The Atari STE was an improved model of the Atari ST. It was widely covered in the consumer Atari magazines and cost a small percentage more for the base model. It was more easily upgradable than the original ST. The inclusion of a “blitter" as standard meant that the user interface was faster for GEM applications.
The machine was originally released very late in 1989, and so 1990 was the year when it had its greatest chance to make an impact.
From my reading of various background stories, it seems that Atari made a great deal of effort from an engineering perspective to create a good product with good compatibility with existing hardware.. however of course games developers who had perhaps not felt the need to follow developers guidelines to the letter (and probably not had a chance to test on the latest machines) were caught out by glitches.
My initial impression was that the big causes of these bugs were changes to the ROM that had a bigger impact than expected. In retrospect it is interesting that just over 12 months later the ROM shipping with TT and Mega STE machines introduced lots of very useful features to the desktop (so it could easily have been more capable but less compatible release). In addition ROM versions 1.62 which came with the STE was broadly the same as version 1.42 which was released with new Atari ST machines built at the same time. Compatibility is a huge topic and Atari’s response to the problem was not the most encouraging to potential purchasers:
“If the programmers have written to our specifications then there won't be any problems.”
ST Format suggested there were 30 games with compatibility problems. The reality was that all STOS based games were broken (Which meant alot of Public Domain / shareware software as well... Later a software program was released which fixed STOS games on a patch per executable). Having recently looked at trying to run games using an emulator and different ROM versions, it seems very likely that a large portion of compatibility is down to the very specific differences in the ROMs.
So: Assuming you might want to decide which machine to buy in 1990, you’d have a choice between (the STE,) an Amiga, an Atari ST, a PC, Macintosh, Super Famicon, and potentially a Sega Genesis/Megadrive - The Genesis had been released the year earlier in USA and reached Europe in Decembr at the end of 1990 as (games) developers were choosing whether to support Amiga/Atari, Genesis or 4 platforms if you include the STE. Given the chicken and egg situation in 1990 (no sales of STE machines, existing ST games will usually just work on an STE…) it seems obvious now why 3rd party developers had little incentive to develop STE enhanced games rather than port their existing titles to the Genesis (which might be easy because the games are 68000 based). Early interviews with people who had developed ST games suggested there was some animosity or anger toward Atari.
So Atari promoted the machine for its ‘serious’ applications. For a short while, Atari had the Mega ST as a 68000 based machine with a blitter running at 8mhz and the STE, the same thing, for less, in a slightly less ‘business style’ case. Eventually they finished the Mega STE which had a 2 times faster processor, but this took until the end of the year. All during 1990 the old ST was still for sale at a slightly reduced price.
Applications enhanced for the STE could benefit from..
- enhanced sound - playback of ‘real’ samples recorded in 8bit stereo at up to 50khz took no processor time. This was useful for ‘tracker’ based sample playback and ‘quartet’
- faster image/screen draw speed - using the blitter - not just for games, but all applications benefitted.
- extra colours could be used in art programs eg; using 16 shades of grey rather than only 8.
Games developers could relatively easily also make use of
- hardware scrolling, as used in vertical scrollers like pinball dreams.
- Extra joystick ports - for games that need more than 2 players.. although almost nothing ever used this.
In total then (possibly with the exception of audio quality) none of this made games better than those on the Genesis/Megadrive or Amiga - because both of those had better ability to display more than 16 colours and to more easily handle ‘sprites’ (assuming developer kits had good code examples -- Although the blitter helps with sprites, without example code in the development kit, it needed manually programming.)
In hindsight, there are a few things that could have been done better, but market forces ultimately played a very big part in giving Atari a hard time. The Amiga was likely selling at least 2x the number of computers relative to Atari, and the STE didn’t mean Atari had caught up with technical abilities (apart from the continued marginal difference of 7 vs 8mhz) The continued availability of the ST as well as rumours and more rumours of other serious machines (TT, Mega STE) and game machines during 1990 (eg; discontinued Panther machine that might have been as capable as an Amiga or very close in capabilities to the Jaguar) created a confusing picture for any developers.
Developer kits and ROMs had very few enhancements to allow standards compliant (esp game) software to be written.
The third thing that could have made a difference, was that thing that the genesis had, that Atari didn’t have: A really great unique thing that was as playable and fun as Sonic or Mario. It looks like Atari weren’t interested in making games. (despite having a really successful arcade business).
In hindsight though looking back over the archives of 1990, the ST impressed me with
a) the huge back-catalogue of games
b) the variety of software that was released each year, from in-depth strategy games to puzzle games that is now associated with causal tablet/phone play.
Atari Really Should have:
- Stopped all sales of the (original) ST. They might have had good reason to continue selling - like excess inventory in warehouses, but the fact that the ROMs got updated (to 1.4 and so made old software break) suggests that it took extra work to keep building these obsolete machines.
- Got to market before christmas 1989 (and got to developers even earlier).
- While waiting for the Mega STE licence/approval, offered STE’s with immediate upgrade to Mega STE when they become available. eg; pay for the Mega STE in installments - half before and the final half on delivery/exchange.
- Offer updated media (or exchange) for any existing software affected by compatibility problems. If a software house isn’t interested in fixing its stuff, retailers will be returning their products.
“ST Power Pack with 24 really cool games,” says Darryl Still, “The pack itself was a monstrous success, selling millions, but also managed to really hack off the developer and publisher sector, because the end user had no need to buy any more games for some time after their initial purchase.