Profile

Cover photo
Joe Pairman
Works at Mekon Ltd
Attended University of Leeds
Lives in London
326 followers|388,106 views
AboutPostsCollectionsPhotosReviews

Stream

Joe Pairman

Shared publicly  - 
 
It's becoming quite common these days to manage XML-based content in Git or Mercurial. It can work very well, but the 3-way merges can catch you out sometimes unless you have an XML-aware diff tool. Here's why.
Using line-based diff and merge tools can result in invalid files or missing content. XML-aware merge tools avoid these problems.
2
Add a comment...

Joe Pairman

Shared publicly  - 
 
WYSIWYG wasn't just a comfort blanket. It helped writers do their jobs. We can't go back, but must make up for the missing components of author assurance.
WYSIWYG-style desktop publishing tools provided author assurance. We must apply three key components of that experience to structured authoring environments.
1
Add a comment...

Joe Pairman

Shared publicly  - 
 
Return of the humans

Technology is stabilising somewhat — yes, there are new development frameworks out all the time, exciting things with IOT, and steady advances in AI, but in the big picture, things are settling and consolidating a bit. What we now need to work on is making the most of human insight and judgement to harness the power of systems in a more transparent way.
1
Add a comment...

Joe Pairman

Shared publicly  - 
 
Enjoying preparing my "Culture Shock! Taxonomy" session for next week's CIDM Online Forum: https://forum.infomanagementcenter.com/forum-2015/agenda/pairman/

The point isn't that RDFa is verbose, by the way. It's that in the world of knowledge organization and the Semantic Web, data-enriched content is often published with its data in a standards-based, search-engine usable way.

I strongly doubt that this meme will go viral. ;-)

#markdown   #ditaxml   #semanticweb  
1
Add a comment...

Joe Pairman

Shared publicly  - 
 
Interesting. Good question Martin raises at the end — is it too late for Mercurial? I see it as a Betamax to Git's VHS: superior but losing the popularity fight. 
 
Lots of good +Mercurial stuff coming out of Facebook these days! You can now get sparse checkouts (creating only part of your working copy — makes "hg status", "hg diff", etc faster) and shallow clones (download only the latest version of your files). This makes it much easier to work with gigantic repositories. There's also a push-rebase extension that does SVN-like merging (rebase, really) on the server side to avoid races on push. Sounds convenient, but it implies a non-pull request workflow.

They've also released a great hgwatchman extension that makes "hg status" instantaneous by using inotify or fsevents to listen for filesystem modifications. Most other commands also get a speed boost since they call status under the hood.

Very exciting developments! The question is now if it's too late to matter for regular developers... Given the dominance of +Git, will Mercurial will end up being the hyper-scalable revision control system used by a few big corporations?
View original post
1
Add a comment...

Joe Pairman

Shared publicly  - 
 
Lots of writers — my dad included — complain about redundant usages. But redundancy often clarified or adds meaning. An unjustified fear of sounding redundant leads writers to adopt unnatural or abrupt phrasing, without the usual nuances & hints.

#WritingStyle #GrammarPolice #Writing
The grammar police may criticise our doubleplusungood writing, but our loyalty should be with more plausible authorities — our global audience.
1
Michael Andrews's profile photoJoe Pairman's profile photo
2 comments
 
Thanks for the comment, +Michael Andrews, and sorry for the late reply. It's an interesting point, and I understand that there's an argument that redundancy could actually reduce attention. If everything is only said once (or for minimalism per Carroll, some things are left for readers to find out themselves), then attention may increase.

I agree that redundancy helps when there are nuanced points to make; points that may not sink in after just one telling, or that require some extra verbal padding to make things really clear.

However, there is also a case for redundancy where different levels of language ability, especially among non-native speakers of the language under consideration, impact comprehension. The hardcore "say it once" proponents also argue that it makes translation easier and cheaper. I would counter that that is only reliably so if we take a mechanistic approach to language where one token = one idea. A more natural approach to redundancy may well help translators understand and convey meaning.
Add a comment...

Joe Pairman

Shared publicly  - 
 
Author assurance is probably the most important factor in structured authoring tool usability and productivity.

This post is the first in a series that I started writing three years ago.
As tools designers, implementers, or authoring team managers, we have to do a better job of assuring authors when they're doing their jobs OK, and provide clear information on how to fix things if not.
1
Add a comment...

Joe Pairman

Shared publicly  - 
 
Very useful info
 
Rich cards are upon us

My survey of all the structured data changes that Google rolled out on the 17th, including the introduction of rich cards.
On 17 May 2016 Google made a number of changes to their structured data site and introduced a successor to rich snippets, rich cards.
View original post
2
Add a comment...

Joe Pairman

Shared publicly  - 
 
Can JSON-LD replace RDFa for mentions in a page's human-readable content?

This is a question for Linked Data folks, such as +Manu Sporny. I understand the advantages of JSON-LD for page-level markup — saying that a whole page is about a particular organization in a particular location; has a certain author; talks about a product's price/rating, and so on. In these cases, JSON-LD is simpler and more developer-friendly than inline markup.

But how about when you want to mark up mentions of inline entities that aren't the main subject of the page, though? For example if you wanted to connect mentions of "Bluetooth" and "Wi-Fi"  to the related Wikidata concepts, or to the standards organizations behind those technologies? http://www.htc.com/us/support/htc-one-m9/howto/604060.html

I understand that you could just separately include JSON-LD markup that would connect to those concepts. But what if you really want to connect the human-readable mentions in the page directly to the concepts? Is RDFa (or Microdata of course) the only way? Or can JSON-LD talk about an inline span, say, by referencing an ID on that element?

Here's a piece I wrote on why you might want to use inline markup in this way (not the JSON-LD bit, but the kinds of applications for doing so): https://medium.com/@joepairman/connecting-with-real-world-entities-3af86b04a0b4

(+Michael Andrews, this relates to that discussion we had the other day on Twitter: https://twitter.com/storyneedle/status/606733122508615680 )

#JSON-LD #RDFa    #LinkedData  
1
1
Joe Pairman's profile photoManu Sporny's profile photo
7 comments
 
+Joe Pairman understood and agreed. Personally, I didn't have much hope for the XML+RDFa profile we put in the spec, but I still hope it'll catch on at some point and help a sub-community get semantics into their XML documents. :)
Add a comment...
Joe's Collections
Story
Tagline
Structured Information Architect
Introduction
My longer postings these days are on Shaping Information.
Education
  • University of Leeds
Basic Information
Gender
Male
Collections Joe is following
Work
Occupation
Lead Consultant
Skills
Information Architecture, DITA, Mobile Applications, Cross-Functional Team Leadership, Project Management, Information Design, User-Centered Design, SDL LiveContent, Mercurial, DVCS, XMetaL, Oxygen, Python
Employment
  • Mekon Ltd
    2014 - present
  • HTC Corporation
    2008 - 2014
  • Linguitronics Co.
    2007 - 2008
  • Lado Management Consultants
    2007 - 2008
Places
Map of the places this user has livedMap of the places this user has livedMap of the places this user has lived
Currently
London
Previously
Taipei - Leeds - Exeter
Links
Other profiles
Contributor to
Good quality, reasonably priced, clean, friendly.
Public - 5 months ago
reviewed 5 months ago
Friendly staff; clean, reasonably sized rooms.
Public - 5 months ago
reviewed 5 months ago
Very tasty food, even if trying a little too hard sometimes (grapefruit with haddock and asparagus didn't quite make sense!) Decor is cosy.
Public - 9 months ago
reviewed 9 months ago
Very family friendly. Tasty fish and chips. Kebab good too. Will definitely go back when I'm in the area again.
Public - a year ago
reviewed a year ago
21 reviews
Map
Map
Map
Simple rooms but clean, and the price is good. Breakfast is quite nice, even if a little overpriced, as so many continental hotel breakfasts are.
Public - 9 months ago
reviewed 9 months ago
Nice friendly staff, great location, clean rooms
Public - 9 months ago
reviewed 9 months ago
I like it. Friendly, welcoming staff and the food's fine for the price. Don't know if other reviewers were expecting Le Gavroche or something?
Public - a year ago
reviewed a year ago