IT Werkz Sometimes

Finding bugs in digital stuff, easy




Archive for January, 2007

Getting rid of the runonce.msn.com page in IE7

Posted by testcrunch on 31st January 2007

Mount Taranaki, New ZealandStart menu -> Run -> regedit and navigate to [HKEY_CURRENT_USER\Software\Microsoft\Internet Explorer\Main] and create DWORD values of 1 for both RunOnceHasShown and RunOnceComplete. Exit regedit and restart IE7

If the above is a bit succinct then try this:

Having navigated to the key, in the right window right click on an empty area and select ’New’ then ‘DWORD’ which will result in a new value with the default values of ‘New Value #1′ in the Name column, Type will be ‘REG_DWORD’ and the value in the Data column will be ‘0×00000000 (0)’.

Now we need to amend it to the two values we need. So right click on ‘New Value #1′ and choose the last option to rename it to ‘RunOnceHasShown’. Right click the now new ‘RunOnceHasShown’ value and select Modify which results in a small window opening. Check the ‘Decimal’ radio button and in the ‘Value data’ window enter 1, then click on OK.

Repeat the last two paragraphs and create the ‘RunOnceComplete’ value. Exit regedit.

Restart IE7 and voila your original tabbed page should be displayed. The next question is how did Microsoft create that mess in the first place.

Quote of the day

‘I believe that a scientist looking at non-scientific problems is just as dumb as the next guy’ Richard Feynman (1918-1988)


Posted in IE7, loadsa tabs, good | 2 Comments »

Formatting a drive over wi-fi, you godda be crazy

Posted by testcrunch on 31st January 2007

The AlpsGot a new external Maxtor 320 mb hard drive yesterday.

Plugged it into a powered USB hub which didn’t work probably because of a juice issue. Tried it on a USB socket on the processor and it was found OK. What was a bit odd was that Windows had assigned it what it thought was the next available drive letter which it thought was M. The M drive I actually have mapped to a shared folder on another pc in the house. Windows explorer showed the M drive with the description of the mapped drive but in the Computer Management Disk Management folder the M drive was the new Maxtor drive as I could see the drive size. Hmmm..

On the box of the Maxtor the label said that it was unformatted. Its been a while since I’ve seen one of those. If I had tried to format the M drive what would have happened? Not much I reckon as Windows was so confused about what was on the M: drive.

Turned out that the label was wrong as the disc was formatted. 

This does raise the horrible idea of what would happen if the whole drive of the remote pc was shared. Would invoking a format over a wi-fi connection work? Just think of the little words ‘Format m: /s’ belting through the ether and unleashing themselves on a poor old share whilst someone’s logged on. Lethal. 

Disconnected the mapped network drive M and also another, N, rebooted and there was the M drive in all its glory correctly labelled and usable. Remapped the 2 shared folders on the other 2 pc’s, as N: and O:, which knocked out the iPod which I changed to Z:, and if that ever fails it means I’ve got too many hard drives.

Posted in XP, you sure you wanna talk about this | No Comments »

Was this an early form of Agile testing?

Posted by testcrunch on 29th January 2007

Eagle NebulaThis may have been the first time anybody did Agile testing or was it just how we did things then? I was working on a project, in the mid 90’s, with a developer who was a bit prone to coding errors. He didn’t really like documentation and he definitely didn’t like the idea of structured testing. But he did like the idea of writing really clever beautiful code. So I knocked up some structured tests with documentation but had to hide it from him or he would have got the screaming hebegebees.

When I found a bug I wrote it down on a simple bug document to fully describe what I expected to happen and what went wrong and gave it to him. Silly thing for me to do because he would immediately stop what he was coding and attempt to clear the bug. Silly thing for him to do as it took his mind off the big picture and his bug clearance code would invariably bust something else way over yonder.

So I stopped giving him the bug docs and batched up a weeks worth. Trouble was because he was sitting right next to me he would inevitably get to know when I had found any defect. No I wasn’t shouting ‘found another freaking bug’. And, of course, he would want to fix that bug as he just couldn’t stand the idea of his code being buggy. We were getting nowhere.

The way forward was for me to tell him what kind of tests I was going to do and when I thought there maybe areas where he may misunderstand the requirements then I would actually tell him to make sure the code will do whatever processing when a particular situation arose and that I was going to do a test specifically for that. We did get a bit side tracked at the start as he started saying things like ‘but you don’t know what the code is doing’.

Gimmee a break. The last thing I wanted to see was his twisted monolithic source code. I told him that whatever it was doing it had better code the requirements.

All I was really doing was reiterating the original requirements back to him before his understanding of them got mangled and resulted in code from hell.

Got the system working though. Was that Agile?

Quote of the day

‘Always listen to experts. They’ll tell you what can’t be done and why. Then do it.’ Robert Heinlein (1908-1988)

Posted in Testing software - watching bits drop off | No Comments »