Email: Password: Remember Me | Create Account (Free)

Back to Subject List

Old thread has been locked -- no new posts accepted in this thread
???
08/04/10 11:04
Read: times


 
#177631 - Not obvious that an ICE is faster to use than print
Responding to: ???'s previous message
For embedded, I do a huge part of on-system debugging using print statements.

It really isn't so easy to say what method is fast or slow. Many times, the thing I'm interested in only happens on the 517th turn of the loop _if_ variable x has one special value and variable y is larger than x. I might have to add a special code block to capture this because the ICE can't do it in hw - but if I do add a special code block, it is often enough to add a print inside it while I'm there.

The bad thing with an using a graphical debugger is that it is so easy to just switch to debugging. So people tends to spend a lot of time in the debugger instead of thinking. Just looking at the code and thinking can get 90% of the bugs, and a couple of print statements can pick a large number of the remaining 10%.

Having a graphical GUI is a bit like having a garbabe-collecting memory manager. People tends to rely on the tool instead of thinking.

I think people should have a reasonably complete tool chest - and the access to an ICE is part of that. But people shouldn't be so quick to go for their ICE.

List of 59 messages in thread
TopicAuthorDate
time and money            01/01/70 00:00      
   My take.            01/01/70 00:00      
      Not obvious that an ICE is faster to use than print            01/01/70 00:00      
         I recall a comment            01/01/70 00:00      
            Good to look at available hw early on in project            01/01/70 00:00      
               I'd love to dump the values from the ADC channels ...            01/01/70 00:00      
                  Always a question of making most use of what you got            01/01/70 00:00      
                  Where there's a will ...            01/01/70 00:00      
                     ... there's a way.            01/01/70 00:00      
   investing in good tools            01/01/70 00:00      
      Cost benefit insanity            01/01/70 00:00      
         "We will save money ...            01/01/70 00:00      
         costs, one-time and repeated            01/01/70 00:00      
            VMware may be an alternative            01/01/70 00:00      
            subversion server            01/01/70 00:00      
               time is money            01/01/70 00:00      
                  A price estimate            01/01/70 00:00      
                     svn cost            01/01/70 00:00      
                        I most often _am_ the IT support            01/01/70 00:00      
                           Then that's the major difference            01/01/70 00:00      
                              Don't you think I try to avoid it too?            01/01/70 00:00      
                        re: svn cost estimate            01/01/70 00:00      
                           Cost for single user            01/01/70 00:00      
                              RE: I'd hope to do it in less than one hour.            01/01/70 00:00      
                                 True, back to time and money though            01/01/70 00:00      
                                    Investment            01/01/70 00:00      
                  revision control when money is scarce and volume is small            01/01/70 00:00      
                     Far too infrequent!            01/01/70 00:00      
                     Directory copies not really fun for version control            01/01/70 00:00      
                        no, I agree            01/01/70 00:00      
                           VCS on your own            01/01/70 00:00      
                              Ditto that            01/01/70 00:00      
                                 Further...            01/01/70 00:00      
                     So inadequate            01/01/70 00:00      
                        that's no difference            01/01/70 00:00      
                  svn benefits            01/01/70 00:00      
                     why?            01/01/70 00:00      
                        Working files, released files, change logs            01/01/70 00:00      
                           a comment on 'revision control' and such            01/01/70 00:00      
                              If a new version does not work find out why            01/01/70 00:00      
                                 the basic premise of my post was ...            01/01/70 00:00      
                                    Thats a good point, but            01/01/70 00:00      
                              Let a pro have good tools - don't worry about fools            01/01/70 00:00      
                              Revert safety net            01/01/70 00:00      
         Budget Insanity            01/01/70 00:00      
            Dilbert            01/01/70 00:00      
   Right Tools            01/01/70 00:00      
      Well, I'm with you there ...            01/01/70 00:00      
      Design            01/01/70 00:00      
   For "professionals" this is good advice, but...            01/01/70 00:00      
      This is exactly what I've been harping on ...            01/01/70 00:00      
         Not "forcing"            01/01/70 00:00      
      You missed Erik's point            01/01/70 00:00      
      you are wrong            01/01/70 00:00      
         professional/critical            01/01/70 00:00      
   Seasoned Professional versus green            01/01/70 00:00      
   Is it either Or?            01/01/70 00:00      
   an addendum            01/01/70 00:00      
   Too simplistic!            01/01/70 00:00      

Back to Subject List