I'm going to address all the questions to me in this single reply.

My original suggestion was that we export the HTML from Confluence, convert
to Markdown and put the Markdown and the images in a git repo. Markdown is
much easier to edit than raw HTML, especially the HTML exported from
Confluence (blech!). The idea was that we could use Jekyll + SAAS to craft
a new website. In fact, Michael Andre Pearce produced a mockup of this
using the Apache Metro website as an example (because it already makes use
of Jekyll + SAAS). It was enough to convince me that we should take this
path, so I started looking into doing a full, new export of Confluence
pages to HTML. If you have not seen Michael's mockup, you should really
take a look.

So, I manually grabbed the raw HTML that is automagically exported from
Confluence and is hosting the current site that we see at
http://activemq.apache.org. I did some testing on it using text2html and
the conversion it does is pretty awful and would require a lot of hand work
to fix it. So, we discussed the point that there are 1600+ pages of HTML to
manually edit. But I later realized that it was only about 950 HTML pages
(from what I can tell so far).

Then, Dan Kulp found a Confluence HTML to raw HTML converter built on top
of PanDoc. So, I have also been trying to export the HTML from Confluence
in order to try out the PanDoc converter (it works based on the Confluence
export function which is different from how the HTML is automagically
converted). Unfortunately, I am running into a NullPointerException from
Confluence. ASF Infra is telling me that the NPE is due to the CDATA in the
search function on the Navigation page and is suggesting that the solution
is to remove the Navigation page. The problem with this suggestion is that
it would fundamentally remove all the navigation on the right-hand side of
the site -- not what we want.

I have also given some thought to the idea that removing the current site
will break all links to old site. This is something that cannot be
overlooked and must be prevented as we do not want to leave users who have
bookmarked a page high and dry. This is a fairly easy problem to solve this
using some mod_rewrite rules, the question is if ASF Infra is willing to
allow us to deploy such custom rules. This should be investigated when we
get to that point, but we are not there yet. First, we need to decide the
best path forward based on what I have described above in the preceding
paragraphs.


Bruce

On Tue, Dec 12, 2017 at 11:39 AM, Martyn Taylor <[email protected]> wrote:

> I was thinking there would be a single css file for all the pages.  But I
> haven't seen the files yet. Let's have a play around when Bruce pushes the
> export.
>
> Cheers
>
> On 12 Dec 2017 5:30 pm, "Michael André Pearce" <
> [email protected]>
> wrote:
>
> > What’s 1600 pages between friends....
> >
> > I agree it will be easier to covert to md than to start doing css styles.
> > It’s all from a wiki anyhow so it’s can’t be that far off.
> >
> > It be good to get some samples (eg 50 pages) if not all just to try and
> > see what it is like.
> >
> >
> >
> > On 12 Dec 2017, at 17:04, Clebert Suconic <[email protected]>
> > wrote:
> >
> > >> Exporting to MD and creating a gitbook seems like a big task, I
> suspect
> > any
> > >> tool we use will cause a bunch of styling/content issues.
> > >>
> > >> At least initially, how about we just create a nice landing page that
> > >> brings the ActiveMQ site and Artemis site together, and refresh/align
> > the
> > >> existing content with some CSS?
> > >
> > > I was just looking for the minimal effort task. I thought that
> > > converting these pages into a doc would be easier than converting them
> > > to another .css...
> > >
> > > if the conversion needed to be done anyways... I thought .md would be
> > > easier and having a better final presentation.
> >
>



-- 
perl -e 'print
unpack("u30","D0G)U8V4\@4VYY9&5R\"F)R=6-E+G-N>61E<D\!G;6%I;\"YC;VT*" );'

ActiveMQ in Action: http://bit.ly/2je6cQ
Blog: http://bsnyder.org/ <http://bruceblog.org/>
Twitter: http://twitter.com/brucesnyder

Reply via email to