Are there any plans for an SPDY  test on the Wikimedia servers?
I'm currently doing some speed tests on a robot and I found out (not
quite to my surprise) that it's much quicker to get whole pages
(hundreds at a time) than to ask the API for each page's last edit
time. I would love to see how would the results compare when using
Have a nice Sunday,
On Thu, Oct 25, 2012 at 10:16 AM, Antoine Musso <hashar+wmf(a)free.fr> wrote:
> I would prefer we revert that change and improve it, possibly using PHP
> DateInterval and definitely with MWTimestamp.
Created a revert patch set in https://gerrit.wikimedia.org/r/#/c/29947/
Some time ago, I've added a new toolbar button on ro.wp for the
Romanian quotation sign. The code is at  (the part below
//ghilimele ). It hasn't changed in 2012. However, today I noticed
that the picture is not displayed any more. On FF I see the alt text
and IE complains about "unsafe content" and does not display anything.
Looking in the page source, I see the image as being . Accessing
that URL directly gets me a 404. Sure enough, if I remove the
timestamp, it works.
I suppose this is somehow linked to the recent deployment of MW
1.21wmf2, but the link is not obvious to me. Can someone help with
advice on how fix the display problem and also the "unsafe content"
warning in IE?
I recently updated packages via PEAR. I don't remember what exactly,
but right now I have
When I try to run tests for Translate I get:
1) SpecialPagesTest::testSpecialPage with data set #3 ('LanguageStats')
Fatal error: Call to undefined function php_timer_autoload() in
/usr/share/php/PHPUnit/Util/GlobalState.php on line 381
make: *** [default] Error 255
We're upgrading the Gerrit and Jenkins boxes to Precise tomorrow.
Downtime is scheduled for 14:00-16:00UTC. During this time,
services provided by these boxes (this includes Gitweb) may be
I do not anticipate the upgrades to take the full 2 hours, but we're
giving ourselves plenty of time just in case.
For the interested: the servers in question are manganese, formey
So I found an interesting paper on SSL validation in non-browser clients:
One of the things is points out is that many clients set
CURLOPT_SSL_VERIFYHOST to true. However, this actually disables certain
validation steps, and the proper value for this should actually be 2. And
in our CurlHttpRequest, the default for sslVerifyHost is true.
I'm going to submit a patch to change the default to 2 so that we actually
perform validation, but I wanted to check and ask if there was any prior
reason that we explicitly didn't set it to 2.
Stevens Institute of Technology, Class of 2015
Major in Computer Science
www.whizkidztech.com | tylerromeo(a)gmail.com
Thanks for everyone who is asking questions! We're broadcsting today
and tomorrow as well and in 2 minutes we will have a talk on OntoWiki!
We're welcome remote participants!
Yury Katkov, WikiVote
On Wed, Oct 24, 2012 at 9:19 AM, Yury Katkov <katkov.juriy(a)gmail.com> wrote:
> Hi everyone!
> The first tutorial day of the SMWCon Fall 2012 has just been started!
> Good news, we have the live translation with Adobe Connect:
> Choose the guest account and join us remotely!
> To see the agenda, go here
> Yury Katkov, WikiVote, Program Chair