Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

BLACK DAY IN HELL my work gets fried

3 views
Skip to first unread message

Tom Bryce

unread,
Jan 1, 1995, 10:25:02 PM1/1/95
to

I believe CodeWarrior FRIED my work. I'm wondering if anyone else has
had a similar experience, and if there is any known bug in CW that
could do this.

I spent the last few days of my vacation here relaxing and programming
with codewarrior. Now, all of a sudden, when I try to open up a source
code file, the contents of that file are COMPLETELY GONE and replaced
with the contents of a different source code file. The file that got
fried is the one I've been working on heinously for the past few days.
I have one backup, which I took today, and which got fried also.

I believe CodeWarrior saved into the wrong file and fried it. I've
observed similar errors and flaws in file management, i.e., where you
remove a header file completely from the project and delete it from
everywhere and every reference to it whatsoever from anyplace in
existence and it still insists on searching for it when you click
"Header files" in the find box. Also, sometimes if you keep backups in
random places and open them while working with a project and the file
has the same name as the backup, some weird things can happen there
too.

I compiled the project this morning, worked fine. Put in a few changes,
then copied the folder with option-copy into the backup folder.
However, the source code must have been FRIED before the option-copy
because both are FRIED.

Has anyone else had any experience with the problem I experienced? If
it is the case that CW (CW 4.5 68k on PPC 7100/66 running 7.1.2) did
fry the hell out of my intense labor for the past few days, and this is
a known bug, why didn't MW send every registered user an email and a
letter in the mail warning them about the bug?

I don't want to point the finger at CW too quickly, but I can't imagine
what else could have caused this, and I have observed plenty of quirks
in the file management of the project manager. I hope it's not CW
because they are doing in general a great job, but if this is due to a
bug in CW it's quite disturbing.

I might as well go crawl into a hole now, because I worked like hell on
this and won't have the time again once school starts. Was hoping to
release a new version of my freeware, much-improved.

------------------------------------------------------------------------

Tom Bryce
for PGP public key finger tjb...@amherst.edu

Tom Bryce

unread,
Jan 1, 1995, 11:26:33 PM1/1/95
to

Okay.

The black day in hell has ended. Obviously, I thought, if it was fine
this morning, the data is still hiding away SOMEWHERE on the disk, and
a plaintext scan yielded a not-terribly-outdated version that I can
update.

However, this does disturb me a little. Besides teaching me a good
lesson about backing up permanently onto locked disks, I wonder if CW
was the problem. I'm inclined to think so, but unless anyone else has
had a similar experience, it could be just my bad luck...

Tom

Tom Bryce

unread,
Jan 2, 1995, 2:57:30 AM1/2/95
to

I've been thinking - this also could be due to a bug in the system
software. I don't use any extensions that patch file io, or at least I
don't think I do.

If there are no known bugs in CW or the system that would cause this,
I'll just chalk it up as a weird unexplainable thing and keep better
backups in the future.

Harvey J. Cohen

unread,
Jan 2, 1995, 11:02:07 AM1/2/95
to
Subject: Re: BLACK DAY IN HELL my work gets fried
From: Tom Bryce, t...@acpub.duke.edu
Date: 2 Jan 1995 07:57:30 GMT
In article <3e8bla$j...@news.duke.edu> Tom Bryce, t...@acpub.duke.edu
writes:

>
>I've been thinking - this also could be due to a bug in the system
>software. I don't use any extensions that patch file io, or at least I
>don't think I do.
>
>If there are no known bugs in CW or the system that would cause this,
>I'll just chalk it up as a weird unexplainable thing and keep better
>backups in the future.
>

Interesting thread. I have used CW 4.5 extensively on both a PowerMac
8100/80 and a Quadra 840av, and have never experienced the problem you
have described herein. I also backup religously (:-)). Now, I work
under system 7.5, and the bugs you describe are potentially due to system
7.1.2, which was a quick and dirty port of 7.1 for the PPC. I suggest
that you cough up the $99 plus s/h and get system 7.5. You may be able
to get system 7.5 for less money if you got your system after July 1994.
Check with a decent mail order house (MacConnection, MacWarehouse,
MacZone, etc...) for upgrade details.

Note: I sure CW 5 will work with system 7.1.2, but I wouldn't bet on CW6
or CW7.
Harvey J. Cohen, Ph.D
<hjc...@covina.lightside.com>

"If at first you don't succeed, try, try again" - popular
aphorism
"If at first you don't succeed, quit. It isn't worth it" - Bill Clinton
"If at first you don't succeed, buy the company" - Bill Gates

Tom Bryce

unread,
Jan 2, 1995, 11:59:15 AM1/2/95
to
In article <3e981v$g...@covina.lightside.com>

Harvey J. Cohen <hjc...@covina.lightside.com> writes:

> that you cough up the $99 plus s/h and get system 7.5. You may be able
> to get system 7.5 for less money if you got your system after July 1994.

Hmm... I called MacConnection, and they said I can get it for $35.
I've heard that 7.5 was just rolling a bunch of patches into 7.1.2, and
that it is a memory hog, and you can get the same functionality if you
want it with freeware extensions... is this not true?

Florian -FDj- Dejako

unread,
Jan 2, 1995, 2:31:11 PM1/2/95
to
In article <3e8bla$j...@news.duke.edu>, t...@acpub.duke.edu (Tom Bryce) wrote:

> I've been thinking - this also could be due to a bug in the system
> software. I don't use any extensions that patch file io, or at least I
> don't think I do.
>
> If there are no known bugs in CW or the system that would cause this,
> I'll just chalk it up as a weird unexplainable thing and keep better
> backups in the future.

Before you post your next article, in which you will probably write that
your mother is guilty for all this coz she ever thought of talking to that
guy who was to be your daddy then... ack, never mind.

Did you analyze your hard disk with a suitable tool to do that? I had bad
experiences with Norton Utilities (including 3.1) so far, so I recommend
you go for MacTools 4.0 Pro, maybe take your HD to a local Apple dealer
who will try and have a look at it with that tool if you don't want to buy
it (you are not gonna steel it, are you? Remember: It's midnight... do you
know what's on TV?).

Did you recently have a system breakdown while saving? Probably with some
sorta write disk cache enabled (I dunno how Apple's disk cache works, but
maybe you formatted your HD with a caching driver). That could have very
well caused the trouble.

Besides, I love CopyDoubler's scheduled background-copies. Just tell
CopyDoubler to make a backup of your project sources every 2 working hours
and you will feel fine. If you only got one HD I recommend you partition
it. For some reasons: on large HDs you get less sector lag, and you can
tell CopyDoubler to keep a backup of your projects on the other partition.
If one partition is corrupt (which happens far oftener than a broken hard
disk) then you have always got your backup on the other volume.

Well, just my 2p, I can understand that you kinda felt lost with your
source code going into something like /dev/nil, but I didn't like the way
you bashed Metrowerks in your first post coz Metrowerks is the last
company not to tell their customers of bugs like the one you experienced
(which I think wasn't due to MW but due to a corrupt file system on your
HD).

Still pretty good year,

FDj

_____________________________________________________________________
Florian -FDj- Dejako Don't feed the people
dej...@informatik.tu-muenchen.de But we feed the machines
f...@muc.de, IRC: FDj on #macdev (Rush)
_____________________________________________________________________

MW Ron

unread,
Jan 2, 1995, 2:59:58 PM1/2/95
to
t...@acpub.duke.edu (Tom Bryce) Writes:

>>I believe CodeWarrior FRIED my work. I'm wondering if anyone else has
had a similar experience, and if there is any known bug in CW that could
do this.

Tom,
I don't want to be an alarmist, however I have seen this reported
twice before with the PPC and in one instance it was the controler and the
other it was the HD. It seems the disk intensiveness of CodeWarrior is
the canary in the mine shaft. In the other cases it steadily worsened and
soon occured on all other files/programs.
Now I hope I am wrong for your sake but I really encourage you to back up
everything and get Norton Disk Doctor to do an intensive disk check and
fix any problems you have.

METROWERKS Ron Liechty
"Software at Work" MW...@metrowerks.com

Patrick Coskren

unread,
Jan 2, 1995, 3:13:57 PM1/2/95
to
In article <3e981v$g...@covina.lightside.com>, Harvey J. Cohen
<hjc...@covina.lightside.com> wrote:

> that you cough up the $99 plus s/h and get system 7.5. You may be able
> to get system 7.5 for less money if you got your system after July 1994.

Much less money. If I recall correctly, it's about $10 to cover shipping
costs and the like. You have to fax them a copy of your receipt, along
with a few other details like your credit card number. When I called,
they said to allow 6 to 8 days for delivery. Call 1-800-SOS-APPLE for
details. Any or all of this may have changed since I called them 2 months
ago, or may be inaccurate in the first place. :-)

-Patrick

Mike Cohen

unread,
Jan 2, 1995, 3:51:04 PM1/2/95
to
t...@acpub.duke.edu (Tom Bryce) writes:


>I believe CodeWarrior FRIED my work. I'm wondering if anyone else has
>had a similar experience, and if there is any known bug in CW that
>could do this.

>I spent the last few days of my vacation here relaxing and programming
>with codewarrior. Now, all of a sudden, when I try to open up a source
>code file, the contents of that file are COMPLETELY GONE and replaced
>with the contents of a different source code file. The file that got
>fried is the one I've been working on heinously for the past few days.
>I have one backup, which I took today, and which got fried also.

Have you run Norton Utilities and/or MacTools? That sounds like a classic
symptom of a damaged directory (cross-linked files and/or a bad volume bitmap
resulting in files getting saved over an existing file).
--
Mike Cohen - is...@netcom.com
NewtonMail, eWorld: MikeC / ALink: D6734 / AOL: MikeC20
Home Page: ftp://ftp.netcom.com/pub/is/isis/home.html
PUSH THE BUTTON, FRANK... OR SOMEONE?

Harvey J. Cohen

unread,
Jan 2, 1995, 6:42:03 PM1/2/95
to
Tom,

I concur completely with all the other posters here. Please run a disk
diagnosis and repair utility (even if it is only Disk First Aid). Either
Mac Tools Deluxe or Norton Disk Doctor (both from that famous purveyor or
vaporware, Symantec) would be excellent. For instance, I run MacTools
Deluxe "Diskfix" twice weekly on both my Quadra 840av and PowerMac
8100/80, and routinely find errors - mostly in the directory. These
errors, if propagated, can a) ruin your backups and b) eventually crash
the system. If this sounds familiar, it is to me. I ran into that
pothole a few years back, and paid dearly for it.

mari...@pulua.hcc.hawaii.edu

unread,
Jan 3, 1995, 12:58:56 AM1/3/95
to
In article <3e7rme$c...@news.duke.edu>, t...@acpub.duke.edu (Tom Bryce)
wrote:

> I believe CodeWarrior FRIED my work. I'm wondering if anyone else has
> had a similar experience, and if there is any known bug in CW that
> could do this.
>

I am using CW/3, and this happened to me. I think it happened when I
renamed a file that was in a project. When I restarted, everything was
back to normal. I ran Norton, disk first aid, ect. Nothing wrong with the
drive. I sespect Now utils.

-Nathan

PS: Sorry for the messy post, keyboard is buried under cables.

Tom Bryce

unread,
Jan 3, 1995, 8:26:30 AM1/3/95
to

I called apple - the $10 S&H upgrade deal is still in effect. I bought
my mac in september, so I'm eligible, though I will have to wait over a
month for 7.5.

Peter Bierman

unread,
Jan 3, 1995, 3:05:35 AM1/3/95
to
In article <3e9bd3$5...@news.duke.edu>, t...@acpub.duke.edu (Tom Bryce) wrote:

> In article <3e981v$g...@covina.lightside.com>
> Harvey J. Cohen <hjc...@covina.lightside.com> writes:
>
> > that you cough up the $99 plus s/h and get system 7.5. You may be able
> > to get system 7.5 for less money if you got your system after July 1994.
>
> Hmm... I called MacConnection, and they said I can get it for $35.
> I've heard that 7.5 was just rolling a bunch of patches into 7.1.2, and
> that it is a memory hog, and you can get the same functionality if you
> want it with freeware extensions... is this not true?

This is not true.

7.5 truely does suck less. It's all of the fixes in Syste Update 3.0, all
of the PPC native code rolled in, all of the system enablers rolled into
the system file, and a few nice extensions thrown in to give marketing
something to hype.

Pull everything you don't want, and your RAM usage shouldn't go up more
than 50k. I installed it on a friend's PB145b, and since he was using all
the extensions that come with 7.5, after the upgrade his disk usage was
DOWN, and his RAM usage was THE SAME.

And the disk cache finally works right. (Tip 'o the day)

-Peter

--
"The other major kind of computer is the "Apple," which I do not recommend,
because it is a wuss-o-rama new-age computer that you basically just plug in
and use. This means you don't get to participate in the most entertaining
aspect of computer owning, which is trying to get the computer to work."
- Dave Barry 2/6/94 I am: bie...@caelab1.cae.wisc.edu

Steve Bryan

unread,
Jan 3, 1995, 12:22:01 PM1/3/95
to
In article <3e9bd3$5...@news.duke.edu>, t...@acpub.duke.edu (Tom Bryce) wrote:

> Hmm... I called MacConnection, and they said I can get it for $35.
> I've heard that 7.5 was just rolling a bunch of patches into 7.1.2, and
> that it is a memory hog, and you can get the same functionality if you
> want it with freeware extensions... is this not true?

No, it is not true.Don't believe the wise guys who are always
pontificating about system software but in most cases don't have a clue.
There are very good reasons Apple releases new versions of system software
and you should always avail yourself of the most recent version. It might
be useful for Apple to always release a list of all the bugs that are
fixed in a new version of system software but I can understand their
reluctance to do so from a marketing perspective. The question of memory
useage is not a real problem as others have pointed out. If you want the
new capabilities (eg Quickdraw GX) then buy more memory and install it.
Simple.

--
Steve Bryan InterNet: sbr...@maroon.tc.umn.edu
Sexton Software CompuServe: 76545,527
Minneapolis, MN Fax: (612) 929-1799

Marc Nimchuk

unread,
Jan 5, 1995, 1:44:58 AM1/5/95
to
Harvey J. Cohen (hjc...@covina.lightside.com) wrote:
: Tom,

Ok, I get directory/volume bit map errors almost weekly on *MY* Quadra
800 + PPC card as well. Under System 7.5. Is there something in S7.5
that doesnt like the PDS card and/or my hard drive?

Yeah, yeah, whats this to do with CW, but the thread caught my attention...

Dave Falkenburg

unread,
Jan 13, 1995, 12:03:47 PM1/13/95
to
In article <3eg4ha$1e...@quartz.ucs.ualberta.ca> Marc Nimchuk,

mnim...@gpu.srv.ualberta.ca writes:
>Ok, I get directory/volume bit map errors almost weekly on *MY* Quadra
>800 + PPC card as well. Under System 7.5. Is there something in S7.5
>that doesnt like the PDS card and/or my hard drive?

Do you hard reset your machine alot? (or use RB in Macsbug?)

One of the things that 7.5 includes is a faster disk cache which probably
flushes the volume less often. The old disk cache has similar behavior,
except that almost everyone I knew used a 32K cache (which always got
dumped).

Whenever you hard reset your machine with files open you have the
potential
to scrag your disk.

-Dave Falkenburg
-Apple Computer, Inc.

ObjFactory

unread,
Jan 13, 1995, 8:08:35 PM1/13/95
to
Dave Falkenburg <fal...@apple.com> wrote:

>Do you hard reset your machine alot? (or use RB in Macsbug?)

>One of the things that 7.5 includes is a faster disk cache which probably
>flushes the volume less often. The old disk cache has similar behavior,
>except that almost everyone I knew used a 32K cache (which always got
>dumped).

>Whenever you hard reset your machine with files open you have the
>potential to scrag your disk.

Perhaps it should be renamed the disk crash? Seriously, did it ever occur
to any of the authors that this behavior might be considered a bug?


Bob Foster
Object Factory
objfa...@aol.com

John W. Baxter

unread,
Jan 14, 1995, 1:56:31 AM1/14/95
to
In article <1995Jan13.1...@guest.apple.com>, Dave Falkenburg
<fal...@apple.com> wrote:

>
> Whenever you hard reset your machine with files open you have the
> potential
> to scrag your disk.

Indeed. Which is why I *always* (well, almost always) run Disk First Aid
to check the previously-mounted volumes after restarting from any such
events. Gives me time to get a soda (Diet Pepsi, not Jolt, sorry). And
to curse Apple's telling me several hundred times that I can't repair the
startup volume. I really did figure that out after the first 37 times.
Dimming the Repair button, and repainting the instructions pane with an
explanation would be plenty. Grrr.

--John

--
John Baxter Port Ludlow, WA, USA [West shore, Puget Sound]
Isn't C fun?
jwba...@pt.olympus.net

Anton Rang

unread,
Jan 16, 1995, 10:20:26 PM1/16/95
to
(This thread no longer really has anything to do with CodeWarrior.)

In article <3f786j$9...@newsbf02.news.aol.com> objfa...@aol.com (ObjFactory) writes:


>Dave Falkenburg <fal...@apple.com> wrote:
>
>>One of the things that 7.5 includes is a faster disk cache which probably

>>flushes the volume less often. [...]


>>Whenever you hard reset your machine with files open you have the
>>potential to scrag your disk.
>
>Perhaps it should be renamed the disk crash? Seriously, did it ever occur
>to any of the authors that this behavior might be considered a bug?

Perhaps, but the problem isn't in the cache, it's in the file system
directory manipulation. Given that HFS uses a B-tree for its
directory, it ought to be possible to implement an atomic update
scheme by always writing changed blocks to a new block, and only then
updating the block above it in the hierarchy. This would give better
protection from crashes, at the cost of much slower directory updates.

Someone could probably write a hack for this, but it'd be cleaner
just to have a separate, more robust file system as an option for
environments where this is really, *really* important. Of course,
then you really want to be only accessing files in a robust manner,
too...database logging and all. (Maybe when the new file manager
comes out, someone will implement this?)

I think you could make a good argument that directory blocks in the
disk cache should be write-through, though, to at least make crashes
which cause directory corruption less likely. (Or they could be
flushed periodically, similarly to System Update 3.0's system file
flush patch. Hmm, that'd be easy to implement as an INIT.)
--
Anton Rang (ra...@winternet.com)

0 new messages