yep, that's what I see
On Tue, 10 Dec 2019 at 12:18, Hershel Robinson
wrote:
>
> What I see on in Firefox 71 is correct, meaning "Carnifex" and here is
> a screenshot:
>
> https://imgur.com/9zZG9Hp.png
>
> and on Chromium "Version 78.0.3904.108 (Official Build) Arch Linux
> (64-bit)" I see "I:Ca
On Tue, 10 Dec 2019 at 12:06, Hershel Robinson
wrote:
> Interesting. On my Linux box, I see it on Chromium but not Firefox, and
> both are the latest versions.
no no, I mean I saw the same behaviour as the original reporter, and
you, are seeing
- d.
___
I saw the problem as reported in Firefox and Chromium, both on Linux.
On Tue, 10 Dec 2019 at 10:46, Brian Wolff wrote:
>
> Well this weird. I was able to reprofuce the I on samsung internet browser
> but not firefox. Also template: was shortened to T: in interface as well
> (in the view source pa
query string is getting stripped. The
> behavior here is configurable--you're looking for the Forwarded Values
> config in TF.
>
> https://www.terraform.io/docs/providers/aws/r/cloudfront_distribution.html#forwarded-values-arguments
>
> -Chad
>
> On Tue, Jan 22, 2019,
I have a work wiki (running 1.27) that I'm trying to put through AWS Cloudfront.
So far it appears to work through Cloudfront! Except that load.php
gives different data through Cloudfront, and I get an unstyled page,
and load.php's entire output seems to be:
/* This file is the Web entry point fo
I'd always recommend MySQL or a highly compatible variant such as
MariaDB - Postgres is great, but Wikimedia runs MariaDB so I do too.
- d.
On Sat, 22 Dec 2018 at 18:18, Jeffrey Walton wrote:
>
> On Sat, Dec 22, 2018 at 12:59 PM Ryan Schmidt wrote:
> >
> > Try installing MariaDB instead of My
We have a couple of internal wikis that are presently 1.27 and will
soon be 1.31, that I will be moving to AWS.
So I can either (a) use a MariaDB RDS or (b) use Amazon Aurora's
version of MySQL, which is a bit cheaper.
Aurora MySQL 2.x is MySQL 5.7 compatible going forward.
So is anyone here act
I remember stats.grok.se with fondness - as I understand it, that checked
the raw Squid logs and tallied up the page hits.
I see we use the "pageviews" tool now - how available is that, and would it
be usable outside tools.wmflabs.org?
- d.
On 27 May 2018 at 19:36, Lewis Cawte wrote:
> I'm no
For this list, it's an obviously good idea. It's already search engine
indexed offsite, there's no good reason not to have our own archive do the
same.
- d.
On 28 April 2018 at 18:51, Brian Wolff wrote:
> Currently https://lists.wikimedia.org/robots.txt bans indexing the
> public archives of t
On 8 February 2017 at 17:03, Chad wrote:
> On Wed, Feb 8, 2017 at 8:30 AM David Gerard wrote:
>> Remember that the server sees *only* http:// connections, it isn't
>> doing SSL at all - SSL is terminated at the external IP.
>> (Can Apache even see if the incoming re
My problem:
I want to open an intranet wiki to the outside world, with Google
login to even view anything beyond the fact of its existence. (I
emailed the list about this before, and it's working great and
everyone is very pleased.)
We have an external IP, which accepts SSL on 443 and http:// on
We have an intranet wiki running 1.27 from tarball. I just installed
the previous version of GoogleLogin (0.3.1, because automatic account
creation works and it doesn't in 0.4). So you now need Google Login on
one of our whitelisted GApps domains to read it, let alone edit it.
What we want is to m
Don't forget to update the wiki:
https://www.mediawiki.org/wiki/Ubuntu
https://www.mediawiki.org/wiki/Debian
- d.
On 22 September 2016 at 07:44, Legoktm wrote:
> Hi,
>
> I'm very excited to announce that Debian and Ubuntu packages for
> MediaWiki are now available. These packages will follow
On 19 May 2016 at 02:52, Bill Traynor wrote:
> I'd love to try it out. Would it make sense to move my wikis to 1.23
> then, or can I jump right to 1.26?
I'd wait a week or two for a 1.27 RC tarball, 'cos that'll be the LTS.
- d.
___
MediaWiki-l m
SolveMedia is an ad-based captcha. I have an ancient 1.19 with it that
works, but I can't get it to work with 1.26 on PHP 5.5.9. PHP in
strict/warnings mode throws up:
Notice: Undefined index: adcopy_challenge in
/home/funsites/www/rwtest.davidgerard.co.uk/w/extensions/SolveMedia/SolveMedia.php
on
GMail is being flaky as hell about accepting or not accepting email
from the RW server, 173.255.233.133 - sometimes works, sometimes hits
spam, sometimes gets 550 refused (with no particular reason given).
Google doesn't do customer service, of course.
We don't *seem* to be in the email blackhole
Yeah ... for a while the Debian package for MediaWiki was distinctly
idiosyncratic (even by Debian standards), enough so that people were
recommended not to use it at all and just use the tarball.
For 1.19 it was less variant from upstream and well-maintained enough
to steer people to, but then th
On 4 February 2016 at 17:14, Brad Jorsch (Anomie) wrote:
> No, it was a change in the underlying image handling code that broke API
> prop=imageinfo in 1.27.0-wmf.12. Tracked as
> https://phabricator.wikimedia.org/T125804, and now resolved.
yep, working for us now. Thanks!
- d.
_
I came here to post about the same problem. It appears to have broken
this morning. Did the API change again?
On 4 February 2016 at 07:33, Dr. Michael Bonert
wrote:
> In September 2014, I described an intermittent recurrent problem associated
> with the InstantCommons images.
>
> I hadn't seen th
On 8 November 2015 at 10:43, Peter Presland wrote:
> Among other things I am considering moving to the Apache event-MPM using
> php-fpm and mod_proxy_fcgi because it is claimed to half (or better)
> process memory usage and provide major cpu-related speed improvements
> over the old pre-fork-MSM
On 26 July 2015 at 10:26, Mathieu Stumpf wrote:
> Now a more policy concern is, why would you ban an IP when it do indeed make
> webhosting? It sounds like a "shoot first at anything that would remotely
> look like a dirty face criminal, and then if it survive let it try to bring
> proofs that it
This should be doable in one jump.
Take a backup, then try the big jump. If it fails, try again via 1.19.
- d.
On Wednesday, 8 July 2015, John wrote:
> As a rule of thumb I would upgrade in +2 version increments to play it
> safe, and make sure you have a full backup first.
>
> On Wed, Jul 8,
RationalWiki uses $wgEnableDnsBlacklist = true; and
$wgDnsBlacklistUrls. We presently have xbl.spamhaus.org,
dnsbl.tornevall.org and all.s5h.net in there. These are pretty good,
but of course run on several hours' delay from the fresh proxy lists.
What we could really do with is one of webhosting
$wgUseNPPatrol = false; appears to have fixed it. Thank you!
On 11 June 2015 at 13:49, Alex Monk wrote:
> What about $wgUseNPPatrol?
>
> On 11 June 2015 at 09:16, David Gerard wrote:
>
>> I have an intranet wiki running 1.23 from tarball, upgraded from 1.19
>> (and
I have an intranet wiki running 1.23 from tarball, upgraded from 1.19
(and previously from 1.15 and originally 1.13). $wgUseRCPatrol =
false; is set because it's an intranet. But "[Mark as patrolled]"
links are showing up. How do I switch these off?
- d.
Upgrading work wikis from 1.19 to 1.23 with slight trepidation. It
went like this:
* Back up htdocs and database, unzip tarball into place
* php maintenance/update.php
* and you're done.
My goodness that was simple. Thank you for something that Just Works that well!
- d.
__
+1 on this. I've found MediaWIki on anything other than Linux (I've
also tried FreeBSD and Solaris) is something you can get to work
eventually, but you're fighting an uphill battle.
(The good bit in Linuxes compared to other Unixes is that they usually
have excellent dependency resolution. Tracki
I would disrecommend it - it's an effectively dead project, and the
problems aren't fixable. rationalwiki.org took it on board, since WMF
said it would be developed ... it still causes infuriating bugs (e.g.
PHP out of memory generating watchlists for one user with a ridiculous
watchlist) that we h
I sysadmin a few internal and one external wiki at work, and also
rationalwiki.org. One of the work wikis is SMW.
For the work wikis, OH GOD I WANT VE. Our users are normal competent
people who aren't so good with computers. But they can work a
word-processorish rich text editor just fine. For thi
https://ftp-master.debian.org/new/hhvm_3.2.0+dfsg1-2.html
Apparently it's in review, but should be present in Debian/Ubuntu in
due course \o/
- d.
___
MediaWiki-l mailing list
To unsubscribe, go to:
https://lists.wikimedia.org/mailman/listinfo/mediawi
On 9 December 2014 at 21:18, Max Semenik wrote:
> On Tue, Dec 9, 2014 at 10:53 AM, David Gerard wrote:
>> 2. What should I expect if I try to run 1.23 on HHVM?
> That it will not work: while HHVM tries to maintain a 100%, we had to fix
> various issues in MW to make the transit
So I'm looking at the CPU graphs for the Wikimedia clusters before and
after the switch to HHVM, and I'm going "oh my goodness, I'd like that
thanks".
rationalwiki.org is on 1.19 and I need to get off my backside and
upgrade it to 1.23 (only been putting it off approximately since 1.23
came out).
On 3 November 2014 18:35, Al wrote:
> Since MW is designed around a crowdsourcing model, I would just get
> volunteers from all parts of the company to help update the wiki. That will
> 1.) point-out the helpful employees of the company, 2.) be very educational
> for the editors about the new
On 12 November 2014 19:14, Rob Kam wrote:
> Yes it works fine through a web page, using Extension:MaintenanceShell:
Yeah, that sounds like a different php.ini for Apache and for
command-line. I've been bitten by this on Ubuntu, where Apache's is
/etc/php5/apache2/php.ini and command-line is /et
Oh, is Wikia now using this visual editor as opposed to their previous one?
On 11 August 2014 12:16, Ed Swing wrote:
> IIRC, Wikia does include the visual editor.
>
> -Original Message-
> From: mediawiki-l-boun...@lists.wikimedia.org
> [mailto:mediawiki-l-boun...@lists.wikimedia.org] On
Excellent, thank you!
On 3 August 2014 23:58, Benjamin Lees wrote:
> It looks like this functionality is actually present, just not documented.
> Set $wgAbuseFilterAnonBlockDuration.
> ___
> MediaWiki-l mailing list
> To unsubscribe, go to:
> https://
If I really wanted I could switch on the proxy blocking, which would
catch Tor too, but community consensus is we only do that when a
concerted troll is really pushing it (we're not popular enough for it
to be a regular thing).
For the moment I've set default abusefilter block duration to 314159
s
The default block duration for AbuseFilter is indefinite.
$wgAbuseFilterBlockDuration can be set to alter this.
My question is: can we make it different for IPs?
The use case here is RationalWiki, which is currently getting a very
tedious troll coming in from Tor. We don't actually want to switch
By default, MediaWiki sensibly blocks upload of SVGs with links.
However, we'd like to upload some to our intranet wiki (1.19). Is
there a setting to allow this?
(We already allow MS Office documents, so we're living on the edge obviously.)
- d.
___
On 1 March 2014 14:57, Dan Fisher wrote:
> How about making a temporary mirror for testing and seeing if you can
> upgrade it straight to 1.22.3, the current? New features in 1.22:
> https://www.mediawiki.org/wiki/Release_notes/1.22#New_features
I'm basically hanging out for 1.23 LTS.
- d.
_
I have discovered to my horror that I'm responsible for a MediaWiki
that's at 1.19.4. Obviously, I want to bring it up to 1.19.12.
Trouble is, not all the old patches are available. Is there an actual
reason for this, or were they just deleted as not useful? Would they
be available anywhere?
- d
On 28 February 2014 21:44, Bartosz Dziewoński wrote:
> You can try using this:
> https://www.mediawiki.org/wiki/User:Remember_the_dot/Syntax_highlighter
Or indeed https://en.wikipedia.org/wiki/User:Cacycle/wikEd :-) There
are a lot of 95% solutions that are quite practical.
- d.
On 28 February 2014 21:29, David Gerard wrote:
> On 28 February 2014 21:26, Dan Fisher wrote:
>> I'm sure this has been asked a number of times before. What is the easiest
>> thing that can be done so that when we click "edit" we see the Mediawiki
>> code
On 28 February 2014 21:26, Dan Fisher wrote:
> I'm sure this has been asked a number of times before. What is the easiest
> thing that can be done so that when we click "edit" we see the Mediawiki
> code highlighted in different colors and various other settings instead of
> it being plain text?
On 18 February 2014 22:11, David Gerard wrote:
> rationalwiki.org is getting hammered again. It looks like MySQL is the
> busiest portion - seriously just doing a lot of work.
Update: so we're finally taking action on this. We're just putting up
a new 4GB VM for Apache, we&#
rationalwiki.org is getting hammered again. It looks like MySQL is the
busiest portion - seriously just doing a lot of work.
Our current arrangement is: one box for MySQL, Apache, Lucene (the
latter reindexing weekly); two Squids; a load balancer. These are all
virtual machines on Linode (who we l
On 3 February 2014 15:38, Mark A. Hershberger wrote:
> Based on the issues raised during a meeting we had at the Architecture
> Summit a couple of weeks ago, those of us involved in MediaWiki release
> management have decided to introduce a monthly point-release cycle. On the
> last Thursday of
I want to add to the rationalwiki.org page header:
- per https://support.google.com/webmasters/answer/35624?hl=en#2
There are several meta tag extensions on mediawiki.org, in varying
states of maintanedness.
Which particular one do others use for this task, or a similar task?
- d.
_
On 3 December 2013 21:51, James Forrester wrote:
> "Not ready" means "not packaged as a 1.22 stand-alone install, and won't
> get back-ported security fixes".
> If you're happy with VE as it currently operates (which is probably "good
> enough" for an intranet unless you have loads and loads of t
On 3 December 2013 21:40, James Forrester wrote:
> On 3 December 2013 21:32, David Gerard wrote:
>> Here's a question: is there a set-out list of steps to enable Visual
>> Editor in this? Or is VE not quite ready for this?
>> Basically, what would the VE experience b
On 3 December 2013 21:26, Mark A. Hershberger wrote:
> We've put together RC3 for 1.22.0. Please test the tarball and report
> any bugs you find on Bugzilla: http://bugzilla.wikimedia.org/
Here's a question: is there a set-out list of steps to enable Visual
Editor in this? Or is VE not quite re
cheers dude :-)
- d.
On 20 November 2013 19:20, Chris Steipp wrote:
> Seems mediawiki-l strips them. I blame marktraceur.
>
> http://pastebin.com/cifbq6w1
>
>
> On Wed, Nov 20, 2013 at 11:06 AM, David Gerard wrote:
>
>> On 20 November 2013 18:52, Chris Steipp w
On 20 November 2013 18:52, Chris Steipp wrote:
> That's odd. I don't think it's intentional that those are missing.
> Here's the core patch for 1.19.7->1.19.8 in the meantime.
Where? No attachment made it through :-)
- d.
___
MediaWiki-l mailing li
I thought "I should probably upgrade our intranet wikis from 1.19.7 to 1.19.9".
I go to http://dumps.wikimedia.org/mediawiki/1.19/ ... and I see only
the patch from 1.19.8 to 1.19.9, not the one from 1.19.7 to 1.19.8?
Was there any reason the previous patches were removed? I'd rather not
just bla
On 12 November 2013 18:29, Chris Steipp wrote:
> I'm opposed to this change. A site administrator with a big enough
> community to address spammy links, and wants to enable this feature, is
> likely savvy enough to change the preference from true to false.
> I think setting this to false by defau
On 11 October 2013 17:33, Daniel Renfro wrote:
> At Vistaprint (www.vistaprint.com), we have an extensive internal wiki, and
> over the last six years, my team has built 50+ custom MediaWiki extensions.
> I'm happy to announce that we are starting to release these extensions as
> open source.
-- Forwarded message --
From: Brion Vibber
Date: 1 October 2013 19:11
Subject: [Wikitech-l] Officially supported MediaWiki hosting service?
To: Wikimedia-tech list
Question for the group:
Would an officially supported general-purpose MediaWiki hosting service be
useful to peopl
On 14 August 2013 23:56, Isabell Alcott wrote:
> ok, that sounds like a clear install.
> is it also possible to autoredirect to the wordpress installation ?
Probably. http://techrights.org redirects to
http://techrights.org/wiki/index.php/Main_Page - but
http://techrights.org/?stories goes to
http://techrights.org sort of does it with a MediaWiki taking the front
page and WordPress URLs working as one would expect.
The MediaWiki does not use nice URLs, it uses /index.php/ .
I chatted with Roy on IRC when he was setting this up ... basically, it was
a string of arcane rewrite rules whi
On 13 July 2013 23:13, David Gerard wrote:
> http://rationalwiki.org runs 1.19 LTS. This is fine. (Barring a DDOS
> over the past couple of days, so please excuse the site being slow.)
> Trouble is, the skin is for desktops and not suited to mobile at all.
> What can a 1.19 LTS s
http://rationalwiki.org runs 1.19 LTS. This is fine. (Barring a DDOS
over the past couple of days, so please excuse the site being slow.)
Trouble is, the skin is for desktops and not suited to mobile at all.
What can a 1.19 LTS site do for a better experience for its mobile
readers?
- d.
__
On 1 May 2013 05:40, Dan Fisher wrote:
> I've noticed Wikia has a nice WYSIWYG editor (FCKEditor, but its no longer
> being developed).
> Any recommendations for the best Mediawiki WYSIWYG editor that works in
> Chrome and Firefox? (and possibly IE too but not required as the market
> share is o
rationalwiki.org is currently serving pages very slowly. It's
intermittent, but when it's slow it's a *slug*. Many users are getting
502 errors from Apache or 503 from the Squids.
We have one Linode doing Apache/MySQL/Lucene. It's an 8GB box with 8
cores. (Was 4GB/4 cores, but Linode just doubled
On 3 April 2013 20:24, OQ wrote:
> On Wed, Apr 3, 2013 at 2:15 PM, David Gerard wrote:
>> MediaWiki plays nice with Squid out of the box:
>> https://www.mediawiki.org/wiki/Manual:Squid_caching
> Does it not require the source patches anymore? Or is that just if you
On 3 April 2013 20:05, Al Johnson wrote:
> From: David Gerard
>> So we got Redditdotted yesterday.
>> http://www.reddit.com/r/funny/comments/1bcqpn/scientific_evidence_of_evolution_being_a_hoax/>
>> This is our traffic:
>> http://rationalwiki.org/w/images/5/53/
So we got Redditdotted yesterday.
http://www.reddit.com/r/funny/comments/1bcqpn/scientific_evidence_of_evolution_being_a_hoax/
This is our traffic:
http://rationalwiki.org/w/images/5/53/Rw_traffic_20130331.png
And the wiki kept right on ticking.
- d.
_
On 12 March 2013 00:47, Krenair wrote:
> Try checking your PHP error log. It's probably somewhere like
> /var/log/apache2/error.log
That's the Apache error log, not the PHP error one (which is
separate), although it might also help.
- d.
___
MediaW
On 12 March 2013 00:45, Steve VanSlyck wrote:
> I had a page on my wiki.
> Created file link on it
> [[File:The_Proud_Father.jpg|right|thumb|250px|The Proud Father. Sam Brown,
> ca. 1920-30]]
> Saved the page.
> Followed the link.
> Uploaded the file (3021 x 3575, 1.55MB)...
> ... and, ever since
Will the assorted modules needed for Lua templates work with 1.19, or
do we need to get an actually recent MW for this? (Does it work in any
tarball release version yet?)
- d.
___
MediaWiki-l mailing list
MediaWiki-l@lists.wikimedia.org
https://lists.w
On 7 February 2013 21:28, Quim Gil wrote:
> I have done a couple of edits to http://www.mediawiki.org/wiki/Jobs
> Now http://www.mwusers.com/forums/forumdisplay.php?43-MediaWiki-Market
> appears at the top and there is a new link to the MediaWiki skills page at
> LinkedIn.
That's a really good
On 10 January 2013 05:41, Dan Fisher wrote:
> I've been recommended to use the following setup in Linode:
>>> NGINX, PHP-FPM with APC and MySQL.
I understand that APC is not nearly as useful with fcgid as it is with
libphp5 - the cache is per PHP process (each of which terminates after
a few ru
On 17 December 2012 20:51, Iry Witham wrote:
> I have been tasked with upgrading a wiki from mediawiki 1.11.0 to the latest
> version. I started with a VM running SLES 11 R2, Mysql Ver 14.12 Distrib
> 5.0.96, for suse-linux-gnu (x86_64) using EditLine wrapper and PHP 5.2.14.
> There have be
On 30 November 2012 03:46, Dan Fisher wrote:
> Thanks, I'm looking into Linode. I've been recommended that once before on
> this list. Does it handle traffic spikes well? Or would that trigger a CPU
> alert? I've gotten those alerts before from my present hosting company and
> it stresses me out.
On 26 November 2012 17:18, Mark A. Hershberger wrote:
> On 11/26/2012 12:14 PM, Quim Gil wrote:
>> That wouldn't be the solution of the tarball problem you are describing
>> but it would be really useful for the majority of MW admins having to
>> search for extensions *after* they miss something.
On 23 November 2012 18:05, Mark A. Hershberger wrote:
> Bug filed: https://bugzilla.wikimedia.org/show_bug.cgi?id=42393
:-D
> Note that there may be more work needed, but we definitely need
> *something* like this.
Notes added, from the perspective of a tarball user.
- d.
___
On 23 November 2012 21:36, Arcane 21 wrote:
> For instance, if an extension is confirmed working for "1.17" and "1.18" but
> there are no tags for 1.19.2 or higher, this still gives incomplete
> information on extension compatibility. However, I do agree it is better than
> nothing, and since
I've just upgraded a couple of our wikis with the 1.19.2 tarball. It
was wonderfully smooth and everything just worked. Nice one.
A suggestion: How feasible would it be for the upgrader to (a) detect
extensions in use (b) check their mediawiki.org page re:
upgradeability [template-scraping] and gi
On 29 October 2012 23:56, Dave Humphrey wrote:
> As a rough guess at what to set MaxClients to use "top" and look at the
> difference between the RES and SHR columns of all httpd processes. I
> believe this is roughly the amount of non-shared memory each of the child
> Apache processes are using
So, rationalwiki.org has been *much* faster and more usable with a
coupla squids and a load-balancer in front of the
Apache/Lucene/database node. (We could probably cope with just one
squid, but Trent wanted to experiment.) The nodes are all Ubuntu 10.04
Linodes, the software manually kept up to da
On 25 October 2012 02:15, Sumana Harihareswara wrote:
> Sometimes we have to shed some backwards compatibility. And it sounds
> like removing support for these particular skins will enable us to
> innovate and move faster. But it's a good idea to publicize this sort
> of thing so legacy skin us
On 19 October 2012 22:29, David Gerard wrote:
> On 17 October 2012 23:25, David Gerard wrote:
>> The problem: In general, rationalwiki.org is melting under the strain.
>> We can't quite afford the next step up right this moment, though want
>> to plan for it, an
On 20 October 2012 22:56, Platonides wrote:
> On 20/10/12 10:03, David Gerard wrote:
>> And we've dropped the hit counters, but the stats.grok.se code is
>> available as GPLv3 (a fact that should be more widely known):
>> https://github.com/abelsson/stats.grok.s
On 19 October 2012 22:29, David Gerard wrote:
> What we've done: Trent's rented a couple more small (512MB) Linodes to
> experiment with, and put Squid on them. And we've abandoned page
> counts, to slight angst from the users. So far, just running what we
> can throug
On 19 October 2012 22:43, OQ wrote:
>> What we've done: Trent's rented a couple more small (512MB) Linodes to
>> experiment with, and put Squid on them. And we've abandoned page
>> counts, to slight angst from the users. So far, just running what we
>> can through a Squid has done *wonders* for t
On 17 October 2012 23:25, David Gerard wrote:
> The problem: In general, rationalwiki.org is melting under the strain.
> We can't quite afford the next step up right this moment, though want
> to plan for it, and need to keep things from intermittently blowing up
> as they are.
On 18 October 2012 00:23, K. Peachey wrote:
> On Thu, Oct 18, 2012 at 8:25 AM, David Gerard wrote:
>> We live on a 4GB Linode. I don't have a full handle on what we've got
>> on it - it's accreted in an ad-hoc fashion with some horrible bodges -
>> but the
On 14 October 2012 23:49, Chad wrote:
> Indeed. You're probably fixing some other bugs as well by upping that
> limit. Really, PHP should have about 100M of memory available to it to
> run MediaWiki. We're a memory hog, but hey, them's the breaks.
yeah, fair enough.
> I don't know of
On 15 October 2012 06:04, Tim Starling wrote:
> That's processes, not memory. I would think that the effect on memory
> would not be so large.
Memory taken appears to be processes * PHP memory_limit . Unless I'm
ludicrously wrong about that.
> Disabling keep-alive should be counted as a
> fai
On 14 October 2012 16:34, Platonides wrote:
> On 13/10/12 00:22, David Gerard wrote:
>> Apache connections are really pretty damn cheap these days. Is
>> KeepAlive actually a good or bad thing for MediaWiki?
> The apache docs say:
>> In some cases this has been shown t
On 13 October 2012 23:50, Krinkle wrote:
> Update LQT to the latest version from Git master[1][2].
> That should make it work again (be sure to make backups first, and run
> update.php if needed).
> [1] https://www.mediawiki.org/wiki/Extension:LiquidThreads
> [2]
> https://gerrit.wikimedia.org/
https://www.mediawiki.org/wiki/Manual:File_cache
Enabled this on rationalwiki.org and the results are amazing. The site
is just ridiculously fast to browse logged-out, and I'm almost looking
forward to our next Reddit-dotting.
Is there any reason this is not enabled by default? (Is the
possibilit
rationalwiki.org was foolish enough to start using the experimental
LQT2. So of course the project was abandoned and we're stuck with a
pile of content in LQT, and it broke in subtle and awful ways when we
upgraded from 1.16 to 1.19.
Symptoms we're seeing:
* The loader gif for the toolbar keeps s
On 13 October 2012 14:35, David Gerard wrote:
> I'm getting reports on rationalwiki.org of occasional blank
> watchlists. The symptoms are that a completely empty page loads
> immediately. It's running 1.19.1 in Ubuntu 10.04 with Apache
> 2.2.14-ubuntu.
> Of course, I
On 13 October 2012 18:43, David Gerard wrote:
>>> I'm getting reports on rationalwiki.org of occasional blank
>>> watchlists. The symptoms are that a completely empty page loads
Here's the on-wiki discussion:
http://rationalwiki.org/wiki/RationalWiki:Technical_s
On 13 October 2012 15:24, Dave Humphrey wrote:
> On 13 October 2012 09:35, David Gerard wrote:
>> I'm getting reports on rationalwiki.org of occasional blank
>> watchlists. The symptoms are that a completely empty page loads
>> immediately. It's running 1.1
I'm getting reports on rationalwiki.org of occasional blank
watchlists. The symptoms are that a completely empty page loads
immediately. It's running 1.19.1 in Ubuntu 10.04 with Apache
2.2.14-ubuntu.
Of course, I changed *two* things last night: (a) cut max memory for
PHP from 256MB to 64MB (b) sw
(I have just accepted root on rationalwiki.org and am looking around
in slight horror. I will be sending a few messages like this.)
Apache comes with KeepAlive on by default. I am unconvinced this is
actually a good idea. I just switched it off and it appears to have no
ill effects, and the server
On 2 October 2012 15:55, Trevor Harmon wrote:
> On Oct 2, 2012, at 7:49 AM, Jeremy Baron wrote:
>>> I am using MediaWiki 1.16.4.
>> That's really old. Are able to upgrade?
> 1.16 is what my hosting provider offers for their one-click install.
> Are you saying 1.16 is causing this problem? I wo
On 28 September 2012 17:01, Mark A. Hershberger wrote:
> I do agree with David Gerard's assessment, though. We need to make sure
> that whatever we use is going to work with package management tools that
> Debian and Redhat and the like already use.
The other reason is, of course, making the d
On 28 September 2012 02:47, Mark A. Hershberger wrote:
> The problem, though, is that there is no way to install, use, or update
> extensions apart from doing it by hand. Requiring the installation of
> multiple modules by hand isn't going to lead to a thriving, modular
> ecosystem. We need a d
1 - 100 of 199 matches
Mail list logo
