me128-aw@kepler.Berkeley.EDU (me128 student) (12/06/88)
Help! I have been having serious problems!
My Amiga has been crashing CONSTANTLY lately, starting about the time I
got 1.3. When it does, I will be flipping thru screens and an interlaced
screen will come up non-interlaced (so that I only see the upper half).
Everything is still fine until I try flipping this screen to back, using
either the gadgets or keyboard. At that point, the machine totally locks
up (keyboard & mouse), responding to nothing non-vulcan. Even the GOMF
button does nothing. This has occured with a variety of programs and
situations, but I can always repeat it with the following:
run Diga! ; Run Diga. Prefs are set to an interlaced screen
MWB -i ; initialize multiple workbench
MWB N n ; Open a non-interlaced workbench
run Memacs file ; run an editor on a non-interlace WB
[ flip this screen to back to reveal diga ]
[ diga screen is non-interlaced ... click...BOOM ]
Please note that I am not always running diga or mwb when this occurs,
this is just one such occurance. What is causing this!!!?
I am running an interlaced, overscan (671x470) workbench, if this is of any
help. It seems to me that the delacing forewarns of the crash, so there
ought to be a way to remedy the problem. I am also using SetAlert in my
startup sequence, an boot up with either KS1.2 or KS1.3 (it has crashed
with both)
2) RAD:
For the life of me, I can't recover with 1.3 and RAD: EVER! Is this a
slow ram problem? I use an 880k RRD. My system is as follows:
Amiga 1000 with 512k
Microbotics 2 Meg Expansion Ram
Chris Erving's 512k memory Hack
Microbotic Stardrive SCSI controller w/ 20 meg hard drive
Interlaced, overscan (671x470) Workbench screen
Gomf button, video hack, audio hack... (Probably irrelevent, but included
for sake of completeness.
My first instinct is to question the slow piggybacked ram. Does Fastmemfirst
recognize non-c00000 slow ram? Shouldn't the 880k size force RAD: to be in
fast ram?
3) How can I get 2 recoverable ram disks? Experimentation shows that 2 RADs
don't get along at all. Even RAD: and VD0: don't like each other. BOOM!
My Idea is: I want to run Rocket Ranger entirely from ram. since RR uses its
own disk-fetching code, it won't work with RAM:. I found that I can diskcopy
one disk to RAD: to speed it up considerably; but I want NO WAITING, PERIOD!
Thanks
-Vincent H. Lee
scotth@harlie.SGI.COM (Scott Henry) (12/06/88)
From article <27024@ucbvax.BERKELEY.EDU>, by me128-aw@kepler.Berkeley.EDU (me128 student): > Help! I have been having serious problems! > > My Amiga has been crashing CONSTANTLY lately, starting about the time I > got 1.3. When it does, I will be flipping thru screens and an interlaced >[... stuff deleted...] > I am running an interlaced, overscan (671x470) workbench, if this is of any > help. It seems to me that the delacing forewarns of the crash, so there > ought to be a way to remedy the problem. I am also using SetAlert in my > startup sequence, an boot up with either KS1.2 or KS1.3 (it has crashed > with both) I have also had this problem, and a similar hardware configuration, BUT, it has only happened once to me, Though I don't use Diga! or MWB. > 2) RAD: >[... second question deleted as I don't know the answer...] > > 3) How can I get 2 recoverable ram disks? Experimentation shows that 2 RADs > don't get along at all. Even RAD: and VD0: don't like each other. BOOM! > My Idea is: I want to run Rocket Ranger entirely from ram. since RR uses its > own disk-fetching code, it won't work with RAM:. I found that I can diskcopy > one disk to RAD: to speed it up considerably; but I want NO WAITING, PERIOD! I have figured out how to make RAD: and VD0: co-exist. It is actually fairly simple: when you first load up the system from a cold boot, VD0: must be loaded first, AND it must have stuff in it (MORE THAN one cylinder's worth of data) BEFORE you mount RAD: for the first time (ie: create it). I copy a small directory's worth of stuff (~6 files, >= 60k) into VD0:, mount RAD:, copy stuff into RAD:, delete the stuff I copied into VD0:. This has worked for me for several weeks (except for run-away programs that have clobbered either/both of RAD and VD0). BTW, I only use a small (9 cylinder) RAD for auto-booting my hard-drive (also a StarDrive). I recently saw a posting about how to make 2 RAD: disks co-exist in memory at the same time, you might look through recent posting history. > Thanks > -Vincent H. Lee You're welcome, Scott Henry -- Scott Henry <scotth@harlie.sgi.com> {or, also on the Internet:} <skywalker@cup.portal.com> #include <std_disclaimer.h>
gmg@hcx.uucp (Greg M. Garner) (12/06/88)
In article <27024@ucbvax.BERKELEY.EDU>, me128-aw@kepler.Berkeley.EDU (me128 student) writes: > Gomf button, video hack, audio hack... (Probably irrelevent, but included > for sake of completeness. > What the heck is the video hack? I have the audio hack and the 512 internal hack, but never heard of a video hack! Please enlighten me..... Greg Garner 501-442-4847 USENET: ...!uunet!harris.cis.ksu.edu!hcx!gmg
bty00298@uxa.cso.uiuc.edu (12/07/88)
The problem with flipping screens is not a new problem with 1.3. It has always been a problem with the OS. It can easily be replicated by opening two interlaced screens and one non-interlaced screen. Your problem will be duplicated as the screens are flipped. (It happens to me often) I heard somewhere that this is an inherent bug in Kickstart 1.2/1.3 that setpatch does not fix. ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Brian Yamanaka a struggling student
sc00250@uxa.cso.uiuc.edu (12/07/88)
I've been using rad: and vd0: in conjunction with each other since early June. And to tell you the truth, I have not had the problems that others have described here. One of the other responses indicated the need to copy files into vd0:, mount Rad:, then copy form vd0: to rad:, and finally delete the stuff from vd0:. Forgive my bluntness, but this is totally convoluted. There is absolutely no reason to do this. My system consists of a 512K 1000 (the only real Amiga :), with a single drive and a 2 meg ram board from Progressive Peripherals. I have been copying my boot-up disk to rad: then just boot off Rad: whenever the need arises. If you still have problems with Rad: then email me and I'll send you a copy startup-sequence and mountlist. Overall, I've been content with rad:, the only gripe I have about it is it sometimes kills my kickstart!!!!!!!! Why ??????????? If you have the answer to this problem then please share the answer with someone who is in the dark.
cjp@antique.UUCP (Charles Poirier) (12/07/88)
In article <27024@ucbvax.BERKELEY.EDU> me128-aw@kepler.Berkeley.EDU (me128 student) writes: >Help! I have been having serious problems! > >My Amiga has been crashing CONSTANTLY lately, starting about the time I >got 1.3. That is coincidence. >When it does, I will be flipping thru screens and an interlaced >screen will come up non-interlaced (so that I only see the upper half). >Everything is still fine until I try flipping this screen to back, using >either the gadgets or keyboard. At that point, the machine totally locks >up (keyboard & mouse), responding to nothing non-vulcan. Even the GOMF >button does nothing. This has occured with a variety of programs and... This is a long-standing bug in the system. The bug causes crashes when you have more than one Interlaced screen plus a Non-interlaced screen simultaneously, and try to flip among them. Occasionally it fails to crash, but don't bet on it. It's safe to drag screens down by mousing the title bar. Sometimes this lets you get to a place from which you can kill one or more screens so that remaining screens can be flipped safely. So remember - don't use TWO interlace plus a non-interlace and then screen-flip. -- Charles Poirier (decvax,ucbvax,mcnc,attmail)!vax135!cjp "Docking complete... Docking complete... Docking complete..."
peter@sugar.uu.net (Peter da Silva) (12/07/88)
In article <22935@sgi.SGI.COM>, scotth@harlie.SGI.COM (Scott Henry) writes: > I have figured out how to make RAD: and VD0: co-exist. It is actually fairly > simple: when you first load up the system from a cold boot, VD0: must be > loaded first, AND it must have stuff in it (MORE THAN one cylinder's worth > of data) BEFORE you mount RAD: for the first time (ie: create it). This is weird, because I have no problem with: addmem 200000 600000 mount RAD: mount vd0: if not exists RAD:c diskcopy df0: rad: endif if not exists vd0:c run copy <nil: >nil: df1: vd0: all endif ... Maybe there's some advantage to non-autoconfig RAM after all? -- Peter da Silva `-_-' peter@sugar.uu.net Have you hugged U your wolf today? Disclaimer: My typos are my own damn busines#!rne
me128-aw@kepler.Berkeley.EDU (me128 student) (12/08/88)
In article <1369@cseg.uucp> gmg@hcx.uucp (Greg M. Garner) writes: > >In article <27024@ucbvax.BERKELEY.EDU>, me128-aw@kepler.Berkeley.EDU (me128 student) writes: >> Gomf button, video hack, audio hack... (Probably irrelevent, but included >> for sake of completeness. > > What the heck is the video hack? I have the audio hack and the 512 internal > hack, but never heard of a video hack! Please enlighten me..... > The "video hack" for A1000s appeared in Amazing Computing Vol2 #7, and involved changing some resistors to improve the color balance. By the way, while we're talking about hacking amigas, I have also grounded my PAL chips, added a filtering capacitor in my starboard, and performed the 1080 monitor hack to fix the interlacing-offset and audio. I have also squished the vertical height and superglued a piece of ferrite to the end of the horizontal coil in the 1080 so I can see the 671/470 wb screen. Yes, I love having a HACK-able computer! -Vincent H. Lee
me128-aw@kepler.Berkeley.EDU (me128 student) (12/08/88)
In article <2450@antique.UUCP> vax135!cjp (Charles Poirier) writes: > >This is a long-standing bug in the system. The bug causes crashes when >you have more than one Interlaced screen plus a Non-interlaced screen >simultaneously, and try to flip among them. Occasionally it fails to >crash, but don't bet on it. It's safe to drag screens down by mousing >the title bar. Sometimes this lets you get to a place from which you >can kill one or more screens so that remaining screens can be flipped >safely. > Thanks for the Help. A few questions: I have found that the it will NOT crash if the non-interlaced screen is opened before the second interlaced screen (wb is the first). I guess this makes sense if the non-interlaced screen is trashing some interlaced bit somewhere when it's opened. 1) Once it did this while I was playing a beta-version of a new game. The game did non open or flip screens, but was just bringing up a requestor in it's interlaced screen. When it did, the screen flashed to non-interlace (BOOM!). How could this occur? 2) Why hasn't anybody written a patch to fix this? It seems to me that this SERIOUSLY undermines the multitasking nature of the Amiga. Shouldn't it be possible to write a little program which monitors the screen ordering and does some magic which the Guru is thinking of dropping by? 3) Could anything make this occur more frequently? I have had an identical system configuration for quite some time now, and have had relatively few crashes until recently. As I recall, I have had interlaced and non-interlaced screens together often with no problems in the past. 4) Is the interlacing a function of having separate odd and even source bitmaps, or just the way the screen is drawn. In other words, if I use SetLace or something like it on the non-interlaced screen, will that prevent a crash?
me128-aw@kepler.Berkeley.EDU (me128 student) (12/09/88)
In article <111400012@uxa.cso.uiuc.edu> sc00250@uxa.cso.uiuc.edu writes: > >I've been using rad: and vd0: in conjunction with each other since early June. >One of the other responses indicated the need to copy files into vd0:, mount >Rad:, then copy form vd0: to rad:, and finally delete the stuff from vd0:. >Forgive my bluntness, but this is totally convoluted. There is absolutely >no reason to do this. > have had problems when i try to diskcopy to rad: with vd0: existing. If I do so, then both ramdisks will show the contents I just diskcopied, not just RAD:. Sometimes vd0: will also lose its name, and the system will just crash. It continues to crash until I turn the machine off and then on again. I haven't tried it in some time, as I wasted a few hours on it some time ago.
gmg@hcx.uucp (Greg M. Garner) (12/09/88)
In article <27064@ucbvax.BERKELEY.EDU>, me128-aw@kepler.Berkeley.EDU (me128 student) writes: > > By the way, while we're talking about hacking amigas, I have also grounded > my PAL chips, added a filtering capacitor in my starboard, and performed the > 1080 monitor hack to fix the interlacing-offset and audio. I have also > squished the vertical height and superglued a piece of ferrite to the end > of the horizontal coil in the 1080 so I can see the 671/470 wb screen. Yes, I > love having a HACK-able computer! > > -Vincent H. Lee Hey, is this starboard hack something I need to do to mine? (I can't resist a good hardware hack). I am sending off for the A1000 schematics so that I can start hacking the expansion bus next. I built a small board with a VIA on it to control things (like solid state relay-type things), but I did something wrong with the handshaking so I never got it working (yet). Soon! If I ever get it to work, I'll post all the particulars right here for all you other dedicated hardware hackers. By the way, if anyone has an idea on how to make a non-autoconfig board talk to the amiga (Autoconfig comes later, I hope), let me know the details. I need to know how to let the amiga know that I have a chip out there, as bringing the vpa* line low is not enough to do the trick. I suspect that I must bring another line low to tell the amiga to stretch the cycle for the 6800 style bus access. Thanks! Greg Garner 501-442-4847 USENET: ...!uunet!harris.cis.ksu.edu!hcx!gmg
me128-aw@kepler.Berkeley.EDU (me128 student) (12/10/88)
In article <1390@cseg.uucp> gmg@hcx.uucp (Greg M. Garner) writes: >Hey, is this starboard hack something I need to do to mine? (I can't resist >a good hardware hack). I am sending off for the A1000 schematics so that .... The starboard hack really just involves putting a filter capacitor on one of the lines on the bus. I did it some time ago on the advice of a friend. It seems that the starsdrive puts too much noise on the power lines that a certain 020 add-on board won't work. -Vincent H. Lee
FelineGrace@cup.portal.com (Dana B Bourgeois) (12/10/88)
Peter says he has no problems with his startup sequence: example deleted Peter, perhaps it is the fact that you have several megs of Ram. I suspect that some of the crashes I have experienced have been memory related. Dana
riley@batcomputer.tn.cornell.edu (Daniel S. Riley) (12/12/88)
In article <12413@cup.portal.com> FelineGrace@cup.portal.com (Dana B Bourgeois) writes: >Peter says he has no problems with his startup sequence: > > example deleted > >Peter, perhaps it is the fact that you have several megs of Ram. I suspect >that some of the crashes I have experienced have been memory related. I had problems with rad:, ram: and vd0: co-existing on my 2.5 Meg A1000. Usually the system would just hang after a write to ram:. After much futzing about, I discovered that, at least on my system, these ram-disks are very sensitive to the order they are mounted in. If I reference ram: (i.e., copy something to it, or dir it), then mount rad: and reference it, and then mount vd0: and reference it, the system seems very nicely stable. If I vary the order (in particular, if I mount vd0: before rad:), it usually won't even complete the startup-sequence (it hangs in a setenv, with env: is assigned to ram:env). This is a bit of nuisance, but I guess I'm getting used to it. I am also totally clueless why it works one way and not the other. -Dan Riley (dsr@lns61.tn.cornell.edu, dsr@crnlns.bitnet) -Wilson Lab, Cornell U.