Community Meeting 2006-08-30-log

Session Start (worldwind chat:#worldwind-meeting): Wed Aug 30 21:56:54 2006 [21:56] *** Initial topic: Agenda - (Link: http://www.worldwindcentral.com/wiki/Community_Meeting_2006-08-30)http://www.worldwindcentral.com/wiki/Community_Meeting_2006-08-30 [21:56] *** #worldwind-meeting: stepman +Bull_[UK] @what_nick @adam_work @TheBean @ChanServ @Nowak [21:56] *** -ChanServ- [#worldwind-meeting] Welcome! Agenda: (Link: http://www.worldwindcentral.com/wiki/Community_Meeting_2006-02-01)http://www.worldwindcentral.com/wiki/Community_Meeting_2006-02-01 [21:56] *** #worldwind-meeting was created on Mon Aug 28 02:21:19 2006. [21:57] stepman: chanserv's greeting is a bit outdated :) [21:57] adam_work: someone going to ping Chris, Viggy and Patrick  [21:59] Bull_[UK]: well I emailed the dev list :p  [21:59] Bull_[UK]: watching eureaka bbl  [22:04] *** what_nick has signed off IRC (Read error: 113 (No route to host)).  [22:19] *** Vigilant has joined #worldwind-meeting.  [22:19] *** Mode change "+o Vigilant" for channel #worldwind-meeting by ChanServ.  [22:19] *** m_k has joined #worldwind-meeting.  [22:20] adam_work: hey randy  [22:22] *** withak has joined #worldwind-meeting.  [22:26] *** nht has joined #worldwind-meeting.  [22:26] nht: Hello  [22:26] Vigilant: yo  [22:26] nht: Hey, did you get a chance to look at the navigator thing?  [22:28] Vigilant: Still in the thinking stages.  [22:28] Vigilant: I made some graphics based on it though.  [22:28] *** T_servo has joined #worldwind-meeting.  [22:29] T_servo: made it  [22:29] Vigilant: One thing I was playing with was the idea of a pan control that resembles your typical console game pad. [22:29] *** TheBean is n=thebean@82-46-94-59.cable.ubr02.hawk.blueyonder.co.uk (TheBeansprout) [22:29] *** TheBean is on: @#worldwind-meeting #worldwind [22:29] *** TheBean is using irc.freenode.net (Link: http://freenode.net/)http://freenode.net/ [22:29] *** End of /WHOIS. [22:29] *** Vigilant is n=8fe8566f@worldwind/nasa/Vigilant ([8fe8566f] chat.worldwindcentral.com) [22:29] *** Vigilant is on: @#worldwind-meeting #worldwind [22:29] *** Vigilant is using irc.freenode.net (Link: http://freenode.net/)http://freenode.net/ [22:29] *** End of /WHOIS. [22:29] nht: Yah. [22:29] Nowak: T_servo: youre 1 hour late :D [22:29] Vigilant: I was wondering if I could / should make it sprite based or an actual 3d model [22:29] nht: Have you seen the GE beta one? [22:30] T_servo: um, 4:30 EST is what the schedule has :P [22:30] T_servo: unless the times have changed yet again [22:30] nht: Mmm...3-D model. That sounds nice but possibly a burden on the older machines. [22:30] Vigilant: I remember the GE one, it has a nifty way of tucking itself away. [22:31] Bull_[UK]: yup its now [22:31] nht: Yah, the smaller footprint is the nicest thing. I think a map with clicking is one of the fastest way for navigating long distances. [22:31] Vigilant: But I'm starting to get the idea that things aren't very intuitive to World Wind overall, so having some base line click control is still necessary. [22:31] withak: the GE one is just a compass until your mouse hits it [22:31] adam_work: also very useful on tablets with just a pen [22:31] Vigilant: Right now, a new itteration of the layer manager has been needing attention. [22:32] nht: Yes. [22:32] withak: steal the one from dapple [22:32] nht: I have the tree widget implemented and the forms implement scroll bars. [22:32] adam_work: are you going to run with the Dapple idea? [22:33] Vigilant: I'm not sure yet. [22:34] *** dumdumhead has joined #worldwind-meeting. [22:34] nht: Hi DDH [22:34] Bull_[UK]: moderator = T_servo ? [22:34] Bull_[UK]: ;) [22:34] T_servo: lol  [22:35] adam_work: people here at work really like the "All" and "Currently Showing" paradigm for Layers  [22:35] T_servo: what is the agenda url?  [22:35] adam_work: in Dapple  [22:35] T_servo: patrick coming as well?  [22:35] Bull_[UK]: (Link: http://www.worldwindcentral.com/wiki/Community_Meeting_2006-08-30)http://www.worldwindcentral.com/wiki/Community_Meeting_2006-08-30  [22:35] nht: (Link: http://www.worldwindcentral.com/wiki/Community_meeting_2006-08-30)http://www.worldwindcentral.com/wiki/Community_meeting_2006-08-30  [22:35] nht: what he said.  [22:36] dumdumhead: adam_work: i'll be adding that functionality into the layer manager, and the opacity adjuster  [22:36] dumdumhead: maybe some extras too  [22:36] adam_work: w00tage!  [22:36] dumdumhead: just be patient... =) [22:36] Bull_[UK]: mmmm [22:37] nht: ddh lets chat about widgets after this...do you have time? [22:37] T_servo: ok, ready to start? [22:38] adam_work: ready [22:38] dumdumhead: nht: i'll try [22:38] dumdumhead: nht: i should [22:38] T_servo: OLD BUSINESS: [22:38] T_servo:   1. What are the results of the various SoC projects? Is there any way to see the results? (withak) [22:39] T_servo: mine is done... but it is hard to test how well it works.. [22:39] nht: Shock finished his. We can demo but release will probably be under 1.4 unless I can get some license issue resolved. [22:39] T_servo: but at least we have a baseline code to work from [22:39] *** Selar has joined #worldwind-meeting. [22:39] adam_work: my student is MIA [22:39] Selar: MIA the Sri Lankan rapper?! [22:39] nht: T_servo is yours the ortho? [22:40] T_servo: UAV [22:40] adam_work: f0urtyfives student completed and integrated (and even change the implememntation at the last minute) :) [22:40] nht: K.  I'll chat with you after to try to test.  [22:40] T_servo: I have the code ... but his test data is somewhere on twobeds  [22:41] Bull_[UK]: 45's was the browser?  [22:41] adam_work: yes  [22:41] Bull_[UK]: cool  [22:41] *** Xenon3D1 has joined #worldwind-meeting.  [22:41] nht: How does it compare with the browser in 1.4/1.3.6?  [22:41] withak: who was the shapefile one?  [22:41] adam_work: what_nick's was the Shapefile and he is done as well  [22:42] adam_work: it is the same - I convinced him to use the ShDocVw control in 1.1 and got rid of HTMLEditor so it is functionally the same in 2.0  [22:42] Bull_[UK]: nht - I think it is tat one  [22:42] nht: Ah cool.  [22:42] dumdumhead: is it part of the kml importer yet?  [22:42] dumdumhead: for the balloons? [22:42] adam_work: in 1.1 I think [22:43] adam_work: he hasnt moved the code over to HEAD yet [22:43] dumdumhead: cuz we need that done, otherwise I'll use whatever method is easiest for me to implement [22:43] nht: The 1.4 one breaks when you enter a URL in the box btw...you end up losing keyboard control. [22:43] nht: And mousewheel... [22:43] withak: the keyboard focus is weird [22:43] adam_work: the problem with WebBrowser is it *cant* load HTML strings without jumping through a few COM hoops [22:44] adam_work: *WebBrowser in 1.1 and 2.0 [22:44] dumdumhead: adam_work: why is that important? [22:45] adam_work: cause the KML doesnt point to URL's they are HTML data in [!CDATA] chunks [22:45] Xenon3D1: KML sort of requires the ability to parse inline HTML for the balloons instead of just fetching from a URL. [22:45] dumdumhead: adam_work: ah...ic [22:45] Xenon3D1: @adam_work: what he said. [22:45] dumdumhead: next item? [22:45] adam_work: the images they point too are usually URL's but no the body [22:47] Xenon3D1: adam: I used this code to do it (in C++) in NatureVIew Express. (Link: http://www.codeproject.com/com/cwebpage.asp)http://www.codeproject.com/com/cwebpage.asp [22:47] * Bull_[UK] pokes T_servo [22:47] Vigilant: Dhur? [22:48] Bull_[UK]: NEW BUSINESS: [22:48] Bull_[UK]: 1. Will all the new code from 1.4 be backported to 1.3.6? (Bull) [22:48] Vigilant: No is my current understanding. [22:48] dumdumhead: we can do our best to port most things [22:49] Bull_[UK]: hehe [22:49] Bull_[UK]: 2 answers [22:49] Vigilant: They're compatible! [22:49] nht: What would be missing? I'm trying to decide which to make my trunk... [22:49] Vigilant: No to all, yes to some. [22:49] dumdumhead: 1.4 is where i commit and develop all new things [22:50] dumdumhead: i probably won't touch the 1.3.6 branch till release time, where perhaps i can do file diffs and port certain things that make sense [22:50] dumdumhead: as time goes on though, the 1.3.6 branch will lag [22:50] withak: maybe get people who will want to use 1.3.6 to make a list of what things might be most important [22:50] nht: Okay...so the converse is a question...has anyone checked in be capability in 1.3.6 without also sticking it in 1.4? [22:51] Bull_[UK]: but there is no real .net 2.0 specific stuff in 1.4 so far? [22:51] dumdumhead: there are a couple things [22:51] dumdumhead: but can be converted to 1.1 stuff for that branch with ease [22:51] adam_work: thats was my question too [22:51] dumdumhead: but, that'll change with time [22:51] stepman: well i've replaced a bunch of deprecated stuff with their 2.0 replacements [22:51] dumdumhead: as 1.4 starts to take more advantage of .net 2.0 stuff [22:52] dumdumhead: stepman: almost forgot about that stuff...yeah, it'll get complicated [22:52] stepman: those changes were fairly localized though [22:52] adam_work: does KMLImport have any 2.0 only stuff in it? [22:52] Bull_[UK]: well maybe adam can get TBOC to pay him to port to 1.3.6 [22:52] Bull_[UK]: ;) [22:53] adam_work: if we are going to really beef that piece up, that would be the one thing we might want to try and keep as portable for the next release  [22:53] dumdumhead: kmlimporter does have changed deprecated functions, but i think we can get it working in 1.1 with a little effort  [22:53] adam_work: ok  [22:53] nht: KML support will be big moving forward.  [22:53] nht: I sent patrick some info as to why...  [22:54] dumdumhead: nht: it's big for many people =) [22:54] adam_work: it is still an internal PI so someone could rip it out and fix it up [22:54] nht: True [22:54] adam_work: the Hurricane Tracker + Webcams is some cool KML :) [22:55] dumdumhead: next?  [22:56] Bull_[UK]: When is the beta release planned? (Bull)  [22:56] Bull_[UK]: roughly  [22:56] Bull_[UK]: ;) [22:57] Vigilant: when it's..... n/m [22:57] dumdumhead: the original plan was to get something out at the end of august (now) [22:57] Bull_[UK]: hehe [22:57] dumdumhead: but stepman has been a coding maniac, and i'd like to get his stuff into 1.4 [22:57] stepman: heh [22:57] stepman: see next question for more on that ;) [22:57] dumdumhead: and that gives me time to do a couple other things i haven't had time to get to yet  [22:57] dumdumhead: =) [22:58] dumdumhead: i do want to redo the layer manager before 1.4 goes out [22:58] Bull_[UK]: well 1.5 is due soon so hurry ;) [22:58] dumdumhead: for now, let's give the 1.4/1.3.6 release date a TBA  [22:58] dumdumhead: and see where we're at in 2 weeks  [22:58] T_servo: think we are up to this now: When is the beta release planned? (Bull)  [22:59] T_servo: ok, that is covered  [22:59] Bull_[UK]: k cool  [22:59] dumdumhead: TBD?  argh..whatever =b  [22:59] nht: I'd like to get the widget code integrated too.  [22:59] Bull_[UK]: lol  [22:59] T_servo: 3  What is the policy on adopting open JIRA issues? (step)   [22:59] dumdumhead: take it =) [22:59] stepman: heh [22:59] adam_work: go for it! [22:59] stepman: well somebody opened the issue, do i just fix it and mark it closed (assuming i can, which i can't right now) [22:59] T_servo: yea, but I think stepman was having problems assigning them to himself [23:00] stepman: or should i try to get the fix approved by whoever opened the issue? [23:00] adam_work: there is a small issue about grabbing new things someone is working on, but those have been rare [23:00] Bull_[UK]: yeah just dont touch wavy flags [23:00] dumdumhead: usually the person who opens an issue isn't the one who is going to fix it (unless it's me, who creates an issue 1 min before I commit a fix) [23:00] dumdumhead: =) [23:00] stepman: well i'm only touching stuff that is unassigned anyway  [23:00] withak: you can still post comments on an issue after it is marked closed, right?  [23:00] Bull_[UK]: yup  [23:01] T_servo: make sure when you commit the change... comment the jira issue  [23:01] dumdumhead: stepman: it's pretty much a free-for-all at the moment due to the lack of active devs  [23:01] stepman: yeah, and reopen if necessary  [23:01] Bull_[UK]: and reopen  [23:01] stepman: ok  [23:01] dumdumhead: there's usually not much toe-stepping  [23:01] stepman: so if i could please get the required access you'll see a few issues closed in the very near future ;) [23:01] withak: didn't jira used to send emails? [23:01] withak: like when there were new comments [23:02] adam_work: yes it did, IIRC [23:02] dumdumhead: yeah, i need to fix the mail server [23:02] dumdumhead: i should make a jira issue for that =b [23:02] nht: heh, i should get my JIRA thing fixed too. refuses to send me a new password. [23:03] stepman: heh [23:03] withak: make an issue [23:03] nht: Haha [23:04] nht: bull make an issue for me :) [23:05] Bull_[UK]: lol  [23:05] *** what_nick has joined #worldwind-meeting.  [23:05] what_nick: morn all  [23:05] Bull_[UK]: hi wn  [23:06] Bull_[UK]: How much refactoring and new work should be targeted for 1.4? What are the priorities for this release? The download infrastructure is in desparate need of an overhaul (see [confluence ((Link: http://issues.worldwind.arc.nasa.gov/confluence/display/WWNET/Download+Infrastructure)http://issues.worldwind.arc.nasa.gov/confluence/display/WWNET/Download+Infrastructure)]), should we try to get this in 1.4? (step)  [23:06] stepman: okay so i just need to know who to pester to look into the JIRA permissions stuff for me.. ddh?  [23:06] stepman: ah well  [23:06] dumdumhead: stepman: i'm an admin, so just let me know what u need  [23:07] what_nick: stepman: send him emails he responds best to those  [23:07] stepman: that download stuff is fairly tricky... the ironic thing is that the call to ultimately *start* the download is blocking anyway, so attempting parallel dl's in several threads is pretty much useless [23:07] what_nick: adamhill is also admin [23:07] adam_work: ?? [23:07] adam_work: cool [23:08] dumdumhead: i would say, if you can complete the refactoring within 2-3 weeks, then ur safe to do it [23:08] adam_work: stepman++ # for actually mapping out what is going on! [23:08] dumdumhead: just be careful of public api changes that might affect plugins [23:08] stepman: ddh: i dunno for sure, from what i've learned in the JIRA docs i am missing some assignable perms or so [23:08] stepman: yeah that is exactly the point [23:09] stepman: so my proposal would be to refactor into a separate class [23:09] stepman: change all internal stuff to use that class [23:09] dumdumhead: well, if u can change the public api (i did for 1.4), just make sure to change all the plugins that are in svn so it works =) [23:09] stepman: and keep the existing stuff for plugins for the time being  [23:09] stepman: perhaps mark them as deprecated so plugins compile with warnings  [23:10] what_nick: so u mean duplication and slower transition  [23:10] nht: How many plugins are currently using widget code?  I know T_Servo's is right?  [23:10] what_nick: ya that would work  [23:10] dumdumhead: stepman: sounds good  [23:10] stepman: okay  [23:10] nht: I've been hesitant to modify any of that.  [23:10] dumdumhead: i would like to finally get the downloader stuff working nicely  [23:10] stepman: that'll keep me occupied for a little while.  [23:10] dumdumhead: and if ur willing to do it, i'm all for it =D  [23:11] what_nick: re WFS stuff  [23:11] T_servo: widget code? the EG plug-ins don;t use that [23:11] T_servo: EQ [23:11] what_nick: i would love a little diagram on the placename logic [23:11] T_servo: not yet at least [23:11] what_nick: i have most of it figured [23:11] nht: Oh...hmm...I thought the legends stuff used that. [23:12] stepman: okay [23:12] Bull_[UK]: yep legends do [23:12] what_nick: and qgis folk one of whom is from brazil is putting the brazil names in a wfs [23:12] stepman: that'll probably break SVN head from time to time though. [23:12] Bull_[UK]: nht-they use dh's widget code [23:12] dumdumhead: WFS is definitely on our list, just haven't found time to write the client side code yet [23:13] nht: I know...if I change the interface underlying the widget code then potentially some derived class will break. [23:13] dumdumhead: using a copy of tiledplacenameset as a base and modifying it is definitely the easiest approach i think [23:13] what_nick: dumdumhead: i am doing that :) [23:13] dumdumhead: what_nick: yeah, i noticed =) [23:14] what_nick: i wish i hadf a whiteboard [23:14] nht: When I say widget code I really mean dh's widget code with my widget code added on top but my underlying structure is slightly different. Ehnough to break things. [23:14] what_nick: i could show u the proposed flow now that i have a tablet [23:15] dumdumhead: nht: i've been doing a lot with my widget code lately, how much different are the 2? [23:15] dumdumhead: otherwise, we can have 2 separate code paths [23:15] what_nick: nht: ddh they need merge [23:15] what_nick: yep [23:15] nht: I'd have to look at your current code closely but two code paths kinda sucks. i'd rather go through the pain of a merge now. [23:16] stepman: good idea.. merging will not become easier over time :) [23:17] nht: Heck, I'd rather throw my stuff away and just hack your stuff if we can't come together but it would be annoying given the paperwork. :) [23:17] dumdumhead: hehe [23:17] dumdumhead: i will take a hard look at it today [23:17] nht: Mkay, I can look with you if you like. [23:18] *** m_k has signed off IRC. [23:18] dumdumhead: next item? [23:19] T_servo: Lots of stuff is not documented (not only source code but also features like HigherResolutionSubsets). Its great extensibility is one of the unique selling points of WW, but it must be well documented to be useful and accepted. This may be more important than adding new features. (step) [23:19] nht: ddh are all your changes in SVN? [23:19] nht: oops. [23:19] stepman: sorry, got carried away writing that ;) [23:19] dumdumhead: nht: most yeah  [23:19] nht: K.  [23:19] dumdumhead: documentation has never been a strong point of the project  [23:19] dumdumhead: =b  [23:19] dumdumhead: but it is important... =) [23:19] stepman: so i've noticed :) [23:20] T_servo: thought you had group to do that now?  [23:20] Vigilant: "ow"  [23:20] stepman: yeah it's hard to argue to people that WW's advantage over GE is its extensibility when you can't point them to the documentatio  [23:20] T_servo: or is randy slacking off again? ;) [23:20] dumdumhead: i would like to solve this by starting a "World Wind Toolkit" of sorts...to help create the content (and expose it) [23:20] nht: You can point them at source which you can't do with GE. :) [23:21] stepman: nht: yeah right  [23:21] T_servo: adam just points them to the EULA ;) [23:21] Vigilant: It's still serious though. [23:21] stepman: nht: tried to read the source lately? :) [23:21] nht: My part is well documented.  I think. :) [23:21] adam_work: ...that you are about to throw away... :) [23:21] nht: Yah well...  [23:22] nht: Can't have everything.  I'll document DDHs widget code as I go along...that it suspiciously ends up looking like mine is coincidence...  [23:22] stepman: yeah well  [23:23] adam_work: maybe we could sell a WW Hacks book to ORA after all  [23:23] nht: documentation is the weakest link in most open source projects.  Plus there's been a lot of change from 1.3.3.1 to now.  [23:23] withak: it might  be outdated before it got printed :) [23:24] nht: ddh what would go in the WW Tollkit? [23:24] T_servo: a lot of info is also scattered over the wiki [23:24] * Bull_[UK] waits for the FEF section to start and notices no 45 :P [23:24] stepman: well there are two aspects of developer documentation.. the one for 'external' users (add-on and plugin developers) [23:24] stepman: and one for those working on the core. [23:24] dumdumhead: the World Wind Toolkit would be a mix of tools used to create content for World Wind [23:25] nht: that would fit the external users well I think [23:25] stepman: i'd say the latter can wait, because there appears to be only a small group working on it anyway... [23:25] dumdumhead: and this would include a content developer's manual of sorts as a guide to what can be created and used [23:25] stepman: ddh: that would be great [23:25] adam_work: we need to get Nowak to put the VS compilable version of dstile up [23:25] stepman: my favourite example are HigherResolutionSubsets.. if it wasn't for the source i wouldn't even know they existed [23:25] dumdumhead: the toolkit might also include server programs for content deployment [23:26] Bull_[UK]: nice [23:26] nht: On the subject...does someone have a HOWTO to stand up a WW server? I have this disk from 45 but no instructions on what to do with it (no funny comments). [23:26] Bull_[UK]: lol [23:26] adam_work: i have a nice Just In Time tiler in .NET to donate [23:26] dumdumhead: hehe, I would suggest using geoserver with a caching wrapper [23:26] adam_work: there is none, nht [23:26] adam_work: but it is pretty simple [23:26] nht: Mgreat. Someone teach me how and I'll write a HOWTO. [23:27] dumdumhead: nht: geoserver/mapserver + cache wrapper is the best choice [23:27] adam_work: you need the PHP script and the files [23:27] dumdumhead: makes the data more useable as well (for other tools) [23:27] adam_work: that is way too slow for Landsat though [23:27] dumdumhead: depends on how it's setup [23:28] adam_work: well once the cache is primed it is HTTP [23:28] adam_work: this is what we use at TBOC now [23:28] nht: Yah...I have the landsat data I need to put into a closed network. [23:28] dumdumhead: i do want to push WMS more though [23:28] dumdumhead: and try to step away from encouraging the tile delivery [23:28] nht: I can try to write up both ways but someone will have to walk me through both ways using small words. [23:28] dumdumhead: allows the WW client to use the data in different ways [23:29] nht: Will we add streaming support a la google? [23:29] adam_work: ooh, JP2! ECWP: ! [23:29] dumdumhead: u can add whatever u want =b [23:30] adam_work: well licenses are an issue for stuff in /trunk [23:30] nht: Yah. [23:30] adam_work: JP2 has none, but I bet ECW does [23:30] dumdumhead: so to sum up this last item, we should put a focus on centralizing a place to develop a World Wind Toolkit and Content Developer's Manual [23:31] nht: Is there an advantage or disadvantage to JP2/ECW vs tiling? [23:31] adam_work: speed [23:32] adam_work: any processing you do when 10K users are requesting tiles will kill most normally provisioned servers [23:33] what_nick: adam_work: yep [23:33] what_nick: ecwp is faster [23:33] what_nick: jp2 would be my choice google uses it [23:33] what_nick: they use a commercial jp2 server from itt vis [23:35] what_nick: (Link: http://issues.worldwind.arc.nasa.gov/confluence/display/SRVR/Home)http://issues.worldwind.arc.nasa.gov/confluence/display/SRVR/Home [23:36] nht: What does ESRI do for their streaming server? [23:36] what_nick: it uses ermapper ecwp imho [23:37] nht: So we can avoid Google hassles if we go ecwp? [23:37] adam_work: ?? [23:37] adam_work: ECW doesnt have an OS license IIRC [23:37] nht: Ah...okay. [23:37] Xenon3D1: ECWP client does, server does not. [23:38] what_nick: adam_work: it does now [23:38] nht: Whew. [23:38] what_nick: they os'ed gpled the sdk [23:38] Xenon3D1: But no O/S server. [23:38] nht: GPL wont work well for us though. [23:38] adam_work: I thought it was lizard tech that made an OS license not ERMapper? [23:39] what_nick: checking [23:39] Xenon3D1: ERM does have a GPL ECW SDK. [23:39] adam_work: there is an OS Kakadu license (JPG2000) [23:39] Nowak: adam_work: lizardtech is mrsid [23:39] adam_work: if it is a lib wouldnt it have to be LGPL? [23:40] adam_work: since you get no source? [23:40] Nowak: adam_work: you get full sources for libecw [23:40] Xenon3D1: Kakadu is not O/S to my knowledge. [23:40] nht: No...there are GPL libraries. It forces you to be GPL to use them. [23:40] adam_work: so we are hosed [23:41] Xenon3D1: ECW SDK requires GPL licensing, no other F/OSS licenses permitted. [23:41] adam_work: unless we can convince them to LGPL them too [23:41] nht: Hence...GPL is viral and mostly hoses other open source developers. [23:41] Xenon3D1: ECW SDK permits read-only access at no charge for non-GPL software (even commercial). [23:41] adam_work: could some build a plugin? [23:41] adam_work: *someone [23:42] Xenon3D1: WW only needs to read ECW, not write. [23:42] adam_work: you are releasing all the source [23:42] Nowak: Xenon3D1: afair you can write but upto 500mb [23:42] Xenon3D1: True, but 500Mb write doesn't go far enough to be worth considering in this discussion. ;) [23:43] Xenon3D1: Effectively it's a read-only license for anything but GPL software.  [23:43] adam_work: we could just do it as plugin and wait for Eblen to come knocking at our door :) [23:43] nht: Plugins are iffy... [23:43] Xenon3D1: But why would WW even need ECW write ability anyway? [23:43] nht: Heh...well that would likely be bad publicity when FSF kicks another OS developer around... [23:44] nht: Just ECW read. [23:44] adam_work: we have Nelson, he can take Eblen :) [23:44] nht: IMHO  [23:45] adam_work: wait ... he's a Quaker... nm  [23:45] what_nick: ecw read and streaming is my choice  [23:45] adam_work: but how do we do the server side?  [23:46] what_nick: +1 that Xenon3D1 does it ..  [23:46] stepman: hmmm  [23:46] stepman: there are three licenses for the ERM SDK  [23:46] nht: You don't do the server side.  It allows WW to hit ESRI servers.  [23:46] Xenon3D1: Server side could be fully GPL licensed.  [23:46] what_nick: the sdk has a server too  [23:46] nht: That too.  [23:46] adam_work: which servers?  [23:46] adam_work: Image Server?  [23:46] nht: Presuming that the ESRI streaming servers are using ECW  [23:46] Nowak: Xenon3D1: but libecw has only client side ...  [23:47] what_nick: i know they do :) [23:47] Xenon3D1: @Nowak: Yeah, I think you're right. It has ECW write, but not ECWP serve capability. [23:47] what_nick: btn is there an agx demo for download [23:48] what_nick: (Link: http://issues.worldwind.arc.nasa.gov/browse/WW-195)http://issues.worldwind.arc.nasa.gov/browse/WW-195 [23:48] Xenon3D1: And I think the GPL licensed version of the ECW SDK prohibits using it to make server software, IIRC. [23:48] what_nick: and there are other servers to  [23:48] adam_work: what_nick:  URL? [23:48] what_nick: australia NSW spot imagery 2.5M colour is on a ecwp server [23:49] what_nick: (Link: http://ermapper.com/downloads/download_view.aspx?PRODUCT_VERSION_ID=288)http://ermapper.com/downloads/download_view.aspx?PRODUCT_VERSION_ID=288 [23:49] nht: lets just say a major imagery provider is making stuff available via GE EntServer and ArcGIS Server. If worldwind can't play in that environment then its google earth's market by default (well okay ESRI too but they seem the underdog to me at the moment). [23:50] what_nick: # no more restrictions on the distribution of 'Server Software' under the GPL-style Public Use License Agreement [23:50] adam_work: what_nick: URL for the AGX demo [23:50] * adam_work nods at nht [23:51] stepman: what_nick: still limited to 500M per image compression for a reasonable license [23:51] what_nick: adam_work: i was asking u ... [23:51] what_nick: is there one ? [23:51] adam_work: no, it is still closed beta [23:51] what_nick: but 500Mb for output [23:51] what_nick: we are not out putting anything ... [23:51] adam_work: haha [23:51] Xenon3D1: Actually, 500MB _INPUT_. [23:51] adam_work: I know how to get around the GPL problem [23:52] what_nick: adam_work: they should stop tantalizing us with screenshots :) [23:52] Xenon3D1: 500Mb input size during compression, regardless of final compressed size.  [23:52] nht: adam_work buy a commercial license? :) [23:52] what_nick: that's restrictive [23:52] what_nick: yep if apogee uses it we will buy a licence [23:52] Nowak: Xenon3D1: thats still enough [23:52] what_nick: i will make a note of it ... [23:52] Nowak: Xenon3D1: you would get like 50MB ecw files for reasonable quality [23:53] what_nick: hmm ok looks good [23:53] Nowak: Xenon3D1: and you can tile from that [23:53] what_nick: u could have ur data set tiled in multiple 500mb ecw's then stream from that [23:53] adam_work: stepman: if you refactor ImageProviders make the Stream abstract, them we can implement a WCF or a TCP/IP pipe that talks to an external GPL'ed app that will fetch tile requests :) [23:53] what_nick: :) [23:53] adam_work: voila1 [23:54] adam_work: *voila! [23:54] stepman: how does GDAL support ECW? [23:54] adam_work: its GPLed [23:54] stepman: adam: easy [23:54] what_nick: via the OS sdk ... [23:54] adam_work: LGPLed [23:54] Xenon3D1: adam: You could do that. Tiles of JP2s seem cumbersome and miss part of the advantage of JP2/ECWP. [23:54] stepman: oh. because GDAL itself has an MIT style license i think [23:54] Nowak: adam_work: but you cant go from GPL to LGPL ? [23:55] what_nick: and frankw told me there is commented out code in gdal cvs to support the ecwp url's as well [23:55] Nowak: adam_work: cause now we could use gdal as an interface to libecw ... [23:55] adam_work: see.. making an abstract pipe to stream data fixes everything :) [23:55] Xenon3D1: step/adam: GDAL uses ECW in read-only mode, I believe.   [23:55] Xenon3D1: Err, no:  [23:55] Xenon3D1: (Link: http://lists.maptools.org/pipermail/gdal-dev/2005-May/005683.html)http://lists.maptools.org/pipermail/gdal-dev/2005-May/005683.html  [23:56] what_nick: ok but generally figuring licences techically it's doable ,,,,  [23:56] what_nick: next issue please ,,,  [23:56] nht: Sorry for the diversion. :) [23:56] what_nick: i would put xenon + adam + step in charge of ecwp [23:56] what_nick: issue WW-195 for u [23:57] Xenon3D1: what_nick: Hey, I gots my own 3d landscape tool I gotta code for a living. [23:57] Xenon3D1: Doubt I could put that much work into WW ECW/JP2. [23:58] adam_work: no, it was a good one, figuring out how to add restrictive protocols was a nice exercise [23:58] what_nick: Xenon3D1: u could release 3d nature world [23:58] what_nick: it will beat anything from leica or any other weaker suites [23:58] Nowak: adam_work: actually, you dont have to add any new code to world wind [23:58] Xenon3D1: We can discuss the distinctions some other time. [23:59] Nowak: adam_work: just xml with localhost server [23:59] adam_work: that too - yea Cassini! [23:59] what_nick: i would leave coding to step who is in "flow" [23:59] what_nick: Xenon3D1: and adam_work can advise [23:59] stepman: hey now :) [00:00] what_nick: it's step's time of the month :D  [00:00] stepman: i think i'll have plenty of work refactoring the download stuff as is  [00:00] dumdumhead: i would say that ECW/JP2 support, while cool, has very little practical use in regards to the already efficient alternatives that are available  [00:00] what_nick: oh well i will do it then  [00:00] stepman: and after that i've got a few upcoming paper deadlines   [00:00] dumdumhead: unless someone wanted to pay for the development of it, it's probably not really worth the effort  [00:01] what_nick: as soon as i have convinced my boss no i dont want to do too much sales work  [00:01] dumdumhead: for me, it wouldn't be good since i'd have to make sure it works with our upcoming java version  [00:01] what_nick: oh oh pay apogee :) [00:01] what_nick: well it will be a bummer in java for sure [00:02] dumdumhead: we don't pay for anything =b [00:02] what_nick: i know [00:02] what_nick: whoever wants it anyway besides the point [00:03] what_nick: how do make a tile gradually sharpen up as the stream flows ? [00:03] Xenon3D1: I kind of agree with Chris -- it's cool, but what does it get you that's a real improvement over what you have now? [00:04] nht: Only interoperability with ESRI and perhaps GE. [00:04] adam_work: interop with other geospatial servers [00:04] what_nick: and a foothold in serious market [00:04] adam_work: bingo [00:04] what_nick: yep [00:05] adam_work: the big boys use ESRI and now days GEE [00:05] nht: Its a defensive move really. [00:05] adam_work: like supporting WFS [00:05] adam_work: GE is feekling the pressure the other way for WMS support [00:05] Xenon3D1: Are you 100% sure that simply by supporting ECWP you're going to be interoperable? [00:06] nht: Which they have for their ent customers. [00:06] what_nick: we would need jp2 and ecwp [00:06] adam_work: but that move has a 'tragedy of the commons" issue [00:06] dumdumhead: when firefox starts supporting jp2 and ecw, maybe i'll be more interested  [00:06] what_nick: jp2 is a superset  [00:06] adam_work: Xenon: no, but I have access to TBOC's internal ECWP: datastores  [00:06] what_nick: and ge uses a particular blend ... which they tested and found to be most efficient  [00:07] dumdumhead: but tom g. is having a hard enough time getting efficient data loaders for the "standard" image formats =b  [00:07] nht: ddh talk to patrick.  If he doesn't feel its an issue then its more than cool by me.  [00:07] what_nick: this is frim the ppl who set up the itt ias servers for ge  [00:07] adam_work: dumdum: they do  [00:08] adam_work: this is how TBOC people access the internal ECWP: data, via plugins that let them save the data out  [00:08] what_nick: dumdumhead: have they looked at ww2d+1 which works ...  [00:08] what_nick: not the very best but works [00:08] what_nick: yes ermapper makes a ffox plugin for ecwp [00:09] dumdumhead: anyways, it hasn't really been something people REALLY want during our meetings, demos and conferences [00:09] dumdumhead: there are like 50 other things that need more attention [00:09] what_nick: please check the nsw imagery site in australia [00:09] adam_work: agreed [00:09] Vigilant: *1000 [00:09] what_nick: agreed [00:09] nht: no argument...oh hey...I may have a line on some bathy for you. [00:09] what_nick: i do want to move to the next thing on the agenda [00:09] Nowak: dumdumhead: i doubt they need vs3.0 flags too ... ;) [00:09] nht: We'll see if its usable.  [00:09] adam_work: if step gets Image* refactored well, we can add codecs at will  [00:10] adam_work: they are all URL + BBOX requests in the end  [00:10] dumdumhead: the flags plugin is a case study for new visualization techniques  [00:10] dumdumhead: and has caused extended work for our d3d based gui  [00:11] stepman: oh yeah i was going to write a water shader (pseudo ocean waves) as another sample for the effects / multitexturing stuff... haven't gotten around to do that yet  [00:11] dumdumhead: the flags plugin is also something to show how we are differnet from GE and others  [00:11] stepman: totally useless but eye candy   [00:11] dumdumhead: stepman: water =)  just don't use the elevation data to determine the coastlines...tried that already =b [00:12] adam_work: flags are cool, but give me some moving arrows for showing flow and I will be more impressed :) [00:12] dumdumhead: i tried something simple a few weeks back  [00:12] stepman: ddh: was more thinking about map hue  [00:13] *** withak has signed off IRC ("NOOOoooooo......").  [00:13] dumdumhead: the could work  [00:13] dumdumhead: i really wanted realistic waves and water in 1.4 =) [00:13] stepman: adam: something like this? (Link: http://www.vrvis.at/scivis/laramee/isa-streamsurface/)http://www.vrvis.at/scivis/laramee/isa-streamsurface/ [00:13] dumdumhead: and volumetric clouds based on real sat data =) [00:13] adam_work: but we got sun and scattering!  [00:13] adam_work: that is pretty col  [00:14] adam_work: *cool  [00:15] nht: ddh we have some algorithms that try to comput cloud base based on sat imagery.  If you're interested I can direct you to the POCs but I dunno how available they are.  [00:15] nht: (ie I dunno who owns the code)  [00:15] adam_work: nht: see your /msgs  [00:17] *** nht is n=80f4ca3f@madmappers.freeearthfoundation.com ([80f4ca3f] chat.worldwindcentral.com)  [00:17] *** nht is on: #worldwind-meeting #worldwind   [00:17] *** nht is using irc.freenode.net (Link: http://freenode.net/)http://freenode.net/  [00:17] *** End of /WHOIS.  [00:17] *** adam_work is now known as adamhill_work.  [00:18] nht: adamhill_work ?  [00:21] dumdumhead: did u guys want to do the FEF section?  [00:23] Xenon3D1: I have something to announce, while we're all here. [00:25] Xenon3D1: Frank Weed has just finished a .x object exporter for our Scene Express tool, that makes output that the landmarks plugin seems to read and place nicely. So, we're closer to a fully functional authoring pipeline for WW from VNS. [00:25] stepman: cool. [00:26] stepman: talking about landmarks, i've got a nice stress test. [00:26] stepman: Graz (Austria) and Philadelphia, reconstructed from aerial images... [00:27] Xenon3D1: Next we'll try to figure out how to incorporate the landmarks-type functionality into an add-on data layer. [00:27] stepman: GE barely manages to load them (and dies in 80% of the attempts) [00:27] Xenon3D1: Step: Cool. I can test it out when we get the exporter further along. [00:28] stepman: looks approximately like this -> (Link: http://tinyurl.com/77vd5)http://tinyurl.com/77vd5 [00:30] *** dumdumhead has signed off IRC ("CGI:IRC (EOF)"). [00:31] Xenon3D1: Sounds like the meeting is over, then. [00:31] Vigilant: vootie [00:32] stepman: i guess the FEF stuff will have to wait until next time, then [00:32] Xenon3D1: Are there FEF people here? [00:33] stepman: dunno really [00:33] * stepman pokes bull [00:33] Nowak: im fef ... in theory [00:35] Xenon3D1: Enough to go on with the meeting or not? [00:36] *** what_nick_ has joined #worldwind-meeting. [00:36] Nowak: should be enough [00:37] *** what_nick has signed off IRC (Nick collision from services.). [00:37] *** what_nick_ is now known as what_nick. [00:37] Nowak: ... if they werent away [00:37] stepman: FEF BUSINESS: [00:37] stepman: 1. Who is setting up the new WWC server? (Bull) [00:38] *** Selar has left #worldwind-meeting. [00:38] what_nick: yep the brazil data is good thing to show off [00:38] Nowak: what_nick: i talked with them ... [00:39] Nowak: but i dont like the response ;)  [00:40] *** withak has joined #worldwind-meeting.  [00:40] what_nick: them = ?  [00:40] Nowak: multispectral ...  [00:48] stepman: hmmm  [00:48] stepman: next topic?  [00:49] Nowak: stepman: dont you see everyone is gone ...  [00:50] stepman: so? other topic postponed until next time? :) [00:50] what_nick: another dawn over australia [00:51] what_nick: poof well then i am gone too [00:51] what_nick: :) [00:51] Nowak: stepman: ask bull  [00:51] *** Vigilant has signed off IRC ("CGI:IRC (EOF)").  [00:51] stepman: yeah me too  [00:51] *** what_nick has left #worldwind-meeting.  [00:51] *** Xenon3D1 has left #worldwind-meeting.  [00:52] stepman: bull has been idle for 85 minutes... i guess he fell asleep :) Session Close (#worldwind-meeting): Thu Aug 31 01:01:56 2006