In the past it was always correctly assumed that "serious” image analysis required serious computing power. A few years ago most of the top image analysis products were written for high end workstation computers, or for PCs which had to be enhanced by additional hardware to give the performance required. "Macs" were largely left out of the arena, especially in electrophoresis applications, simply because the additional cards for the Mac at that time were limited and expensive.This historical development of image analysis on the PC and UNIX machines is why Mac based Gel Analysis software tends to be basic, or a "Mac" version of the PC product.
Now the situation has changed quite dramatically. Until recently PCs were held back by their16 bit operating systems and processor speed. However, with the advent of the Pentium, improved architecture, high capacity hard drives and 32 bit operating systems, the processing power and speed of the PC has caught up in many respects with workstation technology.
Effective analysis of most 1D electrophoresis gels can be done on a 486 PC. A Pentium would only be necessary if the images were large (e.g. high resolution scans of DDRT-PCR gels). Using a UNIX workstation for analysing 1D gels offers no advantages and suffers from several disadvantages (see below). It is rather like buying a jet to go shopping! On the other hand it is advisable to use a fast computer for the analysis of 2D electrophoresis gels because of the high degree of complex processing involved. However the arguments below suggest that a Pentium is far more appropriate than a UNIX workstation.
The Phoretix tutorial experiment contains three 1Mb images with approx. 150 spots per image. On a Pentium 90 it takes an overall time of 33 seconds for Phoretix 2D to identify the spots, measure the spots and match them between each of the three gels using the Phoretix batch processing algorithm. This used to take 15 minutes on a 386 PC (16mhz) - a 30 fold increase in performance over four years. It is unlikely that the same processes could be performed on a workstation computer such as the Sun Sparc in under 30 seconds.
Our own bench tests on the DEC Alpha PC were disappointing when compared to those performed on a Pentium. The Phoretix batch processing in the tutorial experiment described above took 45 seconds on the DEC Alpha compared to 17 seconds on a Pentium 120!
1. Very few of our potential customers are familiar with UNIX systems. UNIX is a very powerful operating system designed for demanding and frequently very technical applications. 2. Handling UNIX is not for the uninitiated and if anything goes badly wrong an engineer may need to be called. 3. Basic software packages for UNIX Machines tend to be expensive and can be inferior in design to their PC equivalents (e.g. Word Processors, Spreadsheets, Slide makers, etc.). 4. Transfer of information from UNIX workstations to "Macs", or PCs may not be easy. 5. Many UNIX machines in the biochemistry lab environment were bought as multi-user facilities and yet tend to be used by only one or two individuals.
Apple has bought NeXT The latest move by Apple to safeguard the future and expand is its $400 million purchase of NeXT. This has been heralded as a strong move resulting in journalistic comments such as “The Mac will survive”. However, the big question is what this acquisition will mean for the current Mac user base. As we understand it, there will be a new operating system called Rhapsody based on NextStep (or OpenStep) which Will describes as “a techies dream”. Initially this is unlikely to be compatible with MacOS applications. This means that Apple will need to market two operating systems in the next year or so. In the medium term Rhapsody will contain a compatibility box which will give it a Mac look and feel and run Mac applications. We suspect that the eventual aim will be to convert all Mac applications to run directly in NextStep so that there is no need to have a “compatibility OS” within an OS.
Currently NextStep works well on Intel hardware. So will NextStep applications work on both the PC and the next generation of Mac computers? It would seem likely. Will this mean a real and effective war between Microsoft and Apple over operating systems and applications as opposed to the lost battle between IBM's OS2 and Microsoft's Windows 95 and Windows NT? Interesting times ahead! The next year is going to be very important. It all depends on the perceived importance of the Mac in the market place. Apple need to convert developers like ourselves. This is where IBM went wrong. Developers will only convert if they believe that the future will require them to do so. Simply having a better operating system is not enough (OS2 was better that Windows). This will be quite hard since we all have knowledge of the OS2 experience. On the other hand there are already a lot of Mac developers who will be easily converted.
Future Phoretix Development
One thing is certain. We will not be producing a Mac version for the current Mac Market. It is too short term and the rewards would not meet the development costs. Note that when Pharmacia were selling competitive software they sold over 800 PC versions and just 5 Mac versions World-wide. However if NextStep is going to become a major force in the market place we would be silly to ignore it. If it is marketed as a multi-platform system and if it becomes a serious competitor to Microsoft in the research environment we will almost certainly have to become converts. At this moment in time there are too many questions and not enough answers. It is very unlikely that any conversion will be necessary over the next year, which means we have a years grace to have all our questions answered. This is where we need the help of all our distributors, customers and potential customers. We need to know the opinions of the Phoretix market place.
Why do we not do a Mac version?
In 1995 we were going to. It was written into our business plan. Our Mac survey revealed a 50% split in the Mac market:-
- Half would accept the PC version if it better helped solve their needs.
- Half would only consider Mac software
This group answered the questions with such extreme affiliation to Macs that we were led to believe that the Mac was more important than their scientific needs. This group did not seem to be likely Phoretix customers. Pharmacia's experience supports this view
The corporate problems of Apple suggested we should wait. None of our competitors have Mac software that is as good as the PC counterparts.
We decided to put all our resources into the PC version in order to put us technically ahead of all our competitors. This we have achieved.