#foswiki 2012-08-08,Wed

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

WhoWhatWhen
gac410Hm Could setting UTF8 or Locales in Foswiki cause some simple code that reads an attachment with Func::readAttachment, and writes it out using binmode; print; ... somehow corrupt the file?
I'm at a complete loss as to how Item12028 - PDF image corruption - is happening.
[00:04]
FoswikiBothttp://foswiki.org/Tasks/Item12028 [ Item12028: GenPDFAddOn not Work Corretly ] [00:05]
gac410SvenDowideit: Ugh - Log::Dispatch doesn't appear to have corresponding read routines. [00:11]
........... (idle for 52mn)
***Cervator has left [01:03]
gac410I suspect that eachEventSince doesn't even make sense for things like logging to syslog. For statistics to work, the logger needs to support eachEventSince [01:06]
SvenDowideitgac410, exactly
don't worry about it - only one or 2 people i know use them (there's a plugin i wrote to show who did what
i may end up thinking about writing UnLog::Dispatch for some backends tho
SvenDowideit runs again :)
[01:12]
gac410UserInfoPlugin, ContributorPlugin, and Statistics fromwhat I can find. [01:13]
SvenDowideitBabar, ooo, cool
yup
[01:13]
gac410TBH it would be better to deprecate eachEventSince and move to offline log processing. [01:14]
............................... (idle for 2h33mn)
Sven - not checked in, I've added a new BOOLGROUP type to configure - **BOOLGROUP info, warn, debug, error** for a group of related checkboxes. Works great, except #@(@@ configure won't correctly report the multiple values in the "your changes" summary [03:47]
......... (idle for 43mn)
***gac410 has left [04:30]
.............. (idle for 1h9mn)
pharveygood day
I've been added to github repo things, hrm. split repos borken. This job calls for the TIME MACHINE!
[05:39]
................... (idle for 1h34mn)
***ChanServ sets mode: +o MichaelDaum [07:14]
MichaelDaumthere's another problem in the rcs store
besides some minor errors that I have spotted
the problem is that you only know the real revision id _after_ the checkin finished ... not before
right now, there is a kind of educated guess relying on rcs to count up strictly. so getNextRevisionID basically replicates this by also counting one up.
the store shouldn't do that
instead it should get some kind of return value from the checkin what the new revision id now is
this will terribly fail trying to implement a git store
there's no way to know the revision id without actually doing the checkin
as it is a checksum on the content being checked in
besides the current getNextRevisionID is broken
as simple as it is, it gets it wrong when theres a checkin pending
getNextRevisionID is computing getLatestRevisionID + 1
seems so simple
well but getLatestRevisionID already counted +1 when there's a checkin pending, thus already anticipating the pending operation.
so getNextRevisionID counts one too much thus producing a gab in the revision history it seems.
not sure what happens then
either rcs counts just fine, or meta:topicinfo is just out of sync.
or both
or rcs is forced to skip one number
in any case other operations are relying on the revision history counting strictly upwards with no gaps in it
so while there is YASQF (yet another storage quick fix) ... the problem is more fundamental as outlined at the beginning
[07:24]
CDotMichaelDaum: it has to do that. I tried to introduce the idea of non-strict-ordering, and even non-numeric revision ids (thinking of git) but it just didn't work; too much of the UI assumes strict +1 [07:37]
MichaelDaumthere's no problem with that for now [07:37]
CDotI think the counting is fine, but the synch to TOPICFON may be flawed [07:38]
MichaelDaumthe problem is getNextRevisionID [07:38]
CDotgetNextRevisionID knows the current revision ID, right? So assuming +1 it can't really get it wrong, can it? [07:38]
MichaelDaumfor now TOPICINFO and rcs rev ids are in sync by accident at best
[09:33] <MichaelDaum> well but getLatestRevisionID already counted +1 when there's a checkin pending, thus already anticipating the pending operation.
which makes +2
[07:38]
CDotwhere? [07:39]
MichaelDaumthe $rev++ some lines above [07:40]
CDotonly getNextRevisionID should be assuming +1. Nothing in core (non-store) should maniuplate rev nos [07:40]
MichaelDaumthe idea as far as I get it behind this is: "a pending checkin is one rev ahead"
so the real current rev is say, 2, then the txt file is botched, the checkinPending detector adds +1, getNextRevisionID also adds +1.
I am about to enhance the tests in VCStoreTests.pm to cover this
what follows is that before entering the RCS impl, a META:TOPICINFO is altered adding the new rev id. Afterwards the real checkin happens, in the hope the rev id returned by the RCS impl is the same as the one added to META:TOPICINFO a few milliseconds before.
in cache language this means
the cache is populated with information _before_ the store underneath actually computed that data
and both values are computed independent of each other
[07:41]
GithubBot[foswiki] FoswikiBot pushed 1 new commit to master: https://github.com/foswiki/foswiki/commit/3c1e4922f64643442f60f6e59c8e17374fe8b3fd
[foswiki/master] Item11983: - MichaelDaum
[08:00]
***GithubBot has left [08:00]
FoswikiBothttp://foswiki.org/Tasks/Item11983 [ Item11983: prevent excessive calls into the revision system ] [08:01]
CDotah, if the .txt file is botched and you are relying on TOPICINFO for the next rev then yes, you are stuffed. That's why i coded it so it never relies on TOPICINFO. But if you can be certain that TOPICINFO is correct, it should be fine. But you must be absolutely certain! [08:14]
GithubBot[foswiki] FoswikiBot pushed 1 new commit to master: https://github.com/foswiki/foswiki/commit/bad2088e4d36bfdf0159fc05157f89946bb85cd6
[foswiki/master] Item11983: fixed unit tests computing the rev id of pending checkins - MichaelDaum
[08:15]
***GithubBot has left [08:15]
........ (idle for 37mn)
MichaelDaumthe tests in VCStoreTests on inconsistent topics are all wrong
they assume the next revision is 3 even though there only was one saved revision before
[08:52]
so the idea behind the performance improvements are
during view, rely on META:TOPICINFO no matter what
during save, detect a pending checkin and properly cache the META:TOPICINFO
so the unit tests in VCStoreTests creating an inconsistent topic with a faked 77 rev id, should expect 77, even on a pending checkin, and not 2, as this would require a deep inspection of the revision history, which is not applicable during view for performance reasons
[09:03]
kip3fMichaelDaum / CDot: if foswiki is using a git backend, then what would happen to the URLs in the action buttons? (Currently showing "r4 < r3 < r2 < r1"). Would it show the internal git checksums? Or is there anything that maps between store-specific IDs and user-friendly ones? [09:08]
CDotthere is no mapping; I was anticipating that such a mapping would be done by the store
it has to be - git checksums do not imply any ordering
[09:09]
MichaelDaumkip3f, r4 < r3 < r2 < r1 should go away completely. Nobody groks these tiny urls. [09:09]
CDotMichaelDaum: I beg to differ; the function of these tiny urls is bleeding obvious [09:10]
MichaelDaumgiggle [09:10]
DummyUserhey gus, I've found something cryptic at the bottom of a wiki page: r4 < r3 < r2 < r1
and I was asking myself: can I remove it?
and: why is 4 < 3?
[09:11]
kip3fit does look a bit 'geeky'. It would be nice if was something like [previous edit] [compare with previous] [more revisions] [09:12]
MichaelDaumthis is what NatSkin shows in a modal dialog clicking on the "Revision History" menu item of the topic actions: http://foswiki.org/pub/Tasks/Item11983/HistorySnap1.png
besides r4 < r3 < r2 < r1 is rendered by a pseudo macro in UI/View.pm called %REVISIONS%...totally undocumented, only works in the bottom tmpl code, not in the content or anywhere else
[09:13]
kip3fTHe first '<' is very useful and deserves more verbosity [09:16]
MichaelDaumit burns cpu too by requesting the list of all revision ids till the beginning of history from the store backend.
on every click
kip3f, yes, something like "See last change"
[09:17]
kip3fbut really, the reason why I asked the question was to know if the 'rXXX' identifiers always correspond directly to the store's identifier or not. I guess now there is some massaging of the RCS id (1.23 -> r23). Another question: what would %TOPICINFO% look like, in a foswiki with git backend?
I mean the string at the bottom "Topic revision: r4 - 08 Aug 2012 - 01:50:25 - KipLubliner"
[09:22]
.... (idle for 18mn)
MichaelDaumkip3f, the numeric id aren't really the store's revision ids
for rcs they only happen to be identical
so even for git we can maintain the 1,2,...,n ids to refer to the cronological order of the revisions
as far as I know git has got the feature to store tags per revision
to use these free-form tags to refer to a revision
thats where I'd put this kind of "mapping" for accessing the revisions by tag quickly
(given we won't ever do branching)
[09:40]
kip3fMETA:TOPICINFO{author="KipLubliner" comment="reprev" date="1344390625" format="1.1" reprev="4" version="4"} [09:44]
MichaelDaumy [09:44]
kip3fI don't see the internal ID here (1.4)
just the publicly visible one
[09:44]
MichaelDaumright version="4" only happens to refer to r1.4
version="4" is computed by foswiki while interacting with rcs. rcs computes r1.4 on its own.
so a git store would compute its checksum and store a tag "4" with it.
_for this one topic_
problems arise when using a single git for _all_ of data+pub
which would be preferable to support reverting a rename across webs
[09:44]
kip3fSo the user would never have to deal with raw git strings (a good thing). What about other code outside the store? Would it always use the "user-friendly" identifiers? [09:52]
...... (idle for 27mn)
gotta go ... bye [10:19]
MichaelDaumnother error: saveAttachment returned and recoreded the change using a self-computed next rev id (not using getNextRevisionID) and on the wrong object (not on the file but on the topic it was attached to) [10:27]
........................... (idle for 2h11mn)
gac410Hi Locale knowledgeable out there. I'm baffled by an attachment corruption issue in GenPDFAddOn. Works fine for me, corrupted by a site using Locale pt_BR.UTF8
Code uses Func::readAttachment to read in a file. Then opens a temp file using binmode and writes it out.
File looks good, but can't be viewed in the generated pdf.
[12:38]
GithubBot[foswiki] FoswikiBot pushed 1 new commit to master: https://github.com/foswiki/foswiki/commit/76c5e9c8a860d78e560f5727fcf42a1cf101ba8a
[foswiki/master] Item11983: - MichaelDaum
[12:45]
***GithubBot has left [12:45]
FoswikiBothttp://foswiki.org/Tasks/Item11983 [ Item11983: prevent excessive calls into the revision system ] [12:45]
gac410except the saved files are identical to the attachments. So that theory doesn't hold water. Why oh why is the pdf broken. [12:46]
MichaelDaum_: and others. I really like the little mini-history 5 > 4 > 3 > ... at the bottom of topics. It is really handy to look at recent checkins without navigating to a formal history screen. I use it all the time.
I would definitely object to it being removed.
CDot: If I check in some experimental code to Configure ... adding a new type - BOOLGROUP - for a group of checkboxes, could you have time to advise me on a bug
And does adding new field types to configure constitute a feature, needing a formal feature request.
;(
[12:54]
GithubBot[foswiki] FoswikiBot pushed 1 new commit to master: https://github.com/foswiki/foswiki/commit/54eb77ad08736c69445732cbde0a88ccccc650a5
[foswiki/master] Item11983: - MichaelDaum
[13:00]
***GithubBot has left [13:00]
MichaelDaumnother error: can't do a normal repRev when there was damage saved [13:10]
gac410MichaelDaum: one thing that concerns me. If you edit a topic with a missing rcs,v file (like editing a "distributed" Main.WebHome) do you get 2 revs in RCS. Rev 1 = the distributed topic, Rev 2 = local changes... on the first save. [13:12]
MichaelDaumyes
that's what I've found in the tests' comments to be intended so
[13:12]
gac410excellent. That was a really bad bug back earlier in 1.1.x [13:12]
MichaelDaumnow there's still tests missing where users are credited for oob changes. [13:13]
gac410The earlier issue was you could never revert damage to distributed topics without referring to the distributed tarbal. [13:13]
MichaelDaumMichaelDaum about to fix that [13:13]
gac410Yeah. that too. the checkin of the oob change should be by UnknownUser, never by the last saver. I thougt that was working. [13:14]
MichaelDaumthere was a lot that did not work out properly
the tests all passed because they used a forecenewrevision
to create the next revision after saving the oob into a rev of its own
without a forcenewrevision a replaceRevision() is done
mungling the oob and the intended changes into one rev
[13:15]
gac410Well it was working in 1.1.5. In addition to unit tests I manually verified that pretty thoroughly.
May have been working by accident. I was trying to fix it at the last minute, without really understanding CDot's changes.
[13:17]
Hm. NatEdit on trunk. it doesn't give you the little cut/paste box for your signature. :( [13:22]
MichaelDaumthere's a button that inserts the signature at your cursor position
curs to me thats better that c&p
[13:26]
gac410Well that's certainly obscure. I spent time hunting in the tabs, and finally gave up and edited with release11 [13:27]
MichaelDaumah in _tinymce_ there's no signature button
in NatEdit it is there :/
[13:28]
gac410gac410 generally hates the little icons esp. for new users. Great for power users, but totally obscure for anyone who doesn't know about it.
Once you know it's obvious. but ...
[13:29]
MichaelDaumgac410, did you know some browsers refused to pick up selected text from disabled input fields. so this feature was busted for a long time already. [13:29]
gac410It's always worked for me. And the "Here is your signature for copy/paste" is a big flag saying ... Sign your work. We'll have nothing being signed with that little butotn. [13:30]
MichaelDaumgenerally, there are so many features in an editor ... you simply can't make them _all_ screaming obvious. [13:32]
gac410CDot: Added new task Item12035 for your perusal. Does this rise to the level of a feature request? [13:32]
MichaelDaumyou end up in screaming for everything thus disturbing normal editing [13:32]
FoswikiBothttp://foswiki.org/Tasks/Item12035 [ Item12035: Add new data type to configure: BOOLGROUP for a group of checkboxes. ] [13:32]
gac410MichaelDaum: Little thing like signing your work are important aspect of a wiki, and the input field was to me an important flag saying: You should sign, and here is what a proper signature looks like. [13:33]
MichaelDaumand the default signature in WebTopicEditTemplate is one of the most hated things. I constantly get a "can we please remove this signature" [13:34]
gac410I agree. In the template is annoying [13:34]
MichaelDaumone thing I like CommentPlugin for is, it adds signatures when required _automatically_.
no need to add them manually
[13:35]
gac410Though I wish it was in the Support web. New questions are never signed. And my responses blend in to the original question. [13:35]
MichaelDaumtrue
thats a problem of these support templates afaics
[13:35]
gac410I am forever going back in and adding the original omitted sig. [13:35]
MichaelDaumsame same
creating a support question should add the signature _automatically_
[13:36]
gac410Agreed. [13:36]
MichaelDaumso as you see, even though there's a c&p box with the signature in there, people dont use it for the support questions-
as they are newbies by definition hitting the support web, they are not used to add their footprints using something to copy and paste from one area of the page to the next as a matter of some good wiki policy.
they wont do that with the natedit buttons either
[13:36]
gac410yeah. But the button is even more obscure. Yes support template should add it. But I *like* to copy/paste.
double-click / chord / done.
[13:39]
MichaelDaumhah: better write something new and unique
it is only a GoodThingToDo on foswiki.org
to sign stuff
in most other corporate wikis it really is a very very minor thing
either people get the signature right automatically, or they dont require it as their name is in the revision history anyway.
or they use better commenting systems
note this is _only_ for commenting, not for refactoring existiing wiki text
[13:39]
gac410yeah that makes sense. [13:44]
MichaelDaum: what would be your though on deprecating the eachChangeSince function. It's only used in a couple of places. Mainly statistics. [13:56]
MichaelDaumhm
not sure if it is still used in SolrPlugin for indexing changes
[13:56]
gac410The issue is moving to a more advanced logger, like Log::Dispatch, Logj4perl, etc. [13:57]
MichaelDaumas far as I remember eachChangeSince didn't work out for me ... could have been due to oob changes not being picked up, for obvious reasons. [13:57]
gac410None of them provide functions to read the log. And in some cases, logging to the Win32EventLog, or Syslogd, wouldn't make sense.
(to provide a read function)
In the case of syslogd, the log might not even be on the same system.
[13:57]
MichaelDaumMichaelDaum recaps http://foswiki.org/Tasks/Item8460 [13:59]
gac410Hm. I'm confusting things then.
It's Log::eachEventSince that reads in the events.log file.
[13:59]
MichaelDaumah and recordChanges() writes to web/.changes
thats two different things
for whatever reason they are separate features ...
[14:00]
gac410yeah. It's the log one I want to deprecate. We need to move to a more modern logger. PlainFile is a disaster.
Busy logs fail to rotate, No locking,
[14:01]
MichaelDaumyea I like that sound [14:02]
gac410Log::Dispatch looks really good. Sven started a skeleton version. Logs to syslogd, plain file, and stderr. (simultaneously for now).
It is pluggable, Has versions that do rotating (like Plainfile) Stamping (like compat), locking for simultaneous access. DBI, Syslog, Jabber, email, Win32Events ...
But it
but it's all write-only.
[14:03]
MichaelDaumoic why you asked for eachChangeSince now
the only "client" hooking into the Logger api is this ActivityStreamPlugin that has got an alternative logger impl
[14:08]
gac410Yeah, except I meant to say eachEventSince. which is part of Func and the Logger API
Grep through trunk - finds the UserInfoPlugin, ContributorPlugin, statistics, and I think ActivityStreamPlugin has eachEventSince commented out.
[14:09]
MichaelDaumeachEventSince...used by UserInfoPl ... gotcha
from the pov of an activity stream thingy, it does make sense to index logged events at some point, maybe _before_ they enter the Log::Dispatch engine.
that way reading from some generic Log::Dispatch sink isn't really required other than system log reader
e.g. you won't be able to read back in from a jabber sink
[14:10]
gac410Log::Dispatch::DBI for ex. will insert each event into a sql database. which would be really powerful for statistics, etc. [14:14]
MichaelDaumtrue
yet needs a lot of cleanup
to make sense of the stuff pumped into the logs
[14:14]
GithubBot[foswiki] FoswikiBot pushed 2 new commits to master: https://github.com/foswiki/foswiki/compare/54eb77ad0873...fc64fec9f310
[foswiki/master] Item11983: repRev degrades to a normal addRev when there's an implicit interim save of oob changes - MichaelDaum
[foswiki/master] Item11983: fixed unit tests on repRev degrading to an addRev - MichaelDaum
[14:15]
***GithubBot has left [14:15]
FoswikiBothttp://foswiki.org/Tasks/Item11983 [ Item11983: prevent excessive calls into the revision system ] [14:15]
gac410So maybe eachEventSince is separated from the Logger API, sits along side to read the logs only when it makes sense. (ie DBI, File, but not Jabber, Syslogd, ...)
I'm putting up a proposal to deprecate eachEventSince. We can hash out a replacement / alternative there.
[14:15]
MichaelDaumor we keep a readable version no matter what backend is used [14:16]
gac410On busy sites, that leaves us writing our own rotate code, file locking code, etc. again. [14:16]
MichaelDaumhm I thought Log::Dispatch can do simultaneous logging [14:17]
gac410Yes, true. it can. [14:17]
MichaelDaumso the idea to preserve readable log events is to always have one logging into a readable backend sim'ly [14:18]
gac410Ah. So we *always* log to a plain file with rotate and flock control. And the other methods are optional. [14:18]
MichaelDaumy [14:18]
gac410I also get concerned about mixing "web" transaction event logging with "store" event logging. One web transaction could result in multiple store events.
They just feel like they should be separate.
[14:20]
MichaelDaumya. but well, it is a log dump: fire and forget. let others make sense of this data. ;)
though the logged stuff should be tagged sufficiently to make sense of each event.
i.e. not being ambiguous: "this log entry was either caused by ... or ..."
[14:24]
gac410yeah. we really need a full code review of logging. I found a place where *nothing* is logged for update events even. :( [14:27]
MichaelDaumthere are also some events that are questionable whether they should be logged
like a rest call
[14:28]
gac410well coming from a mainframe transaction processing background, I'd like to see a *lot* more logging. Sometimes I feel like I'm completely blind to what is actually happening. [14:30]
MichaelDaumah a transaction log for rolebacks ;) [14:32]
gac410well even more than that. precisely where was the time spent. input queue, processing, every i/o, . If someone calls, at 5:pm blah took too long. I'd like to say, well you spent 1.5 seconds, waiting for an apache thread, .5 seconds running,
The %blah% macro took 10 seconds for I/O ... etc..
But that's all wishes, that level of logging just doesn't happen any more.
[14:34]
............... (idle for 1h13mn)
***ChanServ sets mode: +o MichaelDaum [15:48]
...... (idle for 25mn)
JacobEtchesCan the TopicTitle field in nadedit be hidden?
We don't use TopicTitle, and users are confused as to why the TopicName is not changed when they edit that field
[16:13]
MichaelDaumJacobEtches, the TopicName -> TopicTitle mapping will be stored when installing DBCachePlugin [16:15]
JacobEtchesMichaelDaum, Thanks, but TopicTitle won't reset TopicName, will it? [16:17]
MichaelDaumno [16:17]
JacobEtchesthat's what users expect, not realizing they are different [16:17]
MichaelDaumit is there in order to add title and link text to topics without having to change the topic's name (aka database id) [16:18]
JacobEtchesThat way people are very aware of the TopicName
Yes, I understand. But we use descriptive TopicName's and use a default topic that includes the topic name as H1
[16:19]
MichaelDaumhow about displaying the topic title as h1
see the thing in action at http://demo.michaeldaumconsulting.com
[16:20]
JacobEtchesWe could, but I fear the separation of TopicName and TopicTitle will create more problems than it solves [16:21]
MichaelDaumsorry gotta run see ya. [16:21]
........ (idle for 36mn)
CDotgac410: I'd say no. [16:57]
....... (idle for 33mn)
gac410CDot: No for looking at bug? No for needing a proposal? No for deprecation of eachEventSince Or just a general NO to gac410 :) [17:30]
CDotgac410: No. [17:31]
gac410:P [17:31]
CDotI looked at the bug (if that's the one on BOOLGROUP) and I say "no" to needing a feet reqquest
I commented on the deprecation
[17:31]
gac410oh. okay thx. that was my opinion on the feat. Being in configure, and essentially pluggable, not really a mainstream featurre. [17:32]
CDotright [17:32]
gac410Hm. I think that if a logger doesn't support eachEventSince, the parent Logger.pm provides an empty iterator. [17:33]
CDoty, which is a bit crap frnakly
pandering to the extension authors laziness
[17:34]
gac410For each spammer { add to awspi list; delete user; } [17:37]
CDotCDot heads off to cook supper. cya! [17:39]
gac410I'm going to check in my BOOLGROUP ... even though I can't figure out why the change log only shows the first selected value. [17:39]
........... (idle for 50mn)
donbarryGroan. Some Deletionist on Wikipedia wants to yank the Foswiki page again -- while leaving a host of truly insignificant wiki pages undisturbed. [18:29]
GithubBot[foswiki] FoswikiBot pushed 2 new commits to master: http://git.io/kHKGNw
[foswiki/master] Item12035: Implement checkbox groups in configure - GeorgeClark
[foswiki/master] Item12023: Experimental config.spec. - GeorgeClark
[18:30]
***GithubBot has left [18:30]
FoswikiBothttp://foswiki.org/Tasks/Item12035 [ Item12035: Add new data type to configure: BOOLGROUP for a group of checkboxes. ]
http://foswiki.org/Tasks/Item12023 [ Item12023: write a Log::Distch based Logger ]
[18:30]
***ChanServ sets mode: +o MichaelDaum [18:30]
MichaelDaumhttp://en.wikipedia.org/wiki/Foswiki [18:31]
uebera||Is there a new argument on the 3rd nomination page? Shouldn't that be the basic requirement for the nth nomination (no, this is only wishful thinking I don't expect an answer to that)? But I'm curious--someone who followed this... is there actually a summary how different Wiki engine related pages are subject to deletion? [18:39]
gac410Again? yeesh. Now why [18:40]
MichaelDaumplease raise your voice at http://en.wikipedia.org/wiki/Wikipedia:Articles_for_deletion/Foswiki_(3rd_nomination)
here's the previous discussion http://en.wikipedia.org/wiki/Wikipedia:Articles_for_deletion/Foswiki_(2nd_nomination)
[18:42]
gac410I'm afraid that as project members, we will be summarily dismissed.
It's back again to trying to prove notability. ArthurClemens did a lot of work finding mentions, but now someone comes along and dismisses it all
[18:42]
MichaelDaumall arguments from the 2nd nomination to keep the article are still valid 4 months later
thats all I can say
[18:43]
gac410And now they interpret the efforts as "there seems to have been some canvassing going on in the previous AfDs." [18:44]
MichaelDaumreferences have been deleted from the main article and move to the talk page: http://en.wikipedia.org/wiki/Talk:Foswiki [18:45]
gac410yup.
Someone needs to go through that list, "quote" the reference in our article so that it is a *used* reference, and then add it back in.
[18:45]
foswiki_irc1Hi, I am a total newb so please exscuse my ignorance but i created a bunch of webs/topics in the sandbox and im wondering if those will go live when i push the files to my domain
or do i need to put them somewhere else
[18:54]
gac410Which sandbox? On foswiki.org? [18:56]
foswiki_irc1i installed a foswiki on my local machine
and am using the sandbox to show it off to my company
[18:56]
gac410Sandbox is "just another web". so if you push the pub/ and data/ directories then those topics will remain. [18:56]
foswiki_irc1no issues with it being in the sandbox? [18:56]
gac410It's just another web. If you copy the files they will survive. Foswiki doesn't have any sort of automatic deletion, so sandbox is no less or more permanent than any other web. [18:57]
foswiki_irc1thank you
ready for another newb question?
[19:02]
MichaelDaumMichaelDaum added his 2cent to http://en.wikipedia.org/wiki/Wikipedia:Articles_for_deletion/Foswiki_(3rd_nomination) [19:02]
foswiki_irc1I understand these are supposed to extremely easy to edit, but if your pushing the files to the domain wouldnt you have to reupload the foswiki everytime you edit it to your server or am i missing something [19:03]
gac410foswiki_irc1: You would normally edit in place on your domain, not edit elsewhere and push. [19:07]
foswiki_irc1thank you as always gac [19:12]
gac410The complaint wasn't about too many references. It was that they are not linked to quotes within the article, if I understood the process. [19:19]
.... (idle for 18mn)
Yeah... I just tried to add it to the main article
I'm having trouble figuring out how to cite it correctly. Added it to the Deployment section.
It's too bad that pharvey's work isn't published.
really annoying that to properly cite some of these articles, you'd need to spend $20 to purchase access.
[19:37]

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