Path: utzoo!attcan!uunet!pyrdc!pyrnj!dasys1!alexis
From: alexis@dasys1.UUCP (Alexis Rosen)
Newsgroups: comp.sys.mac
Subject: Re: WARNING: Symantec Utilities ('SUM') trashes partitions!!!
Keywords: BOMB sloppy_programming excessive_complaining
Message-ID: <5636@dasys1.UUCP>
Date: 19 Jul 88 11:20:08 GMT
Lines: 301


I recently posted a very explicit warning message about a bug in the Symantec
Utilities Partition DA & INIT. I feel that the vicious, no-second-chance nature
of the bug is such that everyone who might contemplate using it know about it
immediately. The consequences of this bug can be disastrous.

A number of net-folks who (in general) I have a good deal of respect for
reacted very strongly (and mostly negatively) to this posting. One response was
pretty vicious. I think that this is really uncalled for, and I will try to
explain why. My justifiable anger over a disk-recovery product destroying a
large portion of my hard disk should not have been interpreted as a whimsical
temper tantrum.

Please Note: It seems that the author of the bad DA is also the author of
MacZAP Recover, Tools, etc. As I wrote in the very same posting, I have always
had EXCELLENT results with that program.

I wrote:
<Recently I purchased Symantec's Utilities for the Macintosh ("SUM"). It has
<been getting good word-of-mouth, and my previous experience with MacZAP,
<its progenitor, has been good.

     I know a good thing when I see one. I have sold clients copies of MacZAP
     before and will sell SUM (with some careful warnings about the partition
     stuff) in the future.

<There is also a combination INIT and DA which allows you to create parti-
<tions on your hard disk (not SCSI partitions, but invisible HFS files which
<the DA mounts as individual hard disks).  [etc.]
<
<***** DO NOT USE THIS DA AND INIT!!!!! EVER!!!!! *****
<
<On a Mac II with a clean system 4.2, unifinder, and three inits (Suitcase,
<PramFix, and the Partition INIT in question), this DA destroyed partitions
<which I had created only moments after I threw out the other copies of my
<files.

Tim Dierks (uunet!portal!cup.portal.com!Tim_M_Dierks) writes:
>I found that HD Partition didn't work when in a Suitcase file.  After moving
>it to my System file however, it works fine (on a 512e running System 6.0 no
>less!)  You may want to try this (I don't know if you had it in the System or
>not.)  However, I agree with you: it is a bad bug.  Another one I've found:
>try searching for a hex string in Symantec Tools.  Chances are, you won't find
>it, even if you know its in there, because something is wrong with the
>algorithm. (sp?)  Also, the disk-optimizer doesn't move files to maximize your
>contiguous free space, like Disk Express does.  This makes it hard to create
>partitions of signifigant size, at least on my drive.  Also, I've run into
>problems where Tools tells me there are 1900 contiguous free K on the disk
>but the Partition DA won't let me create a 1600K partition, complaining that
>there isn't enough contiguous space.  However, the HD Recover utility seems to
>work like a charm (on the two floppies I've tried it on) and that's what I
>bought it for in the first place.
>
>Tim Dierks
>Address of the day: DIERKS@IRISHMVS.BITNET

And Moriarty, aka Jeff Meyer (moriarty@tc.fluke.COM) wrote:
>1)  Why couldn't it be some interaction with Pramfix or Suitcase?  Sound
>    like more testing needs to be done, especially since it worked fine on
>    my hard disk (Mac Plus, 2.5 Megs, lots of INITs).  How about *asking
>    others* whether they've seen this problem before accusing people of
>    being incompetents?

I am sure that you are both correct. I suspected as much. Suitcase is almost
certainly the deciding factor. However, this is really irrelevant:
1) Suitcase is virtually system software. Testing a DA without testing for
   Suitcase (and F/DA Juggler) is not testing. I guarantee you that any company
   that sold a DA which worked fine with a clean system but bombed under
   suitcase (forget about wiping out files, but just bombed) would go out of
   business. They would be roasted by any reviewers of their product.
2) I don't know if you see the irony here, but my disk recovery product has
   just trashed a large part of my hard disk, and can't bring it back. This is
   during an operation which should not even need to write to this disk AT ALL.
   I feel understandably paranoid. Do you really think I want to put this thing
   in my System File??? The chances of it doing damage are remote, but what
   were the chances of damage in the first place? No less remote. Note that I
   tested it without any other INITs (except Apple's own PramFix).
3) In a later note, Moriarty explains that some of Tim's gripes are correct and
   some are not. I don't know who is correct.

I wrote:
<       This is a 100% reproducible error: After mounting any partition,
<ANY FURTHER ATTEMPTS TO MOUNT ANY PARTITION WILL DESTROY THAT PARTITION
<IMMEDIATELY, WITHOUT ANY NOTICE.

Werner Uhrig (werner@rascal.ics.utexas.edu) wrote:
>while I sympathize with the anguish induced by losing a hard-disk, this
>article is so full of "bad vibes" and so incomplete in information (no
>version INFO of INITs), and so invalid a test (why didn't you try with a
>plain Apple system, i.e. no additional INITs), that the combination of
>this did not make me feel even like responding - much less to bring it
>to the attention of the author(s) of SUM (with whom I'd like to stay
>friends after all).
>
>With the attitude Alexis displays here, I'd not even give him the phone
>number to try his luck himself.
>
>I just thought, that Alexis should know that YOU CAN SHOOT YOURSELF IN THE
>FOOT - even here on the net .....

I can understand Werner attacking me, since he is friends with the programmer.
However, the criticism (here) is not valid. The only version info I didn't give
was for Suitcase. This omission is hardly "so incomplete" (considering the fact
that all Suitcases, especially since 1.2, are pretty similar). I was using
V1.2.

My attitude is one of acute paranoia (see comment above), and not unreasonable.
If I shot myself in the foot, it certainly wasn't here. Not reporting a bug
this dangerous is utterly irresponsible.

<                                 For those of you who have files in a SUM
<partition, you can get them out safely by NEVER mounting more than one
<partition in one invocation of the DA.
<
<To add insult to injury, there is a feature of the new MacZAP HFS Recover
<program which makes it easy to restore damaged partitions... but... when
<run it only recovers a few of the lost files!

Rich Siegel (uunet!husc6!endor!singer) wrote:
>	Do you realize that there can be cases in which a volume is so
>badly munged that no recovery utility can resurrect it? In your particular,
>the fact that only a few filess were recovered is due to the fact that
>the partitioning software trashed the partition, NOT THAT THE RECOVERY
>UTILITY IS DEFECTIVE.
>
>	If you're going to bitch, at least bitch at the right software.

Of course I realize this. I did not imply that the recovery software was
defective and I am very sorry if anyone else thought I meant that. The "insult"
was simply that although the package was supposed to recover disks, it couldn't
in the case that IT inflicted. I am not blaming MacZAP Recover, but it *really*
hurts when one part of the package (NOT the recovery program) trashes your disk
so badly that the recovery part can't restore the lost files.

I wrote:
<>>>>>>>>>> FLAME ON!!!!! <<<<<<<<<
<
<This is the most inexcusable, irresponsible, idiotic, egregious blunder I
<have ever seen from a "reputable" software company. Even worse, I can't
<just say "never buy from Symantec," because they have also produced some
<of the best software ever written for the Macintosh. Lightspeed Pascal
<and Lightspeed C, for example. Q&A for PMS-DOS is a wonderful program (if
<you must use MS-DOS...)

	You should note from this that I am not lashing out blindly at everyone
	associated with the product, just the person responsible for the bug.

Rich Siegel wrote:
>	Have you every hear dof Microsoft Word version 3.0?

Yes I have. No single bug in that disgustingly infested program was as
dangerous as the bug in this DA, not by a long shot. (BTW, "PMS-DOS" was a typo
and not a very bad joke.)

Moriarty (Jeff Meyer) wrote:
>2)  If this is the most "idiotic, egregious blunder" you've ever seen from a
>    computer company, you're forgetting your complaint a few days ago about
>    HyperCard compaction, which was "inexcusable".  How about a little less
>    beating of the chest and hyperbole, and a little more accuate
>    information, eh?

I have not forgotten anything. The bug in HyperCard is very bad. BUT:
1) The HyperCard bug does not occur with all stacks. They screwed up, but they
   didn't show the kind of carelessness exhibited in the Partition DA.
2) No matter haw bad this bug (and you will corrupt your stack if you do hit
   it), it DOES NOT trash your disk when it shouldn't even be writing to it...
3) I gave all the accurate information. It is impossible to miss this bug,
   based on the information in my article. As far as hyperbole, well, you seem
   a little partial to it too, but see my comments below...

By the way, I am now corresponding with a member of the HC development team
about the HC problem. They were very responsive (surprised me, I must admit),
and I hope that I can give them the information they need to kill that bug.

<I can only conclude that this bug is a fluke, and that the programmer
<responsible for this is utterly irresponsible, incompetent, or both. I
<hope that the programmer gets another job in another industry soon, because
<if he (or she) ever works again, it will endanger more people!

Werner wrote:
>	I'd defend your freedom of speech to say that - but YOU deserve
>	to suffer the consequences of a lawsuit for "defamation"...
>
>	as far as I am concerned, the reputation that just got destroyed
>	here is the one of the poster of such irresponsible public
>	utterings, Mr. Alexis Rosen, who with such an article, not only
>	tarnishes SUM and all people and companies associated with it,
>	but also the site (dasys1.uucp) he himself is associated with
>	by virtue of posting from it.

There are absolutely no grounds for a suit. That is ridiculous. As for my
reputation, you are judging me far more hastily than I judged the program. The
information quoted was accurate and NOT irresponsible! (But, see below.) And
Werner goes WAY OVERBOARD coming down on dasys1 (the Big Electric Cat Public
UNIX), which has no control over postings of its subscribers. I didn't trash
Symantec, and they are more responsible for SUM than dasys1 is for me!

In any event, I wasn't any nastier than Werner is to me, and I had a much
better reason to be upset.

And Rich Siegel wrote:
>	You're half right (and therefore, half wrong. I don't know the exact
>circumstances, but I understand that  the defective partition DA *was*
>a fluke, and that it's already fixed. (I'll see if I can post it to
>comp.binaries.mac, because in spite of your idiotic comments, there iis
>something to be said for good software support.
>
>	If Les Herbst quits programming, there will never be another version
>of SUM, EVER. How would you like that?

Ummm. I still have a hard time believing he was responsible. If he was, then in
that respect (only) he _was_ irresponsible. BUT- even if he did write it-

You are absolutely right. He has saved far more lost files with MacZAP than he
will ever cause to be lost with the partition DA. It would be a bad day for
everyone if he quit, and he has built up enough "good karma" to easily
withstand this mistake. I hope he doesn't get another job in another industry,
but sticks around and writes more software. I'm still pissed, but I'm sorry.

I wrote:
<While I'm at it I'd also like to roast Symantec's testing and QA people.
<They really fell down on the job this time!

Rich Siegel wrote:
>	Not as I understand it. Like I said, this particular problem was a
>fluke that made it in between final QA and production. You can't place
>blame anywhere.

Just a minute! _Somebody_ screwed up big-time. So you CAN place it somewhere.
Please set me straight. Was Herbst the only person who touched Partition DA?
In any event, the person responsible really fucked up. No two ways about it.

As Jack S. Brindle (dxjsb@dcatla.UUCP) wrote:
>Come on Rich... It made it in between final QA & production? Sounds like
>y'all were rushing things more than just a bit. An addition to the
>package should have caused a return to QA for more testing. It appears
>that in the rush to get the package out y'all bypassed the proper
>procedures so that it wouldn't be late. Of course, the end user suffers.
>Perhaps someone at Symmantec SHOULD be transferred, or at least made to
>answer the tech support lines.

I wrote:
<I wonder who wrote the guardian program- the author of MacZAP or the author
<of the partition software? If the latter, I wouldn't put too much faith in
<that guardian, at least until a LOT of testing is done. What about the file
<de-fragmenter? That could also do a lot of damage.

Rich Siegel wrote:
>	ANY disk utility can do damage. The QA testing ensures that IF
>CORRECTLY USED, the product will not damage your disk. If it's any
>comfort, I've used HD TuneUp (the defragmenter), and have not had any
>problems. (I still use it now...)

	(Good. I just wish it could do DiskExpress's desktop compaction,
	but I guess you can't have everything (in version 1 :-). )

<(Now if I find out that the author of MacZAP IS the author of the partition
<software, I will really be worried. MacZAP has always worked wonders before.
<I hope they are different people...)
<
<Symantec- Consider this a bug report. I urge you to immediately recall this
<product until a fix is provided. It will cost me hours to recreate what I
<lost. The cost to others may be far worse.

Rich Siegel wrote:
>	Why are you so worried?? Because you've found a bug, you'll instantly
>assume that the whole package is bugridden and unusable? If this is the
>way you think, then I urge you NOT to buy any more software.
>
>	I don't yet know what the fix will be; I suspect a mailing to
>registered owners and a posting to the info services, but since I don't
>make the policies, don't hold me to that.

This too was excessive on my part. I apologize. Remember what I said though-
Paranoia is natural after this kind of disaster. I'm just glad I waited a while
until I was relatively calm before posting (I can just see it: "shoot the
programmer!", etc...).

Rich Siegel says goodbye:
>	I apologize for the tone of my posting, but sometimes....


Well, me too. I didn't want to start a flame war. Can we all just agree on the
following, and get on to more interesting things?

1) My bug report was correct. Someone at Symantec really did screw up.
2) The partition DA is very dangerous. Don't use it until it's fixed.
3) MacZAP Recover is wonderful, and so is Symantec (in general). I've never
   said any different, and I'm a satisfied customer of their other products.
4) Les Herbst is a paragon of virtue. No kidding. If he did screw up here (and
   not some junior programmer), we owe him more than enough gratitude for
   MacZAP Recover to cut him as much slack as he needs on this one...

Alright?

----
Alexis Rosen                       {allegra,philabs,cmcl2}!phri\
Writing from                                {harpo,cmcl2}!cucard!dasys1!alexis
The Big Electric Cat                  {portal,well,sun}!hoptoad/
Public UNIX                         if mail fails: ...cmcl2!cucard!cunixc!abr1
Best path: uunet!dasys1!alexis