[Geysers] Current survey results as of 1/4/13 at 11:50

[Will Boekel]

Gazers,

As of 1/4/13 at 11:50 the tally is 24 to 13 in favor of adding the code. I will probably leave the survey up for two weeks. Below are all of the submitted comments so far.

  a.. User voted YES- So the A is now standing for Approximate instead of Assumed? 
  b.. User voted NO- Too often the current 0000 is used just to force a prediction when someone assumes that an eruption 'must' have happened. 'A' would seem to be just a broadening of that guess/wish work. 
  c.. User voted NO- This needs more discussion and fine tuning. Somehow an "overnight" or "proxy" designation would be helpful, but should not be in the raw data. This would only be used for F&M or perhaps Giant. Double and triple intervals for other features is fine without assuming/entering overnight times. 
  d.. User voted NO- I think adding ANY assumptions to a data base is not a good policy to promote. 
  e.. User voted NO- I prefer the idea of using 9999, a time that can't exist. I think adding unclear, ambiguous letter codes is a mistake. 
  f.. User voted YES- Any token ("A" or otherwise) is fine, and having a token is a good idea, but is "assumed" the best terminology for eruptions that are actually "known" but not known with a specific time? 
  g.. User voted YES- Good idea, but I feel that A for assumed is not the correct wording. If it was actually observed, but the time is unknown, then something like OWT (observed without a time) would be a better wording. Also 9999 is better than 0000 for the reasons Dan gave 
  h.. User voted NO- This issue should be discussed by people who actually use the times to compute statistics, not a general survey. My Excel spreadsheets have been set up to deal with the issue of computing approximate intervals with out using an artificial code such as "A". 
  i.. User voted YES- My thoughts on this are complex. And maybe I'll say something in the list serve, but for now. Here's what I say. If this goes in, it has to be used sparingly, and only for the geysers that we have a good grip on behavior wise. for extremely erratic geysers (Aurum for example) we should not use it at all. 
  j.. User voted NO- I think that it would be better to somehow have it specify between which times it erupted. I think the "A" would be confusing to many people. Geyser gazing jargon is confusing enough as it is! 
  k.. User voted YES- You can always ignore the data point later but you can never go back in time and take it. By adding the "A" it gives the entrant the opportunity to clearly list a data point at the time it is taken. 
  l.. User voted NO- If there is no electronic record or eyewitness to verify that something occurred then IMO it isn't positive enough to merit a data point. 
  m.. User voted NO- not a simple query. I'd say no unless the evidence was compelling for a major eruption. Such as F&M marker missing AND wet boardwalk; Giant log rolled. Melted snow would not be acceptable. And so forth. Recording for the future must be unambiguous for today's and tomorrow's gazer community. any not observed eruption with strong evidence must indicate the evidence in the record. use proxies to massage the data using your nnnn time format. Again, never leave room to misinterpret the raw data. 
  n.. User voted NO- Let the programmers code whatever they think is best. 
  o.. User voted YES- I think we should stick with 0000, though, not 9999, which is meaningless. It could be a range, 0355-0600A for example, if someone was "on shift" until 0355. 
  p.. User voted YES- I like the code to fill in what would otherwise be gaps, but am concerned that it may be misinterpreted as a valid eruption time in some superficial analyses. The whole issue of eruption time codes could use an extended discussion. 
  q.. User voted YES- Assumptions are always made - no matter how some avoid them. I like the idea of A rather than Ass. ;)
Thanks,

Will Boekel
-------------- next part --------------
An HTML attachment was scrubbed...
URL: </geyser-list/attachments/20130104/c305cfa0/attachment.html>