[plug] [link] a lawyer Switches

[Richard Meyer]

Peter J. Nicol wrote:

>>  >From what I understand, IBM plays similar nasty games with AS400 and
S390,
>>  so ...
>>  >bugger em ... I will use the stuff that gives me total freedom ... and
be
>>  more
>>  >than happy when this co-incides with increased productivity as well.

>http://www.midrangeserver.com/tfh/tfh032502-story02.html

>Deliberately limiting how much of your processor you can use?  Bugger em!

Thanks for the answer!

A quote from one of the articles:
                                                                            
 "In effect, IBM sells levels of software performance, which drive user     
 seats, as hardware features because it is simpler. "                       
                                                                            
 It appears from the article that the customer contracts for a machine      
 "tailored" for a specific workload. When his needs change, he phones the   
 vendor who sells him a way to cater for the changed needs - hardware. That 
 the hardware is simply a jumper (or something) to enable software to       
 unthrottle a resource is well known and accepted in the mainframe world    
 and has been used for decades by IBM, Hitachi and Ahmdahl.                 
                                                                            
 It was accepted that it was simpler that the machines were all built the   
 same, with the same hardware specifications. Of course they could have     
 built them in all sorts of different grades - like Intel did with the 486  
 family (explanation below), but then upgrades would have been more like    
 exchanges, with the vendor left with an inventory of unuseable parts.      
                                                                            
 The fact was that the customer was getting top-of-the line hardware, for a 
 lesser price to run the workload he had specified, and when more capacity  
 was needed, it was paid for and supplied. The customer was paying for the  
 capacity, not for the hardware as such.                                    
                                                                            
 If you would prefer it, I'm sure that they could deliberately remove       
 hardware which the customer had not paid for - it's an equally viable      
 approach, except that it complicates record keeping on the vendor side.    
                                                                            
 An analogy would be - person A goes to person B who is a hardware          
 supplier, and says he wants a 500MHz Celeron PC without any CD drives,     
 64MB memory and 5GB disk, right now, but A wants the ability to upgrade to 
 3GHzP4 with 256GB memory and 20 TB of disk with CD drives - burners etc. B 
 then sells him a machine that incorporates all the second lot of specs,    
 but tapes the CD drives shut, disables the extra memory and disk storage   
 on BIOS, and throttles the P4 to Celeron specs, because A hasn't paid for  
 it. And sells it for the price of the first config. Person A then          
 discovers all the other stuff there and cries foul because he can't use    
 all that he paid for, even though he hasn't paid for it. Strange person,   
 this person A.                                                             
                                                                            
 It makes me think of Broadband as well - you contract with an ISP to       
 provide 128/64 to your house after studying all the options provided by    
 your ISP, and pay for that. Then you discover that your line is capable of 
 handling 1024/1024, so you take the ISP to task for deliberately limiting  
 how much of your ADSL line you can use?  Bugger em!                        
                                                                            
 Compare this with this quote :"Deliberately limiting how much of your      
 processor you can use?  Bugger em!"                                        
                                                                            
                                                                            
 I afraid I can't get too excited about this exidence of the perfidiousness 
 of big business.                                                           
                                                                            
 RichardM                                                                   
                                                                            
                                                                            
 Explanation:                                                               
 All 486 chips started equal. After production they were tested at the      
 highest speed they were selling. If the chip passed the test, it was       
 certified as being a 66MHz (or whatever). If it failed, it was retested at 
 the next lower speed, if it passed, it was cerified as being of that       
 speed, and so on down, till they found that a chip couldn't pass any tests 
 in which case it was discarded. Similarly, if the FPU failed, it would be  
 disabled and marketted as an SX instead of a DX.