Recently in blogging tools Category

I've been looking at the relationship between time, content and the flow of history over the past year or so. I've spoken or run sessions about it at eurofoo06, barcamp london 2 and xtech07.

I've been worried about our ability to understand our present after it becomes the past tense. Much of what we read makes sense for a limited timeframe. Time moves on and our context shifts to new current events. So heading back to try to understand the past becomes much harder. The analogy with well commented code springs to mind, maybe my desire is for comments on history. We understand much less of the humour in Shakespeare for this reason. Even moving five years into the past and it can be difficult.

The problem can widen out from here and I'll extend briefly, my context is different to your one, so I want different events and people related together. My family, my work and my friends and interests differ, so unlike code there is not one set of reference points. This is a substantially harder problem, so with that excursion, back to public news content.

I've been researching a proposed microformat hEpoch to deal with time based reference in news stories and I'll admit I've given up. The problem has becomes one of exactitude, stories are written to be read not used as code. Adrian Holovaty has written about this before in his essay on data as journalism. Stories reference past events using terms like "nearby" or "last month", an example story from the BBC. This means that it is currently impossible to make a microformat for linking these stories together. Yet if the data was there then all we need to solve is the time base reference, probably via a bit of POSH.

A good example of this is timelines, they are the pretty, but doomed product I've watched being made many times over. People work hard producing a nice interface and sometimes even separate content from presentation. Then the timeline is often left abandoned once the project or programme finishes. A small bit of care in terms of tags and time formats for news stories and they could auto update.

Time and geographic accuracy need to be flexible, with time we can be somewhat flexible, the hCalendar microformat allows just a year for dstart, which means fixing to a single year is possible, but a approximate date eg 1760s is impossible. I'm not knocking the hCalendar spec, just pointing out a space for new developments. the microformat principles are well defined and do deliver a workable product.

Geographic data is hard too, the geo microformat is superb for specific realtively modern locations, where there is enough context to determine the latitude and longitude for the place. However we frequently need to reference a region, like a city, but cities change over time. The centre of London is now Charing Cross, but it has shifted over time.

This post is entitled "a lament", but really there is a call to action in the tail. We can make this better, all the data exists, it just never makes it out of the editorial process. Similarly when blogging we omit lots of the specifics, as they make it easier to read. eg 06/07/07, how do you unpack that date? It is impossible to know for sure which year, month or day it is.

If you are a publisher then think of the value of being able to automate connecting story threads and managing content by time and location. If you are a tool provider, MarsEdit and MT or WordPress, then offer support for microformats. The same for the more corporate tool providers. If you are an author then think of being able to return to the item of content in months or years and understand the context. This is not easy, but it is possible. Next up, I'll be looking at search, then main user of this data.

Technorati Tags: history, microformat, provenance, time, semantic, journalism, story-telling

You'll have noticed the patchy content creation on take one onion over the past while. So I've decided to start a tumblelog on the lovely tumblr.com. It is called "save for later".
I hope this will mean the flow of content from me will increase and I can then write on here instead of backing up dozens of posts in drafts. So, this means new feeds, the existing take one onion feed continues, but if you want me in terms of twitter, flickr and del.icio.us then you should subscribe to save for later (rss).

Whilst I've been working on talkeuro, several issues have come to the fore. MovableType is great, but it is very focused on single author or publisher setups, at least from the point of view of template management. MT does a good job of managing actually writing new content for your site, the template management side is weaker.

Where MovableType has problems is in two key areas. The first is more generic and is about multiuser access to templates. There is no obvious way to hook something like subversion into MT so that several people working on site templates can do so and not worry about clobbering one another when saving changes. You can imagine a system where all the templates are pulled out of MT and on the filesystem, then managed via subversion. There might be hooks to do this, so that an update to subversion (or CVS) will trigger a rebuild in MT. Ideally I want to edit in my development environment, preview easily and then commit changes.

Secondly a way for MT to allow the same weblog content to appear under two sets of templates. eg www.talkeuro.com and staging.talkeuro.com would be really handy to come from the same database content. Thus allowing easy testing and bugfixing, whilst not live to the world. You can fake this, by having multiple templates and publishing to shadow directories, but it is fragile.

Another issue for me with talkeuro is multi-lingual support. Talkeuro has an english and a french version. To manage both of these I need to maintain two separate sets of templates, where ideally I'd have one template and if french use (Voyez cette section) and if english use (See this section). This is roughly how the internal localization of MT works. So as talkeuro gains languages and I want to add functionality to it, I need to hand edit each template or place each code fragment in a separate template module, which starts to get tedious.

I'd love to see these areas addressed in a future version of MT, 3.16 is a good step along the road, but more is needed to make it really scale for more complex CMS type applications.

I got a new phone at the weekend, a Nokia 6630 with 3G and a quite nice megapixel camera. To this I've added Lifeblog which is super, it collects together your pictures and messages placing them on day by day views. From this simple interface you can then select and directly post to TypePad or Flickr (howto), with others in the pipeline. The combination has really encouraged my desire for mobile blogging. I neither know nor care about the Windows application.

One other clever thing is the data transfer application which ships with the 6630, this lets you copy across your contacts and pictures from your old Nokia phone, a 6600 in this case. Excellent idea, it even copies the application to the old phone over Bluetooth. It'd need to do until Apple can sort out iSync with my 6630, it seems it doesn't even work on Tiger yet. Further thoughts on the 6630 to come later in the month, suffice to say I thought I was happy with the 6600 and only upgraded as it was free, but it is a marked improvement.

Something that struck me over the weekend, traditionally faceted classification is against a controlled vocabulary or structured ontology. Products such as del.icio.us and Flickr use tagging to create a similar effect.
The tags act as a nonhierarchical structure to describe attributes that a photo or url possesses, however the facets are emergent from the weight of the tagging. By this I mean that the number of people who use the same tag to describe the same url in del.icio.us then the more the url can be said to express this property. The knowledge space for these tags is also emergent and describes the interest space of the membership.
The del.icio.us popular page gives an indication of the popular pages by numbers of people listing them. There is some interesting analysis of the tags to be done in terms of their usage, by person, by use of same words across people, by looking at how tagging develops with experience. Other things that might be interesting are number of tags used per person, or per topic area. Now to work out how to do the analysis.
UPDATE: I have since, of course realised that I have described a "folksonomy", I was quite tired, so missed the connections my brain was making.

This posting is a community experiment that tests how a meme, represented by this blog posting, spreads across blogspace, physical space and time. It will help to show how ideas travel across blogs in space and time and how blogs are connected. It may also help to show which blogs are most influential in the propagation of memes. The dataset from this experiment will be public, and can be located via Google (or Technorati) by doing a search for the GUID for this meme (below).

The original posting for this experiment is located at: Minding the Planet (Permalink: http://novaspivack.typepad.com/nova_spivacks_weblog/2004/08/a_sonar_ping_of.html)

One of my desires from a desktop weblog editor is to sync the draft posts from my MT install and then take them away on my PowerBook and finally finish the missing posts of takeoneonion, of which there are twenty or more. Most of them rambling epics which the editing window of MT does nothing to encourage.
Ben Hammersley has answered the first part of this with the release of two extensions to the XMLRPC interface, mt.getStatus and mt.setStatus, which work with MT3.0D. Now I await the efforts of the Mac based weblog editing tool writers...

A few months ago I wrote about the need for a confirmation phase in trackback pinging. I was inadvertantly trackback pinging people when making a list of links entry and wanted to stop some of the pings.
Yesterday, Jeremy Zawodny wrote about an intranet trackback scenario with a company letting the cat out of the bag wrt to a competitor. In the comments it turns out that it is to become a feature of a new version of MT, see comment by Ben, so maybe my article and email did some good.
Maybe then they can also consider the Notify me of comments to this article feature which I wrote about a while back too. The intention is to let people who do not blog, or have commented on a blog article stay part of the conversation

quite fun


via blackbeltjones and foe romeo

Following up on the previous post, I've extended how I think that this third party notification system should operate. Euan comments on the edge of blogosphere, hopefully a mechanism such as this will blur the edge a bit.
I think that the correct mechanism would be to have each post within MT optionally have a series of email addresses associated with it which form a notification list for incoming comments and trackbacks.
eg I write an article about X, based on thoughts by my friend Igor. When someone writes a comment on the article or send a trackback ping, then Igor gets the same email notification that I do.
I think that this would be a great feature for a new Movable Type release and also with TypePad, as not everyone will want to have a weblog. This mechanism gives non-webloggers an ability to at least participate in the conversation, though it does not offer a clickable hyperlinked name. Thus there is no way for the non-weblogger to gain in reputation, except as a comment author on another's weblog.
UPDATE: Of course it is possible to add this to the normal commenting system. So that people could opt into being notified about further responses to a topic that they have already commented on. A sort of "notify me of further comments" checkbox.