[aprssig] New find.cgi mockup
k4hg at tapr.org
Mon Nov 29 08:38:39 CST 2004
On 11/29/04 at 9:06 AM Robert Bruninga <bruninga at usna.edu> sent:
>>First, in order to have any validity, the user's computer
>>must be synchronized to an NTS time server.
>Not at all. This has nothing to do with computer time, or my PC's
>time, it has to do with the viewer of a FINDU web page knowing
>"when" the page was refreshed. Was it now?, 3 minutes ago?
>an hour ago? or an old page from a day ago? Or a screen
>capture from a month ago?
Let me explain, Bob. My servers's clock is very accurate, and is set to UTC. It
has no idea of your local time. The only timestamp I can put on a page is in
accurate UTC. Is your computer set to UTC? Is it accurate to the second? For
most people, the answers to one or both of these questions will be no. If that
is the case, then a mental calculation is necessary to determine when a page was
generated, given a timestamp in accurate UTC.
>>Given that the alternative is simply to hit the manual
>>refresh button, and be assured the data is up to date,
>>your method is a lot of extra work.
>What? What kind of extra work is it for me to read
>"at 1030 on 29 Nov the last posit was 10 minutes old"...
>And I should*not* have to hit a button to get info that
>should be presented to me in the first place...
Again, if I could determine what timezone your computer was in, and what the
error of your computer's clock was, then yes, it would be no additional effort.
In the real world though (unless you live your life in UTC), it requires mental
math to convert between a UTC clock and your local time.
Even in your theoretical world though, it is quite possible that in the time
between when the page was generated and when you are looking at the page a new
position was received. So at best, using the timestamp can only precisely fix a
report in time, you still have no assurance it is the most recent position.
>>the manual refresh is a far superior means of achieving the same
>I see that as Computers and progaammers making us slaves
>instead of helping us see information clearly.... I got more
>important things to do than have to sit like a monkey hitting
>refresh and having to wait for a refresh every time I want to
>know if a FINDU page is valid or not.
It is less than than three minutes old, how often do you really care if a report
is three minutes or three seconds old? And if you do, don't you also care if a
new report has been received in those three seconds or three minutes???
>I guess we disagree ;-)
We (almost) always do!
More information about the aprssig