#foswiki 2012-04-18,Wed

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

WhoWhatWhen
GithubBot[foswiki] foswiki pushed 5 new commits to master: http://git.io/40mFSg
[foswiki/master] Item11746: small spelling fix - SvenDowideit
[foswiki/master] Item11772: replace edit table plugin with edit row plugin - SvenDowideit
[foswiki/master] Item11733: add favicon - SvenDowideit
[01:16]
***GithubBot has left [01:16]
FoswikiBothttp://foswiki.org/Tasks/Item11746 [ Item11746: General documentation task for 1.1.6 release ]
http://foswiki.org/Tasks/Item11772 [ Item11772: replace edit table plugin with the more modern edit row plugin ]
http://foswiki.org/Tasks/Item11733 [ Item11733: create a simple pattern skin theme using the new logo ]
[01:16]
gac410SvenDowideit: I opened a task on EditRowPlugin ... it's not really working correctly for me on 1.1.5. Item11762 Have you seen issues like that? [01:19]
FoswikiBothttp://foswiki.org/Tasks/Item11762 [ Item11762: Issues with pencils and drag operations with EditRowPlugin on 1.1.5 ] [01:19]
SvenDowideitis that a dupe of http://foswiki.org/Tasks/Item11773 [01:19]
gac410:P well 11773 is a dup of 11762
Though it doesn't note the drag issue with is non-functional for me.
And corrupts the display
[01:20]
SvenDowideitactually
they are different issues
mmm, this is frustrating - maybe its not ready
but then, better to report the problemsand get them fixed :)
[01:21]
gac410yes. Much better to fix. If people start using it more mainstream it will improve. [01:22]
SvenDowideitah, but its nice that it doesn't show edit on a topic you don't have perm to edit :) thats new [01:27]
gac410Need to make sure it turns itself off in static render modes (pdf, etc) and diff, [01:27]
SvenDowideitmakeatask so we don't forget / figure out how to automate that [01:28]
gac410we have a context already defined for non-"live" views [01:29]
SvenDowideityup
mmm, i thought that micha made strikeone get loaded forall views
(which i think is the right thing to do)
[01:29]
GithubBot[foswiki] foswiki pushed 1 new commit to master: http://git.io/EPqKKg
[foswiki/master] Item11772: replace edit table plugin with the more modern edit row plugin - SvenDowideit
[01:32]
***GithubBot has left [01:32]
gac410No, I don't think he did it for all views, just made the EditChapterPlugin load it too. [01:32]
SvenDowideitgrrr :)
yetanother feat req
actually, no, i he a better idea ;(
[01:32]
............. (idle for 1h1mn)
kip3fI'm trying to create a bulk dataform creation topic (similar to bulk user registration). But I want the user to be able to put in an XML doc for the bulk upload. Are there any XML parsing plugins? [02:34]
SvenDowideiteeek
look at the import from excel plugin
though micha might have an xml one too
[02:36]
kip3fok thanks [02:37]
.................... (idle for 1h38mn)
***gac410 has left [04:15]
SvenDowideitmmm
isn't there a proposal to allow SECTIONs to set default values for INCLUDE params?
[04:28]
AlexisHazellSvenDowideit: Got saving working .... when i replaced the BUTTON macro with an <a> element of class 'foswikiButton', things suddenly started working as expected. [04:40]
SvenDowideitah :(
i just started to use BUTTON for my thing
[04:40]
AlexisHazellWhich i can live with. :-) [04:41]
SvenDowideitbut i'm contemplating switching back, as its broken for list elements [04:41]
AlexisHazellAh, okay. [04:41]
SvenDowideitusing floats makes a mess [04:41]
AlexisHazelli'd lodge a bug report, but i'm not sure i even understand how it's broken. [04:42]
SvenDowideitit seems to work for my usage - so ergle :)
meanwhile, I'm finding istopic isnt' working in my IF :(
SvenDowideit recons it must be time for lunch :)
[04:42]
AlexisHazellQuoting issues? [04:43]
SvenDowideit%IF{"istopic '%S_ITEMWEB%.%S_ITEMTOPIC%'" then="edit %S_ITEMTOPIC%" else="add New %PREFIX%"}%
somthing :/
i'm thinking something to do with how IF and sectional includes with parameters come together
and its related to a mess micha found
darn
another bug report coming up
food first i recon
[04:43]
AlexisHazellAck.
Good idea!
[04:46]
................. (idle for 1h20mn)
Should an 'action_save' within a <form> element save both form data and topic text? [06:06]
***chriswerry has left [06:07]
SvenDowideitAlexisHazell, yes
mmmm, pharvey / CDot
what is the purpose of the class dropdown in the Insert/Edit tinymce dialog?
[06:16]
AlexisHazellSvenDowideit: Hm, okay, thanks. [06:16]
CDotSvenDowideit: you voke ching? [06:16]
SvenDowideitshould we not hide the options in there that de-tml tables? [06:17]
CDotoh; tinymce dialogs. Probably a hangover from default TMCE behaviour. pharvey has done a lot more work on the dialogs than me..... [06:17]
SvenDowideitgrumple.
course, the q i was asked was: why doesn't the wysiwyg editor make wiki code tables
and the only answer i have, is 'because you used the dialog we provided'
[06:21]
GithubBot[foswiki] foswiki pushed 1 new commit to master: http://git.io/cz4Aiw
[foswiki/master] Item11764: convert to BUTTON, add more tabs to the viewtempalte - SvenDowideit
[06:33]
***GithubBot has left [06:33]
FoswikiBothttp://foswiki.org/Tasks/Item11764 [ Item11764: AddDataFormCreatorWizard feature ] [06:33]
SvenDowideitmmmm
CDot, do you recal if we had a proposal to add defaults to SECTIONs?
and how to do evaluations local to the section when included?
[06:34]
CDotwhat would such a default do? [06:35]
SvenDowideitsame kind of thing as the parametarised vars [06:35]
CDotbut.... what would it provide a default *for*? [06:35]
SvenDowideitthe parameters used in the section
you know INCLUDE{"" PARAM2="as"}
if PARAM2 isn't defined
normally i use a Set inside the section
[06:36]
CDotok, I think I get your drift. No, I've never seen such a proposal. [06:36]
SvenDowideitbut that breaks when the section is used in a linefeed critical place
like tables / bullets
k - maybe i miss read parametarised vars
tho - do those do defaults?
[06:36]
CDotCDot doesn't understand the use case. A Set inside a SECTION? A SECTION that is INCLUDEd? [06:38]
SvenDowideityup
see - one mo
oh crud, trunk's not updated
or i can't spell
http://trunk.foswiki.org/System/WikiApplications?raw=on
the createpopup section
and it'll get much more complex eventually
[06:39]
CDotbut.... a Set in an INCLUDE is a nop for the including topic (I hope)
so why not just move the Set outside the SECTION?
[06:41]
SvenDowideitno, its never been a nop
because the section is used multiple times
and each time, the S_ITEMTOPIC is set for the INCLUDE fragment
ok, yes, its a nop for the topic that its included into
but not a nop for the included fragment
[06:41]
CDotCDot doesn't see how that works, because Set is evaluated outside the macro expansion/rendering cycle [06:42]
SvenDowideitits not [06:42]
CDotok, capiche [06:42]
SvenDowideitSvenDowideit wonders how few people grok the INCLUDE function prefs stack :( [06:43]
CDotCDot still doesn't see what the value of a Set inside a SECTION is (sorry for being stupid here, but....) [06:43]
SvenDowideitok, you need to
these sections are basically function calls
the Set inside is a local variable to the INCLUDED fragment when instanciated
[06:43]
CDot^these^the sections SD is writing [06:44]
SvenDowideity
most sections that micha does are this too
[06:44]
CDotbut what evaluates these Set's? [06:44]
SvenDowideitfoswiki [06:45]
CDotwhen? [06:45]
SvenDowideitwhen the section is included into something
INCLUDE has a prefs stack
any use of INCLUDE puts all its params onto the stack, and then evaluates the section
then it pops that stack
its been this way for a very very very long time
[06:45]
CDotCDot is reading code. He never uses this meme, so hasn't encountered it before. [06:47]
SvenDowideitI could write this without the Set's but I'd be duplicating the IF statements inline
which is woeful when things get even more interesting
[06:47]
CDotCDot can't find anuthing in the code that would expand a Set in an INCLUDEd SECTION. Not disputing what you say, I just can't find it :-( [06:48]
SvenDowideithttp://foswiki.org/Development/ParameterizedVariables <-- i can't find the docco for this either :(
so i have no idea what was implemented
[06:49]
CDotSystem.PreferenceSettings [06:49]
SvenDowideitnot mentioned in that topic at all - and its hard to work out which of the things were coded [06:50]
CDothttp://trunk.foswiki.org/System/PreferenceSettings section "Parameters" [06:51]
SvenDowideiter
that mentions DEFAULT - is it also true for other param names
meh - I'll kick it and see if it gets me close to where i need to go
won't help me for the INCLUDE case tho
darn, gotta go pick up kids
I'm presuming you do use INCLUDE params with SECTIONS
[06:52]
CDotI do, yes [06:54]
SvenDowideitjust that you c&p the same IF or whatever when needed? [06:54]
CDotand yes, it's true for other param names; I just used DEFAULT in the example
note that each usage of a param can have a different default
y, I c&p IF quite a lot
[06:54]
SvenDowideitok, whereas i don't
or when i do, i squirm
[06:55]
CDotcan you show me an example of what you are doing? I want to explore this further, to understand where this expansion is done [06:56]
SvenDowideitthis _is_ an eg
git svn rebase and goto the System.WikiApp's topic
create a new app - that then uses those sections
(or better said, a copy of those sections, as i have not yet implemented the INCLUDE a list of topics thing
[06:56]
CDotCDot doesn't use git with trunk; is it in svn yet? [06:57]
SvenDowideity
i thought i used this meme in the original bugs contrib too
but meh - we weren't obvious about it
its what imo makes SECTIONs and INCLUDE useful as functions
be back in a few hours :/
[06:57]
CDotCDot doesn't see any SECTION in either WikiApplication* topic [06:59]
SvenDowideithttp://trunk.foswiki.org/System/WikiApplications?raw=on
there are several named SECTIONs
[06:59]
CDotok, found it, thanks
CDot will investigate
[07:00]
SvenDowideitthat topic is used as a templatetopic to put those sections into the newly created topic (that goes into Sandbox web
lates
[07:00]
AlexisHazellWhen a topic gets saved, does FW mess around with the order and layout of Set statements?
Because if not, it must be SetVariablePlugin.
[07:06]
.... (idle for 16mn)
CDotAlexisHazell: no; by default a save does not touch the topic content. [07:25]
..... (idle for 20mn)
GithubBot[foswiki] foswiki pushed 1 new commit to master: http://git.io/iGnMLQ
[foswiki/master] Item11690: forgot to close the table cell :-( - FlorianSchlichting
[07:45]
***GithubBot has left [07:45]
FoswikiBothttp://foswiki.org/Tasks/Item11690 [ Item11690: FormPlugin: global variables, double Location HTTP headers a problem under mod_perl ] [07:45]
AlexisHazellCDot: Hm, SvenDowideit said otherwise .... [07:46]
CDotI doubt he was answering the question you just asked, then. [07:50]
SvenDowideitAlexisHazell, i did?
I didn't see _that_ question
just '<AlexisHazell> Should an 'action_save' within a <form> element save both form data and topic text?'
[07:50]
AlexisHazell(16:06:41) AlexisHazell: Should an 'action_save' within a <form> element save both form data and topic text?
(16:07:19) chriswerry left the room.
(16:16:20) SvenDowideit: AlexisHazell, yes
[07:51]
SvenDowideitexactly
nothing to do with Set's
[07:51]
CDotSvenDowideit:
%STARTSECTION{"memepickme"}%
Passed %PARONE%
* Set LOCAL = %IF{"defined PARONE" then="%PARONE%" else="NOT DEFINED"}%
LOCAL is %LOCAL%
%ENDSECTION{"memepickme"}%
---+ Include
%INCLUDE{"%TOPIC%" section="memepickme" PARONE="value"}%
That works as *I* expect. So what am I missing?
[07:52]
AlexisHazellCDot, SvenDowideit: Ah i see now, my apologies for the confusion. [07:52]
SvenDowideiti don't know what you expect [07:52]
CDoti.e. the Set LOCAL is a nop [07:52]
SvenDowideitthen why does the stuff i do work? [07:52]
CDotI don't know! That's what I'm trying to understand [07:53]
SvenDowideitas the eg's i showed you do exactly that same thing [07:53]
CDotCDot is trying to find a simple example
CDot thinks this should be the subject of a FAQ - perhaps even a blog post - as it may be an important meme
[07:53]
SvenDowideiti spent all day crafting simpler eg's that work
it is
but it also breaks when you put the INCLDUE in a table / bullet list etc
[07:54]
CDotyeah, I hear that. But I'm trying to get to first base, here [07:54]
SvenDowideitfoswiki.org/Tasks/Item11776# [07:54]
FoswikiBothttp://foswiki.org/Tasks/Item11776 [ Item11776 ] [07:54]
SvenDowideitoh, the set's were removed over teim :() [07:55]
CDotteim? [07:55]
SvenDowideitoh no, there they are
time
i'm tired and some little shit abused me for wanting to ride on the footpath they parked on
but yes, that topic does the same thing your eg does
maybe the problem is that my section is in a verbatim?
[07:55]
CDotmaybe - trying that [07:56]
SvenDowideitjust look at the link wouldya?
maybe you can simplify that down and get something that works still
or tell me what i did that is so wrong that it works
cos that's a good possibilty
[07:56]
CDotyes, the <verbatim> did the trick. Now, I have to work out why that works [07:57]
SvenDowideitits shite that it needs it
mind you - it thus probly also works without - if you don't INCLUDE it into the topic the section is defined in
i hope micha stops by
[07:57]
CDotdo you know where the FAQ is? It might explain [07:58]
SvenDowideiti'd like to confirm that this is the same has he uses
what FAQ??
i'm doing this wikiapps thing to fix the docco
cos 99% of what i do daily is unknown
and i didn't realise
[07:58]
CDotok, I understand why it works with a verbatim and not without [07:58]
SvenDowideity? [07:59]
CDot(08:54:00) ***CDot thinks this should be the subject of a FAQ - perhaps even a blog post - as it may be an important meme
....
(08:54:04) SvenDowideit: it is
I thought you were saying there is a FAQ
[07:59]
SvenDowideiti messed that ;) i agreed it it imprtant enough to need doc [07:59]
CDotCDot admits 4s is a pretty tight window :-) [07:59]
SvenDowideitas its an important meme :) [07:59]
CDotthe reason the verbatim is needed is to suppress expansion of the %IF [08:00]
SvenDowideiter that makes no sense
as INCLUDing the section should beget a fresh IF
[08:00]
***ChanServ sets mode: +o MichaelDaum [08:00]
SvenDowideitas INCLUDES are newly evaluated
yay MichaelDaum
http://foswiki.org/Tasks/Item11776
[08:00]
MichaelDaumsvn: OPTIONS of 'http://svn.nextwiki.org/trunk': Could not resolve hostname `svn.nextwiki.org': Host not found (http://svn.nextwiki.org) [08:01]
CDotyeah, but the Set LOCAL sets LOCAL to the %IF, and it's *that* pref that is expanded when the section is instantiated [08:01]
FoswikiBot[ / - Revision 14634: / ] [08:01]
MichaelDaumyay SvenDowideit
how do I fix my svn?
[08:01]
SvenDowideitCDot was surprised that you could use Set's in a function SECTION
fix the domain to one that works
[08:01]
MichaelDaumhow do I do that? [08:02]
SvenDowideiti finally allowed those to lapse after 3 years of warning
git svn clone (is the facetious answer)
the svn answer is doccoed somewhere
but i can't recal where
[08:02]
MichaelDaumokay will find out [08:02]
SvenDowideit(sorry, i need to go look at girls) [08:02]
CDotCDot still doesn't understand how it works, but is getting there [08:02]
SvenDowideitCDot, i think that the non-verbatim version is a bug
can you talk to MichaelDaum about it a little tho, as i have to go
[08:03]
CDotnot sure. I'm working through the expansion steps.
CDot suspects the * Set is a red herring, and the mechanism actually being used is different. Experimenting.
[08:03]
MichaelDaumMichaelDaum doesn't use * Set inside sections.
not sure why anybody should
[08:10]
SvenDowideitreuse [08:11]
CDotMichaelDaum: it makes a certain amount of sense
and somewhat to my surprise, it actually seems to work
[08:11]
SvenDowideiti'd rather have it in the SECTION{declaration [08:11]
CDotbut I can't work out why! [08:11]
SvenDowideitand will need it there in 1.2 [08:11]
MichaelDaumthis should be done using a %SET macro [08:11]
CDotgood grief no! [08:12]
SvenDowideitwhy, don't need it
the core has doen this since cairo or so (i think)
[08:12]
MichaelDaum%SET should push vars onto the prefs stack following the normal parse order [08:12]
SvenDowideitso does this [08:12]
MichaelDaum* Set simply is a stranger from outer space for the parser. [08:12]
SvenDowideitand its been in core for.... [08:12]
MichaelDaumbesides values not executed/cached unless patched in
there are lots of cases where I have to fall back to %CALCs $SET() and $GET() ... which I hate to do
still their use case exists
[08:13]
CDotCDot loves the fact that we have independently manage to find solutions to the same problem, while failing miserably to communicate that fact to eachother! [08:14]
SvenDowideitdont' need them much, as the core has Set
y :(
thats what 1.2 is for (for me)
[08:14]
MichaelDaumCDot, it was you who rejected an already accepted %SET proposal [08:15]
SvenDowideiti intend to make STARTSECTION allow defaulting like Parametarised Vars
because %SET is un-necessary
[08:15]
MichaelDaumSvenDowideit, why is it unnecessary? [08:16]
SvenDowideitbecause the core already has that functionality
you just insist that you need to ignore it
[08:16]
CDotMichaelDaum: I have not yet seen a convincing use-case for %SET [08:16]
MichaelDaumhere's one [08:17]
SvenDowideit(i actually have a use case for it, but i want to do it differently)
as STARTSECTION{value="%IF{....}%"} to default 'value' unless its set - same as Set Params
thus reducing the use of inline Set's as i've been doing
[08:17]
MichaelDaum%SET{"found" value="false"}% %SEARCH{"query" format="$percntSET{\"found\" value=\"true\"}$percnt"}% .... %IF{"$found" then="found results" else="nothing there"}% [08:18]
SvenDowideitthats wher i use * Set in a SECTION
function style - with more SECTIONs
[08:18]
MichaelDaumthis is the most general pattern. it applies in various variations in wiki apps. [08:19]
SvenDowideit%SEARCH{"query" format="$percntINCLUDE{section="somethning$count"...
y
and as CDot pointed out
we have all found totoally different ways to do the same thing
i retentivly use core only, so that core is as good as i can help make it
you use whatever you have at hand
both are right
[08:19]
MichaelDaum%CALC is what I use in these cases [08:20]
SvenDowideitbut we need to start doccoing
gotta go
laters
[08:20]
***ChanServ sets mode: +o OliverKrueger [08:21]
MichaelDaumthe pattern is all about creating sideeffect. that is affect other code in multiple places. information is transported via global variables.
while there is a variable namespace in foswiki, you can't set it actively as part of the parsing process.
only static
[08:21]
***OliverKrueger sets mode: +oooo CDot Colas ColasHome AlexisHazell
OliverKrueger sets mode: +oo padraig_lennon uebera||
[08:22]
MichaelDaumor as part of an INCLUDE [08:22]
SvenDowideiti use INCLUDE and Set's in Sections as local vars [08:23]
MichaelDaumthe main difference between variables pushed on top of the stack as part of a parametrized include and using some way to %SET variables is scope obviously [08:23]
SvenDowideitas they are basically a stack
ah, ok, i avoid the global scope
as i'm trained (like a friggen monkey) to avoid globals
[08:23]
MichaelDauma global scope has its value if used with taste [08:24]
SvenDowideiti agree - but my conditioning makes me not think of it much [08:24]
MichaelDaumbest example: sempaphors [08:24]
SvenDowideitactually - i think of using goto's more than globals, which is scary
semaphores are not global
that will kill you
they are _resource_ bound
if you code them globally, and then someone gives you a thousand of those resources, you have to re-write
(yes, been there, slaved at it)
not important tho -_that_ is an interesting thing to use %$SET for
but means its not what i need
heya OliverKrueger :)
[08:25]
MichaelDaumsemaphores are _globally_ visible flags indicating whether that resource is available. if it wasnt visible globally, the resource isnt proteced. [08:26]
SvenDowideitbut they are not global (unless oyu code simplistically)
in the simple case, one uses a singleton to abstract it
so that later you can make n resources without having to change everything
[08:27]
MichaelDaumanyway [08:28]
SvenDowideit(yes, splitting hairs a bit, but real global values are not the same thing
and our prefs are more like real globals - in the bad way
so you use CALC to do local to a Function vars?
[08:28]
MichaelDaumMichaelDaum can finally svn up after svn switch --relocate http://svn.nextwiki.org http://svn.foswiki.org finished [08:28]
SvenDowideitand I use Set?
excellent - that was it
i use CALC about once every 4 years :/
[08:29]
MichaelDaumI use it for my timetracking and billing [08:30]
SvenDowideiti have an app that does that too
it only uses CALC for summing the cols
not for local vars
but i wrote that in... er 2004
too many things to ignore :)
i'm seriously surprised that you use CALC for local vars in Function INCLUDEs - i thought we'd discussed the other ways
[08:30]
MichaelDaumthere is so much cruft silo'ed into CALC
$SET is one of it
[08:31]
SvenDowideitthat makes local SECTION * Set's the biggest secret thats been around forever [08:32]
OliverKruegerIts from the creator of "nonoise=on"... ;) [08:32]
SvenDowideiti know about $SET, even used it once or twice
but i seriously use it only to store things i need in another CALC
this will make the ParametraisedINCLUDEDefaultValues proposal even more interesting!
[08:32]
MichaelDaumanother use case for %SET is list values [08:33]
SvenDowideiti thought you said %SET was global
i'm talking entirely about local vars
[08:33]
MichaelDaumbasically %SET pushes a Foswiki::Attr object onto the prefs stack [08:34]
FoswikiBothttp://trunk.foswiki.org/System/PerlDoc?module=Foswiki::Attr [08:34]
SvenDowideitso does * Set in a SECTION [08:34]
MichaelDaumand %GET would have formated that back in the open [08:34]
SvenDowideitbut its automatically poped off when you leave the function [08:34]
MichaelDaum... which is not what I am talking about [08:34]
SvenDowideitgood
i'm trying to find out wha tyou think about what I _am_ talking about
[08:35]
MichaelDaumwho was that guy that wanted to work on Foswiki::Attr being able to store list values? forgotten. [08:35]
SvenDowideitbut i guess that can wait until i write the featreq for 1.2
kip?
(he's here when you're asleep :)
SvenDowideit has to go - laters again
[08:35]
MichaelDaumCDot, why don't I have JSEditableContrib in my svn?
ah it is called JEditableContrib
[08:41]
OliverKruegerMichaelDaum: shall I checkin? [08:46]
MichaelDaumy please [08:46]
OliverKruegerk [08:46]
CDotMichaelDaum: because it's called JEditableContrib? [08:51]
MichaelDaum^^ [08:51]
CDotCDot is catching up :-) [08:51]
MichaelDaumthere's something wrong with the ui buttons of ERP [08:53]
CDot%SET *is* global. It's just global from the point of definition onwards, where "onwards" means inside-out-left-right order [08:53]
MichaelDaumthats the point [08:54]
OliverKruegerHi CDot [08:54]
MichaelDaumgives you much better control than having to crawft a parametrized include [08:54]
CDotOliverKrueger: good morning Oliver [08:55]
SvenDowideitdifferent control - not better
parametrized include being a _function_
so local vars are useful
i'll write up that feat req later this week
i need to explore parametarised variables first
[09:00]
GithubBot[foswiki] foswiki pushed 1 new commit to master: http://git.io/IZTb9Q
[foswiki/master] Item11777: Make FlexFormPlugin aware of the urlparam formtemplate - OliverKrueger
[09:01]
***GithubBot has left [09:01]
FoswikiBothttp://foswiki.org/Tasks/Item11777 [ Item11777: Make FlexFormPlugin aware of the urlparam formtemplate ] [09:01]
SvenDowideitMichaelDaum, shame about LabJS etc
i had the same impression
i really want to get away from loading js globally when a user might only need it for one dynamically loaded popup
[09:01]
MichaelDaumSvenDowideit, yep. cool you filled that proposal. gets us started our brainz
what about LabJS? seems to fit for the purpose...
[09:09]
CDotCDot has explored use of * Set inside SECTION, and it is fraught with problems [09:22]
SvenDowideitCDot, ... will you dooco them, or will you just leave this 'fraught' functionality that I've been using in the core as is?
MichaelDaum, i don't know enough about the options yet :) i dont really think i know enough about JQREQIRE yet either, but clearly that'll have to change
[09:34]
CDotI don't see anything wrong with the core functionality; it's the use case that's a bit suspect. There are some very delicate constraints on when it works and when it doesn't, and how it works. [09:34]
SvenDowideitand they are? [09:35]
CDotCDot is writing it up [09:35]
SvenDowideitsweet - that was the q :)
i hope we can make it redundant by adding parameter defaults to STARTSECTION to be honest
but that presumes that those params can be delayed until INCLUDE time, and that is a tad risky too
SvenDowideit goes again :)
[09:35]
MichaelDaumusing LabJS theres not that much that'd have to change
RequireJS enforces lots of redesign
[09:36]
AlexisHazellArgh. Why isn't my topic template text getting copied into new topics? i'm calling the edit script with templatetopic being passed the correct template topic, and the action on the <form> element of the new topic is a save. Am i doing the right thing? i've been trying various incantations, nothing seems to work. [09:37]
SvenDowideiti was wondering if LabJS also gives less benefits, but i just don't know yet - it feels like the easy choice [09:37]
MichaelDaumHeadJS does lots of Modernizr but still not on par with it [09:37]
SvenDowideity, i don't want to import huge other things too [09:37]
MichaelDaumlets experiment a bit with LabJS to get to know the material bit better [09:38]
SvenDowideit(90% of the reason i'm peppering feat req's atm is to get a chance of getting some into 1.2 [09:38]
MichaelDaumthen we might end up having more requirements and look at options [09:38]
SvenDowideiti thought we might read a little more - but then extract the code to tmpl's
so we can switch easily :)
[09:38]
MichaelDaumthe hard work is what you called manifest [09:39]
SvenDowideityup [09:39]
MichaelDaumcasted on "modules" in RequireJS, a main.js for HeadJS or a $LAB script for LabJS [09:39]
SvenDowideitCDot, I most often use the Set from SECTIONS that are called from SEARCH format= for every item
but mostly its not done on the topic that the section is defined on
y - more places to generate the js from spec file based hashes
its a blooming build system again
SvenDowideit goes to try to get the girls into the shower >:}
[09:39]
MichaelDaumit kind of feels like a special %RENDERZONE{"script" format="<script> $LAB... whatever thingy
MichaelDaum wished it was that easy ... maybe it is
[09:41]
CDotSvenDowideit: findings so far: http://foswiki.org/Development/UseOfSetInsideIncludedSections [09:45]
SvenDowideityou need a verbatim even when the INCLUDE is in a different topic from the section ? [09:45]
CDoty [09:46]
SvenDowideitmmmm, ok, i don't think i've used the same var in 2 sections that are defined or used in the same topic
i think your caveats are probly too wide for the test you have?
but interesting find
[09:46]
CDotno, so long as the vars are unique it seems OK. Just exploring if the Set needs to be in the SECTION at all [09:46]
SvenDowideit_most_ times, i have a topic that defines lots of SECTIONs
and then use some nested in those sections
and then they get actually used on a totally different topic
so i would expect that they do need to be in the section to make sense
for eg SiteChanges would have a delayed INCLUDE in the format for a web's SECTION definition topic that then defines how different topic types get rendered
[09:46]
CDoty, the * Set doesn't have to be inside the SECTION
.... which kinda solves your prob with newlines, dunnit?
[09:48]
CDot has added an example of the * Set outside the SECTION [09:53]
SvenDowideitits not actually useful tho is it
as that means you can only use the section on the topic its defined on
hey, you're not actually using the sections are you?
[09:57]
CDotexperimenting further [09:58]
SvenDowideitdepressing tho this experiment is
as it means i need/want actual prefs stack locals not some *** that kinda works
[09:58]
CDoty. So %SET is looking more constructive. [10:00]
SvenDowideitits lookin g just as destructive really [10:00]
CDotthe * Set outside the SECTION doesn't work in another topic (of course) [10:00]
SvenDowideitas MichaelDaum keeps saying %SET is global
y, and the 99% use of these function sections _is_ call from other topic
[10:01]
CDotI don;t know what it is; I didn't look at the code. It depends on where it's pushed on the stack [10:01]
MichaelDaumthere _is_ no %SET [10:01]
SvenDowideitthere is a %SET - its whatever you say it is [10:02]
CDotI *suspect* it's pushed on the session stack, in which case it's global *and* local [10:02]
SvenDowideituntil its impled [10:02]
MichaelDaumthere isnt even an impl [10:02]
SvenDowideitMichaelDaum, exactly - and thus its whatever you say it would be [10:02]
MichaelDaumah hum [10:02]
CDotCDot added a Func to set a variable, which pushes it on the session stack [10:02]
SvenDowideitjust as any competing proposal would be whatever the proposer says [10:02]
MichaelDaumCDot, right. [10:03]
CDotso the trivialest impl would call that [10:03]
MichaelDaumI'd love to see that happen [10:03]
SvenDowideitthe trivialest timpl of local vars would push to the session stack? [10:03]
CDoty [10:03]
SvenDowideitor does no-one else want true local vars [10:03]
MichaelDaumthats the thing most in line with what we have atm [10:04]
CDotthe session stack is the topmost stack during macro expansion [10:04]
SvenDowideitwhat i have works essentially like a local var - though it might be due to how i use it [10:04]
CDotto have "local" vars, you have to have the concept of "scope"
what is "scope" during an INCLUDE?
[10:04]
SvenDowideitlocal - as in in the SECTION's scope
remember, the INCLUDEed text has a different value for %TOPIC%
[10:04]
CDotthere is no such think. A SECTION is a textual substitution, not a function call [10:05]
MichaelDaumINCLUDEs arent processed until after inserted in the caller's scope [10:05]
SvenDowideitwhich is why %BASETOPIC% exists
there )_____is__ this a local scope for sections
no
they are not
if they where %TOPIC% would always == what was requested in the url
[10:05]
CDotwell, there *is* the topic context, I guess [10:05]
SvenDowideitand it doesn't [10:05]
CDotwhich is where the BASETOPIC etc is stored [10:06]
SvenDowideitSvenDowideit goes again - this is difficult [10:06]
MichaelDaumthats the only difference TOPIC and BASETOPIC and INCLUDINGTOPIC: these are maintained during insertion [10:06]
CDotyou *could* process the section to add local *Sets to that
but.... when do those sets go out of scope? When the topic scope closes, I guess
actually, he's right, that *could* work
[10:06]
SvenDowideitits how %TOPIC% works
when the INCLUDE is processed we change that and INCLUDINGTOPIC
which is why i always presumed the SECTION's sets worked that way too
[10:09]
CDotnah; the included text is only macro-xpanded - there is no line-by-line extraction of * Set [10:10]
SvenDowideiti do not want these var's to leak [10:10]
CDotno, i can see that [10:11]
SvenDowideitbut i'm very happy to change it to be [10:11]
CDotCDot has a good idea now what SvenDowideit is trying to do, and is thinking about it [10:11]
SvenDowideitSTARTSECTION{LOCAL="STUFF..."
as that gives us real local params
and the stuff in the section becomes the 'result' only
[10:11]
CDotCDot wonders is * Local might be more consistent [10:11]
SvenDowideit_that's the proposal i want to write after i poke your stuff
as using the bullet syntax has rendering issues
SvenDowideit runs
er CDot
[10:12]
CDotaye [10:16]
SvenDowideitso your findings are essentially that Set in INCLUDED SECTIONs implement MichaelDaum's %SET [10:16]
CDotno [10:16]
SvenDowideitah, so it only works that way through leakage if the section is in that topic?
SvenDowideit goes
[10:17]
CDota %SET is evaluated at macro expansion time; so two %SETs in the same topic have two different values
a *Set has only one value
(the last one defined)
so e.g. "My favourite colour is %SET("fave", "blue"}% %fave% no I meant %SET{"fave", "green"}% %fave%" gives "My favourite colour is blue no I meant green"
whereas using * Set it would always be green.
[10:17]
MichaelDaumimho, relying on the tml parser is more safe than parsing * Sets out while processing a section. [10:19]
CDotof course quite what %SET("fave", "$percntSET(\"fave\", \"blue\"}$percnt"}% does is too much for my small brain [10:20]
MichaelDaumthese two stages - first get all * Set, then expand macros - is bound to fail in strange cases [10:20]
CDotthe advantage of parsing * Set is the handling of newlines, I guess.
and also avoiding $percnt :-)
[10:21]
MichaelDaum%SET{"foo" value="some \n lines \n of \n code"}% is fine as well [10:21]
CDoty, but escapes get people tangled [10:21]
MichaelDaumalso: %SET would effectively allow to _cache_ values [10:21]
CDotCDot doesn't want to think about caching until the user-facing bit of it is understood [10:22]
MichaelDaumTML in * Set variables are recomputed on every insertion ... also quite unexpected for users
caching is just a fact of inside-out-left-right
evaluation
%SET{"results" value="%SEARCH...."}% does all of the trick
compare this to * Set results = %SEARCH
typing %result% %result% %result%
^result^results
using %SET, there's only one computation of %SEARCH
using * Set, the same %SEARCH is performed as many times as there are %results% invocations
[10:22]
SvenDowideiti suspect thats actually an argument against %SET
as mixing vars with cache is conflagrating things a bit
not a strong arg perhaps, but still
[10:41]
CDotCDot is trying to understand whether %SET should be a macro, or some other construct
I would like to see some explanatory documentation on it
%SETVAR in the SetVariablePlugin is quite....complex
it conflates %SET with some %IF functionality
[10:46]
MichaelDaum%SETVAR is not what %SET is supposed to be
%SETVAR is an actor to change meta preferences conditionally
[10:51]
CDotCDot thinks qwe need a featrue rqeuets for %SET
CDot thinks he should not try to type with his left hand without looking at the screen
CDot increasingly likes the idea, but really wants to see what it would look like to an end-user
[10:54]
...... (idle for 26mn)
SvenDowideiti have the feeling that we can implement what MichaelDaum wants SET to be using what you found about INCLUDE
well, partially
[11:21]
MichaelDaumit seels trivial to impl %SET
seems
[11:23]
SvenDowideity, without code even [11:23]
MichaelDaum:) [11:23]
SvenDowideitthe sad thing for me is that i can't think of very many times where i want to use a global SET
as i want to _increase_ modularisation and decrease coupling
CDot, are you still working with the local vars on sections thing?
or have you dropped what i care about to think about global var stuff MichaelDaum wants
CDot> so e.g. "My favourite colour is %SET("fave", "blue"}% %fave% no I meant %SET{"fave", "green"}% %fave%" gives "My favourite colour is blue no I meant green"
[11:25]
***ChanServ sets mode: +o OliverKrueger [11:26]
MichaelDaumSvenDowideit, your %STARTSECTION{fave="blue" ...}% feels good [11:27]
SvenDowideitmmm, _or_ have you worked a way to make global and local scope the same thing
its consistent with what we have on parametarised Set's
and (so i just realised trying to get the girls to bed) could separate the function from the result
as our sectional include functions currently mash the 'code' into the result
_but_
i want %STARTSECTION{fave="%IF{}%" ...}% to be late evaled
whereas your %SET would be early eval
maybe escaping crap could save us there :(
and as CDot noted
maybe we can have * Set and * Local do the 2 scopes
[11:27]
MichaelDaumproblem with * Set remains: setting and using the vars are done in two isolated stages [11:39]
SvenDowideitthats the advantage for my use
and not entirely
as a Set that was inside a section would happen at the begining of the INCLUDE of that section
which is why i wonder if i could implement your %SET using that (except for the stupid linefeeds mess that i'll ignore as an implementation detail for now)
[11:44]
GithubBot[foswiki] foswiki pushed 1 new commit to master: http://git.io/N_Ey_g
[foswiki/master] Item11777: Fix the formtemplate patch - OliverKrueger
[11:46]
***GithubBot has left [11:46]
FoswikiBothttp://foswiki.org/Tasks/Item11777 [ Item11777: Make FlexFormPlugin aware of the urlparam formtemplate ] [11:46]
pharveygood evening
what-the-set-bomb?
[11:48]
SvenDowideitpharvey, grin
lotsa bad ideas coming together today :)
[11:49]
pharveyoh wow. http://foswiki.org/Development/UseOfSetInsideIncludedSections is fascinating [11:50]
SvenDowideitand doesn't cover the fascinating way i was (mis) using it as locals
WOW
i just added to that topic
[11:51]
pharveyAnd Paul.A fixed my CSV exporter code at http://foswiki.org/Support/BestPracticeTip27 so it now properly escapes formfield content's " as "" [11:51]
SvenDowideitand er - seems that Set in a section is like a __member__ var [11:52]
pharveyI'm learning a lot about Foswiki today :} [11:52]
SvenDowideitish - as we will have to think about chaning this stuff
now, lets make a copy of the topic.. and use Local instead of Set
gosh i hope the results are the same
[11:53]
pharveydo we have a html=off url param? [11:59]
SvenDowideitno idea what that is [12:00]
pharveypharvey thinks Paul.A is writing more advanced wiki-apps than himself :P [12:00]
SvenDowideitCDot, i just changed your horrible confusion
and i think it now changes your finding
rather than using PASS and FAIL in both INCLUDEs
i use obviously different values
and the values look different to me
not that i can see why your first include thingks LOCAL has 2 values at the same time
but at least the test no longer obscures the result
so atm, i think your assertion 2 is wrong (why i'm not speculating on)
oh, iswym - ok, roger
[12:02]
SvenDowideit shakes his head, as he's just shown that what he's had working today doesn't work
http://foswiki.org/Development/UseOfSetInsideIncludedSectionsTranscluded
[12:16]
well there you go - seems i've been lying all along and what i think i use doesn't work
friggen wonderful
[12:30]
...... (idle for 28mn)
CDotSvenDowideit: I'm not working on parameterised includes; I did some research into related matters, reflected in open feetrequests, but no more than that (all my patches are there)
CDot was hoping to get some feedback on http://foswiki.org/Development/PreferenceValuesInDifferentContexts, but accepts that people are busy
[12:58]
SvenDowideitCDot, :)
i did some more investigations based on your topic
and none of what i said seems to be true
which leaves me with alot more investigation to do
ffs, that comma breaks the url :(
[13:02]
CDotdoesn't break the URL for me. I didn't really understand your new topic, but I'm sure you'll tidy it up/merge it as appropriate. [13:15]
SvenDowideitmaybe if i ever get mor esleep
ok, commented for you
not that i stand by my initial reaction - as i seem to be more tired than i thought
[13:17]
........... (idle for 54mn)
HenroRitchieHi - i am struggling with an EOF error when building a table [14:12]
CDotHenroRitchie: source painful. What is an EOF error (yes, end of file, I know, but there are no files in Foswiki, only topics)
^source^sounds
[14:18]
................ (idle for 1h17mn)
***HenroRitchie has left [15:35]
........................ (idle for 1h58mn)
ArthurClemens"Blog" in top menu now points to blog.foswiki.org. Not ideal, but at least we point to the latest articles again.
And user FosWiki000 has been deleted for spamming prada
[17:33]
........................... (idle for 2h13mn)
foswiki_irc6having problem with NewApacheConfigGenerator
server is a recent install of ubuntu server
apache reads the foswiki.conf in conf.d/
but no rewrite rules are executed and no actions logged, despite log level of 9
log file is created
the only error anywhere is: [warn] NameVirtualHost *:80 has no VirtualHosts
since ports.conf has the line: NameVirtualHost *:80
i'm not an apache pro, as you might guess from tye question
was hoping to both shorten the URL and also send attachments through viewfile
[19:47]
................................................ (idle for 3h59mn)
***card.freenode.net sets mode: +o julian_w1rk [23:51]

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