|
From: Scott C. <sc...@sc...> - 2016-09-01 19:24:37
|
Dave, This is good stuff and brought up points I hadn't thought about--thank you! Do you happen to know what the instance size of the Galaxy wiki server is? I suspect that the Galaxy wiki gets more traffic than the GMOD wiki, but they might be at least roughly comparable. Thanks, Scott On Thu, Sep 1, 2016 at 1:24 PM, Dave Clements <cle...@ga...> wrote: > Hi Scott, > > We are in the process of migrating the Galaxy Wiki from MoinMoin to a > GitHub hosted pseudo-wiki (still looking for a good term). The content will > be in GitHub Flavored Markdown (GFM) plus YAML for metadata and structured > content. We'll then use MetalSmith (http://www.metalsmith.io/) to > generate the actual web site, and it's MetalSmith that will process the > YAML. > > Editors will be able to edit local copies of the doc and generate the > website locally, as well as edit doc directly in GitHub. This model will > fully leverage the power of GitHub. > > Moving GMOD.org to straight github would also leverage that power and > community involvement. > > *However, I think this would be tremendously difficult,* and the > advantages of moving from MediaWiki to GitHub are far less than moving from > MoinMoin to GitHub. > > 1) MediaWiki supports a ton of extensions such as Templates, Categories, > and Extensions. GFM does not support any of that. A lot of content on > GMOD.org would have to be disentangled. > > 2) GFM is deliberately crippled. The theory (I think) is that you should > be using CSS for formatting instead of markup. I like that goal, but the > current GMOD.org is chock full of MediaWiki / HTML hacks to achieve > formatting (I put a lot of it there). Translating that to pure GFM / CSS > would be a nightmare. > > GFM does support directly using HTML, but it does not support mixing the > two in many circumstances. For example, tables are all GFM or all HTML. > And GFM tables are painfully limited. Only the simplest tables can be > translated to GFM. If you want row headers, or colspans or rowspans, or > right or center alignment in a cell, or ... you have to translate the whole > table to HTML. And that defeats the whole purpose of having a wiki. > > *So, my vote is stick with MediaWiki.* I don't think there is any hope > of migrating to GFM without seriously mangling 90%+ of the pages. > > I've CC'd Dannon Baker who is also working the Galaxy Wiki move. He's not > familiar with the current GMOD.org implementation, but he knows about > migrating wikis to GitHub. He may have more to add (and might even > contradict me :-). > > And, we've been running our wiki on AWS for years. We've been very happy > with it. > > My 98¢ > > Dave C > > > On Thu, Sep 1, 2016 at 8:58 AM, Nathan Dunn <nat...@lb...> wrote: > >> >> Scott, >> >> I would advocate moving to GH pages just for the long-term ease of use, >> permission management, and cost. There is no server to setup, but you’ll >> have to configure a Jekyll theme (http://jekyllthemes.org/) and add the >> content, which is a PITA unless you create a single-page app (which I don’t >> advocate). >> >> My guess is that actually converting the content will be easier (and I’m >> sure folks in the community will be glad to help). >> >> https://github.com/philipashlock/mediawiki-to-markdown >> >> >> Nathan >> >> On Sep 1, 2016, at 8:46 AM, Scott Cain <sc...@sc...> wrote: >> >> Hi all, >> >> I'm addressing this to a few of the more heavily trafficked GMOD mailing >> lists in hopes of getting some good advice. The server that runs >> gmod.org is being retired at the end of the month, so I have to find it >> a new home. One option includes leaving it at OICR on a new virtual >> machine, but it will require me to port the existing MediaWiki instance, >> which I already know won't be fun. >> >> There are other alternatives though--it was suggested recently that the >> content in gmod.org could be migrated to GitHub pages, which sounds >> appealing but I have no idea how much work would be involved. I've also >> considered migrating to AWS, which I think will be a similar amount of work >> as the internal migration at OICR but would likely be cheaper in the long >> run. >> >> Thanks in advance for any insights! >> Scott >> >> >> -- >> ------------------------------------------------------------------------ >> Scott Cain, Ph. D. scott at scottcain >> dot net >> GMOD Coordinator (http://gmod.org/) 216-392-3087 >> Ontario Institute for Cancer Research >> ------------------------------------------------------------ >> ------------------ >> _______________________________________________ >> Gmod-tripal mailing list >> Gmo...@li... >> https://lists.sourceforge.net/lists/listinfo/gmod-tripal >> >> >> > > > -- > http://galaxyproject.org/ > http://getgalaxy.org/ > http://usegalaxy.org/ > https://wiki.galaxyproject.org/ > -- ------------------------------------------------------------------------ Scott Cain, Ph. D. scott at scottcain dot net GMOD Coordinator (http://gmod.org/) 216-392-3087 Ontario Institute for Cancer Research |