Want to find the most elite twitterers… Or even those in your home town?
Archive for February, 2009
Topaz Filer 2009 – launched!
It’s been a *lot* of work, but I am pleased to say that we have today launched Topaz Filer 2009, the e-mail filing system that works with what you already have.
The Web site can be found here – www.topazsw.com. Please check out the fun video on the home page (we had a lot of fun putting it together).
The other thing we have done that we think is really special is created a “hosted trial” system. We have a bunch of VMs on a server in Kent with a full end-to-end version of the system on it. To try out the system, go to the Web site, click on Test Drive then Hosted Trial. This will give you instructions on how to get connected. When you’re done, the server you used is wiped and restored to factory settings. (I’ll blog more on how to do this at a later time.)
Please do take a look and contact us with any feedback.
Microsoft’s severance pay debacle
Microsoft have apparently overpaid severance to a batch of fired employees… and want their money back.
Psion/netbooks
Personally, I think Psion were awesome having been a major fan as a kid of Psion Organisers and then into other devices like the Series 3, but I am not a big fan of them looking to protect their “netbook” trademark. Personally I think that ship has sailed/bird flown, etc.
So why not spend some time over at Save the Netbooks that is looking to gather support in keeping the netbook name in the public domain.
WSS 3 and unbelievable database space consumption
I am sure that from time-to-time I come up with lousy, unscalable designs, but here’s a precis of a problem I have just run into with WSS 3.
For our new e-mail filing product I wanted to create a decent size sample database, which I duly did – 1,000 customers and 2,500 projects. For the demonstration to work against the sample I needed to create a SharePoint site per customer and another per project, 3,500 in all.
Obviously I wasn’t going to manually create these, so I write something that uses the SharePoint API to do it for me.
With 1,000 customers created and about 200 projects I started to get timeouts. Looking at the machine it was thrashing like mad. Thinking that I knew SharePoint wasn’t brilliant with masses of sites, I dug into it a little.
Having already decided to dump 800 customers and go with a smaller sample, I thought I would find and delete the content database. By default WSS uses “Windows Internal Database”, which is a quasi-hidden instance of SQL Server 2005. When I found the files, I had a 1.7GB temp file, a 2.8GB data flie and a 2.7GB log file – i.e. 7.2GB of data being used to store 1,200 sites. (Albeit all by 2.8GB of that is scratch.)
Honestly, can it really me that each SharePoint site uses 2.4MB of space? And these were not complicated sites – they were the basic team site with *no data*.
Exchange ActiveSync licensed to Google
A very good article by Paul Thurrott on licensing Exchange ActiveSync to Google.
Especially apposite for as at the moment as I think about moving away from Exchange and moving our e-mail platform over to Gmail.
What’s particularly interesting about the way that Microsoft has done this is that they have licensed the technology to allow Google to plug into Exchange on the back-end, cf. to most licensees of this technology who look to get their devices to talk to Exchange.
Random company names
So, for Filer we needed a sample database of customers and projects. To create a random list of 1000 customers we took our marketing database, split up the company names into words and mixed them all together and randomly appended Ltd, plc or LLP to the end.
The weird thing is that you get a list of very legitimate sounding names, for example:
Auerbach Biosciences Ltd
Mini Aircargo Ltd
Eveden Leisurewear plc
Dellstone Paul plc
Alphatasknet Psychological Ltd
Clayton Spencer Ltd
Rosjohn Edwards plc
Lennox Hess Ltd
Wishing Pharma Ltd
Brownings Faithful Ltd
Essentis South Ltd
Newage Pharmaceuticals plc
Hifi Nelson Ltd
Multimatic Searches Ltd
Primasolve Dene plc
Winchester With Ltd
Cowan Northampton plc
Dawsons English Ltd
Hopefully the names that we are using aren’t actually real company names, but I was struck by just how you read that list and go “they sound familiar” when actually it’s just a whole load of random names.
We also needed a list of fake projects, and so to do this we took a random list of (clean) English words and did the same, but the results were even weirder, this time creating what look like Haiku, viz:
Question Interest Friend
Structure Support Delicate
Whistle Violent Science
Flower Circle Curtain
Against Pencil
Orange Straight Expansion
Against Kettle
Harbour Winter Powder
I presume this is lighting up the bit of my brain (and your brain) that looks for patterns. You assume there must be some sense in it and approach it from that angle.
domaintools.com – new to me…
Pretty impressed with the data that you can get out of DomainTools – e.g. http://whois.domaintools.com/mbrit.com.
ClamWin and anti-virus comparisons
We just took “delivery” of another dedicated server for a project we were doing and this time used a differnet hosts to the ones we usually use, mainly because for this project we needed cheap. (I’ll post who they were once I have some confidence they are any good!) The install ClamWin free-of-charge, which I had never heard off.
ClamWin bills itself as “open source anti-virus”. For me, anti-virus seems like the sort of thing that *needs* to be commercial because surely you need to pay a team of developers to research and build fixes for new malware. I wasn’t, and am not, immediately sure how you do that on a free model. Googling around for them I came across their Wikipedia article.
At the bottom they provide a link to Virus.gr (http://www.virus.gr/portal/en/), which runs a comparative test of the different products. ClamWin comes in 37th with a 55% detection, which I guess is better than 0%, but still… not good.
What surprised me more was the rest of the list. The Top 10 were:
1. G DATA 2008 version 18.2.7310.844 – 99.05%
2. F-Secure 2008 version 8.00.103 – 98.75%
3. TrustPort version 2.8.0.1835 – 98.06%
4. Kaspersky version 8.0.0.357 – 97.95%
5. eScan version 9.0.742.1 – 97.44%
6. The Shield 2008 – 97.43%
7. AntiVir version 8.1.00.331 Premium – 97.13%
8. Ashampoo version 1.61 – 97.09%
9. Ikarus version 1.0.82 – 96.05%
10. AntiVir version 8.1.00.295 Classic – 95.54%
But, wow, basically what’s the target here? I would say that a 2% miss rate was too high for this problem. Now, most of those I had not heard of, but what really surprised were:
20. McAfee Enterpise version 8.5.0i – 86.57%
23. Norton 2008 – 83.34%
…and Trend, which we use in the office:
33. Trend Micro Antivirus+Antispyware 2008 version 16.10.1079 – 67.28%
…and seeing as Trend have just started nagging me about our yearly subscription being elapsed, goodbye to them. Oddly on our production servers we use Kaspersky, mainly because I found Trend a nightmare to manage and irritating and Kaspersky not so.
The latest Virus.gr tests are here:
http://www.virus.gr/portal/en/content/2008-06%2C-1-21-june. Others can be found from the top navigation. Digging around, Kaspersky has been around the top in previous years.
Makes you think thought, because you assume AV from the big name providers would perform the best…
REST vs Web Services
Most of the projects that we undertake at MBRIT involve integration of some sort between two systems. Whenever I write up the technical proposals for the systems, I’m always faintly embarassed to recommend hooking the two systems together using XML documents submitted over HTTP/HTTPS. It always *feels* like the wrong way of doing it, which I presume is way back in 2001 when .NET was getting started Microsoft pitched it primarily as a way of building Web services. That primal part of my brain that remembers that early history has a reaction that using non-SOAP methods for moving data about in that way is unprofessional.
However, given that almost eight years of water has passed under the bridge, and actually a lot of end-points are very non-SOAPy. I can’t remember the last time I came across a non-Windows end-point that uses SOAP. Most of them use what I have recently discovered is called a REST interface – i.e. post a document over HTTP/HTTPS.
I also happened across this article with Tim O’Reilly on Web services cf. REST. Worth a read… http://www.theregister.co.uk/2006/04/29/oreilly_amazon/.
