#foswiki 2016-05-05,Thu

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

WhoWhatWhen
GithubBot[distro] gac410 pushed 1 new commit to master: https://git.io/vwFfc
distro/master 097e76e George Clark: Item000: Merge branch 'Release02x01'
[00:00]
***GithubBot has left [00:00]
FoswikiBothttps://foswiki.org/Tasks/Item000 [ Item000 ] [00:00]
....... (idle for 33mn)
gac410vrurg ... got time for a question? [00:33]
.................................... (idle for 2h56mn)
foswiki_irc0Hi! I'm completely new to this. None of the links (Main.WebHome, Installation Guide, Upgrade Guide, or Release Notes) seem to lead anywhere. What can I do? [03:29]
gac410How did you configure your server. Tell us a bit about your instalation [03:30]
foswiki_irc0Okay. Well I'm a part time volunteer guy at a charity and when it comes to Wikis, a complete novice. I essentially unpacked the foswiki zip file into a subdomain and that's as far as I went. [03:31]
gac410is it your own private server or a hosted domain somewhere [03:32]
foswiki_irc0It's on a hosted domain
A subdomain of our website
[03:33]
gac410Is the web server apache? on a linux or *nix server?
And is it accessible on the internet? Could I browse to your default page (where the links don't work)
[03:34]
foswiki_irc0I'm unsure about the first question [03:35]
gac410okay [03:35]
foswiki_irc0and yes: http://wiki.haringeymsc.org [03:35]
gac410okay ... hang on. [03:35]
foswiki_irc0Thanks [03:35]
gac410okay. So browsing to http://haringeymsc.org/wiki/bin/view ... I'm seeing perl code. So the web server is not configured to run the perl CGI code. [03:36]
foswiki_irc0Okay
How do I configure it as such?
[03:37]
gac410So now I'm guessing, but typically on a web hosted server, you use .htaccess files copied into the directories
You appear to have expanded 2.1.2, so it has a simple example .htaccess. Look at http://haringeymsc.org/wiki/INSTALL.html and scroll down to the header that starts "Method 3: "
[03:38]
foswiki_irc0Ah yes, I see it [03:41]
I wonder if this is a bit beyond my skillset [03:48]
gac410hard for me to know
The step by step instructions are all there, but the do assume some knowledge of apache
[03:50]
foswiki_irc0Okay
I'm gonna give it a whirl
Thanks for your help :)
[04:03]
***gac410 has left [04:03]
.................. (idle for 1h28mn)
ChanServ sets mode: +o CDot [05:31]
...................................................................................... (idle for 7h7mn)
ChanServ sets mode: +o gac410 [12:38]
................. (idle for 1h23mn)
gac410vrurg: Do you have time for a question? [14:01]
........ (idle for 37mn)
vrurggac410: hi,
Yes, have some.
[14:38]
gac410I've been trying to figure out how to get your branch running.
Calls to webExists and topicExists. Some use Plugins::SESSION Some use $Foswiki->app Not sure what the correct calling convention should be
The Foswiki::Request code that you merged from me still uses Plugins::SESSION
[14:39]
FoswikiBothttps://trunk.foswiki.org/System/PerlDoc?module=Foswiki::Request [14:40]
gac410I ended up changing it to Foswiki::Func:: calls, which got closer. But ran into other issues. [14:41]
FoswikiBothttps://trunk.foswiki.org/System/PerlDoc?module=Foswiki::Func [14:41]
gac410expandStandardEscapes. was not found. Now in Macros and claims to be static but uses $this-> Had to remove the $this
Just wasn't sure if I start trying to fix this stuff if I'm going to disrupt you.
Got past all them, and had Preference expansion issues.
[14:42]
vrurgProblem one: I often forget to change docs. So, expandStandardEscapes is ObjectMethod and is been used this way some other places. But frankly, I think this was a mistake on my side as it's more of a static in nature. [14:44]
gac410Yeah. It really is purely static. probably ought to be part of render.
Or a utility function
[14:44]
vrurgPlugins::SESSION cannot be used. I overlooked its use in parse while merging. Strange it does something. [14:45]
gac410well complains about undef ;)
The reason the landing URL works is because there is no path to parse.
[14:45]
vrurgSo, as in fact it has to be $Foswiki::app->store->xxxExists
expandStandardEscapes – I think it's better be moved to Foswiki.pm then.
[14:45]
gac410Ah... okay. actually then Func is wrong too. It uses Foswiki::app->webExists($web) [14:47]
vrurgA lot of code still using Plugins::SESSION but due to some methods moving into another namespaces it's not possible to just replace it with Foswiki::app.
So, they are acting as markers where to pay special attention.
Func needs fixing then, right.
[14:47]
gac410What might be helpful would be a table of | OLD | NEW | that lists the old conventions and the new convention. I was getting confused finding 3 different calls. ;)
If I start to fix up some of these would it be helpful, or are you still making big structural changes that I'll conflict with.
[14:49]
vrurgI know and thinking about a topic about changes. But cannot find enough spare time this week to make it. Or add a section to NewOODesign. [14:51]
gac410I don't mind trying to start something. Probably a separate topic would be best. Some of the other topics are pretty big already [14:51]
vrurgI foresee some changes yet. But waiting for them will slow down the whole process. [14:52]
gac410I wonder if the change to PSGI would fix a rather vexing issue we have on Foswiki.org. Our FCGI engine keeps leaving zombie processes, and we've had implement a cron watchdog to restart apache [14:53]
vrurgFor example, engine rewrite isn't done. I want remove any references to request and make engine pure I/O driver which doesn't care about core's architeture. [14:53]
gac410If as Jomo suggests, the engine could actually go away, maybe psgi can manage it
The FCGI engine tries to implement an automatic process restart of the FCGI handlers, but it is buggy.
[14:53]
vrurgWhat jomo proposes means a user will need additional modules. What's worse – he would need them in hosting environments or it won't work.
This is where engine come in handy. Plus, I'm thinking of a test engine which would simulate working environment for unit tests.
[14:54]
gac410Anyway, our FCGI engine appears to be broken. I'm not sure how much better mod_perl is either. [14:56]
vrurgPSGI itself may introduce another problem: memory hog. I would need to be restarted once in a while too, I'm afraid. [14:56]
gac410That's fine. It's not that FCGI needs to be restarted. It's that the code that handles it has issues. [14:56]
vrurgAnyway, just to put it all together: my engine-is-a-driver concept will make it easier to get rid of engines later anyway simply because it will be a lightweight layer between actual environment and the core. So, for now we need it anyway. [14:57]
gac410Good. we can cross that bridge later.
gac410 is hoping that if PSGI framework is well tested, it might solve some issues that we've had in our own attempts to be cross-platform.
But that's down the road.
[14:58]
vrurgExactly. Plus, whatever jomo dreams of might not be completely true and different PSGI wrappers for CGI, FCGI, mod_perl may happen to be less compatible than we would like them to be. This might make engines necessary to reduce those discrepancies. [14:59]
gac410I'll let you go. I'll try to work on a table of before/after code snippets. And will try to deal with the *exists routines for now.
I'll let you deal with any refactoring - moving things around.
[15:00]
vrurgThanks a lot! Unfortunately I have a meeting today and won't have time again. :( [15:01]
gac410n/p I have stuff to do too. [15:01]
vrurgAs to exists – I think they solely belong to store, I think. As any other web/topic manipulation code. [15:02]
gac410makes sense to me. [15:02]
vrurgAnd if you gonna move/changes expandStandardEscapes – try to grep thru places where it's been called as macro->expandStandardEscapes. Or change your code to call it as a method and I will take care of it tomorrow. [15:04]
gac410Okay. It was in Macros/SEARCH.pm calling $this->expandStandardEscapes was not found.
I changed it to simply expandStandardEscapes() and removed the $this = shift from the routine.
Since SEARCH is in Macros namespace
But It doesn't really feel right. But I'll let you refactor ;)
[15:06]
vrurgOk, I'll fix it. I knew it was a bad decision. ;)
Will be gone for now. cu tomorrow!
or may be tonight. Not sure yet. :)
[15:11]
.......... (idle for 45mn)
***ChanServ sets mode: +o CDot [15:56]
gac410vrurg: See Foswiki:Development/Foswiki3CodeChanges ... This is sure going to need your guidance, but I'll add to it as I encounter changes [16:06]
FoswikiBothttps://foswiki.org/Development/Foswiki3CodeChanges [ Foswiki3CodeChanges ] [16:06]
................................... (idle for 2h54mn)
gac410whoops thanks for catching my typo vrurg [19:00]
vrurgnp. My other plans for today are broken, got some time. [19:09]
....................... (idle for 1h51mn)
gac410vrurg, Noticed for getScriptUrl some places you call $Foswiki->app->getScriptUrl, others it's $Foswiki->app->cfg->getScriptUrl
gac410 assuming the latter is correct
[21:00]
...... (idle for 26mn)
vrurgYou're right. It's config's methods by nature. [21:27]
Fixed. Will be commiting soon. [21:34]
gac410okay. I'm not editing, so I won't get in your way. [21:34]
GithubBot[distro] vrurg pushed 1 new commit to Item13897: https://git.io/vwNro
distro/Item13897 3ebf44b Vadim Belman: Item13897: Moved expandStandardEscapes to Foswiki.pm...
[21:40]
***GithubBot has left [21:40]
FoswikiBothttps://foswiki.org/Tasks/Item13897 [ Item13897: Implement ImproveOOModel proposal. ] [21:40]
gac410vrurg: for my learning. ;) When you access $app->webName, and there is no method of that name is it automatically accessing the instance data {webName} ? [21:47]
vrurgNah, no way! Great way to locate mistypes. ;)
webName is dead. Long live $app->request->web! :)
[21:47]
gac410hm. Will request->web return a default? (in my version of Request it did not.) it strictly returned what was requested in the url / query not what we actually use.
So foswiki->{webName} had the defaulting applied.
[21:50]
vrurggac410: Yes, it will. Look at 'has web' default attribute. [21:51]
gac410Ah... very clear. Thanks. [21:52]
vrurgThe requested web is in _pathParsed as a key. But if we really need it then we better have requestedWeb attribute added. [21:52]
gac410hm The only time we need it is in error handling if we want to say what you requested (blah) was invalid
I need to get hopping on Request::Json to get the web/topic out of the posted json data.
[21:53]
vrurgThen we will need the original uri, I guess.
For the error report.
Sorry, didn't get it about Request::Json. What's the problem?
[21:54]
gac410I didn't write it yet ;)
For JSON requests, the Web / Topic is actually contained in the POST, it does not have it in the query path or params.
[22:02]
vrurgAh, right. I remember this. In my branch it would be useful if _pathParsed gets filled in with parsed data so that other attributes are initialized without keeping in mind where the info comes from – uri or JSON. [22:05]
gac410Right. It was the last part of Foswiki::Request that I had not written yet. It's important that the JSON POST web/topic be used as the base web/topic for the request, otherwise the wrong defaults (from web prefs, etc) can be applied. [22:06]
FoswikiBothttps://trunk.foswiki.org/System/PerlDoc?module=Foswiki::Request [22:06]
vrurgAnd I just refreshed that code in my mind. _pathParsed get initialized from Request::parse, thus as soon as parse returns consistent hashref the rest of the code wouldn't pay a damn. [22:08]
gac410Right. and Request::JSON::Parse would examine the POST to get the web/topic and extract the "Namespace" and "verb" It's part of Item13969 and Foswiki:/Development/MoveQueryPathParsingIntoFoswikiRequest [22:12]
FoswikiBothttps://foswiki.org/Tasks/Item13969 [ Item13969: rest and jsonrpc scripts should default the web/topic if -topic query param is not provided ] https://foswiki.org//Development/MoveQueryPathParsingIntoFoswikiRequest [ MoveQueryPathParsingIntoFoswikiRequest ] [22:12]
gac410er Namespace/Method for JSON, [22:12]

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