[Bioperl-l] New Jekyll theme at bioperl.github.io

Fri Mar 11 13:53:01 UTC 2016

Chris and Peter,

A bit late with this, but you may have noticed: bioperl.github.bio now has search capability, it’s ready to take the bioperl.org address.

Peter, we talked about this a bit. Turns out it is not hard at all. Jekyll is file-based, no database there. There’s a Liquid template (/search.json) that’s used to convert all the HOWTO’s (and misc. other files, 111 pages in total) to JSON, in a single file (/_site/search.json). Then some simple JS (Jquery) searchs this file. The performance is good, on the other hand the JSON file is not large, 963 lines. 

There is remaining work upcoming. In May GitHub will disable all the Markdown converters except for kramdown, and we’re using redcarpet. I did some testing, it appears that our tables are not converted by kramdown, they will all have to be fixed. My guess is that this will be straightforward.

Thanks again,

Brian O.

> On Jan 10, 2016, at 10:05 AM, Fields, Christopher J <cjfields at illinois.edu> wrote:
> 
> BioJava and BioSQL would be the only other ones I can think of.  EMBOSS is static web content for the most part if I recall; is that hosted on the AWS server or is it a redirect?
> 
> chris
> 
>> On Jan 10, 2016, at 2:26 AM, Peter Cock <p.j.a.cock at googlemail.com <mailto:p.j.a.cock at googlemail.com>> wrote:
>> 
>> On Sat, Jan 9, 2016 at 7:03 AM, Brian Osborne <bosborne11 at verizon.net <mailto:bosborne11 at verizon.net>> wrote:
>>> Peter,
>>> 
>>> It looks like you have a very nice plan to migrate this content that is
>>> close to - if not fully - automated. Good idea.
>>> 
>>> Personally, the way I was thinking about the BioPerl Wiki was a bit
>>> different. The BioPerl Wiki is quite old, quite large, and a lot of its
>>> content is dated, or wrong, or now superfluous (e.g. each module in BioPerl
>>> has a Wiki page, which was written by code, and there are many Wiki pages
>>> about applications or terms or formats that are now handled by Wikipedia).
>>> So this migration, again my personal view, was an opportunity to remove a
>>> lot of content yet preserve what was truly useful. Also bear in mind that
>>> resources are very stretched at BioPerl, the less content we have to
>>> maintain, the better. So we hand selected the pages we wanted to migrate,
>>> and we also did some amount of editing of _each_ selected page, and
>>> consolidated pages. So that process gave us a set of some 150 GFM pages. No
>>> versions.
>> 
>> Fair enough - I'd have opted for a full automated conversion (to
>> capture the contributor history), and then done the clean up -
>> but the key thing is you've done this. Chatting with Chris Fields
>> there is certainly a lot of good stuff on the old BioPerl wiki which
>> would be better placed somewhere central (e.g. file formats on
>> wikipedia?).
>> 
>>> Then the question I puzzled over was “what look?” Check these out, if you
>>> want to get a sense of your Jekyll options:
>>> 
>>> http://jekyllthemes.org <http://jekyllthemes.org/>
>>> https://github.com/jekyll/jekyll/wiki/sites <https://github.com/jekyll/jekyll/wiki/sites>
>>> 
>>> There is necessarily some editing involved in the migration. Hopefully most
>>> of GFM does not have to be touched, (and the “front matter” section that
>>> Jekyll wants can be added by script to the GFM pages) but the migrators will
>>> have to look at the “topmost" pages (and maybe there’s just one of these,
>>> index.html or index.md).
>> 
>> Yes - and likely categories would work that way too.
>> 
>>> You also will want search capabilities in some of these sites. Some themes
>>> have it already, most of these above do not.
>> 
>> Interesting - I'd not looked at that at all.
>> 
>>> There also appears to be an issue around Google searching if sitemap.xml is
>>> not right, have not looked into this yet:
>>> 
>>> http://gon.to/2015/03/03/f-percent-number-ck-github-pages-for-jekyll-why-i-decided-to-use-digital-ocean/ <http://gon.to/2015/03/03/f-percent-number-ck-github-pages-for-jekyll-why-i-decided-to-use-digital-ocean/>
>>> 
>>> Anyway, having said all of that, I can assist with the migration. What is
>>> the full list of Wikis? Is it something like this?
>>> 
>>> [bosborne at ip-10-116-249-158 ~]$ ls /obf/websites/
>>> biodas.org           biopathways.org  bioruby.org  cgi.biodas.org
>>> helpdesk.open-bio.org  www.arareko.net
>>> biojava.org          bioperl.org      biosoap.org  cross-site-stuff
>>> lists.open-bio.org     www.open-bio.org
>>> biolib.open-bio.org  bioprolog.org    biosql.org   doc.bioperl.org
>>> news.open-bio.org
>>> biomoby.org          biopython.org    blipkit.org  emboss.open-bio.org
>>> obda.open-bio.org
>>> 
>>> Some of these are not Wikis, like biomoby and open-bio.org.
>> 
>> BioMoby has some complicated stuff, but www.open-bio.org <http://www.open-bio.org/> does
>> have a wiki (which I'm looking at as the first migration) and some
>> static HTML pages like old BOSC conferences (hopefully they
>> can be moved into GitHub as static HTML pages).
>> 
>>> And do we have owners for all of these? Are some of no interest?
>> 
>> Hopefully - migration is a good chance to verify all that.
>> 
>>> Finally, I think that each of these needs its own GitHub account
>>> in order to get a GitHub Pages site. Or were you thinking of only
>>> migrating some of these? A bunch of questions for OBF here.
>> 
>> Yes, but most of the active projects already have GitHub accounts.
>> We're looking at BioPerl (you), the main OBF wiki and Biopython
>> (me) as the first trial migrations to GitHub pages.
>> 
>> Regards,
>> 
>> Peter
>> 
>> _______________________________________________
>> Bioperl-l mailing list
>> Bioperl-l at mailman.open-bio.org <mailto:Bioperl-l at mailman.open-bio.org>
>> http://mailman.open-bio.org/mailman/listinfo/bioperl-l <http://mailman.open-bio.org/mailman/listinfo/bioperl-l>
> _______________________________________________
> Bioperl-l mailing list
> Bioperl-l at mailman.open-bio.org
> http://mailman.open-bio.org/mailman/listinfo/bioperl-l

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.open-bio.org/pipermail/bioperl-l/attachments/20160311/2575e366/attachment.html>