Latest revision as of 12:22, 31 October 2008

Musings on the next OpenOffice.org

This page is dedicated to foster ideas about the next major version(s) of OpenOffice.org. It should be seen as a way to discuss and outline some key feature of what we wish OpenOffice.org could be in two, three or five years. As such, this page covers product marketing, but also technological considerations and overall project strategy.

1. General Wish List

(Random items be written below)

• Give OOo "online capabilities"
  • Have OOo sync documents with webdav service
  • Publish OOo directly to various filesystem
  • Send documents to IM (kopete/pidgin)
  • Document sharing through XMPP (or others)
  • Enhance Blogging and CMS capabilities
  • Use OOo as a framework for AJAX webapps (CRM, ERP, HRS)
• Proposing online services on the OOo website
  • Mailing list Subscription through OOo Help menu
  • Support through Jabber chat (watch www.sameplace.cc)
  • RSS reader for forums (keep track of questions)
• Mind maps
  • Use draw as an interface to mindmap a paper (link document to shapes)
• Document builder framework
  • Generate and manipulate complex documents through a database
  • Database backend integration support
• JavaScript-based scripting support (allow easy scripting by web developers)
• User Interface able to be customized with skins

2. Product Positionning

3. Technological/Development Goals

4. IRC sessions / Work items

IRC session on the 21st of July, Monday, 2008. 5 pm CET, 4pm UK time, 11am ET (NYC, Boston, DC, Atlanta...) IRC: Freenode, #odf@www

Meeting minutes can be found below.

Next IRC session: 19th of September, 2008, 5 pm Hamburg time, (same time).

Work items: create a concept document. The concept is called Pinneberg. Kr investigates on the server requirements for a wiki integrated with OOo + resources, etc. The document will first start on a wiki, then text will be collated, presented and if possible will give birth to specification documents kr continues working on the odf@www as an extension.

5. Meeting minutes

(16:43:37) dmpop-afk est désormais connu sous le nom de dmpop

(16:47:38) paveljanik [n=Pavel@unaffiliated/paveljanik] a rejoint le salon.

(16:47:47) paveljanik: Hi

(16:47:52) dtardon a quitté le salon (quit: "Leaving")

(16:53:03) kr a quitté le salon (quit: Remote closed the connection)

(16:54:07) thorsten [n=me@pD953999D.dip0.t-ipconnect.de] a rejoint le salon.

(16:55:16) kr [n=Kay@sd-socks-197.staroffice.de] a rejoint le salon.

(16:56:27) chs [n=chs@30.245.72-86.rev.gaoland.net] a rejoint le salon.

(17:01:20) chs: Hello everybody

(17:01:26) chs: Hello Kay

(17:01:34) chs: should we start now?

(17:02:15) chs a chang√© le sujet en : OOo_Next

(17:02:33) kr: my watch is 17:04 - so let's start

(17:02:39) chs: yup

(17:02:58) chs: kr: you want to say some first introductory words?

(17:03:12) kr: just wanted to discuss that :-)

(17:03:18) kr: so, yes ...

(17:03:29) mmeeks [n=michael@mmeeks.gotadsl.co.uk] a rejoint le salon.

(17:03:46) chs: kr: go ahead

(17:04:03) kr: I assume some of you have read my blog and may even have taken a look at the video, right?

(17:04:10) mmeeks: nice music ;-)

(17:04:43) kr: Hi Michael, may be we should join the music business ... :-)

(17:04:53) mmeeks: kr: the music video business ? ;-)

(17:05:09) kr: seems you liked the background music ...

(17:05:28) mmeeks: it was a neat demo too of course.

(17:06:06) chs: kr: was the video a starting point for something broader?

(17:06:23) kr: actually yes ...

(17:06:48) kr: What I have in mind is, to bring ODF more to the server ... an ODF Wiki only being the beginning ... hopefully :-)

(17:07:41) kr: The current implementation is very simple, it is an apache running on the sever, converting ODF documents to HTML on request

(17:08:10) chs: indeed. What is being considered here is perhaps more than "just" an ODF wiki as in "let's export my wiki content to ODF".

(17:08:50) flr [n=flr@pD95FFD95.dip.t-dialin.net] a rejoint le salon.

(17:08:51) chs: Something that we could, for instance, not just have as an extension but as an actual feature of OOo?

(17:09:43) kr: I suggest to start slowly ... though this might become a feature of OOo

(17:10:22) kr: The nice point about the ODF Wiki respectively the ODF@WWW is, that everybody I tell of it immediately jumps on it ...

(17:11:04) mmeeks: jumps on it to beat it to death ? or swallow it whole ? ;-)

(17:11:05) chs: sure. So let's perhaps discuss this in two ways: 1° what the actual roadmap for odf@www could be according to you 2°if we can think about such online-offline capabilities as a way to achieve a future milestone of OOo

(17:11:48) mmeeks: kr: perhaps step back a little: do you want to attract developers to help out in this meeting, or attract marketing support ?

(17:11:59) kr: mmeeks: both ...

(17:12:24) kr: without marketing nobody is going to know about it

(17:12:25) chs: mmeeks: we need everybody here but we should realize we talk about two successive things

(17:12:32) thorsten: ok - then I guess the hype to ride is collaboration currently, in whichever form

(17:12:42) chs: thorsten: yes

(17:12:46) thorsten: so, wiki is just one instantiation of clllaboration

(17:13:03) kr: you understand me :-)

(17:13:12) chs: thorsten: but remember; hype is sometimes more than just "hype". Products get built around it

(17:13:34) ***thorsten would like to see something that's as easy as google docs in collaboration for OOo

(17:13:56) chs: so perhaps one way of looking at it, both on marketing and technical terms would be this: how do we make OOo and the wiki work together?

(17:13:59) thorsten: as in, have the OOoCon organizers use OOo again for planning ;-)

(17:14:17) chs: thorsten: how come I didn't see that one coming :-) ?

(17:14:50) thorsten: it's an obvious pain point

(17:15:30) kr: actually, Wiki editing is more or less everybody pain point

(17:15:37) chs: yup

(17:15:43) thorsten: oh, definitey

(17:15:44) chs: and of course, this is just our wiki

(17:15:56) chs: other wikis out there, with other issues, albeit similar

(17:16:12) chs: But let's have a look at the list there: http://wiki.services.openoffice.org/wiki/OOo_Next

(17:16:46) chs: Pretty much every desired feature on there has to do not just with wiki but also with collaboration et al.

(17:17:05) flr: kr: one thing is still unclear to me with your odf@www idea. I guess when I understand correctly you wanna use ODF as the "Wiki" language. But what will be the presentation language: HTML or ODF too?

(17:17:34) mmeeks: kr: I'm interested to hear the plan you propose really

(17:18:30) kr: flr: the presentation language can be anything, HTML is attractive because everybody has a browser

(17:18:41) kr: the plan ... OK

(17:19:01) flr: kr: well then we have to ODF to WWW conversion problem obviously

(17:19:11) flr: s/WWW/HTML

(17:19:57) kr: flr: yes, but ODF is much more expressive than HTML, so it should be the master format

(17:20:37) chs: kr: let's spell out the plan :-)

(17:20:48) kr: OK, here comes the plan ... :-)

(17:21:20) kr: first, provide good installation instructions for what I currently have, others need to be able to see it working and to play with it

(17:21:45) kr: this will likely be ready by end of tomorrow

(17:22:29) kr: second, I would like to add support for drawings and presentations

(17:23:00) kr: I don't have a real clue yet, how long that takes, though a week or so should be sufficient

(17:23:58) chs: kr: how about tables/cells?

(17:24:05) kr: third, I would like to get rid of the indirection resolver script, that means a slight change to the office itself

(17:24:33) kr: chs: spreadsheets are already kind of working ;-)

(17:24:47) chs: kr: oh great :-)

(17:25:19) kr: the overall goal must be, to reach "production" quality as soon as possible ...

(17:25:19) thorsten: kr: ok, that's rather short-term. what about the long run, i.e. OOo 3.x?

(17:25:45) chs: thorsten; perhaps it will be even further

(17:25:49) chs: like OOo 4

(17:26:04) kr: in the long run, I would like to separate the documents contents on the server side

(17:26:29) kr: for example to let documents share stuff, such as images, paragraphes, tables etc. ...

(17:27:07) kr: ideally a user would be able to e.g. use the gallery to select parts of documents to be included (either as a copy or be reference) into the current document

(17:27:24) thorsten: kr: what about other collaboration features?

(17:27:41) kr: thorsten: ??? Concurrent editing?

(17:27:45) ***thorsten sees chat & shared editing pop up in other FLOSS software a lot

(17:27:57) chs: perhaps we need to have a collaborative effort on defining collaboration features

(17:28:00) thorsten: kr: yeah, for example

(17:28:01) chs: :-)

(17:28:18) chs: there is a need to have some "Q concept" document about this

(17:28:38) chs: Ultimately what kr and thorsten describe are features

(17:28:43) chs: but how about the whole package?

(17:28:58) kr: I think that ... while starting ODF on the server ... the other things (chat ...) are falling naturally in place

(17:29:03) chs: if we include those features, the change won't be trivial

(17:29:18) chs: at least in terms of what we distribute as a software

(17:29:19) thorsten: there are some obvious architectural changes that most of these features entail

(17:29:35) chs: So let's call make a concept out of this

(17:29:48) chs: it does not have to be developed in one specific milestone

(17:29:54) kr: ... ahhhh ... this sounds so heavy weight

(17:30:26) chs: what does sound heavy weight? :-)

(17:30:28) mmeeks: we should write a set of interlocking specifications,

(17:30:43) mmeeks: and get buy in for each one, with competitive review, screenshots, user-experience input at every stage ...

(17:31:13) thorsten: :-D

(17:31:14) chs: mmeeks: we should :-)

(17:31:36) kr: hey, this is currently a kind of a pet project only ... don't make the first scope to broad ...

(17:32:04) chs: kr: sure. It will be called... Pinneberg . Just a concept, really

(17:32:18) chs: :-)

(17:32:35) thorsten: well, the scope of this chat is clearly broader than just the wiki thing -

(17:32:52) thorsten: and it would be cool if people allocating resources would share

(17:33:02) thorsten: the idea that this is really important

(17:33:11) chs: I am all for it

(17:33:37) chs: I can volunteer some time effort and mind processing power for the document editing and throwing in some ideas

(17:34:09) ***thorsten otherwise sees other, more agile projects eat increasingly larger parts of OOo's lunch ;-)

(17:34:29) chs: ssuch as?

(17:34:38) thorsten: abi word, inkscape, etc.

(17:35:28) thorsten: and the web2.0 projects from the other end

(17:35:38) chs: *chs does not criticize those projects in public :-)

(17:36:14) chs: thorsten: perhaps all there is to understand here is that it is time for OOo as a product to make its own Copernician revolution :-)

(17:36:17) thorsten: in fact, this is no critique, but rather acknowledging that _they_ did it right there ;-)

(17:37:19) kr: I actually would like to see ODF@WWW to escape the OOo ghetto and to interact (bidirectional) more with other FLOSS projects

(17:37:31) chs: for instance (that will be discussed later on our works) : how about we let OOo users interact / collaborate on the OOo web site?

(17:38:01) kr: the moment the ODF Wiki becomes usable I am going to deploy it on OOo ...

(17:38:02) thorsten: kr: the whole ghetto problem is something that really bothers me, yeah

(17:38:58) kr: I already talked to the hardware guys :-)

(17:39:09) chs: but doesn't the componentizetion of OOo make it easier for OOo to have parts reused?

(17:39:13) chs: kr: cool?

(17:40:01) thorsten: chs: the real meat (e.g. layout/document engines, filters) are still ~monolithic

(17:40:05) kr: I fear the componentization (is that an English word?) is not coming forward fast enough ... it seems that there is not enough real life demand

(17:40:20) thorsten: or, rather: even if they are components, they need the whole stack to run against

(17:40:58) thorsten: the drawing layer rewrite has some promise there, though

(17:41:10) chs: thorsten: I see. Well. The wiki and its interaction - should I say- integration? would provide a solution then

(17:41:25) kr: Just imagine that the EDIT button of the ODF wiki installs an OOo in case non is locally available ... that would address the deployment problem while enabling the user to RICH editing

(17:41:40) chs: first we could end up shipping a new wiki platform reusable by all

(17:41:41) thorsten: chs: solution what for?

(17:41:44) chs: kr: +1

(17:41:51) kr: ... and would create more demand for componentization ...

(17:41:58) chs: thorsten: the monolithic issue

(17:42:00) chs: kr : yes

(17:42:14) chs: 2°: it would completely change the value of what we propose

(17:42:19) thorsten: chs: 'k - would maybe provide some incentive for internal changes, right

(17:42:30) chs: 3° it could even provide some revenue streams

(17:42:43) chs: but here I'm venturing far out, as I realize :-)

(17:43:12) kr: servers typically provide more revenue that clients ... though that probably shouldn't be in the foreground

(17:43:40) chs: kr: see revenues you can get from deikiwiki or wikidot, etc.

(17:43:53) chs: there is the community version, free as always

(17:44:07) chs: but if you want to run the whole shabang on our servers then you pay

(17:44:22) chs: but of course, source code and all are free as in beer and as in speech

(17:44:32) chs: and of course you still have OOo as a standalone product

(17:44:41) chs: but some of it, or perhaps the whole of it

(17:44:53) Rail: Sorry, I need to leave. Please, post the log to the list.

(17:45:00) kr: I am currently more interested in promoting ODF and OOo ... to foster the ecosystem

(17:45:00) chs: has become what is sometimes referred to as a RIA (Rich Internet Application)

(17:45:04) chs: see you Rail

(17:45:14) chs: an offline and online client

(17:45:32) kr: chs: Are you going to take care of the log and to post it? May into the traditional wiki?

(17:45:36) Rail: chs, bye, all bye

(17:45:39) Rail a quitté le salon (quit: Remote closed the connection)

(17:45:43) chs: ah, sure.

(17:46:07) kr: see ... I actually already try to involve everybody :-)

(17:46:19) chs: :-)

(17:46:26) ericb2 [n=eric@aud25-1-88-166-8-11.fbx.proxad.net] a rejoint le salon.

(17:46:32) chs: but you should have told that to me earlier

(17:46:44) chs: my client will only record the log from now on

(17:46:53) chs: oh, okay, I can still do it from the beginning

(17:46:57) sophi: chs: I've got them if you need

(17:46:58) kr: I send you the beginning

(17:47:15) chs: what wouldn't we do without sophi? :-)

(17:47:30) chs: anyway

(17:47:34) sophi: chs: lol, thanks :)

(17:48:20) kr: ok, another thing on my list is a "browser" mode for OOo ... already talked to Andreas Schluens and Carsten Driesner about it

(17:48:25) chs: so yes: perhaps we should have several collaboration streams on this

(17:48:59) chs: You have odf@www as an extension

(17:49:00) dmpop: kr: a "browser" mode?

(17:49:39) chs: you have the componentization

(17:49:45) kr: dmpop: Currently clicking a link in OOo opens a new window, I would like it to replace the current document ...

(17:50:09) kr: dmpop: adding a forward/backward button and a history -> you can browse ODF :-)

(17:50:30) dmpop: kr: i see. it's a nifty idea.

(17:50:34) chs: you have the wiki/OOo web site/server

(17:51:05) chs: all this have to be worked on, discussed and put forth on a concept document, even if some will exist as extensions from day one

(17:51:19) chs: but then, we make it specs out of the Pinneberg document

(17:51:22) kr: dmpop: you suddently get high fidelity and printable documents :-)

(17:51:29) chs: does that sound good?

(17:52:09) thorsten: it would be cool if some of this could put priority to long-overdue internal changes,

(17:52:16) chs: such as?

(17:52:30) thorsten: e.g. /me cringes when imagining 20 or something ooo instances run on the wiki server,

(17:52:48) thorsten: to convert odf to html, and being restarted every other half an hour,

(17:52:58) thorsten: because mem gets depleted

(17:53:03) kr: chs: are you taking care of the documentation?

(17:53:05) chs: ah...

(17:53:13) chs: kr : what documentation?

(17:53:18) chs: the Pinneberg document?

(17:53:20) chs: yes

(17:53:26) chs: but you're in too

(17:53:28) chs: :-)

(17:53:29) flr: one of the power of current WIKI is the interaction between HTML and CSS. ODF is less powerfull here....

(17:53:31) kr: thorsten: that is a feature ... otherwise nobody would use all that cheap memory :*)

(17:53:49) thorsten: kr. nah. that's a huge scalability issue

(17:54:07) ***thorsten wants OOo to become agile & lean, to cope with future challenges

(17:54:07) chs: flr: what do you envision?

(17:54:09) flr: so in order to be successfull with an ODF/OOo Wiki you need to work on stylesheets

(17:54:11) kr: flr: that can be addressed by separating the document content, ODF has styles as HTML

(17:54:27) flr: not really

(17:54:35) kr: flr: ???

(17:55:28) kr: flr: IMHO ODF styles are even more useful than CCS ...

(17:55:32) flr: The problem here is that people tend to produce dos with hard formatting in OOo

(17:55:46) flr: this is a killer for style-sheets and thus for the WIKI stuff

(17:56:30) kr: flr: that is unavoidable anyway ... but IMHO not a stopper

(17:56:36) flr: ODF styles more usefull??

(17:56:47) flr: ODF has not even a selector concept..

(17:57:16) flr: In fact there is no way to affect the layout with ODF styles as you do with CSS

(17:58:26) kr: mmmhhh ... I think that is what I mean, that constructs of CSS don't seem to be local enough ... thus hindering composition ... though I am not an expert ;-)

(17:58:40) chs: flr: this can be worked later on and besides, there is the OASIS ODF TC about these issues... :-)

(17:59:01) kr: flr: I come back to you later about it ... after I have found it again :-)

(17:59:33) chs: so documentation: am ready to work on it. Who else is interested?

(17:59:41) flr: Just as a use-case lets take www.openoffice.org. One thing you really want is that the branding of the side of enforced.

(17:59:50) kr: chs: this is just another motivation, I think we can make ODF more versatile if bringing the right requirements

(18:00:11) flr: When people start doing WYSIWYG editing in OOo you have a problem..

(18:00:32) chs: kr: yes but don't forget: we're not the only ones working on ODF. There are a whole bunch of players with a whole bunch of requirements

(18:00:48) kr: flr: what is it, that you can not achieve if using ODF shared styles ?

(18:00:57) chs: flr: would you like to help out on these issues at the ODF TC?

(18:02:09) flr: well the main problem IMHO are that you can not make "semantic markup" as you can do with HTML and with OOXML custom schema ;-)

(18:02:20) kr: flr: WYSIWYG editing is not the problem ... just using hard formatting may be problem, but only for the pages involved

(18:02:22) flr: what I'm missing in ODF is a thing where I can say:

(18:02:44) flr: <mytag>.....</mytag> and then have CSS render it..

(18:03:12) flr: So users need to learn to make semantic annotations instead of style annotations...

(18:03:29) kr: flr: that is out of scope ...

(18:03:54) flr: kr: ;-)

(18:04:14) flr: kr: OK then. Have fun. Bye

(18:04:21) flr a quitté le salon (quit: "Leaving")

(18:04:32) chs: so, where do we stand on all this?

(18:05:08) kr: in German I would say this is "ueberkandidelt"

(18:05:42) chs: which means ? :-)

(18:05:48) thorsten: over the top

(18:05:51) chs: ah

(18:05:54) chs: :-)

(18:06:07) chs: so we need to dig or go down in the mud :-)

(18:06:11) kr: thorsten: thanks for translation

(18:06:17) thorsten: kr: well, not sure - after all, it's a competition of formats,

(18:06:31) thorsten: kr: and if one format lacks an important feature, well...

(18:06:56) chs: and you cannot expect to achieve full roundtrip interop :-)

(18:07:02) kr: thorsten: I don't know yet of anything important missing

(18:07:21) kr: don't need a roundtrip -> HTML is readonly anyway :-)

(18:07:25) thorsten: well, I know a bunch of stuff in svg that is not representable in odf ;-)

(18:07:37) thorsten: from hands-on experience :-)

(18:08:08) kr: thorsten: the moment we need we may go and add it :-)

(18:08:23) thorsten: that would be: now :-P

(18:08:45) thorsten: svg is _the_ standard for web vector graphics

(18:08:46) kr: chs: we are already after hour - I think we should start closing the meeting - at least I have to leave in a few minutes

(18:08:51) thorsten: but I digress...

(18:08:55) chs: kr: me as well

(18:09:01) chs: so let me summarize:

(18:09:09) chs: kr works on an extension

(18:09:19) chs: and checks server requirements

(18:09:27) chs: chs: starts document

(18:09:36) chs: first starts it on the wiki

(18:09:44) sophi: chs: I'll help you with documentation

(18:09:44) chs: this is a collaborative effort

(18:09:48) chs: sophi thanks

(18:09:59) chs: torsten: er, what do you do? :-)

(18:10:27) thorsten: chs: lobby for resources :-)

(18:10:32) dmpop: kr: i'd be happy to write some end-user documentation and other odf@www-related stuff.

(18:10:35) chs: excellent

(18:10:44) chs: dmpop: you're on board

(18:10:55) chs: mmeeks: what do you want to do?

(18:11:19) mmeeks: chs: right now - eat my fist & jump out of the window...

(18:11:26) kr: dmpop: perfect ... I keep you posted about virtual boxes and much better installation instructions

(18:11:35) mmeeks: chs: you've got me right where you want me ;-)

(18:11:53) chs: mmeeks: oh, things start to become hot around here :p

(18:12:08) chs: okay, then as a closing point: the roadmap

(18:12:25) chs: let's see what we have in, say, september on this channel.

(18:12:29) dmpop: also, i plan to write an article on odf@www for linux.com -- assuming they accept the idea.

(18:12:36) chs: status reports, etc.

(18:12:47) chs: dmpop: perhaps not that early in the process?

(18:13:06) kr: dmpop: I would love to see that :-)

(18:13:27) dmpop: chs: not at all, release early, write early :-)

(18:13:45) chs: dmpop: right; was afraid kr didn't want too much noise around it at first

(18:13:54) chs: but anyway

(18:14:00) kr: chs: au contraire mon ami :-)

(18:14:19) kr: chs: make as much nose as you :-)

(18:14:39) jsc a quitté le salon (quit: "ChatZilla 0.9.83 [Firefox 3.0.1/2008070206]")

(18:14:40) chs: kr: okay, am switching to Steve Jobs now

(18:15:00) chs: how about the 19th of September, same time here?

(18:15:22) kr: chs: sounds good

(18:15:50) sophi: chs: ok for me too

(18:16:09) chs: by that time, a small but clearly written document and some technical info concerning server requirements, and componentization, (or whatever that is called) should be available

(18:16:12) chs: ideally of course

(18:16:18) chs: but you know Steve Jobs

(18:16:33) chs: what is impossible will thus be achieved :p

(18:17:11) khirano: kr: you said the good installation instruction will be ready by the end of tomorrow? Where will you upload it?

(18:17:44) kr: khirano: Yes, I will do so. I also plan to briefly blog about it

(18:17:58) chs: OKay, should we adjourn the meeting?

(18:18:03) khirano: kr: ok, thanks

(18:18:15) khirano: thanks, chs :)

(18:18:36) kr: chs: Yes, let's close it. And thank you very much for support and everything :-)

(18:18:41) chs: thanks to everyone who attended the meeting

(18:18:53) chs: thank you kr for believing in all this!

(18:19:05) chs: and thank you to everyone who wishes to contribute!

6. Relationship to the ODF@WWW project

This page was created pretty much at the same time Kay Ramme started to work on the ODF@WWW's proof of concept, and now belongs to this project's category. Future IRC session are expected to be posted on this page.