#foswiki 2013-01-09,Wed

↑back Search ←Prev date Next date→ Show only urls(Click on time to select a line by its url)

WhoWhatWhen
gac410Item12337 - jast ... any comments on how or if to fix? [00:31]
FoswikiBothttp://foswiki.org/Tasks/Item12337 [ Item12337: Item11755 added an undocumented dependency on HTML:Treebuilder >= 4.0 ] [00:31]
gac410So... should Item12337 block 1.1.7? [00:43]
.... (idle for 15mn)
SvenDowideitgac410, um, i would fear so
given what a patch should be
SvenDowideit doesn't really feel he has a right to an opinion atm
[00:58]
gac410yeah. That's my thought. Though the problem is in 1.1.6.
why not?
[00:58]
SvenDowideitits not like 1.1.5 works with utf8..
cos i'm not really here
[00:59]
gac410yeah true. So just revert the 1.1.6 fix and 1.1.7 == 1.1.5 at least for CompareRevisionsAddOn.
Ugly... in that some sites might have taken advantage of the 1.1.6 fix. :(
[00:59]
SvenDowideitat some point the utf8 people need to help get a release happening [01:00]
gac410yeah... I'm not really in a position to test this either way... or even fully understand it
Reading through Item11755 i'm not sure reverting is good either, as the code is pretty badly broken.
[01:01]
FoswikiBothttp://foswiki.org/Tasks/Item11755 [ Item11755: compare breaks charset encodings ] [01:02]
SvenDowideitgoing from 'installed and working' in 1.1.5 to installed and crashing
but the code is less broken - it just crashes
[01:03]
gac410Part of me really despises that the distributions hang so far behind and don't provide any easy way to keep up to date. CentOS shipped in July 2012 has cpan from 2010. [01:03]
SvenDowideitSvenDowideit thinks that we should push foswiki to cpan
and then re-do all instructions to use cpanminus
and use plack
[01:04]
gac410Then we'll have distributions running foswiki 1.0.4 if we are lucky. [01:04]
SvenDowideitwhy?
the installation would be
[01:05]
gac410redhat? Perl from punched card days?
<humor>
[01:05]
SvenDowideit1. either download bootstrap-foswiki.pl and put in cgi dir, or goto ssh and type cpanm foswiki [01:06]
gac410Seems like boiling the ocean because of a small dependency issue. [01:09]
SvenDowideitSvenDowideit fails to see, and does not agree, but meh [01:10]
gac410move all of foswiki to cpan because we accidentally depneded on HTML::TreeBuilder >4.0 vs. 3.23 ? [01:10]
SvenDowideitSvenDowideit has no idea why you'd think this is only because of the one latest issue [01:11]
gac410tbh I don't even see how they are related. I have no end of grief with CPAN... why would I ever want to move Foswiki there. All I want to do is get 1.1.7 out so we have the CVE's integrated into a release.
If CentOS can't install HTML::TreeBuilder >= 4.0, how does putting Foswiki into CPAN help,
[01:13]
SvenDowideitif installing foswiki === cpanm foswiki
then 'centos' and its packages / lack are irrelevant
and 'what OS' becomes a non-question
[01:15]
gac410Does CentOS even have cpanm? Why not just have them "cpanm HTML::TreeBuilder" and we get 1.1.7 out. [01:16]
SvenDowideitimo 1.1.7 has a fixed set of dependencies - the same ones as 1.1.0
but then again, i/we broke that by not getting a 1.2.0 out 6months after 1.1.0
[01:17]
gac410Well we can't "un-ring" the 1.1.6 bell. It's already out with changed dependencies. (And I suspect there have been others as well.)
So either we revert the 1.1.6 fixes for 1.1.7, to fix the dependencies, and break the characterset fixes in 1.1.6 or find some other solution. Or expect users to install the dep.
[01:18]
SvenDowideitor have a conditional [01:20]
gac410If it's okay to install foswiki using cpanm, why is it not okay to install a dependency using cpanm. [01:21]
SvenDowideitso that 1.1.7 can work with new and old, and anyone that encounters the bug can have the fix by doing the hard work
if foswiki is installed via cpanm, then cpanm is a dependency, and can be relied on further
[01:21]
gac410I need the urf8 guys to fix that, since I can't really test to know if whatever I write works or breaks things further. [01:21]
SvenDowideitbut... in a patch release that fixes a new dependency to add yet another dependency?
so if utf8 people need to do work, the only incentive we can use, is to remove the fix from 1.1.7 unless they fix it better fast enough
you should not have to perform magic on their behalf
[01:22]
gac410but if the reason we have dep. issues is because of the commandment "Thou shalt use the package manager"
yeah. Probably it was my mistake. I'm the one that cherry picked the fix over from 1.2
[01:23]
SvenDowideitno, we have depds to allow installers to use their own policies
and some people have narrower rules than we do
[01:23]
gac410well it's jast's fix for MichaelDaum's issue. So I'll leave it to them to figure it out. Otherwise I just revert. Unfortunately anyone who installed 1.1.6 to get CompareRevisionsAddOn working will be broken again. [01:26]
***SvenDowideit has quit IRC (Quit: ZNC - http://znc.sourceforge.net) [01:31]
gac410Okay, I made Item12337 awaiting feedback from Michael and Jan. [01:32]
FoswikiBothttp://foswiki.org/Tasks/Item12337 [ Item12337: Item11755 added an undocumented dependency on HTML:Treebuilder >= 4.0 ] [01:32]
gac410SvenDowideit: I made the task waiting for Jan and Michael. [01:34]
SvenDowideitpersonally i'd be more passive aggressive and revert the fix too
with the timeline being obvious
but then as i say, i'm not really feeling it
[01:37]
gac410I would if the change was in 1.1.7. But given we shipped 1.1.6 with a working fix for a set of users, that seems a bit draconian to those who need the fix. [01:38]
SvenDowideitit might seem draconian to reval a bug for a small number of utf8 people
but to crash foswiki for the others seems worse
if i grok it right, its a fix for users of an undersupported option (non-ascii) which on patch update breaks all users that have an older but correct (as per pre-patch and listed deps) pre-req
so its draconian for both parties
[01:39]
gac410y. At minimum we should doc the dependency. How hard can it be to install a cpan module. Well maybe I'll add a "can ..." test and make the code conditional.
So at least it won't crash
[01:41]
SvenDowideitand given the possibility that some may not have upgraded to 1.1.6 due to crash, and then try to do the same because of the rather more importatn CVE fixes [01:42]
gac410Well we **claim** support for languages. wtf are we bothering to translate if our code languages code is totally unsupported, and we can't fix bugs that corrupt encodings. [01:42]
SvenDowideitSvenDowideit can't see 'how hard can it be' as a rational response to the question of 'i only updated the patch due to a CVE, and suddently have to de-sabelise my setup' [01:43]
gac410but y. I agree with you. unfortunately. [01:43]
SvenDowideit(its not like this is my full opinion, but it felt like you were taking one side of the argument)
SvenDowideit is trying to figure a way to reduce foswiki to 2 minimal server components: store&query accelerator
and to move everything else into the browser
[01:44]
gac410Looking at commit log, we definitely have made other dep. changes along the way. Added version(), SHA1 dep change, Added Crypt::PasswdMD5, etc.
So it's not like we've tried really hard in the past to freeze dependencies
[01:45]
SvenDowideitfairysnuf :) [01:46]
.... (idle for 19mn)
gac410So after all that... I'm looking at HTML::TreeBuilder 3.23 and it documents no_expand_entities() [02:05]
https://metacpan.org/module/JFEARN/HTML-Tree-3.23_1/lib/HTML/TreeBuilder.pm Seems to be there. [02:11]
.............................................. (idle for 3h48mn)
SvenDowideitOHPOO
so I thought i'd have some fun with modern html5 stuff
and for the last i dont' know how long, i've been failing - turns out foswiki is protecting me from 'non-standard' html
SvenDowideit commits .... quricide
or is that curicide
mmm, and crawford would commit curricide?
pharvey, do you know where the &gt; etc escape code is?
[05:59]
MUMBLE. [06:15]
........... (idle for 52mn)
pharveySvenDowideit: lots of places [07:07]
...... (idle for 28mn)
SvenDowideitpharvey oh wonnerful
so i guess that leaves me back to thinking about not using foswiki to convert from {tml}.txt to {widget}.html
as using foswiki's rendering has so many legacy bits
SvenDowideit wasted 3 hours, and now is surrounding this html5 magic in literal tags
thus it works
oh wow. and to boot, debian stable's chrome has bugs that make it less likely to work
[07:35]
pharveyworse than the bugs of latest chrome on win8? [07:40]
SvenDowideitprobly
if you have a class definition with an element named 'default' you get an error
that one hit me yesterday when i changed computers
[07:41]
pharveythat is interesting [07:42]
.... (idle for 15mn)
SvenDowideitawcrud. i guess i'm going to start working with the development version of crhome
i think i just dropped off the insane side of webdev
oooooooo, whataterribleidea
[07:57]
when the user hits 'inline wysiwyg edit' convert transcluded sections to a canvas element the same size, so they become static for normal editing
i wonder what gets saved
oh dear, /me goes further down yet another rabbit hole
[08:08]
.............. (idle for 1h8mn)
jastare we okay with a workaround for Item12337? [09:18]
FoswikiBothttp://foswiki.org/Tasks/Item12337 [ Item12337: Item11755 added an undocumented dependency on HTML:Treebuilder >= 4.0 ] [09:18]
RockumSockumhello
anyone on here?
blast
[09:21]
SvenDowideitowo
if you'd stuck around just 10seconds more
mmm, does anyone know if there's a way to get the canvas element to show its fallback html?
ie, tell a canvas element to pretend the browser does not support canvas
(ie, roll over, play dead)
[09:22]
jastis that just the contents of the canvas tag? (haven't played with canvas yet) [09:26]
SvenDowideitjast ya
i have a mad idea, based on todays curi-cide
'death by curiosity'
[09:28]
.... (idle for 16mn)
jastwith jQuery, I'd guess something like $('#mycanvas').replaceWith($('#mycanvas').children()) [09:45]
***ChanServ sets mode: +o MichaelDaum [09:59]
GithubBot[foswiki] FoswikiBot pushed 1 new commit to master: http://git.io/-XeMHw
foswiki/master 55c4c68 JanKrueger: Item12337: remove dependency on version 4.0 of HTML::TreeBuilder...
[10:00]
***GithubBot has left [10:00]
FoswikiBothttp://foswiki.org/Tasks/Item12337 [ Item12337: Item11755 added an undocumented dependency on HTML:Treebuilder >= 4.0 ] [10:00]
MichaelDaumhey guys
Sven, do you plan to fix WikiDraw anytime soonish?
as far as I see we don't have a workable solution for this feature atm, right?
[10:01]
SvenDowideitMichaelDaum pharvey tells me that wikidraw does actually work [10:06]
MichaelDaumreally? [10:06]
SvenDowideitthough i've not reviewed it at all
thats what i'm told
i don't recal having trouble the few times i've used it
[10:06]
MichaelDaumit failed badly for me. opening it up showed a lot of javascript yadda needing some loving care [10:07]
SvenDowideitnoidea, too vague :) [10:07]
MichaelDaumpharvey, do you have WikiDrawPlugin up'n running atm? [10:08]
SvenDowideitatm, i'm trying to work out some rather different fun, still on holidays :) [10:08]
MichaelDaumSvenDowideit, there's a bug report on it with more
MichaelDaum hopes Sven doesnt get his a
burned off
[10:08]
SvenDowideitits happening [10:08]
MichaelDaum:( [10:08]
SvenDowideiti spent 3 hours today on something not working
i finally relased foswiki was escaping the <
pretty much the office temp being too high i think
[10:09]
MichaelDaumany fires near u? [10:09]
SvenDowideitcourse :)
but thats normal
our house is safe enough tho
and thankfully the girls don't have any breathing issues
[10:09]
MichaelDaumhow do you prepare defending the house? or any escape planes? [10:10]
SvenDowideitwe bought somewhere where its pretty much 90% safe
where the last 10% would really assume no avenue or time to escape anyway, so why worry
same wrt flood - though as we're at 60m, thats a fricken huge tidal wave
[10:11]
MichaelDaumstrange country. burns or floods. ;) [10:12]
SvenDowideityup
and once in a while an iceberg floats by
[10:12]
MichaelDaumwow [10:14]
SvenDowideiti think its happened once in my lifetime :p [10:14]
jastMichaelDaum: if you have some time, I'd appreciate if you could do a quick test of <http://git.io/-XeMHw>, too, especially if you have an older system with, say, perl 5.8 handy
Unicode handling changed with pretty much every new 5.x series
for some reason I can't get my own 5.8.9 (via perlbrew) to work; crashes somewhere in Foswiki.pm
[10:16]
SvenDowideitdarn, i ca'nt find any way to get a canvas element to act as though canvas was not supported [10:19]
MichaelDaumjast, ya will do. the compare rev thingy?
perl 5.8? never seen this one for a while.
jast, the patch looks familiar. am using a similar pattern in other plugins as well.
[10:19]
jastyeah... [10:22]
MichaelDaumall would be easier if anything < utf8 (or better utf16) was banned from planet earth [10:22]
jastSvenDowideit: rough testing seems to suggest that my approach works
but then again I just inserted a dummy canvas element into a blank page
utf16 is no improvement over utf8 IMO
but I agree, of course
perl 5.8 is what's in RHEL and such
in any case, Foswiki is supposed to support it :)
[10:22]
SvenDowideitjast 'your approach'? [10:24]
jast110118 < jast> with jQuery, I'd guess something like $('#mycanvas').replaceWith($('#mycanvas').children()) [10:24]
SvenDowideitoh [10:24]
jastjust tried it without jQuery (on about:blank) [10:24]
SvenDowideityeah, that would look like what i asked
but not be what i asked :)
i still want the canvas element to be there
[10:24]
jastwell, best I can think of is to temporarily replace it with a dummy element [10:25]
SvenDowideitie, not dissappear, just play dead
i'm hoping for more - i was hoping to fake up a shadowDOM
but i guess i'll have to try something more replacy
[10:25]
jastMichaelDaum: using Encode was pretty much the first thing I thought of in the first place, but I didn't figure out exactly the right places to do it before now :) [10:28]
wow, deep purple temperatures? that's rough [10:33]
MichaelDaumbtw what are the reasons configure has to run as a pure cgi, i.e. not as fastcgi? [10:44]
jastprobably not fortified against accidentally persisting values [10:45]
SvenDowideitlack of code
(seriously)
[10:45]
MichaelDaumpure cgi is a neidmehr unter nginx [10:45]
jastI suspect you could make a FastCGI wrapper fairly easily
one that pretends to the actual script that it's in a normal CGI env
something something fork something something
[10:45]
MichaelDaumyea I know: a fastcgi stub that forks the real configure everytime [10:46]
jasthardly ideal, of course [10:47]
MichaelDaumnothing I'd deploy for a customer. lighttpd is much better in that respect.
only drawback of both (nginx, lighttpd): no kerberos sso
MichaelDaum would send hugs to anybody knowing how to kerberos sso using lighty
[10:47]
jastneeds a module, right? [10:50]
MichaelDaum... no virt-hug for that one ;) [10:51]
.... (idle for 19mn)
SvenDowideitmmm
so - anyone wanna take bets on how much foswiki rendering is stuffed up if one adds a html element wrapper around every %INCLUDE render output?
[11:10]
jastouch
<a href="%INCLUDE{"MakeMeAUrl" foo="bar"}%">Boom!</a>
[11:11]
SvenDowideitthat is correct, except for one detail
its not used as often as i thought
[11:12]
jastnot by Foswiki core, perhaps [11:12]
SvenDowideitgo grep for all uses of INCLUDE in your massive wiki [11:13]
jastperhaps not even by extensions [11:13]
SvenDowideitand compare
i was expecting major death
but instead i'm finding its a small proportion, not a big one
i'm still trying ton convince myself that we should code conveniences for the main uses, and work out better ways to do the others
[11:13]
jastwhat kind of wrapper were you thinking of? [11:15]
SvenDowideithtml5 widget :)
SvenDowideit is still working on pushing all rendering to the browser
[11:16]
jastwill that work in old browsers, though? [11:16]
SvenDowideitwho cares [11:16]
jastwe do
if foswiki breaks compat with IE7/8 we'll have to fork it
[11:16]
SvenDowideitthis is not likely to see the light of day without thousands of hours work
i would also fork with you
i'm interested (in my holidays) in exploring what comes next
[11:17]
jastplus, personally, I hate web apps that are unusable without javascript [11:17]
SvenDowideitas foswiki code and concept is >10 years old [11:17]
jastyeah, exploring is good, definitely
I've had some ideas myself, but they're tricky to implement given today's web ecosystem
[11:18]
SvenDowideitand in the last few days i've seen some interesting things :)
its much less tricky than it was 5 years ago
you said ie7
i still come from a world of ie5
[11:18]
jastI remember those days [11:18]
SvenDowideitand i wrote stacks of dhtml
and used contenteditable (or whatever it was called then)
[11:18]
jastI started web stuff back when PHP 4 just started getting popular [11:19]
SvenDowideitphp4?
SvenDowideit started writing his first dynamic webapp in perl4 in ... er... 1991
[11:19]
jastperl4 was before my time :} [11:20]
SvenDowideiti wish it was before mine [11:20]
jastin 1991 I had a Commodore 64 so I used BASIC and 6510 assembler [11:20]
SvenDowideiti used to think it was, but then i looked at the release dates
hehe
[11:20]
jastalso I was 7 :} [11:20]
SvenDowideitstarted uni in 1990, so got access to more
trs80, vic20, 8086, then onwards&upwards
i wonder if anyone uses the webrss and webatom topics now
[11:20]
jastI mostly rely on notify [11:23]
SvenDowideitaha. WebSearch is broken due to my mucking [11:23]
jastbut then again I'm not as much into RSS as some other people are [11:23]
SvenDowideiti remember using rss lots, but um, that was when twiki first started [11:24]
jastthere's just too much XML in the world if you ask me :} [11:24]
MichaelDaumargh a nerd to nerd race on legacy crap [11:25]
SvenDowideityup, its almost bed time for me
so fun to be had
[11:25]
jastMichaelDaum: yeah, that's a nice way to think about XML [11:25]
SvenDowideitits either that, or i install the chrome developer version :( [11:25]
MichaelDaumtheres other things being a lot more wrong than xml. [11:25]
jastsure
almost everything is wrong
[11:26]
SvenDowideitna, rdf is just xml in disgust [11:26]
MichaelDaumxml is okay. [11:26]
jastxml is a bad solution to a non-issue IMO [11:26]
SvenDowideiti'm so glad few people agree with you MichaelDaum :)
else i'd have to write more xslt's
[11:27]
MichaelDaumthere are a lot of severe problems that are solved by xml
for one otd
[11:27]
jastabout the only good thing about XML is that it's transformable, and even that is solved badly [11:27]
MichaelDaum^odf I mean [11:27]
jastxml isn't the thing that makes odf a solution [11:28]
SvenDowideitmmm, i wonder if i compare odf to hitler, if that might bring an end to it :p [11:28]
MichaelDaumthe real problems is not on a tech level. it is on a political and market level of organisations crapping standards [11:28]
jastany other well-defined and easy to parse data representation 'language' could have done the same
crapped standards are a bigger problem than XML, yes
[11:29]
SvenDowideitjast if you bring up netcdf, there will be sniggering [11:29]
MichaelDaumproblem is good standards being sabotaged by corps. thats what we are in today. [11:29]
jastbut XML promoted the illusion that making good standards is a simple problem [11:30]
MichaelDaumno
xml is just xml
[11:30]
jastif only the data transport mechanism is flexible enough, everything will magically work out!
the same attitude existed before XML, of course
and it pounced on XML, and here we are
the reason we don't have a better de-facto standard representation for structured data is that XML was in the right place at the right time
I'm not sure the new-fangled JSON is much better, though at least it's slightly easier to parse
[11:30]
MichaelDaumpeople have sabotaged standards being expressed in xml by explicitly _not_ agreeing on specs and make it all optional and pluggable beyond recognition. thats when good standards degrade to a hydra then being called framework [11:32]
jastspeaking of: sharepoint doesn't support CMIS via REST. yay! [11:33]
MichaelDaumjson shmason. the tech does not matter. corps will f* it up for you.
cmis. good example.
oauth nother one
[11:33]
jastoauth2 in particular, or so I've heard [11:34]
MichaelDaumshould be called ooutch
reading cmis specs it is full of phrases like "this may be implemented by a repo. use the capability discovery channel to find out more"
I mean specs are there to _require_ stuff to be implemented, not making it a mere eventuality
[11:34]
SvenDowideitmmm, thats not the experience i have
specs tell you how it should be implemented should you choose to do so
and if you're lucky recomend a fallback response if they don't
discovery being the avoidance of fallback spec
[11:36]
jastit's a good thing to be able to support extensions
it's not a good thing to offload core functionality into extensions
in a spec, that is
[11:39]
SvenDowideitunless you are in the real world
where the specs are written after several products exist
[11:41]
jastit's perfectly possible to write a good spec for a new protocol, and then add support for it in existing products
just not all that likely in the corporate world, for the most part
[11:53]
SvenDowideity, possible, if only it were realistic [11:54]
jastjust depends on the area [11:55]
SvenDowideitthe entire http auth area is a depressing morrass [11:55]
jastCMIS? very unlikely
HTTP? too many implementations to fix
[11:55]
SvenDowideiti can't think of an area
as adding support in existing products isn't that simple if the standard does not take the product's fundamental design issues into accout
like writing a simple transclusion spec that works for all wiki's
given that foswiki's and mediawiki's transclusion mech are basically opposites
[11:57]
jastyeah, but that's a difficult example
there are things with much more overlap than that
[12:00]
SvenDowideitreally?
SvenDowideit hopes you think of one eg to make my life easier
[12:01]
jastthe thing RSS set out to do, for example
a list of updates isn't *that* hard to spec
[12:01]
SvenDowideitmmm, a data format spec not a function spec, true [12:01]
jastI was thinking mostly of data format specs, yeah
but function specs can work, too, if the function is coordinated across multiple projecrts
and that *does* happen
but most of the time, proprietary projects simply aren't interested in cross-compatibility
so then of course it does *not* happen :)
[12:02]
........................... (idle for 2h12mn)
GithubBot[foswiki] FoswikiBot pushed 1 new commit to master: http://git.io/n5yK4Q
foswiki/master 8c8fdbe TimotheLitt: Item12180: Item11633: Expose execution environment variables in configuration audit...
[14:15]
***GithubBot has left [14:15]
FoswikiBothttp://foswiki.org/Tasks/Item12180 [ Item12180: Implementation for AJAXOnDemandCheckersForConfigure ] http://foswiki.org/Tasks/Item11633 [ Item11633: Run checkers from a Foswiki page ] [14:16]
...... (idle for 28mn)
GithubBot[foswiki] FoswikiBot pushed 1 new commit to master: http://git.io/XE9qhQ
foswiki/master 8bd536c TimotheLitt: Item12180: Audit webserver: decode execution undefs correctly; display (existing but) undefs correctly...
[14:44]
***GithubBot has left [14:44]
................ (idle for 1h16mn)
orev has left [16:00]
...... (idle for 27mn)
Simon_DHGreetings, I'm new to foswiki and am hoping for some help with a problem. I have setup FosWiki 1.1.6 and installed the latest version of the EditChapterPlugin. However, when I click on the edit chapter icon I receive a popup with the text "Invaild JSON-RPC request - no method". [16:27]
Has anyone seen this error both? [16:36]
........... (idle for 51mn)
mtHi, first of all: great work! I am using Foswiki for about 2 years now and it works pretty good. I am now moving to another system and in the first step installed Foswiki from the Debian package. I configured Apache using the proposed apache.conf and created a .htpasswd file with the user who will configure the system. When calling the foswiki page in the browser, however, I receive an "Internal Server Error". The apache error log does not
provide any information and the foswiki log does not yet exist. Which steps can I do to find the problem?
[17:27]
................... (idle for 1h33mn)
Can anybody help? [19:00]
...... (idle for 29mn)
***Demosthenex has quit IRC (Ping timeout: 245 seconds) [19:29]
...... (idle for 29mn)
mt has left [19:58]
.............. (idle for 1h8mn)
SvenDowideitmt huh?
installed Foswiki from the Debian package
means that you do not need to put anything in the apache conf
its already done for you
[21:06]
gac410SvenDowideit: Did you see the q. about the broken regex in the deb lsc?
Foswiki:Support/BestPracticeTip32
[21:07]
SvenDowideitgac410 another one? [21:08]
FoswikiBothttp://foswiki.org/Support/BestPracticeTip32 [ BestPracticeTip32 ] [21:08]
SvenDowideitSvenDowideit has coniptions [21:08]
gac410Noticed it pop up in the rss feed while I was away.
Same one? The email regex?
[21:08]
SvenDowideity, thats the same one
and the solution proposed is a good one
excellent :)
[21:09]
gac410Why is it broken in the deb. It's not broken in the tgz. [21:09]
SvenDowideitits odd
i've experienced it with the tgz too
[21:10]
gac410Strange. [21:10]
SvenDowideitits not due to something simple, and it doesn't seem to be in the deb itself [21:10]
gac410Is it a perl version thing? I've never run into it. [21:10]
SvenDowideitit seems to happen due to some combination of perl mucking i think
i suspect it due to how i hand modify the cfg using sed & perl -e
i'll poke it when i do 1.1.7 one day soon :)
in fact, thinking about what i have to do when i go back to work, very soon
[21:10]
gac410Jan posted an alternate fix for Item12337 - [21:12]
FoswikiBothttp://foswiki.org/Tasks/Item12337 [ Item12337: Item11755 added an undocumented dependency on HTML:Treebuilder >= 4.0 ] [21:12]
gac410So I guess I'll do a 1.1.7-RC3 and see if we can get the original reporter to test, if he hasn't updated TreeBuilder to the new version. [21:12]
SvenDowideitdid he have the conviction to commit it ? :) [21:12]
gac410To trunk :P [21:13]
SvenDowideitexcellent work :) [21:13]
gac410Which in some ways it good, since trunk has locales enabled. damn... not any more. I need to re-enable locales. It was disabled due to the maketext bug. [21:14]
SvenDowideitgibber [21:14]
Simon_DHGreetings, any chance either of you have an insight into the EditChapterPlugin problem I have run into? [21:14]
SvenDowideitSvenDowideit doesn't use it :/ [21:15]
Simon_DHlol, this is true, at least in my case. [21:15]
SvenDowideitbut i wonder if its missing a dependency
ok, have you enabled all the pre-req's
RenderPlugin i mean :)
the others are marked optional?
mmm, and um, what skin
micha used to test mainly with natskin, though lately i think he's tested with pattern to
[21:16]
Simon_DHJQueryPlugin (v.4.45), RenderPlugin (v.3.1), and JsonRPCContrib are all installed and enabled, I'm checking on the skin [21:19]
gac410Hm. SvenDowideit I've enabled i18n on trunk. Quick test of bin/compare, and the inline diff doesn't show. Side by side works fine.
No idea if that's a new issue or not. Damn.
[21:22]
SvenDowideit:)
actually, i think its not
doing diffs on trunk was busted when i was testing last time
[21:23]
gac410wunerful.
I guess I'll cherry pick the fix into release11, build another RC, and then beg / plead for feedback. I've not seen one iota of comments on the last two RC's
[21:24]
Simon_DHPatternSkin [21:25]
gac410Which means... is it just not tested, or is it working fine. I have no idea. [21:25]
SvenDowideitSimon_DH only one other possibility - are you using fcgi / mod_perl and need to bounce the process?
otherwise - good chance its a bug
[21:27]
gac410SvenDowideit: I think the trunk diff issue is CSS. Looking at the html source, I'm seeing css classes indicating added blocks, but nothing visual. [21:27]
SvenDowideitexcellent :) [21:28]
Simon_DHSvenDowideit, mod_perl is not loaded into Apache [21:33]
gac410Looks like the css in the template never makes it into the rendered output. So lots of div's with classes, no classes defined.
gac410 is clueless.
[21:33]
***donbarry has quit IRC (Remote host closed the connection) [21:48]
SvenDowideityes.
8am, 30 degrees c
somehow i doubt the 32 max forecast today
[21:59]
.... (idle for 15mn)
Simon_DHThanks for the suggestion as to where to look for the trouble! [22:15]

↑back Search ←Prev date Next date→ Show only urls(Click on time to select a line by its url)