From 8ec8ffd3a10bd4325c8775a0e5f1b88d59a76610 Mon Sep 17 00:00:00 2001 From: Richard Burgess <> Date: Thu, 2 Dec 1999 03:35:32 +0000 Subject: [PATCH] autocommit for file dated 1999-12-02 03:35:32 --- problems.txt | 133 +++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 133 insertions(+) create mode 100644 problems.txt diff --git a/problems.txt b/problems.txt new file mode 100644 index 0000000..954f56f --- /dev/null +++ b/problems.txt @@ -0,0 +1,133 @@ +This file relates to version 1.0 of the MMURTL +computer operating system that accompanies the +book: + + MMURTL-V1.0 + +This discusses known problems trying to run MMURTL, as well +as bugs, and portions of the code that are incomplete + +1) LOADING AND RUNNING MMURTL + + If MMURTL will not run on your system, you may have to + remove EMM386 from your Autoexec.BAT or Config.sys file. + EMM386 doesn't take kindly to hostile memory take-overs. + + If you receive an error during file system initialization or + hard disk device driver installation: + + a) Your disk controller may not be supported. + MMURTL has some problems with some of the newer 32 bit + disk controllers that move data directly between the + drive and memory. + All ISA driven controllers seem to work fine (I have + tested more than 12 different ISA IDE controllers). + I have also tested many PCI based controllers and they + seem to work ok too. I have no problems running MMURTL + on my 233 Mhz Pentium MMX with a PCI controller so it + can be done! I have no idea what effect the UDMA + 33 and 66 PCI controllers will have. + + b) SCSI disks and other SCSI devices are not supported. + (no SCSI driver) + + c) Your data may be compressed by MS-DOS or another product + such as Stacker. MMURTL does not support disk data compression. + + d) If you are running OS/2 V2.x or Warp (V3.x or V4.x), + MMURTL doesn't recognize the multiboot partition correctly. + You can copy the MMSYS directory to a floppy and run it + from there to test it. + +2) WHAT'S NOT DONE?? + + FILESYSTEM: + + a) The file system call to remove a directory is not completed + so I didn't include it. Use MS-DOS to get rid of directories if + needed. + + DLLs: + + a) DLLs are not supported. AND NEVER WILL BE - I hate them... + + LOADING DEVICE DRIVERS: + + a) Device drivers can't be loaded dynamically. If you need to add + a driver you will have to build it into the OS (sounds like Unix...) + That's actually quite easy to do, so I didn't strain myself + getting it done. + + DASM LIBARARY SEARCH FUNCTION + + a) The search function (.SEARCH command) is not fully implimented. + The work around is quite easy, and really there is not much + library code to search through at this point in time. + Simply include the ASM modules in the ATF file. All of the sample + programs and the operating system are built this way right now. + They should provide you with good examples. + + COMMAND LINE INTERPRETER + + a) There is no wildcard feature in the file system or any + template matching code in a library. This means *.* has no + effect. The copy, rename, and delete functions all take single + filenames. + +3) BUGS + + OPERATING SYSTEM + + a) On occasion, when a program is beating the filesystem to death + and you attempt to load and run another CLI, the system crashes. + It doesn't even make it into the debugger. It is apparently a + processor double fault. Still looking for this one.... + + b) The CTRL-ALT-DELETE function is a rather nasty way to shut down + an application. It was a late addition, but was necessary. On + occasion is causes some problems. The system still functions and + tasks appear to be running fine - but every once in a while when + you use it, you can't switch to the other applications. This IS + a problem I'm still looking for. + + CM32 C Compiler + + a) Occasionally when you have accidentally "nested" a function + (which is not allowed in C) the compiler goes into a loop... + It's churning away somehwere. I just haven't found out where + that is yet. + +4) An updated Filesystem is included for download but not incorporated + - you have to do it by replacing the file. I did not want to +mess with the code so it folloows the book exactly. Otherwise I +would have create a confusing mess! + +This is STILL version 1.0. You WILL find many other problems. +Let me know if you get the time. Send E-Mail to rburgess@aol.com. +(Please, bugs only - and yes, the AOL email address is still mine +after 6 years.) + +Quite obviously, neither myself nor the publisher has time to really +support a product of this complexity. You're on your own. BUT at least +you DO have the source code. + +Thanks, and I hope you get something out of all this! + +Richard Burgess +Alexandria, VA. +1995-1999 + +P.S. Most of this text is 6 years old but still applies. +HOWEVER, with a publisher like Sensory willing to keep a BBS +up and running for user exchange and support - maybe MMURTL +will take off like I wanted it to in the first place... +I'm not sure how much time I'll have to help, but it is tempting +and MMURTL really is a sweet OS. I've never run it on a PII +or PIII - let's hear it out there.... whoa... who's got the guts? +But it does run on my IBM Aptiva 233Mhz MMX... Hmmm. +Not bad for 5 years later. It does cough on the bigger disk +quite often though. + + + + -- 2.40.0