Category Archives: Windows

Eventcountdown.com

I have just launched at new website at eventcountdown.com.  This website contains free countdown clocks,  for both Windows and web, to allow you to count down the days, hours, minutes and seconds to any event. It can be a wedding, a birthday or a major cultural or sporting event. Are your children wondering how many days, hours, minutes and seconds it is to Christmas ? Wonder no more! You can also use it to count up from an event e.g. giving up smoking or Perl.

web countdown clockweb countdown clock for Mac, iPad, iPhone and Android

windows countdown clockWindows countdown clock

The hope is that a significant number of people interested in planning events will go to evencountdown.com for a free clock and a small percentage will click through to PerfectTablePlan.com and buy my software. I have no idea how successful this will be. I certainly don’t expect a quick payback on the investment. But hopefully it will pay for itself in a few years, and then anything after that is pure profit.

This project has also been a small scale experiment in outsourcing that might lead on to greater things.

  • The web design and CSS/HTML coding was done by Sergey Pozhilov of μISVStyle.
  • The Windows countdown clock was done by Milan Marusinec of VectorGraphica.
  • The Javascript for the web countdown clock was done by Paul Kossowski of Dolphin Futures.

They all did a great job and the total cost was less than I would have paid for a couple of ads in event industry newsletters (which I tried recently, with fairly miserable results).

I have quite a few ideas about how I can improve eventcountdown.com, but I wanted to get something out there ASAP. After all if you aren’t embarassed by v1.0 you didn’t release it early enough. I would be interested to hear any feedback. Backlinks to eventcountdown.com would also be very welcome. ;0)

milan@crossgl.com

Speccing my dream development PC

My main development machine is a 4 year old Dell desktop, PC running 32 bit Windows Vista. Time for a new PC. I have been doing a bit of research, mostly by looking through back copies of PC Pro magazine (which I recommend,  by the way). I am speccing it out here on my blog in the hope that I, and anyone else thinking about a new PC, can benefit from my research and (more importantly, given my limited knowledge of hardware) the assembled wisdom and experience of readers of this blog.

A system failure could lose me several days work and a lot of mental energy to sort out, so reliability is my number one requirement. After that I am interested in performance, particularly speeding up compilation and linking, as this has a significant effect on my productivity (my table planner software is now well over 100k lines of C++ code). Getting a fast machine also means it will last longer before it becomes obselete.

Form factor

Tower desktop. I don’t like using laptops for extended periods and there is plenty of space under the desk for a tower case. Having a bigger chassis means more options for upgrading and hopefully less chance of overheating. I like to have my PC physically locked to a ground anchor to reduce the possibility of theft. So the case needs to have a Kensington-compatible security slot I can attach a security cable to. Some LED bling wouldn’t hurt.

OS

Windows 7 64 bit Professional (or Ultimate is if I decide to use bitlocker drive encryption). Windows 7 is increasingly what my customers are using. 64 bit will allow me to address >4GB RAM.

As I develop for both Windows and Mac, I could use a Mac for my main development machine. But I am more familiar with Windows and its associated apps, you get more bang for your buck with a PC and I’m not keen on Apple’s authoritarian attitude to developers. So I prefer to use Windows for my main development machine and use a Mac Mini for porting.

RAM

8GB of DDR3 RAM. Maybe more, depending on cost. But I am not sure whether having more than 8GB of RAM will make much of a difference to real world performance. DDR3 memory seems to be fairly standard on high end PCs. I can’t imagine the clock speed of the memory (e.g. 1,333 MHz vs 1,600 MHz) affects development related performance much. I haven’t seen any data on this.

Storage

Possibly a 128GB Crucial RealSSD C300 SSD for storing \Windows and \Program files and a 600 GB 10,000 RPM WD Velociraptor HDD for storing everything else. SSDs seem to be increasingly the way to go for storage that is predominantly read-only and the prices are coming down. 128 GB should be ample for Windows and lots of applications. But there have been issues over the reliability and performance of some SSDs, notably those with Indilinx controllers. anandtech.com, Jeff Atwood and PC Pro all rate Crucial SSDs well. The WD Velociraptor is one of the fastest HDDs around and 600 GB is currently the biggest one you can buy. It is pricey though at around £0.30 per GB, with 7,200 RPM drives around £0.05 per GB. Especially if I buy an extra one to keep as a hot spare.

I am still making up my mind on this though. Having two separate drives means an extra possible point of failure. Also some people say that, while benchmark results are impressive, the difference in performance between SSD and fast HDDs isn’t that great in real world scenarios. I also wonder whether a 10,000 RPM HDD is going to be as reliable as a 7,200 RPM HDD, such as the Samsung Spinpoint F3. I don’t know of any independent data on HDD reliability between different makes and models. In the absense of such data I guess the next best bet is to look at how long the manufacturer warranties are. You need to be fairly confident on reliability to give a 5 year warrantly.

I am also considering encrypting the drives for extra security. This will keep my data secure in case of theft and it also means I can return a defective drive under warranty without worrying about the security of the data on it. The main alternatives are Truecrypt and Windows 7 Bitlocker. Tests show Truecrypt AES encryption adding a 5-12% performance penalty. Anecdotal reports seem to show that the difference is hardly noticeable in real world use, especially with a fast CPU. A comparison of Truecrypt and Bitlocker from April gives Bitlocker a small advantage for speed, but favours Truecrypt for its flexibility. Also Truecrypt is free, whereas Bitlocker is only available if you buy Windows 7 Ultimate.

My current PC has 2 disk HDDs in RAID1 (mirrored) configuration. This was intended to decrease the chance of data loss, but it has been a huge headache. I have had to replace one or other of the RAID1 HDDs 4 or 5 times while I have had this PC. I can only assume that RAID was responsible for this catastrophic failure rate, so I certainly won’t be going for RAID again.

Media

Any reputable make of DVD drive that can write dual layer DVDs should be fine. The 8.5 GB capacity of dual layer DVDs is useful for backups. I don’t see any need to pay a premium to get Bluray.

Motherboard

A Gigabyte or Asus motherboard with support for USB 3.0 and SATA/600. I have just bought a USB 3.0 external HDD for backup so I would like to make the most of the additional USB 3.0 performance and a fast SATA connection is necessary to get the most from a fast HDD. Gigabtye and Asus motherboards seem to do well in the PC Pro magazine tests.

Power supply

Any respectable make should be fine. If a power supply fails the power surge can wreck the whole machine, so it definitely isn’t worth risking a cheap and nasty power supply to save a few pounds.

USB ports

Lots, including some USB 3.0 ports.

Graphics card

AMD Radeon 6850. I don’t play a lot of computer games, but I want a graphics card fast enough to give me the option. I would also like to have the option of multiple monitors, so it needs to have at least 2 DVI outputs. PC Pro magazine are recommending the AMD Radeon 6850 as exellent value if you don’t need the fastest possible card.

CPU

Intel Core i5-750 (4 cores). Raw processor speed is obviously important, but you pay quite a premium to get the very fastest chip. The i5-750 seems like a good balance between price and performance. Multiple cores are useful for running VMs and compiling (apparently Visual Studio 2005 requires a tweak to use multiple cores for compilation). Quiet fans would be nice. I’m not interested in overclocking as I worry about the effect this could have on reliability.

Networking

Gigabit Ethernet.

VDU, keyboard and mouse

I will re-use my existing monitor, keyboard and mouse.

Software

The less crapware pre-installed with the OS the better.

Warranty

The longer and more comprehensive the better.

Manufacturer

I have neither the time, the inclination nor the aptitude to build a PC myself. So I will be looking for a UK company that can build a PC close to the above spec. I had a terrible experience trying to buy a media PC from gamingpc-guys.co.uk earlier this year (I had to do a chargeback to get my money back when neither PC or refund had appeared after a month), so I certainly won’t be approaching them. Thankfully they seem to have gone out of business anyway. Given the problems I have had with the current Dell and the ludicrous cost of Dell replacement parts, I won’t be going for Dell again either. UK PC manufacturers that regularly do well in PC Pro tests include:

So I will be investigating how close a system they can provide me to the above spec and for what price. Interestingly all these custom PC companies seem to have products aimed very much at gamers, not developers. Given that there are a lot of developers in the UK that seems surprising. I guess most developers get their PC bought for them by IT departments and don’t have the luxury of a custom built PC. I think these companies are missing a trick by not having a PC aimed at developers amongst their base specs.

If you had roughly £1000/$1700 to spend on a development machine, what would you choose differently from the above and why?

App stores set to dominate future software sales?

Following the success of the iPhone app store (over 6 billion downloads to date), app stores are becoming more and more of a feature of the software landscape. In case you missed it, Apple announced yesterday that there will be an App Store for Macs  ‘within 90 days’. In summary:

  • The Mac app store will be tightly integrated with Mac OS X, including automatic install and update.
  • There will be restrictions on technology, for example Java apps will not be allowed.
  • Apple will keep 30% of any revenue from sales.
  • $99/year subscription for developers.
  • Developers will still be able to sell their software outside the App store.

It is easy to see why Apple would want to do this:

  • A potentially huge new revenue stream from third party Mac software sales.
  • They get even more control over the customer experience.

And this could have advantages for Mac users:

  • Simpler payment and installation.
  • Screening out of low quality apps and malware.

And potential advantages for Mac developers:

  • Mac users might buy more software if it is easier to do so.
  • One main channel to concentrate your marketing efforts on.
  • Some of the boring infrastructure of selling software (licensing, shopping cart etc) can be taken care of by Apple.

But the disadvantages are all too obvious:

  • Your app could be rejected outright. And you won’t know until you submit it for approval. Apple are judge, jury and executioner. The iPhone app store has been infamous its capricious and opaque approval process.
  • 30% is a huge chunk of revenue. Typical payment processors take 5-10% of revenue. Where the new app store cannibalises existing sales (and it is hard to see that it won’t) vendors will lose 20-25% of existing sales revenues.
  • New apps and updates will be delayed by days or weeks as they go through the app store approval process.
  • A single centralised app store is likely to make it harder for niche/long-tail apps to make any sort of living. Certainly this is what seems to be happening in the iPhone App store.
  • Apple are control freaks and have traditionally taken a rather heavy handed approach with developers, including the liberal use of NDAs. The app store will give them even more control.

And worse might follow:

  • Apple makes a lot of their money from selling over-priced hardware. It may be in their interest to drive software prices down so they can sell more hardware. $5 is considered expensive in the iPhone App Store.
  • This could be the first step to making Mac OS X a closed system, like iPhone, where only Apple approved apps can be installed.

I guess they can’t piss off developers too much – a computer without third party applications isn’t going to be very attractive to customers. But I am finding it hard to work up any enthusiasm for a Mac app store. If it is successful I can either be in the store and give up a lot of freedom and cannibalize exisiting sales at a much lower margin, or stay out and be shut out of a large chunk of the market. It isn’t an attractive choice. As my app is written in C++/Qt, rather than Objective-C/Cocoa, I am not even sure that it will be eligible for inclusion in the store. I could just abandon Mac OS X, but Microsoft is also rumoured to be working on their own app store (despite the failure of DigitalLocker). That is a truly terrifying prospect given the awfulness of their ‘Works with Vista’ approval process (I speak from personal experience).

Suddenly web apps are looking more interesting.

Why I won’t be bothering with the Windows 7 logo program

not compatible with Windows 7Am I the only one being totally bombarded with ‘Give your application the green light’ and related emails from Microsoft and its minions? I must have had at least 30 so far. I took a few minutes to list my product in the Windows 7 compatibility guide (beware, cheesy audio). But that is all I intend to do.

I went to the trouble of getting the ‘works with Vista’ logo in 2007. The process was very broken:

  • The winqual and Partner websites give me “certified by unknown authority” warnings.
  • The Winqual website didn’t work at all in FireFox.
  • There was a complete lack of clear guidance about what you needed to do next at each stage.
  • The documentation was very poor, with broken links and much of it completely out of date (i.e. “more information will be available in 2006”, this was 2007).
  • It took me about an hour of rummaging around on the winqual site and an email to tech support to work out that you can ‘sign’ the legal documents online (you have to tick a permissions checkbox on a separate page and do a few refreshes).
  • You had to download a signing tool. It was in a zip file with a password. They didn’t tell you what the password was! Luckily I already had signcode.exe installed.

It was easily the most frustrating thing I have done in my career as a microISV. For that I ended up with an ugly ‘works with Vista’ logo (that probably just made customers think my software didn’t work on Windows XP), an entry in Windows marketplace (I already had one) and a Verisign authenticode certificate that I wasn’t allowed to use to sign my software.

Hopefully Microsoft have cleaned up their broken logo process since the Vista launch. But the benefits of the Windows 7 logo program seem slim:

  • a “Compatible with Windows 7” logo (prettier than the ghastly “works with Vista” logo admittedly)
  • 30 Partner Points for use in the Microsoft Partner Program (I have no idea what I would want those for)
  • some PR templates (when did having a near identical press release to thousands of other companies become a benefit?)
  • priority Listing in the “Windows 7 Compatibility Center” (I doubt any of my customers know or care about this)
  • Windows built-in error reporting (I rolled my own, thanks)

Worst of all I would have to buy another overpriced Verisign certificate to authenticate myself to Microsoft winqual, even though I already have a perfectly valid authenticode certificate from Comodo.

I think I’ll pass.

Getting ready for Windows 7

windows-7I am currently downloading the new Windows 7 release candiate. It is 2.36 GB and the ActiveX download control predicts it will take around 10 hours in total (NB I am also doing a large FTP upload, which is probably slowing it down considerably). Once the download is complete I intend to install it in a VM to test my table planning software.

You can download the latest Windows 7 release candidate here. It is free to download and use, but  it expires on 01-June-2010 and will shutdown every 2 hours starting on 01-March-2010 (insert your own joke here). A couple of points to note from the Windows 7 FAQ:

While the RC is stable and has been thoroughly tested, it’s not the finished product. Your computer could crash and you could lose important files. So please back up your data and please don’t test the RC on your primary home or business PC.

When you use the RC, your PC sends information to our engineers to help them check the fixes and changes they made based on Beta tests.

Microsoft hasn’t officially announced when Windows 7 will be for sale. But there are rumours that it is planned for October 2009. Will you be ready?

Mac OS X market share accelerates in 2008

2008 was a good year for Apple and Mac OS X. According to netapplications.com data (via sharewarepromotions blog) Mac OS X’s share of the OS market increased from 7.31% in Dec 2007 to 9.63% in Dec 2008. That is a 32% increase in market share during 2008, compared to a 22% increase during 2007.

macosx_market_share_2007_2008

Windows market share fell from 91.79% to 88.68% in the same time. While Mac OS X’s annual gains are impressive, it has a long way to go to catch Windows. 15 years if you project the 2008 gains forward.

macosx_vs_windows_market_share_2007_2008Of course, it is highly questionable to project 15 years from a single year of data, but it gives an idea how much work Apple still has to do.

I sell table planning software for Windows and Mac OS X. Mac visitors to my website have followed the general trend, up from 7.41% in 2007 to 8.5% in 2008 and accounting for around of 10% of visitors at the end of 2008.

macosx_visitor_percentage% Mac visitors to http://www.perfecttableplan.com

My data also shows that Mac users are twice as likely to purchase my software as Windows users (I have heard similar figures have reported by others). So Mac users currently account for 20% of my sales. I wouldn’t want to live off my Mac sales, but it is very useful additional income. Given the disparity in cost between Windows and Mac hardware it is hardly surprising that Mac users are more ready to reach for their credit card.

My software is built on top of the Qt cross-platform toolkit. The recent porting of Qt 4.5 to Cocoa gives me the opportunity to further improve PerfectTablePlan’s Mac look and feel and to release a 64 bit version. Hopefully this, coupled with increasing Mac market share, will further improve my Mac sales.

A beta of Windows 7 has just been released.  It will be interesting to see if it can repair some of the damage caused by Vista and slow the growth of Mac OS X. Personally, I doubt it – the Windows 7 feature list certainly doesn’t set my pulse racing.

CoverageValidator v3

The nice folk at Software Verification have done a major new release of Coverage Validator, and the new version fixes many of the issues I noted in a previous post. In particular:

  • The instrumentation can use breakpoint functionality to get better line coverage on builds with debug information enabled.
  • Previous sessions can be automatically merged into new sessions.
  • The default colour scheme has been toned down.
  • The flashing that happened when you resized the source window has gone.
  • It is now possible to mark sections of code not to be instrumented. I haven’t had time to try this yet, as it was only introduced in v3.0.4. But it should be very useful as currently I have a lot of defensive code that should never be reached (see below). Instrumenting this code skews the coverage stats and makes it harder to spot lines that should have been executed, but weren’t.

There are still a few issues:

  • I had problems trying to instrument release versions of my code.
  • It still fails to instrument some lines (but not many).
  • I had a couple of crashes during testing that don’t seem to have been caused by my software (although I can’t prove that).

But the technical support has been very responsive and new versions are released fairly frequently. Overall version 3 is a major improvement to a very useful tool. Certainly it helped me find a few bugs during the testing of version 4 of Perfect Table Plan on Windows. I just wish there was something comparable for MacOSX.