Heyho, I'll try to take a step backwards to look at the large picture here for a second.
In general I really love what you have done Chris! Having everything in an asciidoc file is great and building with Maven, I think most people can get behind that (certainly more than for any other solution :) ). Before talking about specific implementations for charts, graphs, etc. I think we should maybe try and discuss the overall design of what we will be trying to build and how we want to organize it. As I said, I'd be on board with Maven and Asciidoctor as vehicles. Reveal.js I'll need to look at further, but the target format is probably a less final decision than the format that we maintain our content in anyway. For graphs etc. I personally think that the target should be some sort of plugin system (Asciidoctor already has one, maybe we don't need anything beyond that) that we can use to slowly grow a curated list of acceptable formats to create content in. For diagrams there is a huge list of services that we could look at out there: cacoo, draw.io, dia, yEd, gliffy, LucidChart, ... We should probably come up with a more or less formalized way of accepting new formats to avoid needlessly growing the list, but I also see no need to be too restrictive here. Best regards, Sönke On Fri, Mar 1, 2019 at 7:59 PM Lars Francke <lars.fran...@gmail.com> wrote: > > Thank you for sharing Christofer. I need a quiet minute to look at that. > > What I like is that it builds easily using Maven, content in Asciidoc, > easily versionable etc., the speaker notes are good > I see that reveal.js also does PDF export. > > The only thing I don't immediately like are the diagram options. I think > they are pretty...ugly. > > I used the same example[1] to play around a bit and newer versions work as > well. > > One thing I haven't tried yet is how to "depend" on other content. If we > have a ZooKeeper Training which we'd like to include in a Hadoop ecosystem > training for example. And would all our content be in Maven projects? > > I believe it'd be great if we could have "content-only" projects and then > other projects that do the packaging/converting/distributing part. > > Cheers, > Lars > > [1] < > https://github.com/asciidoctor/asciidoctor-maven-examples/tree/master/asciidoc-to-revealjs-example > > > > On Mon, Feb 25, 2019 at 7:53 PM Dmitriy Pavlov <dpav...@apache.org> wrote: > > > Yes, for me it works. Thank you > > > > пн, 25 февр. 2019 г., 20:54 Christofer Dutz <christofer.d...@c-ware.de>: > > > > > Does it work now? > > > Even if I said "everyone with a link" slak keeps on asking me to grant > > > permissions :/ > > > > > > Chris > > > > > > > > > > > > Am 25.02.19, 18:47 schrieb "Dmitriy Pavlov" <dpav...@apache.org>: > > > > > > Hi Chris, could you please add view permission for the google > > document? > > > Thank you. > > > > > > пн, 25 февр. 2019 г. в 20:44, Christofer Dutz < > > > christofer.d...@c-ware.de>: > > > > > > > Hi all, > > > > > > > > I'm not suggesting to build something ... it's sort of already > > there: > > > > Here an export of one of my current presentation template: > > > > > > > > https://drive.google.com/open?id=1pZ5l9X__gTM4vg2PJRbc-0GXuEf058aI > > > > > > > > It uses Asciidoc and I quite like that in general for all sorts of > > > use > > > > cases. > > > > Markdown to me appears a lot less powerful and extensible (but that > > > just > > > > might be me dropping the ball on that quite some time ago) > > > > Doc-book and Latech I remember being quite low level and I don't > > know > > > > reStructuredText. > > > > > > > > Regarding images I started adopting PlantUML and DITAA quite some > > > time ago > > > > and quite recently am updating to SVGBob > > > > https://github.com/ivanceras/svgbob > > > > > > > > Regarding your format ... just have a look at the content of > > > > src/main/asciidoc/index.adoc > > > > In the Zip file ... that's pretty much what you describe. > > > > Most of these require some installed open-source tools to render > > > images > > > > correctly and I have started setting up some init scripts to > > install > > > > missing things, but that still needs quite a lot of love to be in a > > > > releasable state. > > > > Currently it's just something I use myself and the scripts are > > more a > > > > reminder to myself of how to install things. > > > > > > > > Please have a look. > > > > > > > > Chris > > > > > > > > > > > > Am 25.02.19, 17:42 schrieb "Sönke Liebau" < > > > soenke.lie...@opencore.com > > > > .INVALID>: > > > > > > > > I agree with Mirko, I don't think we should head down the route > > > of > > > > creating a full blown publishing framework or similar. > > > > > > > > The issue, at least to my mind, is divided into two main > > things: > > > > - text content (which I consider to include tables, lists, > > etc.) > > > > - graphical content > > > > > > > > For text content there are quite a few good options out there, > > we > > > > probably just need to conduct a hunt for the main competitors > > and > > > > agree on one that meets most needs. Otoh the main ones are > > > probably: > > > > - asciidoc > > > > - markdown > > > > - doc-book > > > > - latex > > > > - reStructuredText > > > > > > > > For graphical content my personal opinion is that the > > > possibilities > > > > are simply endless and we should not necessarily be trying to > > > restrict > > > > what people may want to use either. For the "compiled" > > > presentation in > > > > the end I think the common denominator will always be "a > > > picture" (no > > > > other way to express a photo or a logo) and I personally think > > > it is > > > > fine. > > > > The way of getting at this image is what I think we should be > > > focusing > > > > on, so the basic idea would be to have a text representation of > > > the > > > > image in version control and at "compile" time create the > > actual > > > image > > > > that is then part of the release. > > > > For the "text representation" part there are a lot of possible > > > > options, what I have used a lot in the past is for example: > > > > - websequencediagrams [1] > > > > - draw.io [2] > > > > > > > > But since there are so many services out there that offer > > > something > > > > similar I think this should really be something extensible so > > > that > > > > people can develop converters for their own formats. For the > > > Apache > > > > training content we should then probably have a rule that only > > > > converters that are part of the official repo may be used for > > > content, > > > > which allows us to curate a little. > > > > > > > > So basically in version control slides might then look like > > this: > > > > > > > > == Slide One > > > > > > > > * Foo > > > > * Bar > > > > * World > > > > > > > > == Slide Two > > > > >>> imageContent(websequencediagram) > > > > User->Server: Connect > > > > Server->User: Respond > > > > <<< > > > > > > > > Whereas the content of the wsd part would be replaced by the > > > > corresponding picture when building the actual slides. > > > > > > > > > > > > Hope that made a little sense? Otherwise I am happy to > > elaborate > > > > further :) > > > > > > > > Best regards, > > > > Sönke > > > > > > > > > > > > [1] https://www.websequencediagrams.com/ > > > > [2] https://www.draw.io/ > > > > > > > > On Mon, Feb 25, 2019 at 5:17 PM Mirko Kämpf < > > > mirko.kae...@gmail.com> > > > > wrote: > > > > > > > > > > Hi, > > > > > regarding content versioning, I suggest to search formar like > > > > doc-book xml > > > > > (it can be anything which allows Separation of content and > > > Style). > > > > > With this, we can generate PDF, PPT, Google-Presentations for > > > final > > > > > customization. > > > > > > > > > > The issue is, how to convert a result from a creativity > > session > > > > incl. media > > > > > content / sketches / fotos back into such a fundamental > > format. > > > > > > > > > > I suggest not to try to build another CMS or publishing > > > Framework, > > > > but > > > > > rather Focus on the process of content creation/Update. > > > > > > > > > > Cheers, > > > > > Mirko > > > > > > > > > > Lars Francke <lars.fran...@gmail.com> schrieb am Sa., 23. > > Feb. > > > > 2019, 16:23: > > > > > > > > > > > On Sat, Feb 23, 2019 at 7:31 PM Sharan Foga < > > > sha...@apache.org> > > > > wrote: > > > > > > > > > > > > > > > > > > > > > > > > > > > On 2019/02/22 23:12:29, Lars Francke < > > > lars.fran...@gmail.com> > > > > wrote: > > > > > > > > During the DISCUSS and VOTE threads I tried to postpone > > > any > > > > discussion > > > > > > > > about the actual content and technical bits but now > > > would be a > > > > great > > > > > > time > > > > > > > > to start. > > > > > > > > > > > > > > > > I know that Dmitriy was eager to get started and > > > Christofer > > > > also > > > > > > > explained > > > > > > > > his workflow briefly. Maybe you could go into more > > > detail? > > > > > > > > Christofer demonstrated his own tooling to us and I > > > really > > > > liked it. > > > > > > This > > > > > > > > could be a great start. > > > > > > > > > > > > > > > > I'm sorry this is going to be a bit longer and maybe a > > > bit > > > > "rambling". > > > > > > > Take > > > > > > > > it as you will. I just needed to write it down once :) > > > > > > > > > > > > > > > > When we've done trainings so far they usually consist > > of > > > a > > > > couple of > > > > > > > things: > > > > > > > > > > > > > > > > * Slides (for us usually in Powerpoint) > > > > > > > > * Whiteboard sessions (usually the most interesting > > parts > > > > because they > > > > > > > > usually are the result of attendee feedback/questions) > > > > > > > > * Labs (the actual content, things that attendees need > > to > > > > "solve"/do) > > > > > > > > * Lab setup (especially for the larger distributed > > > systems > > > > getting a > > > > > > > > realistic setup of the tools itself for all attendees > > > isn't > > > > trivia > > > > > > > > > > > > > > > > I'm sure I'm missing something. > > > > > > > > > > > > > > Thanks Lars - this is good. Off the top of my head a > > > couple of > > > > things > > > > > > came > > > > > > > to mind - the first is testing (to see how much attendees > > > have > > > > learned > > > > > > and > > > > > > > this could be linked to certification which I think was > > > > mentioned in one > > > > > > of > > > > > > > the threads) and the second was a way of collecting > > > feedback > > > > about the > > > > > > > training - so perhaps a survey > > > > > > > > > > > > > > > > > > > Those are good points I didn't think of. > > > > > > > > > > > > Tests we have never done by choice but I see that people > > > might be > > > > > > interested in them and surveys are something that we > > probably > > > > should have > > > > > > done ourselves a long time ago already. So: Definitely. > > > > > > > > > > > > > > > > > > > > What should our scope be? > > > > > > > > Our initial idea centered around Slides and Labs. It > > > would be > > > > great to > > > > > > > also > > > > > > > > have something that makes the Labs setup easier but in > > > our > > > > experience > > > > > > > > that's pretty hard (e.g. corporate firewalls don't > > allow > > > > access to X or > > > > > > > Y) > > > > > > > > to make generic (that shouldn't stop us from trying!) > > > > > > > > > > > > > > > > Slides: > > > > > > > > I'd love to have a workflow where I can design slides > > > > entirelly in > > > > > > > > Asciidoc. That makes them easily versionable and > > > composable. > > > > Should we > > > > > > > > allow multiple formats? If we decide on a text-only > > > format and > > > > someone > > > > > > > > donates a bunch of courses in Powerpoint. Would we deny > > > that? > > > > > > > > > > > > > > I think that we would want to accept contribution that is > > > > relevant. There > > > > > > > may be an overhead to convert the content into a more > > > generic > > > > format but > > > > > > > that's doable especially if it encourages contributions. > > > > > > > > > > > > > > > > > > > I assume you meant "any contribution"? > > > > > > In general I agree but any binary format (e.g. Powerpoint - > > > I'll > > > > call it > > > > > > binary even though it's really XML now but it's pretty > > > useless for > > > > what I'm > > > > > > going to mention or PDF) has the problem that doing reviews > > > is > > > > tedious to > > > > > > impossible. There's no good way (I know of) to create diffs > > > for > > > > example and > > > > > > people on Linux are left out entirely for Powerpoint. > > > > > > > > > > > > I currently believe having "one true format" for all of > > them > > > is a > > > > good idea > > > > > > (I am happy to be convinced otherwise), maybe with a kind > > of > > > > "staging" area > > > > > > of accepted contributons that have yet to be converted and > > > are not > > > > coverd > > > > > > by "quality guarantees". > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > Labs: > > > > > > > > Similarly for Labs we've had a good experience with > > > (e.g.) > > > > > > > > https://antora.org/ which also allows to create > > > documentation > > > > in > > > > > > > Asciidoc > > > > > > > > and create a website out of it. But there's lots of > > > ideas on > > > > how to > > > > > > > improve > > > > > > > > this (e.g. Notebooks in Zeppelin) and it'll also be way > > > > different > > > > > > > depending > > > > > > > > on the training topic. > > > > > > > > > > > > > > > > Audience/Customizability/Composability > > > > > > > > I would assume that our trainings will also be used by > > > > non-commercial > > > > > > > folks > > > > > > > > or people needing to give a training in-house at their > > > > companies. For > > > > > > > them > > > > > > > > a prepared "deck" with ASF branding is fine but others > > > might > > > > want to > > > > > > > > incorporate these slides into their own work (see the > > > Legal > > > > thread) and > > > > > > > > also compose their own out of smaller "components". > > > > > > > > So for me a good thing would be if we produce smaller > > > > "chapters" of > > > > > > > things > > > > > > > > that can then be composed however one would like and to > > > make > > > > our > > > > > > product > > > > > > > > customizabile (e.g. custom header, footer, background > > > colors > > > > etc.) > > > > > > > > > > > > > > > > Apache vs. non-Apache // Product vs. non-product > > > > > > > > I wouldn't want to limit us to Apache products. I don't > > > see a > > > > reason > > > > > > not > > > > > > > to > > > > > > > > also talk about 3rd party tools. Especially if they are > > > tightly > > > > > > > integrated > > > > > > > > into the ecosystem (e.g. the ELK stack is often used > > > alongside > > > > Hadoop). > > > > > > > > > > > > > > +1 I like the idea and it also could make our content > > > valuable > > > > to others > > > > > > > outside the ASF > > > > > > > > > > > > > > > > > > > > > > > I also don't see a reason to only focus on > > > > < > > > > > https://maps.google.com/?q=%3E+I+also+don't+see+a+reason+to+only+focus+on+&entry=gmail&source=g > > > >single > > > > products. A training > > > > > > > > could focus on "IoT" and cover lots of products. > > > > > > > > > > > > > > +1 this will also give the Apache projects visibility of > > > others > > > > in the > > > > > > > same domain. I'm not really sure how cross pollinated our > > > > projects are. > > > > > > > > > > > > > > > > > > > > > > > In a similar vein it doesn't always have to be > > technical > > > > products. I've > > > > > > > > already been approached from multiple people about "The > > > Apache > > > > Way" > > > > > > > > presentations. Now whether they make more sense in > > > ComDev is > > > > to be > > > > > > > decided. > > > > > > > > Maybe Sharan can weigh in? > > > > > > > > > > > > > > I think Training would be a great place for managing the > > > Apache > > > > Way > > > > > > > content. In ComDev we've tried to gather and collate this > > > type > > > > of content > > > > > > > and have ended up with a page of different presentation > > > slides. > > > > Each > > > > > > person > > > > > > > has a different spin on it - so creating something > > > standard as a > > > > nice off > > > > > > > the shelf template that anyone can use will be great. And > > > I'm > > > > happy to > > > > > > > ensure we maintain a link and communicate with ComDev > > > regularly > > > > so > > > > > > > potential contributors know about what we are doing here > > in > > > > Training. > > > > > > > > > > > > > > > > > > > Okay, that's good! > > > > > > As you said: There's a dozen of those out there now. > > > > > > > > > > > > Lars > > > > > > > > > > > > > > > > > > > Thanks > > > > > > > Sharan > > > > > > > > > > > > > > > > > > > > > > > Thanks, > > > > > > > > Lars > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > -- > > > > Sönke Liebau > > > > Partner > > > > Tel. +49 179 7940878 <+49%20179%207940878> > > > > OpenCore GmbH & Co. KG - Thomas-Mann-Straße 8 - 22880 Wedel - > > > Germany > > > > > > > > > > > > > > > > > > > > > > > -- Sönke Liebau Partner Tel. +49 179 7940878 OpenCore GmbH & Co. KG - Thomas-Mann-Straße 8 - 22880 Wedel - Germany
