Loading ...
Sorry, an error occurred while loading the content.

Re: [XP] How do you handle new ideas to implement a story dentified in mid iteration?

Expand Messages
  • Paul Michali
    ... Agreed. The nagging thought I had was that although the new method doesn t offer any significant advantages, it would (if it works) be way faster to
    Message 1 of 4 , Aug 12, 2004
    • 0 Attachment
      Mark Striebeck wrote:

      > In these situations, we usually have to remind ourselfes to go back to
      > the fundamentals of agile methods: delivering customer value.
      > If the new implementation doesn't have any significant advantages
      > (performance, stability, ...) we have to remind ourselfes that we should
      > focus on customer value. And the customer is usually not interested how
      > this piece is implemented.

      Agreed. The nagging thought I had was that although the
      new method doesn't offer any significant advantages, it
      would (if it works) be way faster to implement.

      In addition, it changes, somewhat, the direction of the
      rest of the implementation. Specifically, if I stay with
      the original method, I'd read the spreadsheet using POI
      or JDBC/ODBC and then I'd write out the database table
      to MySQL (trimming out fields, merging others, etc).

      If I use a macro or script, I'd run it on the spreadsheet
      and then would need to "load" that file into the database
      and then extract (say using SQL SELECT) out the fields
      that I want for the final table. This is much quicker
      in effort to implement, from what I can tell so far.

      So, while I had already started the POI (I did a short
      spike to make sure it would work and then started
      implementing it), I was wondering if the other way was

      So, I was wondering what the procedure is in this case,
      where you've already started a story and see a potentially
      much faster way. In other words, how to deal with change,
      in mid iteration.

      I didn't know if you grin and bear it through the story
      and then afterwards do a spike and possibly another design
      in the next iteration, or if you paused and did a spike
      and possibly change the approach in this iteration.

      > On the other side, this usually opens up the possibility of an
      > engineering refactor card - and I know how some people think about those
      > (I just had the discussion with Ron several days ago :-)

      Maybe it's my inexperience with TDD, accessing databases,
      refactoring, etc., but I didn't get the feeling that I could
      easily refactor from the one way to the other.

      Any thoughts on this?

      Was there things I could have done differently, with the
      way I approached this whole project? Like maybe spending
      more time on spikes and looking at alternatives, given
      my inexperience in this area? I was anxious to get going
      and maybe didn't explore all the alternatives.

      BTW, I haven't had much success with the alternate method
      yet (I've been trying things, in my spare time). With a
      script, I can't get the file to write out in tab delimited
      text mode. With a Excel macro, I can't seem to easily
      apply the macro on other spreadsheets (I have to record
      or cut and paste it in each spreadsheet).

      PCM @ WORK (Paul Michali)
    Your message has been successfully submitted and would be delivered to recipients shortly.