, there is also the furor over the racist/sexist material in the jokes group that has, due to threats of legal action, caused it to be banned in some areas. I am not sure there is any solution to all this. The wider the audience the more likely there is to be someon who will not just be offended and not just flame, but actually use the assorted judicial and administrative monkey wrentches available for gumming things up. -- +1 617-969-9570 Donald E. Eastlake, III ARPA: dee@XAIT.Xerox.COM usenet: {cbosg,decvax,linus}!cca!dee AppleLink: D2002 Box N, MIT Branch PO, Cambridge, MA 02139 USA
kent@lloyd.camex.uucp (Kent Borg) (05/19/89)
In article <1968@ccnysci.UUCP> alexis@ccnysci.UUCP (Alexis Rosen) writes: > >Finally, a Peer-to-Peer non-dedicated distributed AppleShare! ... >file service. You can read all about it on page 1 of next week's >MacWeek, but here is a summary: > >1) Peer-to-Peer AppleShare runs on Macs, PCs, and Unix machines. ... Do you have any more details on the Unix machines aspect? Can they publish big, cross-mounted file systems? Can the Unix copy be a client too, or only a server as with Tops? Do they need a Unix<=>Mac beta site?? Also very curious about speed and robustness. How easy will it be to crash? ...I'm pretty good at crashing things... Kent Borg kent@lloyd.uucp or ...!husc6!lloyd!kent
tim@hoptoad.uucp (Tim Maroney) (05/20/89)
In article <2360@molokai.sw.mcc.com> werner@molokai.sw.mcc.com (Werner Uhrig) writes: >In article <7319@hoptoad.uucp>, tim@hoptoad.uucp (Tim Maroney) writes: >> In article <1980@ccnysci.UUCP> alexis@ccnysci.UUCP (Alexis Rosen) writes: >> >Mention that you saw my posting on the net- It won't get you a deal, but >> >the more respect companies have for the net as an influence on their sales, >> >the better it is for all of us... >> >> Wrong as usual, Alexis. Those of us who share negative information in > ^^^^^^^^^^^^^^ > RUDE Let's talk rudeness. Alexis came out saying that after "extensive discussions with technical personnel at TOPS", he knew that TOPS did not do any kind of range locking whatsoever. I pointed out that this was false; TOPS has in fact had range locking in accord with the specification in Inside Mac volume IV since before Inside Mac volume IV was even published. This was not followed by any correction or retraction from Alexis; instead, he simply said in his next message that it had "major bugs which made it practically useless" for any database application. (These are not exact quotes, since the messages have expired, but anyone who still has them can verify that I have not changed his meaning.) At the time he sent this, I had already pointed out to someone else that these alleged bugs were in fact correct implementations of the advisory range locking described in Inside Mac volume IV. Again, no correction or retraction from Alexis, and this rudeness was compounded by a statement of intent to violate journalistic ethics by publishing this false information in MacWeek. Damn straight I'm annoyed, and I have a perfect right to be. >> our possession are more than accustomed to overly sensitive companies >> threatening us with legal action and economic sanctions. As far as I'm > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > ALARMIST Are you accusing me of lying? I assure you that these threats are real. I would love to describe the most recent round in detail; long time netters are probably surprised that I haven't, after the UNC affair. If the threats were merely against me, I would in fact describe them, but they are against my current client as well. I do not feel that I have the right to make this moral decision for him, as it could cost him a great deal of money if these threats are fulfilled. Threatening one's friends is a particularly scummy tactic, but in this case, it is an effective one. I am not familiar with the details of the Joel West case, but I have little difficulty believing that, as described by Donald Eastlake, it was in fact caused by legal threats resulting from a critical message he posted. I nearly lost my job at TOPS after I posted a message putting the output of MPW C next to that of Consulair C and showing that the former was much better than the latter, and this permanently harmed my relationship with management. All because the president of Consulair complained to the president of TOPS; the merits of the case didn't enter into it at all. I don't know what ivory tower you live in, Werner, but out here in the real world, this kind of stuff is common. >> Increased marketing awareness of the networks can only lead to > ^^^^^^^^^^^^^^^^ > WRONG >> increased pressure to suppress critical messages. > > where is your data to proof these allegations, Tim? I see, you are accusing me of lying after all. I will refrain from the appropriate epithets. > ---Werner (fed up with articles without > meaningful contents!) Then why did you send this? -- Tim Maroney, Consultant, Eclectic Software, sun!hoptoad!tim "Women's wages are 56% of men's -- but that's not necessarily evidence of discrimination in employment." -- Clayton Cramer in news.groups and soc.women
alexis@ccnysci.UUCP (Alexis Rosen) (05/22/89)
It seems that my posting about peer-to-peer AppleShare has caused a minor flame war. The particularly weird thing about it is that I have not seen *ANY* of the followup articles! (No, it didn't mysteriously wind up in my kill file.) I wouldn't know about it at all if my friend David hadn't mentioned it in passing, and mailed me a copy of one article. Since I only have this one article (by Tim Maroney) from this message thread (though it quotes two others), I don't have the complete picture, but I'll try to clear up a few things. First of all, Tim was quite pissed off when I said that TOPS had no byte-range locking at all. Well, that was in a different article that was not directly related to this one. But, he was right, and what's worse is that I knew this- I just forgot it at the time I was writing. That's why I kept quiet when he posted a correction to that other article- he was right, and I had nothing to add. I also can understand that he still feels angry that Apple changed the rules of the game, but that's hardly my fault. The statement that TOPS is incompatible with AFP, _as it now stands_, is correct. In addition, it has other problems (which are getting lots of press right now) that have nothing to do with this- they're just bugs. These bugs do in fact make TOPS useless for major database work, since my goal when maintaining 60 MB databases is NOT to have to repair corrupted files every week or two. I am NOT going to go into this- if you want to know more, check in last week's PC Week, or MacWeek from a few weeks ago, or talk to TOPS tech support. Anyway, I said that using TOPS with databases was "difficult to impossible." This is absolutely true- it can't be done with FoxBase, and using design mode in 4D is likely to corrupt your database. (It also had major problems with OMNIS for a long time, but that was a bug in OMNIS that was since fixed, after months of complaints, by a fellow who used to work for TOPS [Tracy Lakin].) This comes from Tim's note: >Again, no correction or retraction from Alexis, and this rudeness was >compounded by a statement of intent to violate journalistic ethics by >publishing this false information in MacWeek. This is rather heavy-handed, I think. In regard to databases, my original posting was correct. In any event, why would I want to publish false information??? Tim also warns about negative messages bringing down the wrath of some companies on the net (and the posters). If anyone is likely to be the subject of a lawsuit, it's them! They made fairly specific claims that version 2.0 would be compatible with AppleShare, which as we all know (and regardless of IM-IV) it's not. I can only suggest that in the event of such nasty tactics by vendors, we net.folk return the favor by boycotts, negative word-of-mouth, letters to magazines, and whatever other typical grass-roots activist tactics seem necessary (how about a sit-it at corporate headadquarters :-) :-) ). To get back to my original posting: It was not meant as just another potshot at TOPS (they're too easy a target). They simply got mentioned extensively in the historical information because they make up a large part of macintosh networking history. The message was primarily about the IPT product, and my comment that more exposure for the net was good was in reference (obviously, if you read it again) to my favorable mention of IPT, not my jabs at TOPS (they already know what I think of them). If there were other points made (or other shots taken), please let me know and I'll try to respond to them. I still wish I knew why I haven't seen any of these articles in news. --- Alexis Rosen alexis@ccnysci.{uucp,bitnet} alexis@rascal.ics.utexas.edu (last resort)
alexis@ccnysci.UUCP (Alexis Rosen) (05/27/89)
After some delays, I finally received the IPT software today. Its official
name is "IPT Personal Server Network."
I have just started to play with it. It is very very rough around the
edges, but the important point is that it appears to work! Sometime next
week I will post my benchmark suite, and results for AppleShare and IPT-PSN.
Novell results will follow a bit later.
The title of this (and my earlier) article says "this month." From the
state of the software, I'd say that that's an overoptimistic estimate, but
not by a whole lot. I'd guess early June.
So far, there don't seem to be any conflicts with the DeskTop Manager INIT
with it running on _both_ the server and the client. I don't know how it
would behave in a mix-and-match environment but I expect it would be OK.
Memory consumption on the server machine appears to be minimal. I've noted
RAM usage increasing by 30-50K with the server active (this is quite a bit
less than TOPS, for example) but I have not done extensive tests on this yet.
The user interface of the administrator program is where the rough edges
come in. It's quite ugly and there are a few visual anamolies. Still, I am
very encouraged by their response to my bug reports, since they are generally
of the form "Wow, I'm glad you found that, we'll fix it right away." For
those of you who remember my article about Beta-testing, this is a company
that apparently knows how to do Betas right.
Have a good weekend, more will follow next week.
---
Alexis Rosen
alexis@ccnysci.{uucp,bitnet}
alexis@rascal.ics.utexas.edu (last resort)alexis@ccnysci.UUCP (Alexis Rosen) (05/31/89)
In article <400@lloyd.camex.uucp> kent@lloyd.UUCP (Kent Borg) writes: >Do you have any more details on the Unix machines aspect? Can they >publish big, cross-mounted file systems? Can the Unix copy be a >client too, or only a server as with Tops? Do they need a Unix<=>Mac >beta site?? This software, called uShare, was IPT's first product. It has been shipping for a long time (over a year, I think). So they don't need any betas for that. I'm pretty sure the unix boxes can mount Mac file systems as well as serve Unix FSs to Macs. >Also very curious about speed and robustness. How easy will it be to >crash? ...I'm pretty good at crashing things... Not much hard data yet, but so far things look pretty good. The server code itself hasn't crashed once, though the administrator program (for setting up users and groups) has died twice. But relax, it's a Beta, not shipping code. They are going to do a lot of work on the administrator before they release it. --- Alexis Rosen temporarily at spector@vx2.gba.nyu.edu alexis@rascal.ics.utexas.edu (last resort)