Google aquires JotSpot, JotSpot Server product dropped

Google has acquired JotSpot (coverage also on the Google Blog)

I’ve been evaluating wikis for rollout at work for a good while now, and the overwhelming impression I came away with was one of disappointment. Almost all of the hosted wikis are very poor indeed. You see the editing on Wikipedia? That’s almost as good as it gets. JotSpot essentially provided hosted wikis with wikipedia-like features (i.e. not many) but with added enterprisey bolt-ons like MS Word import, project management, bug tracking, spreadsheets and so on. This is where the value in Jot lies, not in the actual wiki software, which is distinctly average.

Jot had just put out a beta version of JotSpot Server – a VMWare-based system you could download and use to run your own hosted JotSpot install. We tried it out as part of our evaluation of many the different wiki tools. Other than being very hard to set up and actually get going, it really did provide companies with a way to get JotSpot inside their firewall. No more. From an email today:

At this time, we are no longer planning to offer JotSpot Wiki Server.
You can continue to use the beta version for the 90-day Term you
agreed to in the click-through JotSpot Wiki Server Beta Agreement
but we are no longer providing support for the product. (The Term may
be different for those of you who signed a paper version of the beta
agreement.) Once your Term has expired, per the agreement, we ask you
to destroy all copies of the software you have installed.

So now, if you want something inside your firewall, it’s pretty much Confluence or nothing (we chose Confluence anyway btw – it’s amazing and far, far ahead of everything else).

According to a JotSpot employee:

I work for JotSpot and just wanted to make a quick clarification that the JotBox product line has not been cancelled. All new business has been suspended for now as we join Google including sales of our JotBox hardware appliances. That being said, we currently have a number of customers actively using JotBoxes which we will continue to upgrade, support, and maintain.

Well, that doesn’t quite match up with the email I received from the Sales team as I quoted above: we are no longer planning to offer JotSpot Wiki Server. I can understand some confusion, given what’s going on, but this is kind of important to both Google and Jot’s customers, surely?

Real ads – probably not worth it

2lmc comment on the current hatred of the new Sony Bravia advert, and link to the Media Guardian article (behind a regwall, so you’ll need BugMeNot) which tells us that

… the “Paint” commercial cost about £2m to produce, including the cost of relocating residents, 60 people cleaning up paint for five days afterwards and an unprecedentedly large fee for the talented director Jonathan Glazer.

I thought it was CGI.

Atom slugs

9.6 The Slug: Header

Slug is a HTTP entity-header whose value is a “slug” – a short name that can be used as part of the URI for a Member Resource.

When posting an entity to a Collection to add a new Member, the server MAY use this information when creating the Member URI of the newly-created resource, for instance by using some or all of the words in the last URI segment. It MAY also use it when creating the atom:id or as the title of a Media Link Entry (see Section 9.5.).

Servers MAY ignore the Slug entity-header and MAY alter its value before using it. For example, the server MAY filter out some characters or replace accented letters with non-accented ones, spaces with underscores, etc.

What a terrible idea.

I will be recommending internally that we never, ever use this header.

TV gets lazy

Amy whine whine Winehouse picture-in-picture

Television is so bloody confusing and irritatingly designed.

I finally left the 30% and got a Freeview box with built-in 80GB PVR at the weekend, and gosh isn’t it exciting? Apart from the new-found ability to watch Batfink whenever I like, I can now watch Torchwood on broadcast night. This is very exciting, but I suspect the raft of channels which have now been opened up to me and the ways of navigating between them are a giant wave which will shortly break on the beach of my mind.

So far, I’d give it four out of five. Very impressive for £100.

Wiki Comparison Matrix

Initial wiki comparison table.

Some notes:

  • Can’t be Windows-platform, otherwise technology-neutral
  • Er, some other things.
JSPWiki MediaWiki Confluence JotSpot XWiki
support multiple wikis (or equivalent paradigm, like wikifarming) Unsupported Unsupported Supported Unsupported Supported
WYSIWYG editing plugin Unsupported Supported Supported Unsupported
page history Supported Supported Supported Supported Supported
LDAP authentication Supported Supported Supported Unsupported Supported
export into at least one of: XML, PDF, DOC, plugin Supported Supported Supported Supported
email notification plugin Supported Supported Supported Unsupported
fine-grained access control (private, editor, reader for wikis and pages) Supported plugin Supported Supported Supported
users can grant and revoke access to their wiki in a per-wiki administrator role Unsupported Unsupported Supported Unsupported Supported
system-wide skinning/themeing Supported Supported Supported Supported Supported
create page Supported Supported Supported Supported Supported
delete page Supported Supported Supported Supported ?
page diff Supported Supported Supported Supported Supported
attach file Supported Supported Supported Supported Supported
delete file Unsupported Supported Supported Supported ?
file versioning Supported Supported Supported ? ?
ability to use API for automatic group creation Supported Supported Supported Unsupported Supported
add user to a group wiki Unsupported Unsupported Supported Unsupported Supported
remove user from a group wiki Unsupported Unsupported Supported Unsupported Supported
make user group wiki admin Unsupported Unsupported Supported Supported Supported
Pricing Free, open source Free, open source Prices, commercial (for our needs, c. £2000) Unannounced price for server product (c. £100/month hosted), commercial Free, open source

Sorry about the whole “paradigm” thing.

Dreamhost as a backup provider

The more I read, the more tempted I am to bite the bullet, buy “web hosting” space on DreamHost, and just start rsyncing all of both my and my wife’s data up there.


  1. ADSL lines have terrible upload rates
  2. Remote backup means we may not have access if the network is down (which does happen quite regularly)

Overall though, I think remote backup is the way to go. I am far, far too lazy to do regular backups to DVD, and definitely too lazy to copy my backup disks every few years in case of media failures. Outsourcing the lot makes a lot of sense to me. I’d probably try and keep a local DVD-based backup as well, but that would be just for convenience rather than the master copy.

Encouraging wiki usage in your team

Anthony just asked a great question in one of my recent blog posts, and I thought my hasty answer was worth making into its own post (given that I’ll soon be giving my second talk on this subject):

If you’re trying to make more people use the wiki, then that’s exactly what you need to do – I’m guessing you’re in a small team, so get one other person interested and get them to start putting information on the wiki. For example, do you have a list of projects plus a description of what they are, who is on them, their status and so on? That should be on the wiki.

Are you doing any work that most people don’t know about? Like you’ve posted that Javascript code on your blog recently – put it on your wiki linked from a “code snippets” page.

Depending on your team and exactly what you do, maybe a “tips and tricks with PHP” page. Make it an internal point of reference for the rest of the developers and encourage them to add their own advice to everyone else.

The world’s your oyster in theory, but the content does need to be compelling in the first place to get people looking at it.

I do of course also maintain a fabulous list of wiki-related resources in which are well worth checking out

Listening to OPML changes

At work we’re putting together a wiki page which lists which feeds we should probably all be subscribed to. This mirrors the increasing amount of RSS we’re creating both in the team and across the University as a whole.

We’re writing the list of feeds as a bulleted list which links to both the main site and the feed you’ll need to subscribe to. Obviously this is just an XOXO list.

What this means is that we can use one of the many tools (hopefully Les Orchard will have done this for me) to convert this list to OPML which anyone can then subscribe to in their aggregator.

The problem is this:

  •  we have to use desktop aggregators because many of the internals feeds use HTTP authentication and everyone also maintaine their own subscriptions
  • the list of feeds we should be subscribed to will change occassionally (especially over the next few weeks)
  • a change in the feed list means either each person will have to re-import the entire list, or identify the new/changed feed(s) and add them to their subscriptions.

What we really need is an aggregator which can subscribe to an OPML file (yes yes, this is the OPML Reading List meme from ages ago).

At the moment, that means BlogBridge. BlogBridge is an open-source Java-based desktop aggregator (which actually responds quickly enough to feel native, although like all Swing apps it doesn’t look anything like a native application).

I’ve tried to use BlogBridge in the past and not really liked it, but back in July I promised D’Arcy that I’d take a look at it again, and this looks like the time to make good on that promise (especially now that he’s been promoted to expert guide).

An initial go shows that it imports my FeedReader OPML just fine (although it breaks up each folder of feeds into its own representation which it calls “Guides”) but that it doesn’t support “river of news” style reading. That is, I can’t select one of my “guides” and instantly see all the unread articles – I have to navigate between individual feeds (even if it’s just by pressing space). This is definitely not the behaviour I want, but this blog post suggests that I should be able to create a SmartFeed of unread items for each guide and read those instead, but again this probably isn’t what I want. I don’t even want to see the list of feeds – I just want to click each ‘Guide’ and view the unread items from all the feeds in that guide.

I guess it’s lucky I’m a Java developer then, right?