about this? Great.
The answer .. ahem .. would be no. Not really. But you knew that probably.
I think James has a point in saying that is not so easy to see what might
affect requests, I certainly agree given the e-mails I see about the
effects of this banner or other once in a while.
It will probably be useful to keep a log of issues that have affected
pageviews in the past in wikitech, it will serve as a reference for "if you
are about to change something like ....<blah> let us known", pageviews and
request are "too abstract" whereus concrete examples might go a long way to
help people grasp possible effects.
We keep similar logs for other data streams.
On Mon, Feb 23, 2015 at 1:26 PM, James Forrester <jforrester(a)wikimedia.org>
wrote:
On 23 February 2015 at 12:12, Oliver Keyes
<okeyes(a)wikimedia.org> wrote:
On 23 February 2015 at 15:02, James Forrester
<jforrester(a)wikimedia.org>
wrote:
On 23 February 2015 at 11:50, Oliver Keyes
<okeyes(a)wikimedia.org>
wrote:
Well, specifically on the pageviews problem; is your patch going to
change what a request looks like? Then let us know. I'm interested in
(c), although that's probably for a different thread, because that's
not the perception I have from inside the greenhouse.
OK, so hits to api.php are always not-page-views? Even when we do
action=parse to preview what an edit will look like? What about hits to
Parsoid? Even when they're powering a read-like experience? What about
hits
to the forthcoming RESTbase system? Even if those
hits are "just"
asking for
meta-data as part of a search tool? What about
hits to tools in Labs?
Even
if they're a reader-focussed tool with
"readable" output like
Reasonator?
What about hits to the forthcoming Wikidata Query
Service? Even if we're
using the results to paint a "readable" page? Etc.
And that's why I said the test is "is your patch going to change what
a request looks like?" and not "is your patch going to change what a
pageview looks like?" - I don't consider it the responsibility of
every engineer to keep constantly up to date on what we consider a
pageview, but presumably people know what things they're building look
like when they hit the varnishes (at least, I'd hope so).
Aha, so if we never hit the read-mode Varnishes we can ignore anything
about this? Great.
[Snip]
Also, I'd point out that e-mailing this list
is a perfect example of
inside-baseball – people reading here probably
already know about this
stuff, or aren't engineers building things that impact it, so it doesn't
matter. What's the strategy for making sure everyone knows? Staff
channels
are insufficient – e.g. I believe the change to
redirects was done by
one
volunteer writing a patch and a second volunteer
merging it, which is
something we should continue to encourage.
Well, I sent this to the public analytics list for a reason ;). Do you
have suggestions for a more useful venue? Is this the sort of thing I
should be throwing at wikitech, for example?
Definitely post about it on wikitech-l. And, repeatedly, via follow-ups
each time a big whoopsie happens to remind people. Keep banging away at it
until we're bored to death of the message. Only then will you achieve your
goal.
J.
--
James D. Forrester
Product Manager, Editing
Wikimedia Foundation, Inc.
jforrester(a)wikimedia.org | @jdforrester
_______________________________________________
Wmfproduct mailing list
Wmfproduct(a)lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wmfproduct