I tried installing AMouse and now I don't have a mouse object anymore

Some of the latest AMouse installation packages (in the 2.7x release numbers) have a problem when registering the class replacement for the mouse object. As a result, after installing AMouse you can find you don't have anymore a mouse object in the "System Setup" folder.
Paul Ratcliffe posted a simple solution to this problem: copy the following lines into a text file with the name mouseobj.cmd (or whatever.cmd), then execute the script from the command line or simply by double-clicking the icon.

/**/
call RxFuncAdd 'SysLoadFuncs', 'REXXUTIL', 'SysLoadFuncs'
call SysLoadFuncs
class = 'WPMouse'
title = 'Mouse'
location = '<WP_CONFIG>'
setup='NOCOPY=YES;NODELETE=YES;NOPRINT=YES;
DEFAULTVIEW=SETTINGS;OBJECTID=<WP_MOUSE>;'
call SysCreateObject class, title, location, setup, 'R'

Please note that the "setup=..." line, which is wrapped here due to space limits, must be joined back to form a single line of text.

Many thanks to Dario Ragazzini for the suggestion!

Trap E in windowed DOS and Win-OS2 sessions

With certain video cards, you can experience a trap every time you attempt to run a windowed DOS or Win-OS2 session. Your system crashes to the black screen: Trap E, "consult your sales representative ..."

You can do something better than consulting the representative: replace file VIOTBL.DCP in the \OS2\BOOT directory by the VIOTBL.DCP from the unpatched Warp 4. Restart. The problem with DOS and Win-OS2 sessions should be gone.

When the replacing can help:

• You are running Warp 4 with fixpacks or eCS.
• Windowed DOS and Win-OS2 sessions always crash, full screen sessions *run normally*.
• It's always Trap E.

What to do:

• Get VIOTBL.DCP from the Warp 4 CD.
• Go to \OS2\BOOT and rename the file VIOTBL.DCP there to something else.
• Copy VIOTBL.DCP from the Warp 4 CD to \OS2\BOOT
• Restart.

The DOS and Win-OS2 sessions should run now. If they don't then you can easily return to your original setup: delete the Warp 4 VIOTBL.DCP, rename the original file back to VIOTBL.DCP.

This tip was contributed by Radek. Thank you!

I am new to eComStation. Where can I find software, info, etc.?

There are a lot of resources for eCS (and OS/2), but if you're used to the abundance of the Windows or Linux world you'll have a hard time finding them.

First of all: if you find infos and/or software for OS/2, you have a very high chance that the software will work correctly in eCS, and a slightly lower chance that the informations will still be valid for eCS.

The first repository of eCS and OS/2 software that you should check is Hobbes. It is by far the most used and the most complete repository. Another good repository is OS2Site.

There is a site that is to eCS what SourceForge is to Linux: it's Netlabs. The main site may not be very functional or friendly at this stage, but its FTP repository is full of software, and you can find a lot of infos on software in development in its SVN/Trac pages.

A good (probably the best) site for eCS-related news is V.o.i.c.e, with its Voice-News mailing-list and its newsletters.

Speaking of software, you can find a good categorized compilation of OS/2-eCS software here. Each page includes informations, screenshots, tips, dependencies, etc.

If you are a developer, you can find a lot of infos in this wiki.

What is this LVM thing and how do I use it?

LVM stands for Logical Volume Manager. It is the standard way of managing disks and partitions since eCS 1.0.

With LVM the user can:

• assign drive letters the way he/she prefers
• make eCS "see" foreign partitions even if/when there's no native IFS driver (this means that you can assign a drive letter to the partition, not that you can actually READ its contents. Why should you want to do it? For example you may want to keep the same drive lettering across multiple OSes)
• make a single JFS volume spanning more than one physical disk
• extend a JFS volume by adding another physical disk

With LVM you have to distinguish among partitions and volumes.

In an LVMized eCS everything on disk is a volume: there are no more partitions. Volumes are the native unit of work for LVM.

You can use HPFS, FAT or JFS to format a native volume. The only file system that can be extended btw is JFS.

If you want to access partitions created with legacy FDISK or by other OSes you have to create a "compatibility volume" and assign the old partition to it.
With LVM you can also create old-styled partitions, but then you need to assign them to a compatibility volume to be able to use them correctly (e.g. you need it to create bootable volumes).

If you want to have access to all the advanced features of JFS and LVM (jounaling file system, disk spanning, etc.) you should create an "LVM volume", that is a native volume. Then format it with JFS.

The "LVM" utility included in eCS (either the text-based or the graophcal one) is powerful, and is the only one that can harness all the power of LVM. But if you're accustomed to the classical "FDISK" approach, the LVM utility can be quite confusing.
Recent eCS releases (from 1.2 onward I believe) include the MiniLVM tool which is much easier to use for most common tasks.
You can find it in the "System Setup" folder. It is titled "Maintenance & Installation Volume Manager".

If you want to use the more complete LVM utility you can find it in the "Programs" -> "Utilities" folder, under the name "Logical Volume Manager (Advanced)". Or you can go to the "Local System" -> "Drives" folder, right click, and choose "Manage Volumes" from the popup menu.
You'll find a lot of information on LVM by using the included online help.

I want to use FAT32 on an external drive but CACHEF32 gives an error at boot

The default behavior of CACHEF32 (the cache daemon for FAT32) is to abort loading if there's no FAT32 partition found at boot time. So if you're using a USB stick, and the stick is not inserted at boot time, you'll get the error message.

The error is not fatal: you will still be able to access FAT32 partitions, but you will lose some performance on reading from the device.

The solution is to add the "/F" option on the CACHEF32 line in config.sys. This will force CACHEF32 to load even if there's no FAT32 partition found at boot time.

e.g.:
CALL=X:\TOOLS\SYSTEM\BIN\CACHEF32.EXE /L:OFF /F /Q

FAT32 driver is SLOW when writing

eCS FAT32 driver is somewhat slower with respect to native performance under Windows.

But some people is finding that writing to FAT32 external devices (such as most USB sticks) can be slow as a dog. This is due to a (yet unsolved) problem with the lazy writer.

The solution is to disable lazy write (this is already the default if you downloaded the package linked above).

May I access NTFS partitions from eCS?

eCS 1.2 comes with an NTFS driver installed by default. You can use it to access NTFS partitions, but the driver is READ ONLY.

The driver has a switch to allow limited writing, but DON'T USE IT unless you're ready to face the possibility of losing data.

More specifically, NTFS driver development hasn't advanced since months, and it is not a priority. Better using FAT32 to exchange data with Windows.

You can find the latest FAT32 driver here.

Can't install eCSMT Utilities package

eCSMT (eCS Maintenance Tool) consists of two packages: ecsmtxxxx.wpi and ecsutyyyy.wpi, where xxxx and yyyy are version numbers (not necessarily the same).

Many people try to install the "ecsut" package directly, and fail with an error that's sometimes difficult to understand.

The solution is simple: put both the "ecsmt" and "ecsut" packages in the same directory, then start installation of the "ecsmt" package. It will find the "ecsut" package, and it will install both.