🚀 go-pugleaf

In article <365823c3$0$12759@nntp1.ba.best.com>, Tom Woolf <woolf@shell5.ba.best.com> wrote:

>As near as I could tell this old copy was coming from the server because I
>cleared the cache and could see the data received light flash as the page
>downloaded, the delay to bring up the images, etc. yet no sequence
>of reloads or cache clearing would fetch the new copy.

Several others have reported some strange persistent cache issues
with updates. Some browsers don't always clear cache properly and
for a definitive client .v. server diagnosis, it would be worth
checking with a non-caching browser (e.g. lynx from shell).

-- 
|~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
| Malcolm Hoar           "The more I practice, the luckier I get". |
| malch@malch.com                                     Gary Player. |
| http://www.malch.com/               Shpx gur PQN.                |
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

In article <3658646c$0$12762@nntp1.ba.best.com>,
Tom Woolf  <woolf@shell5.ba.best.com> wrote:
>I don't see any other explanation except those pages were downloaded
>copies of the older, apparently server cached text.

Is your computer's clock accurate?

Malcolm Hoar <malch@malch.com> wrote:

:... it would be worth checking with a non-caching browser (e.g. lynx from
: shell).

Malcolm, I hate to be argumentative but I can see the difference between a
downloaded page and a locally cached one in the LED pattern of my
33.6 external modem and the delay time before the page appears.

Additionally, I was doing my checking in MSIE 4.0 (I know, I know, <g>) 
and fired up Netscape 4.5 to see if it was indeed a local cache problem.
(As Netscape would use a completely different set of cache folders and had
never seen this particular page anyway.) It displayed the old copy and
then the new one at a later time just as MSIE did. (The reload/refresh
brought in the new source at the same time on each. IOW no combination of
refresh/reload, shifted or unshifted, and local cache clearings would
bring in the new source I could see from shell until suddenly, for no
apparent reason, it was available and visible in both.

I don't see how that could be a local cache problem. . . .

I don't see any other explanation except those pages were downloaded
copies of the older, apparently server cached text.

Tom

In article <3658646c$0$12762@nntp1.ba.best.com>, Tom Woolf <woolf@shell5.ba.best.com> wrote:

>Malcolm, I hate to be argumentative but I can see the difference between a
>downloaded page and a locally cached one in the LED pattern of my
>33.6 external modem and the delay time before the page appears.

I'm 99% sure I've seen Netscape download a fresh document but still
render a cached version when it gets into some sort of confused state.
The only way I found to clear that was to quit Netscape and physically
blow away the cache.

>Additionally, I was doing my checking in MSIE 4.0 (I know, I know, <g>) 
>and fired up Netscape 4.5 to see if it was indeed a local cache problem.

Well, that certainly makes it more conclusive but Todd has a point
about the time thing.

-- 
|~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
| Malcolm Hoar           "The more I practice, the luckier I get". |
| malch@malch.com                                     Gary Player. |
| http://www.malch.com/               Shpx gur PQN.                |
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

I'm out of real ideas.  I've suggested in the past that this could be
a filesystem bug, but Matt was quite certain it wasn't.  Ah well.

Within a minute of WWV. I'm only compulsive with my watch

Tom

Todd Michel McComb <mccomb@medieval.org> wrote:
: In article <3658646c$0$12762@nntp1.ba.best.com>,
: Tom Woolf  <woolf@shell5.ba.best.com> wrote:
:>I don't see any other explanation except those pages were downloaded
:>copies of the older, apparently server cached text.

: Is your computer's clock accurate?

Here was the sequence:

I ftp'd the update.

Requested it in MSIE to check the change. Normally I don't have to do
anything to see the updated page. a simple request to the URL brings the
new version in as I have the cache set to refresh each session.

Didn't see the change.

Logged onto shell and checked the file timestamp. It had a logically
correct value and was the newest file in the directory.

I reloaded, cleared caches etc. Didn't see it.

Started Netscape which had never seen it online and _had_ to fetch a new
copy from the online URL. It didn't see it. 

Cleared, reloaded, etc. didn't see it.

Went back and double checked the time stamp and looked at the source to
see the change was in the source on shell. It was.

Went back to Netscape reloaded the page and the change was visible.

Task switched to MSIE, hit reload and got the change without clearing the
cache.

How on earth could it be related to my system clock?

And again, during all of this I could tell exactly when I was getting a
local copy vs a DLd one because of the delay in image formation, page
completion and duration of the 'Receive data' LED falshes on my modem.

Tom