IRC 2007/03/21

16:25 < jfhv> OK, guys, I think we can start now :)
16:25 < lewiscot2> Ok
16:25 < danielr> ok
16:25 < Stellaris> Right
16:25 < jfhv> ok
16:25 < jfhv> 1) Development progress
16:25 < jfhv> 2) Selection API
16:25 < jfhv> 3) Safari implementation
16:25 < jfhv> 4) ???
16:25 < jfhv> About 1)
16:26 < jfhv> Many things have been uploaded to the trunk today
16:26 < jfhv> Selection API works for FF
16:26 < jfhv> I've been working on Opera, too
16:26 < jfhv> Thanks to Volker's jQuery fixes,
16:26 < jfhv> jQuery 1.1.2 is now in the trunk
16:27 < jfhv> There were issues/bugs in MSIE
16:28 < jfhv> that needed to be fixed before we can use 1.1.2
16:28 < jfhv> so now we can focus on the selection API
16:28 < jfhv> ok?
16:28 < lewiscot2> Excellent.
16:28 < lewiscot2> Good job Volker.
16:28 < danielr> :)
16:28 < vmx> the selection api should work in opera almost "as-is"
16:29 < jfhv> yes, I've copied/paste the FF code to the Opera one
16:29 < jfhv> and it seems to work well
16:29 < jfhv> that's good :)
16:29 < jfhv> Anything to add for 1) ?
16:29 < vmx> jfhv: have you also uncommented my small "tests" to see if it works?
16:30 < jfhv> ah no
16:30 < jfhv> I've added console.log
16:30 < vmx> ok
16:30 < jfhv> in Wymeditor.prototype.dialog or something
16:30 < jfhv> so when you click on 'Link' button
16:30 < jfhv> you see info about the selection
16:31 < jfhv> (so we're switching to 2) I guess)
16:31 < vmx> yes
16:31 < jfhv> most calls work in FF
16:31 < jfhv> though there's a problem with cursorToStart and End
16:32 < vmx> the firebug console is really nice
16:32 < jfhv> yes
16:32 < jfhv> About cursorToStart:
16:32 < jfhv> when you click on the Link button, I think the iframe looses the focus
16:33 < jfhv> so the cursor doesn't actually go to start
16:33 < jfhv> there's no cursor anymore
16:33 < lewiscot2> JF, you might try changing the 'click' handler to 'mousedown' (required for Safari anyway).
16:34 < jfhv> perhaps
16:34 < jfhv> (trying)
16:34 < vmx> probably the selection goes to the start, but you can't see it as it looses the focus
16:35 < vmx> but that's a button problem, and not a selection api one
16:35 < jfhv> yes, probably
16:35 < jfhv> mousedown works :)
16:36 < lewiscot2> It is probably a good idea to use mousedown rather than click in all cases because of the order in which the browser (all browsers) interpret the input device events.
16:36 < lewiscot2> Excellent.
16:36 < jfhv> that's nice
16:36 < jfhv> and pretty simple :)
16:36 < lewiscot2> The best kind of solution!
16:36 < vmx> i always use down/up, for keys and mouse. it seems more logical to me
16:36 < lewiscot2> I agree.
16:37 < lewiscot2> As far as I can tell, 'click' = 'mouseup'
16:38 < vmx> has anyone of you tested the event handling? i'm not quite sure if it works for all browsers at the moment
16:38 < jfhv> uploaded, so you can test yourself
16:38 < vmx> thingsd like "keydown" in the text area
16:38 < lewiscot2> Test event handling in what? Selection API?
16:39 < lewiscot2> Suggestion: can we refer to Selection API as "SAPI" so I don't have to type the whole thing? I'm kind of lazy :)
16:39 < jfhv> brb
16:39 < vmx> you'll use the selection api within the event handling. e.g. you press enter, the selection api will tell you the current situation (e.g. cursor is at the end), and then you will go on and do waht has to be done
16:40 < lewiscot2> ok, I see.
16:40 < lewiscot2> Perhaps before we write more code on other things, we should write some test cases for SAPI and have everyone run the test cases on their browser.
16:41 < lewiscot2> This way, we can make sure that we get consistent behavior in all browsers before going on to the next thing.
16:41 < vmx> with the selection api it should be possible to write all the code for event handling only once for all browsers
16:41 < lewiscot2> SAPI is going to be critical for Safari.
16:42 < vmx> i think the "click on link to get information" is already almost a test
16:42 < lewiscot2> Does the sapi capture things like currentNode, parentNode, nodeOffset (within the DOM)
16:43 < vmx> generic tests for all cases could take a lot of time. i haven't written any tests for javascript yet, so i don't know how much work it will be
16:44 < danielr> yes, but the tests can be made by several people
16:44 < lewiscot2> I think it will be time well spent, though. I think the best way to handle the test cases would be for you to write a short description for each function within the sapi that gives the following:
16:44 < lewiscot2> - expected inpu
16:44 < lewiscot2> input
16:44 < lewiscot2> - expected output (return value)
16:44 < lewiscot2> - what event triggers the sapi (if any)
16:45 < lewiscot2> - potential problem states (e.g., nothing selected)
16:46 < lewiscot2> Testing may take some time but it is better to do it now when it is easy to fix errors than to wait until other code is written that is dependent on the sapi.
16:46 < danielr> most these tests can be made by people like me, who don't know JS enough to develop.
16:47 < vmx> is the a common way to write tests for javascript?
16:47 < jfhv> (I'm back)
16:47 < vmx> perhaps the code from jquery can be used (they also use a test suite)
16:47 < lewiscot2> I don't know. I would assume it is the same as for other languages.
16:47 < jfhv> There's a framework:  http://www.jsunit.net/
16:48 < jfhv> And Selenium IDE, too
16:48 < jfhv> But the problem is iframe/designmode
16:48 < lewiscot2> Just so we are using terminology clearly ...
16:49 < lewiscot2> by test cases, I mean parameters like expected input, expected output, error returns, etc. These are ideas, and are not code-specific.
16:49 < lewiscot2> A test suite seems like more of a tool for executing test cases (a debugger)
16:50 < jfhv> JsUnit is a Unit Testing framework for client-side (in-browser) JavaScript. It is essentially a port of JUnit to JavaScript. Also included is a platform for automating the execution of tests on multiple browsers and mutiple machines running different OSs.
16:50 < jfhv> That's what you mean?
16:51 < lewiscot2> That is a test harness.
16:51 < lewiscot2> Technically speaking, test cases are the individual test descriptions
16:51 < lewiscot2> test suite is the entire collection of test cases
16:51 < lewiscot2> test harness is the tool that executes automated or manual tests
16:51 < lewiscot2> I think
16:52 < vmx> i agree
16:52 < lewiscot2> there are two common approaches to testing:
16:52 < vmx> but if i think of "tests" for wymeditor i think of "test harness" (although i haven't heard tat word before ;)
16:52 < lewiscot2> - Test at each milestone
16:52 < lewiscot2> - Test at the Alpha version
16:53 < lewiscot2> - Testing at each milestone makes more sense to me because you catch potential problems earlier.
16:53 < lewiscot2> It is also easier to find bugs
16:54 < lewiscot2> It saves a lot of rewriting of code that is dependent upon code that may have a problem in it.
16:54 < vmx> for the sapi the best would be a test harness, for the rest tests can be done manually
16:55 < lewiscot2> It seems like the sapi is going to be a critical component of Wymeditor. A lot of other code will be dependent on it. We should make sure that it behaves as we intend before adding a lot of event handling.
16:56 < vmx> i thought about something like that:  http://jquery.com/test/
16:59 < jfhv> that's interesting
16:59 < jfhv> what mean the numbers?
16:59 < vmx> i'm not sure, i just took a short look
17:00 < vmx> but coding the tests seem looks quite easy (once you understood what you need to do)
17:00 < lewiscot2> I think the numbers mean:
17:00 < lewiscot2> (num of errors, actual result, expected result)
17:01 < jfhv> I guess you're right
17:01 < lewiscot2> On FF Mac, test 30 fails
17:01 < lewiscot2> The result is:
17:01 < lewiscot2> core module: is(String) (2, 24, 26)
17:02 < jfhv> on FF Linux too
17:02 < lewiscot2> 2 errors, actual result 24, expected result 26
17:02 < jfhv> the same
17:02 < lewiscot2> At least it is consistent
17:02 < jfhv> and 57, 63, 73, 77 to 84...
17:02 < lewiscot2> consistent failure is easier to debug than inconsistent failure
17:02 < jfhv> beautiful
17:03 < lewiscot2> It means that the bug is probably in the code and not in the environment
17:03 < lewiscot2> JF, did you get 12 failed tests?
17:03 < jfhv> 23
17:04 < lewiscot2> Aie!
17:04 < danielr> on FF windows: 4 tests failed
17:04 < vmx> someone should take a look at the code and decide whether it could be used for wymeditor or not
17:04 < lewiscot2> JF, since Safari will likely be on hold, do you want me to be in charge of testing?
17:04 < jfhv> (CTRL+F5 and I get 4 errors)
17:04 < lewiscot2> I can test on Mac and Windows
17:05 < jfhv> yes, good idea
17:05 < lewiscot2> Ok. Great.
17:05 < jfhv> so we need to know what to test
17:05 < lewiscot2> I will put together some notes on how I think we should approach testing and post it on the wiki.
17:05 < vmx> as daniel said, writing test should be quite straight forward even for javascript newbies
17:06 < vmx> scott, what do you think about jquery automatic framework?
17:06 < vmx> brb
17:07 < lewiscot2> I agree, volker. It is not important to know how to write the code. We all know what should happen with each event. Example: I select some text and click the "link" button, a dialog should pop up, etc.
17:07 < lewiscot2> I will take a thorough look at the jquery test harness and let you guys know if it will work for our purposes.
17:08 < jfhv> OK, please keep in mind we use an iframe + designmode
17:08 < lewiscot2> I will.
17:08 < jfhv> This is not 'simple' JS
17:08 < lewiscot2> Understood
17:08 < jfhv> e.g. cursorToStart works in FF
17:08 < jfhv> not in Opera
17:09 < jfhv> although sharing the same code
17:09 < lewiscot2> Ok. When I test, I will try to identify the source of the problem(s). I will not try to fix the code, just track down the bugs.
17:09 < jfhv> sorry, phone call
17:09 < lewiscot2> I will let the author of the code fix the bugs.
17:09 < lewiscot2> Daniel, I have a cool jQuery plugin you might be interested in.
17:10 < danielr> yes ?
17:10 < Stellaris> Can I sneak in a quick question what editors you guys are using? Barebones or some IDE?
17:10 < danielr> personnaly, i use PSpad for all my coding needs
17:10 < danielr> + some jedit
17:10 < lewiscot2> I use a text editor (TextWrangler)
17:11 < lewiscot2> Sometimes vi (linux text editing tool)
17:11 < Stellaris> I'm currently using Komodo Edit and looking at Aptana
17:11 < Stellaris> Yeah, heard of vi, Scott
17:11 < lewiscot2> Thorsten, do you use linux or windows?
17:11 < danielr> haven't heard of komodo
17:11 < Stellaris> Linux, I'm on Ubuntu Edgy
17:11 < lewiscot2> I've heard of komodo but never looked at it.
17:11 < lewiscot2> Cool. Linux rocks!
17:12 < Stellaris> Right on!
17:12 < vmx> i'm back
17:12 < jfhv> Yes, I agree.
17:12 < lewiscot2> Daniel, I'll send you an email about the plugin I mentioned.
17:12 < danielr> ok
17:12 < lewiscot2> I'm also working on a hacked version of thickbox that will allow you to add custom themes.
17:13 < danielr> cool
17:13 < jfhv> OK. Anything to add for 2) ?
17:13 < danielr> thickbox is very useful
17:13 < lewiscot2> JF, I have nothing to add. I will read up on testing and let everyone know what I find. It will probably be sometime this weekend.
17:14 < lewiscot2> Oh, I do have one thing to add:
17:14 < vmx> scott: about the testing: if you take at jquerys test harness keep in mind that you might add my query-fixes there too (for the iframes)
17:14 < lewiscot2> Since I am going to be in charge of testing, perhaps we should post on the site/forums that bugs should be reported to me.
17:15 < lewiscot2> Ok, I will do that Volker.
17:15 < jfhv> I think it's a good idea to use the Trac :)
17:15 < danielr> yes
17:15 < vmx> trac is really nice
17:15 < lewiscot2> Can users report bugs on the Trac?
17:15 < lewiscot2> I agree.
17:15 < vmx> espacially code tags (that could be 4))
17:16 < jfhv> yes, registered users, though
17:16 < jfhv> they can register online
17:16 < lewiscot2> Ok. I was thinking that the general public can report bugs to me and I can add tickets to trac.
17:16 < lewiscot2> But that may be too much work for one person.
17:16 < jfhv> well, we can also use wymeditor-user
17:17 < jfhv> this is a new mailing-list I've set up
17:17 < danielr> ?
17:17 < danielr> ok
17:17 < Stellaris> Having to register will possibly be a hindrance to some/a lot of people wanting to report bugs.. at least that's the way I personally tick
17:17 < jfhv> I'll announce it soon
17:17 < Stellaris> (as for the trac ticketing)
17:17 < jfhv> Thorsten: the problem is spam
17:17 < Stellaris> jfhv, I know, that's no excuse for forcing people to register though :)
17:17 < jfhv> yes, it is :)
17:18 < lewiscot2> LOL
17:18 < jfhv> I prefer to work on the project instead of fighting against spam
17:18 < lewiscot2> I agree.
17:18 < jfhv> (and I know this problem pretty well)
17:19 < Stellaris> Right, if it doesn't lower the number of useful tickets, I agree
17:19 < Stellaris> We'll see
17:19 < lewiscot2> I think Thorsten has a point about registering being a nuisance, but spam is a big problem.
17:19 < jfhv> see jQuery
17:19 < jfhv> you need to register
17:19 < vmx> what about (i forgot the name) the images with letters you need to type (to prevent spam)
17:19 < danielr> I think most user will subscribe to the forum without problem. So we could have a special discussion for bugs. Then we could create tickets on the trac when bugs are reported on the forum.
17:19 < Stellaris> Daniel, good point
17:19 < lewiscot2> That is probably a good compromise
17:20 < danielr> This again is a job that can be done by "newbies" like me
17:20 < Stellaris> vmx, you mean captchas, these have accessibility issues though
17:20 < danielr> vmx: captcha
17:20 < lewiscot2> It also means people don't need to go to two different places
17:20 < lewiscot2> Should we move on to #3
17:20 < lewiscot2> ?
17:20 < vmx> stellaris, sure, but i don't think a blind person or someone with a text edtior needs a wasiwym editor
17:20 < jfhv> sorry, but I prefer Trac for tickets
17:20 < danielr> there are accessible captchas, based on text
17:21 < lewiscot2> LOL!
17:21 < jfhv> Trac is a bug/issue tracking system
17:21 < jfhv> phpBB is a forum
17:21 < lewiscot2> Ok. Let's just use Trac.
17:21 < Stellaris> You've got a point there <blushes>
17:22 < jfhv> :)
17:22 < vmx> but bug forum and add ake tickets aut of them sould be also ok
17:22 < vmx> can perhaps trac and phpbb accounts be merged?
17:22 < jfhv> users can use whatever they prefer to submit bugs/issues/requests...
17:23 < jfhv> but it should be nice to use the Trac for tracking them
17:23 < vmx> or a different forum than phpbb, one that incooperates with trac?
17:23 < jfhv> I don't know such forum
17:23 < jfhv> in Trac you can comment tickets, too
17:23 < lewiscot2> Ok, I will write up some simple rules for submitting bugs and requests.
17:23 < vmx> yes, but onw will need 2 accounts
17:24 < jfhv> Some users prefer a forum
17:24 < jfhv> some other mailing-lists
17:24 < jfhv> some other Trac
17:24 < vmx> ok, can we go on to 3) (or can we even skip it)
17:24 < jfhv> the forum has RSS
17:24 < jfhv> the Trac, too
17:25 < jfhv> and you receive messages from the m-l
17:25 < jfhv> OK. I'll let Scott speak about 3) Safari
17:26 < lewiscot2> Sorry. I was getting coffee.
17:26 < lewiscot2> Ok, Safari:
17:26 < lewiscot2> Safari has a lot of issues that relate to 2 major flaws.
17:27 < lewiscot2> First, Safari has no native execCommand support for Link, Unlink, Lists or images.
17:27 < lewiscot2> Second, styleWithCSS = false, does not seem to work correctly in Safari.
17:28 < lewiscot2> The problem with the first issue is that custom event handlers will be needed for those functions.
17:28 < lewiscot2> This means that the SAPI is a must ...
17:28 < lewiscot2> plus, a virtual copy of the selection and it's parentNode will need to be maintained an manipulated.
17:29 < lewiscot2> The second issue means that the code will look like this:
17:29 < lewiscot2> <p><span style="font-weight: bold">Bold text</span></p>
17:29 < lewiscot2> The in-line styles will have to be cleaned up manually.
17:29 < lewiscot2> Both of these problems will require a lot of work.
17:30 < lewiscot2> I can do the work, but I think it is not realistic to get it done for the next release of Wymeditor.
17:30 < lewiscot2> I think it is more important to get a core set of functionality working really well in FF and MSIE at the very least and build from there in future versions.
17:31 < lewiscot2> It is also important to get the API for Wymeditor to a point where it is very easy for 3rd party devs to implement Wymeditor into their applications.
17:31 < vmx> will it be better to change the inline tags it on-the-fly or at the end on-submit?
17:32 < lewiscot2> I am proposing that we put Safari compatibility on hold until the next stable version of Wymeditor is released.
17:32 < lewiscot2> vmx: Probably on submit.
17:32 < lewiscot2> The cleanup could even be handled by PHP or ASP rather than JS.
17:33 < vmx> but then manipulating will get hard e.g things like parent()[0].nodeName=='B' won't work
17:33 < vmx> as it is a <span>
17:33 < lewiscot2> Hm, that's a good point.
17:33 < lewiscot2> Does anyone have a problem with putting Safari on hold until after the next stable release?
17:34 < vmx> wrapping that into an api would mean a lot work for all browser just to get it work wit safari
17:34 < lewiscot2> I know how JF feels about it. He would prefer to have a Safari version.
17:34 < vmx> i think safari should do the additional work, if it needed
17:34 < lewiscot2> I agree.
17:35 < lewiscot2> That just means the APIs will need to keep that in mind (i.e., be set up so their methods can be overloaded)
17:35 < vmx> doesn "putting on hold" mean you stop developing atm, or that you'll go on with coding, but know that it needs that long
17:36 < lewiscot2> I will continue working on some of the bigger problems. I am hoping to get in touch with the Safari dev team to get their advice.
17:37 < lewiscot2> Waiting to release a Safari version that truly works seems better than releasing a version that "kind of" works
17:37 < vmx> if the safari bugs will be fixed it is worth a wait imho
17:37 < lewiscot2> JF? Daniel?
17:38 < Stellaris> I'm not developing, but I second Scott's approach, my 2c
17:38 < vmx> with all the bugs it would be a waste of time
17:38 < danielr> i'm here
17:38 < lewiscot2> What is your opinion on Safari?
17:38 < danielr> i'm ok with that
17:38 < lewiscot2> JF, what do you think?
17:39 < jfhv> I guess there are more bugs in MSIE than in Safari :)
17:39 < jfhv> I'm kidding.
17:39 < lewiscot2> Perhaps, but that makes the case for waiting on Safari even stronger. MSIE = 85% of the browser market.
17:39 < lewiscot2> Safari = 2%
17:40 < jfhv> I know some WYMeditor users wait for a Safari version
17:40 < vmx> and i think the safari devs fix bugs, ie devs ...
17:40 < lewiscot2> create more bugs?
17:40 < lewiscot2> :)
17:42 < danielr> maybe the safari devs think nobody needs these features anyway... so maybe getting in touch with them will give them some motivation :)
17:42 < jfhv> Well, I just found  http://tinymce.moxiecode.com/punbb/viewtopic.php?id=666
17:42 < jfhv> It was in March 2005 :(
17:43 < vmx> may we go to 4) volker has some things to talk about
17:44 < jfhv> OK. Anything to add for 3) ?
17:44 < lewiscot2> Great link, JF.
17:44 < lewiscot2> So we are agreed that Safari will have to wait?
17:45 < jfhv> I'm OK about that :)
17:45 * vmx agrees
17:45 < danielr> yes, users with macs will need to use FF...
17:45 < danielr> not so terrible
17:46 < lewiscot2> Ok, so #4?
17:46 < jfhv> OK. About 4) ???
17:46 < jfhv> Volker?
17:46 < vmx> yes
17:46 < danielr> ok
17:46 < vmx> have you heard of "codetags"?
17:46 < lewiscot2> Not sure.
17:46 < vmx> i think they are quite useful, i use them all the time
17:46 < lewiscot2> What is it?
17:46 < vmx>  http://www.python.org/dev/peps/pep-0350/
17:47 < vmx> you write notes directly to the source
17:48 < vmx> there's also a nice plugin for trac  http://sopra.vmx.cx/cgi-bin/trac.cgi/codetags
17:48 < lewiscot2> Nice.
17:48 < vmx> this way you can easily what needs to be done
17:48 < vmx> without writing everything into extra files
17:49 < vmx> if you take a look at my branch i used them already
17:49 < lewiscot2> Makes sense to me.
17:49 < vmx> for a project at university i used them together in a team and it worked nice
17:50 < jfhv> isn't it redundant with trac tickets?
17:50 < vmx> i can post something in the wiki. i wrote some information for it for that project
17:50 < vmx> jfhv: think of tags like "note" or "ugly"
17:50 < vmx> you won't create a ticket for them
17:51 < vmx> hacking the codetag plugin for trac is quite easy (i've done that). we could even at things like:
17:51 < vmx> FIXME v.mische  trac:#234
17:51 < Stellaris> (I need to run folks, thanks for having me, see you next time)
17:51 < vmx> and if you take a look at the codetags page, you'll have a link to the ticket
17:52 < vmx> cu stellaris
17:52 < lewiscot2> Ok, see you Thorsten.
17:52 < danielr> Bye Thorsten :)
17:52 < lewiscot2> Thanks for attending.
17:52 < jfhv> bye bye :)
17:53 < lewiscot2> Why don't we do this: Volker, write up an explanation of codetags and how we would use them and submit it via the mailing-list.
17:53 < jfhv>  http://www.trac-hacks.org/wiki/CodeTagsPlugin
17:54 < vmx> i'd use a stripped down version, we don't need all the tags proposed in pep-350
17:54 < vmx> you could also take your tiem and think about it
17:54 < vmx> this has also time, it isn't a major issue, just an idea
17:55 < vmx> ok another thing, which is connected to it: the comment format
17:55 < lewiscot2> Do you want to write up a sub-set of the tags you think we should use and post it on the wiki?
17:55 < vmx> yes, i've written it already in german for another project
17:55 < vmx> so it'll be easy
17:55 < lewiscot2> can you translate it easily.
17:55 < lewiscot2> ?
17:56 < vmx> (and made my thoughts about it already)
17:56 < vmx> yes
17:56 < lewiscot2> Ich sprache eine kleine deutch ... but only very little.
17:57 < vmx> probably as good as i speak french
17:57 < lewiscot2> Probably worse :)
17:57 < vmx> about comments within the source: i don't like the javadoc commenting
17:58 < danielr> sorry, phone call brb
17:58 < lewiscot2> Volker, what do you suggest?
17:58 < vmx>  http://naturaldocs.org/
17:59 < vmx> i try to find the exampel page
17:59 < vmx>  http://naturaldocs.org/documenting/walkthrough.html
17:59 < lewiscot2> JF, While he is looking, I have a non-wymeditor question.
17:59 < lewiscot2> Never mind.
18:00 < jfhv> Explain.
18:00 < vmx> again, i haven't really take a closer look to it, but y first imperssion was "that loos much nicer"
18:01 < lewiscot2> I agree with Volker. It is a lot easier to learn and read than JavaDoc.
18:01 < vmx> if you take a look at jquery comments, it looks really ugly
18:02 < jfhv> Example here:  http://naturaldocs.org/documenting/walkthrough.html#OurFirstFunction
18:02 < lewiscot2> I vote "yes"
18:03 < jfhv> Yes, 'Natural' docs is a well-chosen name
18:03 < lewiscot2> agreed
18:03 < jfhv> I don't have any preference atm
18:04 < jfhv> What does jQuery uses?
18:04 < lewiscot2> JavaDoc
18:04 < jfhv> ok
18:04 < lewiscot2> NaturalDoc supports JavaDoc though (according to the website)
18:04 < vmx> the only thing i just see: it seems you can't define the type of a parameter
18:05 < lewiscot2> We could do something like this:
18:05 < lewiscot2> Parameter:
18:05 < lewiscot2> x (int) Description of the parameter
18:05 < vmx> yes, right
18:05 < jfhv> Perhaps it'd a problem if we don't use JavaDoc as jQuery, since WYMeditor will be a jQuery plugin...
18:07 < vmx> i think wymeditor won't ever be a "official" plugin
18:07 < vmx> so it should matter so much
18:07 < vmx> if is easier for us to use this code we should go our way
18:08 < jfhv> well, I think it's important for the project to gain the interest of jQuery users
18:08 < lewiscot2> I agree
18:09 < jfhv> I don't see any rule about code comments on their site, though
18:09 < lewiscot2> On that same note, how difficult would it be to make Wymeditor also compatible with Prototype?
18:09 < jfhv> 'compatible' ?
18:10 < danielr> scott: i think you can already use them side by side without any conflicts
18:10 < lewiscot2> Prototype is very similar to jquery. I think many of the plugins for jquery will work in Prototype. It would be really good if Wymeditor worked with both.
18:11 < jfhv> There's a function in jQuery for that: jQueryNoConflict
18:11 < jfhv> So you don't have the $ issue
18:11 < lewiscot2> Ah, very nice.
18:11 < jfhv> var $j = jQueryNoConflict()
18:11 < jfhv> I used it before switching back to 1.0.x
18:11 < jfhv> Now we can use it again
18:12 < LAV> Sorry for jumping in guys, I've been on the phone for 2 hours othervise I'd introduced myself sooner. I mailed JF today about this project, so that's how I found you. I'm Albin by the way :).
18:12 < jfhv> Welcome, Albin :)
18:12 < danielr> welcome Albin :)
18:13 < lewiscot2> Welcome Albin.
18:13 < LAV> I'm working on a project where we needed an XHTML2 editor, so I'm we decided for WYMeditor. Since the project is ongoing now, I'm naturally very interested in how the development of Wymeditor is going :)
18:13 < lewiscot2> Cool.
18:13 < lewiscot2> Where are you from?
18:13 < LAV> (we do some pre and postprocessing to get it to XHML2)
18:14 < LAV> I'm located in Stockholm, Sweden
18:14 < lewiscot2> Excellent.
18:14 < lewiscot2> JF, I need to go but I wanted to mention 2 things before I go.
18:15 < jfhv> I'll need to go soon too, sorry.
18:15 < vmx> sorryi was away, i'm back
18:15 < lewiscot2> Sorry, Albin. You got here just as we are about to finish. Can you come back next Wednesday at 15:00GMT?
18:15 < LAV> yes, sure :)
18:15 < lewiscot2> Great.
18:16 < LAV> I know. I was planning to be here at 1500, but I got hold up on the phone
18:16 < LAV> I thought I'd introduce myself at least :)
18:16 < lewiscot2> JF, I wanted to mention that I think we should start thinking about how to market Wymeditor.
18:16 < vmx> lav: do querys (private chats) work?
18:16 < jfhv> (Albin has already subscribed to the mailing-list)
18:16 < lewiscot2> Excellent. I'm glad you were able to introduce yourself.
18:17 < vmx> which comment system will we use? further discussion? decision?
18:17 < lewiscot2> Perhaps we should continue the discussion on commenting next week.
18:17 < vmx> ok
18:17 < lewiscot2> So we don't rush it.
18:17 < vmx> ok
18:17 < jfhv> ok
18:17 < danielr> ok :)
18:17 < jfhv> Albin ?
18:18 < vmx> and take a look at code tags pep-350
18:18 < lewiscot2> Ok.
18:18 < jfhv> ok
18:18 < LAV> yes?
18:18 < lewiscot2> JF or Daniel. I have an IE7 CSS question. Can one of you help me after we wrap up on Wymeditor?
18:18 < danielr> yes ?
18:18 < jfhv> yes?
18:18 < danielr> of course
18:19 < lewiscot2> Are we finished on Wymeditor?
18:19 < vmx> a last thing:
18:19 < vmx> the mit/gpl dual license makes no sense
18:19 < vmx> we just need mit
18:19 < jfhv> no
18:20 < jfhv> I had to add the GPL for Drupal
18:20 < jfhv> Drupal needs GPL code
18:20 < vmx> it can use it even if its mit
18:20 < jfhv> no, jQuery needed to add GPL for Drupal too
18:20 < vmx> mit is like (new) bsd, gpl can jsut use it
18:20 < danielr> vmx: it's a requirement coming from drupal :)
18:20 < vmx> ok i just take a look at it last night
18:20 < jfhv> MIT and GPL are very different
18:21 < vmx> so perhaps i missed something
18:21 < vmx> jfhv, yes but bsd is gpl compatible
18:21 < vmx> (but not the other way round)
18:22 < vmx> but it's ok for me to have mit/gpl :)
18:22 < jfhv> well, I won't go on that slippy ground :)
18:22 < vmx> i just thought it would be easier without gpl
18:22 < vmx> sure
18:22 < vmx> lav: have you doen any xhtml->xhtml2 work yet?
18:23 < LAV> vmx: i'm using Miranda as a IRC-client, so i'm not completely sure how it handles messages ;) but I tried to reply to your private message.. did you not get it?
18:23 < vmx> no i haven't
18:23 < vmx> have you registered your nickname at freenode?
18:24 < LAV> ahhh... yeah, it said something about that
18:24 < LAV> nope
18:24 < LAV> will do
18:24 < lewiscot2> JF & Daniel. I emailed my IE7 question to you both. I need to go now.
18:24 < lewiscot2> Bye everyone.
18:24 < danielr> ok
18:24 < danielr> bye scott
18:24 < jfhv> Sorry guys, I really need to go know. That was a very nice meeting, BTW. The logs will be on the dev wiki.
18:24 < jfhv> bye Scott :)
18:25 < lewiscot2> Ok. a bientot.
18:25 < LAV> anyway, the answer (if i'm not interupting something now) is: Yes, we built a SAX-parser in Python to transform from XHTML -> XHTML2. I'm afraid it's not very general though... We're using it for a specific task, to create legal documents and embed RDF-tags.
18:25 < jfhv> (I let my IRC client open)
18:25 < danielr> à demain, JF
18:25 < jfhv> bye bye :)
18:25 < LAV> Bye JFHV
18:26 < danielr> won't it be easier to use XSL ?
18:26 < vmx> cu jfhv
18:26 < danielr> to transform XHTML -> XHTML2
18:26 < vmx> it isn't that easy in xslt
18:26 < vmx> therfore i asked
18:26 < LAV> We're actually planning to use XSLT for the XHTML2->XHTML part
18:26 < vmx> i write a stylesheet for basic transformation
18:26 < vmx> cool
18:27 < vmx> perhaps we can incorporate
18:27 < LAV> for the specific structure we are going to use, this seemed like the best choise.
18:27 < vmx> i want to do it for my cms too
18:27 < LAV> it was really easy to do XHTML->XHML2 with SAX
18:27 < danielr> we often use a lot of XSL to post process WYMeditor generated XHTML, very useful
18:27 < danielr> ok
18:27 < vmx> the biggest problem is <br/> to <l></l>
18:28 < LAV> but it would be much harder the other way, due to how we must rearrange the tags to make it XHTML-compatible
18:28 < LAV> so, we'll use XSLT for that
18:28 < danielr> i see, grouping problem
18:28 < danielr> difficult also in xsl
18:28 < danielr> but feasible
18:29 < LAV> hehe, I think at the moment we're just removing all br-tags ;) but hey it's just a prototype.. or something :P
18:29 < danielr> IMO it's easier to transform <l></l> to <br/> than the other way around in XSL
18:30 < vmx> lav, will your project be open source?
18:30 < LAV> kind of, we hope
18:30 < LAV> it's a development project for a swedish government agency...
18:31 < LAV> so we're hopeing they will agree to make the whole thing open source...
18:31 < danielr> why did you choose tu use XHTML2 ?
18:32 < LAV> well, to make it "future-proof" i guess
18:32 < LAV> the decision was made last year, before I entered the project
18:33 < danielr> ok
18:33 < LAV> the other part of the project is to develop an RDF-editor, and a standard for how to tag swedish legal documents
18:33 < danielr> that's interesting :)
18:34 < LAV> (RDF = Resource Description Framework, in case anyone was wondering)
18:35 < danielr> Sorry, but I'll have to go too...
18:35 < LAV> so what we are modifying the editor to do right now, is to be able to tag specific parts of the text.. like to say, THIS text here is the title of the document, and this text here is the date it was publihsed. and then make it possible for people to search using this meta data.
18:35 < danielr> I think classes will be very useful for you
18:35 < vmx> lav, if you like i could give you the stylesheet, what i've done so far
18:36 < vmx> (though it isn't much)
18:36 < LAV> yes, thanks :)
18:36 < LAV> can never hurt! that was my task for tomorrow ;)
18:36 < vmx> at least in converts <br/> to <l></l>
18:37 < LAV> ah, your transformation was XHTML->XHTML2?
18:37 < vmx> yes
18:37 < vmx> xhtml2 is named "hdl"
18:38 < vmx> as it won't be pure xhtml2 but based on it
18:38 < LAV> not sure this'll work... the firewalls here can be pretty agressive
18:38 < vmx> so if your code will be open source i'd really like to work together with you on it
18:38 < vmx> it doesn't seem to work
18:38 < vmx> i can also mail it to you
18:38 < LAV> you can email it perhaps
18:38 < LAV> yep
18:39 < vmx> send the mail adress with /msg
18:39 < LAV> hehe,
18:39 < vmx> ok, we need to cut it from the meeting logs if jfhv doesn't ;)
18:40 < vmx> daniel: haven't you talked about valtech before (en francais)
18:40 < danielr> don't worry, he doesn't blindly copy/paste :)
18:40 < danielr> yes, with JF
18:40 < LAV> hehehe, ok :)
18:40 < LAV> yeah, Valtech is a french company
18:41 < vmx> so you are swede who talks french?
18:41 < danielr> JF said me he knows it for a long time
18:41 < LAV> hehehe, not a word ;)
18:41 < danielr> i discovered it today by the websites
18:41 < LAV> the swedish office is really an older swedish company that was bought by Valtech
18:41 < danielr> seems a pretty cool company :)
18:41 < LAV> so, the office here feels very swedish ;)
18:43 < LAV> hehehe, yeah it definately seems ok ;) i've only worked here for 2 months though
18:44 < vmx> ok, i've sent the mail
18:44 < LAV> thanks!
18:45 < danielr> I have to go
18:45 < danielr> bye :)
18:45 < vmx> so you'll be here on wednesday again? or perhaps we can stay in touch via mail
18:45 < LAV> bye
18:46 < LAV> I'll try to be here next wednesday..! shouldn't be any problems
18:46 < vmx> i've to go. i'm really interested in an xhtml->xhtml2 conversion with xslt (and back of course)
18:46 < LAV> :)
18:46 < vmx> working together is always much more fun (and less work)
18:47 < LAV> yep :)
18:47 < vmx> ok, cu then
18:47 < LAV> i'll begin with the XHTML2->XHTML part tomorrow :)
18:47 < LAV> yep, cya!
18:47 < vmx> that should be easier though