|
 |
 |
|
|
|
|
|
 |
|
 |
|
 |
|
 |
|
|
SYSTAT RESOURCES |
|
|
|
|
|
|
|
|
|
|
SigmaPLot : SIGMASCAN : TableCurve2D : TableCurve3D : PeakFIT : AutoSignal |
The British Psychological Review of SYSTAT Software |
|---|
A Review of the SYSTAT v.10.2 software package
On behalf of : the British Journal of Mathematical and Statistical Psychology (augmented web-version) Minimum machine specifications are: a Pentium/equivalent processor, Microsoft Windows 95/98/NT 4/2000/ME/XP operating system, 32Mb RAM (minimum), 30Mb hard disk space (add another 42.3Mb in case the system Manual Acrobat pdf files are to be installed), and an SVGA monitor. In the bginning, there was the mainframe. The literally power-hungry, paper tape and card-guzzling behemoth that sat in computer departments, shepherded over by its operators and systems analysts. Great lumps of metallic core memory sat in huge cabinets along with disc drives and the 0.5" reel-to-reel magnetic tape drives. Three statistical software packages reigned supreme, SPSS, BMDP, and SAS. With manuals the size of small encyclopaedias, fantastically intricate I/O commands, and 15" wide fanfold output that could fill a suitcase if not carefully controlled, these were mighty beasts indeed. Then came the humble PC - first the Apple, then the first CPM and DOS machines - with less RAM than the average calculator of today, and disc drives the size of old gramophone records. Along with these came some ingeniously written statistical utility programs - but nothing to match the giants of the mainframes. A year or so went by whilst we all fiddled with these little PCs - mainly for experiment control and some limited "word-processing". Then along came the 512k RAM machine - a hard disc (10Mb) - and Microsoft, BASIC, the Sirius, and IBM PCs. Then SYSTAT for the PC arrived - on 10 or so 5¼" floppy discs, and a small box of manuals. Let there be no mistake, this package from Leland Wilkinson, sold from Tulsa in the US, served notice on the giants that their days of market domination were over. And so it proved to be. SPSS and BMDP released PC versions of their products - but these were unbelievably poor implementations of what were still essentially mainframe packages. STATISTICA CSS meanwhile was yet another revolution in the PC domain - building from SYSTAT's wonderful graphics and modules - it created a completely new, colourful, and spreadsheet-based output user-interface for its routines - and drove the final wedge with SYSTAT between the mainframe and PC packages. Then Windows arrived with PCs that were now becoming "powerful". That was the end of mainframe statistical computing. It was also the end of every MS-DOS computing package - apart from a few niche products. STATISTICA produced the first truly Windows-oriented system. SYSTAT blinked, paused, and finally paid the price for not doing the same. That is, even though they produced something which tried to look as though it was a Windows program -it was a ghastly sight to behold. This is where I finally parted company with what had almost been "a friend" to me for many years in the Biosignal Lab. SPSS eventually bought out the company (as it did with the now languishing BMDP software), did some remedial work on its interface, and tried to sell it alongside the similarly "but not quite there" BMDP PC system. But, by this time SPSS for the PC was looking mean and slick, STATISTICA 5.0 has been released and was taking SPSS on head-to-head. All that was left for SYSTAT and BMDP was the "loyalty" market, and some discerning users who fully appreciated the high quality specialist routines that were still lurking within both packages. Frankly, SPSS lost interest in SYSTAT and BMDP - its attention now firmly fixed on what it perceived as its biggest threat - Statsoft Inc. and STATISTICA. Not surprisingly, some entrepreneurial businessmen bought SYSTAT from SPSS - and Systat Software Inc. was born - again - along with SYSTAT 10.2. What has become of my old friend? |
 |
 |