« Web standards workshops with Russ Weakley and John Allsopp | Main | Free Seminar from Web Essentials »

June 07, 2005

Biting bullets

For several years, we developed or apps at westciv using a wonderful development tool/language called Prograph. A genuinely revolutionary, and to this day still unique programming language, in which you visually developed "object flow diagrams" that became your executable applications. It had unbelieveable features, such as the ability to write code, including develop classes and their methods while the application executed, and was a joy to work with. It also had the best cross platform class framework I have ever seen. Applications written on the Mac required very little work to run on Windows. And vice versa.

The first 2 versions of Style Master, for Mac and Windows, as well as Layout Master, and an earlier Mac only app we wrote, Palimpsest, were all written using Prograph.
But as time passed, and Prograph did not have the success that its developers, Pictorius International, expected, it became less and less of a focus, and slipped further and further behind.

The final straw came for us when after a lot of work by the Prograph community, particularly a smart amazingly hard working guy named Daniel Taylor, it transpired that the Prograph PPC compiler misused the stack, and so clashed with Aqua. All Prograph apps were effectively dead on Mac OS X.

To keep Style Master alive, we needed to rewrite the whole app from the ground up.

Given our druthers, of course we would not have done that. If somehow we could have got Style Master for Mac OS X to compile and run properly, we would have done that. And by now I imagine Style Master would no longer be viable.

Sometimes you need to bite the bullet, for long term viability. That year or so of pain while we rewrote Style Master gave us long term viability. Mac OS X for PPC (with Mac OS X for Intel really not that costly as far as I can tell from this distance) Windows and even the possibility of Linux.

My completely uninformed guess, based only on my experience with our own dependence on software development tools is that Apple realized sometime in the last 6 months that IBM and Freescale, the developers of the PowerPC chip family, simply weren't going to deliver the performance improvements and feature sets they need. At that point, they needed to bite the bullet. The G5 powerbook was probably never going to happen. Even the long awaited, and for Steven Job's probably increasingly embarrassing 3GHz PPCs might never happen. IBM had much bigger fish to fry - with the use of IBM chips in PlayStation 3 and XBox 360, these devices would likely consume 10s of times the number of chips compared to the number of units Apple would purchase, giving Apple little or no power in their relationship with IBM. Even though Apple will command no more power in their Intel relationship, one of Intel's main focusses is desktop and mobile chips. At least Apple will be on a level footing with their main competitors.

I'd suggest this is a reactive strategy, but one Apple seem to have long been prepared for (with apparently secret versions of Mac OS X running on Intel all along). Maybe I'm wrong, maybe Apple had planned this transition all along, once the migration to Mac OS X was effectively complete, a it most certainly is now.

Is it good, bad or neutral? With complex real world situations, Hamlet's rather depressing observation, "nothing is either good or bad, but thinking makes it so" is probably most apt.
Apple may suffer a crippling downturn in unit sales, as customers feel all PPC Macs are being End of Lined.
But, presumptuous as it is to compare the experience of one small software development company with one of the great visionary technology companies of the last quarter century, our experience of biting the bullet gave us much greater flexibility, and many more options once that transition was complete. If Apple pulls off what in some ways might be the easiest major technical transition of its three major seismic shifts (developers should at least find the effort of porting to Mac OS X Intel much less arduous that porting from Mac OS 9 to Mac OS X) then it opens many possible strategic options, particularly by way of licensing.

Who knows.

June 7, 2005 | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference Biting bullets: