digitalsean-paint

research for the new year

Been doing more reading and research than I've done in a while. Kate and I haven't really posted much about the impending event. Probably because it has been easier to take things one step at a time - get the practical stuff ready, and not worry about what happens next.

But eventually it comes time to do some reading and start to think about what happens next. My recent brush with doing reading around Nanophysics was not very successful - so here's hoping that bringing up kids will be an easier topic.

So You're Going to be a Dad, by Peter Downey, was a gentle introduction - friendly and comic view of impending daddy-hood to help over the denial and panic stage after the news.

Among the various leaflets from NHS was Ready Steady Baby, which I had been avoiding reading because of the large print and many bright coloured Q&A boxes. I was pleasantly surprised to find that there is a Ready Steady Baby website too.

Things are a lot closer now, and for some much more detailed advice I've had some rather heavier bed-time reading. The Baby Book, a gift from Barry, weighing in at 700 and odd pages has the appearance and layout of a university textbook. This is a complete approach. Reassuringly so, what with human bodies being complex things, and little ones seem to be doubly so.

digitalsean-paint

geek stuff - git and version control

A few weeks ago Mathie mentioned git, a distributed version control system. It's working really nicely from the command line, and is lightweight and really fast.

There's a video of Linus Torvalds talking at Google about why distributed version control is a good thing.

This whole entry is a reminder that I need to fix my broken geek project log - so that I can write entries like the following and not annoy everyone reading it.

You can probably stop reading about now, unless you really want to know the rest.

<geek level="95%">

It allows version control to be added in to track an existing directory structure. In practice git adds just one directory .git, into the top level of a project. All the version tracking data is held in ordinary files within this directory, so no more version control directories scattered all over the project. It also means if you take a copy of the files in .git you have a copy of the whole repository.

I've been able to add version control tracking onto existing projects to track what is going on when other people are changing files by hand, also been able to quickly pull in tarball from external projects and put the files under version control in a really pain-free way.

<geek level="98%">

<div lang="TLH" xml:lang="TLH">
tar xvf whatever-v1.78.tar
cd whatever-v1.78
git init
git add *
git commit

The distributed aspects are nice. Just tried a remote clone operation via ssh - and pulled a copy of the entire repository for our main project from the server onto my laptop in 132 seconds. Not just the current files - the entire history of the files.

git clone ssh://user@host/home/someone/somedir/myproj

Not only that, the copy of the laptop is a complete repository in itself without reference back to the server it can handle commits, branching and merging. Even cloning onto another computer independently.

Commited a couple of changes into my local copy and pulled the changes back and merged then with the copy on the server. It took a few seconds to transfer the changes over, and a quick git checkout to see the changes on the server.


</div>

</geek>

Radio Times style-waiver: Other distributed version control systems are available

Mercurial looks nice too. Ironically one of the dependencies for git is ASCIIdoc, which is being developed using Mercurial. Of course I am sure that the other external libraries, like most other software development projects, are still going to be managed by CVS or Subversion ... so it doesn't really matter.

Anyway, if Git is good enough for Linus to manage the entire Linux kernel and all the branches and patches,... then it's more that good enough to track a couple of thousand files on our project.

</geek>
digitalsean-paint

Gmail and Spam

Managed to grab a screenshot as my gmail spam folder ticked over to 9999 spam emails in the last 30 days. A few minutes later it was up to 10006. But of course that doesn't look nearly as good.
Gmail has caught a lot of spam this month

Spam
digitalsean-paint

Quechup and FOAF : taking control of the sharing contact details

Edit: this was the last blog post I had written in a while, it was written on 3 September 2007:

This was the day when Gordon brown had been in post for about a month, and we were still wondering about a snap autumn election. Acording to The Scotsman, the Jamacan opposition won the election, and Sighthill drew Angus in a preliminary tie of the 2008 Scottish Inter-County Top Ten bowls championship.

Wikinews tells me that on the same day that Iraqi peace talks between between Sunnis and Iraqi Shi'ites in Finland ended, and a yorkshire man completed his quest to sleep on 163 Scottish Islands

Oh ... and Sproglet was about 19cm, and Kate had a much smaller bump.

Instead of worrying about the large or small events in the world, it was also the day that I had a lot of Quetchup spam, and wrote a random rant about websites which ask for people's email address and password.

... and then I forgot to post it.




There seem to be quite a few spam emails, and emails discussing a social website called Quechup. The only problem is that they didn't mean to send these invitations. There may also be a question of what the website do with any third party contact information in the long term. There seems to be a problem with trust, addressbooks and social websites.

What are you expected to do when facebook or quetchup offer to provide an easy way of finding all your friends from your address book. They provide a nice looking interface and all you have to do is hand over your email password. We all know that this is wrong,giving your password to third party it is much easier for users than other methods.

Facebook seems to be doing it properly, not spamming your friends unless you tick the box asking them to, whereas Quetchup has sent email without permission to people's entire list of contacts including mailing lists. And I note that facebook makes the password based option the first choice and hides the 'import a file' option under 'other email clients'.

I guess we all need to think about whether we trust a service, particularly when giving them contact details of friends and business contacts. Should we do this at all?

It is more of an effort to export a CSV file of contact details, and more effort again to clean it out and then upload it to a site. But perhaps this is a more sensible way to go keeping control of what you give to such sites. Perhaps even the proper way of doing it is to have existing social websites allow you to export a friends file (in FOAF format), and allow you to import it again. Issue still is that the obvious key identifier is an email address, and sharing an email address allows unscrupulous websites to email other people without your permission.

Fortunately the blogging world is good at picking this sort of thing up, and a google search for quetchup is full of blog entries like this one at the moment.
digitalsean-paint

keeping up with my own sites - rss readers

The sites that I am working with are not providing any news feeds at the moment. Oops. This is a problem which I have meant to look at for a while, and after my news feed post this morning I really should do something about this. Without useful feeds, his actually means that I am not keeping efficient track of all of the news and course activities that we are maintaining at work.

I would really like to fix this as soon as possible, especially the public facing news page. We cannot expect people to visit the news page regularly just on the off-chance that something new has happened. I think it is essential to allow people visiting a site to subscribe to information in a flexible way. Tony Hirst from the Open University has some useful thoughts on over at OUseful Info on how education institutions can create opportunities for users to access material in interesting ways using RSS. (including a longer piece entitled 'we ignore RSS at our peril')

For the main site, the next step is to set up a quick screen scraper to grab the current news page, and then to look at how to make this more manageable for the main site.

For the VLE, I need to figure out how to get mooodle to give out private feeds of course activity, forum posts and calendar events in a useful way. It does not do this well at all at the moment. There is a technical and social issue to consider whether to and how to provide semi-private feeds for closed courses, and forums. I note with interest that there is a student project for moodle looking at this in the google summer of code which should be completed in the next few days. I look forward to seeing how far the project has gone.
digitalsean-paint

keeping up with all the noise - rss readers and video

Again this week I was surprised to find a couple of technically literate people who spend a lot of time working and away from work using computers who were not using news feed aggregators (commonly known as feed-readers). For me, the RSS feed is an essential way to keep in touch with a growing information space in which I am able to keep track of the things that a lot of different people and a lot of websites are publishing. The common craft show have a really nice video explaining rss in plain english

Through one of my news feeds I see that the news feed aggregator bloglines is getting a little bit of a ajax makeover which you can see on the beta site. Nothing too radical, just making use of the start page in a nicer way, and making things a bit faster. I like bloglines because of the integration with both a mobile friendly site, and also the fact that the offer an API which allows integration with third party clients like egress for pocketpc. Really nice only having to read the same news once no matter how I read it. Also while travelling it was really nice to be able to load up all my feeds onto my phone, and work my way though all my news offline.

Kate still makes a lot of use of netvibes and the tabbed interface, and ability to browse over a lot of feeds is looking quite tempting, especially with their newly released mobile friendly version.

By making sure that your site provides news feeds for articles, comments and all kinds of other activities you will allow people to subscribe and track things in a flexible way. They can get the information on their own terms, and you get an opportunity to feed things out to them and draw them back into the site again when there is something relevant and new.



<!-- #include RSS tutorial post -->

There are a lot of different tools out there to help you manage news feeds, and they allow you to track things on your own terms - on the web they include : bloglines, google reader, netvibes.

On the desktop newsgator has a major presence with products for Windows, Mac, and standalong and outlook integrated feedreaders. Open source side : RSS Bandit looks interesting (includes NNTP reading, synchronisation across different machines), and for the mac Vienna might be worth a look. There's also RSSowl which is cross-platform.

Of course the list on wikipedia entry on news aggregators is a lot more comprehensive.