Well, it's a lot prettier, but I wouldn't say I like it any more than I do the original. Then again, I think that the original, barring a few issues, is pretty good; I have used it nearly exclusively for over a year now. The following are the notes that I took during the install and initial testing:
The install waits for a disturbingly long time before giving any option screen or warning of a prior install. The amount of activity that goes by without even a ``Please wait while Install examines the system'' is enough to strike fear in the heart of anyone who has had a virus attack or experienced a Microsoft DOS 6.22 install on a system shared by OS/2. Once a screen does come up, sensing a prior version, the install wants you to remove the old drivers from CONFIG.SYS by hand (i.e. cancel the install, edit, and re-boot) rather than allowing a copy of the configuration with the suggested modifications to be created. You don't have to do so for the install to proceed, but it breaks up the flow of thought.
The choice of tape devices supported is outstanding. I am aware of only one or two very rare boxes that are not covered by the included drivers. On the down side, the drivers for the older devices, for example my CMS Jumbo 250, are the same code that came out with version 1.0, right down to the file size, creation date, and internal build information. In other words, if your existing Sytos driver is lousy, you won't get any improvement. Any increase in performance has to come from the Sytos user interface and that's NOT the place to look for a solution to shoe-shining! Another ``gotcha'' is that if you indulge in a lapsus digitus and accidentally install the micro-channel driver on an ISA machine, the install does not notice it, nor does the driver itself until it is MUCH too late. (See horror story below!)
The overall speed of the install is quite satisfactory and it seems that the code has been slimmed down some from prior versions. The one glitch found that took up a lot of time is that there is no retry mechanism in case of a disk read error. Since I was running the install from a DISKCOPY done on an uncertified diskette in a drive about which I know nothing, an I/O error or two would not disturb me, but having the install bail out entirely rather than let me pop the diskette out and in for a retry was quite disconcerting when it happened at 92% of the way into the install. I found that while Sytos' install speed is fine, it does NOT multi-task worth a darn. Nothing else could get any mouse attention while it was running, although other tasks did not crash or shut down; i.e., if they were running, they got OS/2's time-slice OK, but any manual attention had to wait until Sytos was finished. I have installed a lot of other packages that let me start writing up my notes in DeScribe while the install was running, not so Sytos.
Configuration of the software after install was just as ugly and painful as ever. The default IRQ, DMA, and I/O address have the same conflicts with the floppy and CD that they always did and the driver insists on exercising them nearly to death before petulantly informing the user that it cannot find the device. It spends a little over a minute banging and grinding every non-SCSI device in the system before relenting and allowing one to set the correct values.
Running the system the first time is also just as awful as it ever was. It's really cute; I LIKE the new appearance, but, until the complete device and library set up is complete, the old feeling of having to fight the interface for every task is still there. First, the horror story promised above: The MCA driver runs just fine (it IS Intel code, after all) until you create any sort of an interrupt or DMA event. At that point you go down, hard. Trying to open a floppy while Sytos was reading tape destroyed EPM.INI, OS2.!!!, OS2SYS.!!!, VX-REXX, DeScribe, the Sytos folder, the A: drive object, the Sytos .INI and .CFG files, a couple of .INF files, and the contents of CMOS. After rebuilding CMOS, several CHKDSK /F runs, and some file restores from the old Sytos Plus, I figured out the difference between the Q80-M driver and the Q80-A driver and got back to testing. During the reinstall I made further observation that the install screen is ugly; it takes up most of the screen and then wastes it by showing nothing but a lot of grey space.
The Sytos window can be minimized to run tape operations in the background, but it cannot be resized to run in a small window; it's either all there or not there. It read a Sytos Plus created tape for two hours doing an ``Add to Library'' run, reported no files found on the first volume of the set, never asked for the second volume, but properly added the names of the files and the volume to the library anyway! It won't release the mouse during rewind or long tape movements (this is a driver issue, not the new interface); mouse motion is jerky or stopped.
The scheduler doesn't seem to like to kick off a job when Lockup is on and you have to change the incremental backup job for a new tape so that it does an overwrite on the first time and an append on subsequent runs. It makes no assumptions that a blank tape is OK to append to; a purist on the definition of append, I suppose, rather than allowing the first ``append'' to be a ``write initial''.
Once the system is set up and all the old tapes are added, it is not unpleasant to run. The new interface makes the selection of files to be dumped or restored much more intuitive than the old screens, except for one thing: disk files show up on the directory with an icon of a paper document. I find this disconcerting, preferring an icon that would make me think of spinning platters, not round-file fodder. I'd like more progress indicators, but since I USUALLY run tape in the background, I suppose that this feeling was just because I was trying to watch a new product rather than just sending it off to do my bidding and forgetting about it until it reported completion. In the final analysis, I still think it is a GOOD system, but only AFTER it has been installed and tuned. Rather like Warp itself: installation can be trying, but the results are probably worth it.
Tests of the ``Drop Dead'' system ``Rebound'' were not performed because of a lack of documentation. The REBOUND.EXE file got bent out of shape over my E: drive being removable, claimed my tape was write protected when it was not, and I did not have the ``User Guide'' that the miserably small help screen referred to to attempt to make it ignore the CD. (An advisory to the user group: supply the ENTIRE box when asking someone to test software.)
Return to Reviews index
Return to TAMU OS/2 Users Group home page