<Martani/> Blog

Update:
Much of the search queries to this blog are about this post, or how to access the old start menu on windows 8.

There are great applications that mimic the old start menu. One of the preferred features is that once a user logs in, she is taken directly to the standard desktop. Personally I have been using Stardock's Start 8 (the same guys who developed the awesome Fences application) which I find excellent, it is not free however (costs over $5). Following are screenshots of Start 8 on my Windows 8 laptop.

Start 8

Start 8

Another option is to use the free "Classic Shell" which can be found here http://classicshell.sourceforge.net/. Following is a screenshot of  Classic Shell's start menu:



Old post

If you have already tried the Windows 8 consumer preview, you have surely noticed that there is no start button anymore, which means no easy way to access your programs using the start menu as we used to do since forever.

Start Button missing in Windows 8

This is impractical since it means: whether you have to put some shortcuts on the desktop, or go to the metro start screen each time hoping to find the application icon there.

However, you can do a simple trick to access the whole start menu again.

To do this, right click the Super Bar and choose Toolbars -> New toolbar. In the dialog, choose the folder C:\ProgramData\Microsoft\Windows\Start Menu\Programs and click "Select folder".

Start Menu items without a start button on Windows 8

• The first one keeps the exponent vector of each smooth number (impractical due to the huge memory needed to keep the vectors).
• The second one sieves at an interval centralized at sqrt(N); according to my testing, this has no visible improvements over the standard one. 
• The last version which is the same as the first one but which uses only a binary exponent vector (saving one bit for each prime in the base) for the smooth numbers.

Distributed version (60 intel i7 nodes):

I'll be developing a program for factoring numbers (especially RSA numbers), the goal is to have a parallel quadratic sieve program running on GPUs (using CUDA or OpenCL) to factorize RSA numbers.

I have just started playing around GMP so I implemented a naive version (in C) of the Pollard's rho factoring method, it uses the optimisation technique proposed by Pollard and Brent, however it doesn't check for cases that may cause the algorithm to fail.

In order to compile this program, you need to have GMP installed.
You can invoke the program with ./pollard-rho NUMBER or ./pollard-rho p q where the number to factorize is p*q.

Compile : gcc pollard-rho.c -o pollard-rho -lgmp -lm

If you had a project that was developed before the Mango SDK was available, or a project that was created with the Windows Phone 7.0 as the target platform and wanted to upgrade to 7.1, then there is no way to roll back to 7.0 again, which means that your app will be available only for those who have Mango updated devices.

On creating a new Windows Phone project with the Mango 7.1 SDK tools installed, Visual Studio prompts you for the target platform:
Because the referenced projects are not upgraded with the app, there is actually a way to roll back anyways, even if Visual Studio warns it is not possible.

Bloginto 2.1 is available for download from the Google Chrome Web Store, the 2.1 version is a patch to the 2.0 version after the (abrupt) changes of Bloginy Algeria.

Bloginto is a Chrome extension that brings Bloginy Algeria and Morocco feeds to the browser. With Bloginto you can:

• Read the live feeds of Bloginy Algeria and Morocco
• Keep track of the new feeds and get notified whenever newer news are available
• Keep track of the read and unread feed entries
• Vote for the feed entries directly from the browser with 1 mouse click
• Tweet directly through the extension

Working on some Amazon EC2 benchmarks I run through few problems of having my processes to run with a high time precision where a more than 1ms shift between operations is not acceptable.

My program accumulates work time and sleep time, ideally, it would work for time t, sleeps for T - t microseconds and loop again. The problem is that in the sleep routine, a shift of few microseconds can be accumulated to become a shift of milliseconds after few loops.

After investigating the problem a little, I did some tests on how precise are the sleep functions in C, here is the code I used to test nanosleep(), usleep() and sleep().



Running these tests on my machine with Fedora 15 gives me these results :


As you can see, there is always an overhead of 100µs to 180µs, a call to printf inside the function will take about 70 µs more to execute. The call to the function itself (the nsleep function) takes at worst 1µs. With this accumulation of  microseconds, a program would get to a state where it is late by some milliseconds after a dozen of iterations and calls to these sleep functions.

On the other side gettimeofday() behaves pretty well considering precision, the benchmark below (found here) gives the following results:

$ ./gettimeofday_benchmark

50000000 cycles in 2315879542 ns = 46.317591 ns/cycle

Which is satisfying for its precision.

A solution (that is not applicable in all cases) is to have a global program independent clock regarding to which the sleep period is adapted. For example instead of waiting T - t microseconds, one will use every second tick on the system as a landmark, and thus any microseconds lost in the last loop iteration are discarded in the next one since the wait time is not dependent on the time the loop has started.