Firstly, create a new file called /etc/group.login.allow (it can be called anything - you just need to update the line below to reflect the name)

In this file, pop in all the groups that should be able to login

Edit /etc/pam.d/common-auth (in ubuntu), it might be called /etc/pam.d/system-auth or something else very similar. At the top of the file (or at least above other entries, add the following line:

For the record, found this little tidbit over at the centos forums\

There has been a solution for this for several years and the first time I did this, I installed cygwin with X11 but there is a far simpler way to accomplish this.

Install XMing. I then used putty, which has the forward X11 option. Once logged in, running xeyes shows the window exported onto my Windows 7. Ah.. so much better.

I actually used this to run terminator to connect to a number of servers. Over local LAN, the windows didn't have any perceptible lag or delay. It was more or less like running it locally.

It is possible to set up shortcuts to run an application through putty and have it exported to your desktop. I haven't played with this enough to comment though.

This of course only worked because I have another box which is running Linux. If that is not the case for you, then you might want to try VirtualBox but since the linux kernel developers have described the kernel modules as tainted crap, you might want to consider vmware instead which is an excellent product.

Luckily, working in Linux means that I can spread each stream of work into the various workspaces.

Now GNOME 3 makes things a little more complicated with the dynamic workspaces but I’m learning to use it to my advantage

However, with Ubuntu 11.10 Oneiric Ocelot and GNOME 3, I seem to be running into an issue regularly…If I leave my computer for a while, it doesn’t unlock correctly. The screen remains black and I can’t move the mouse to my second screen and the unlock screen doesn’t show up.

Thinking about it, it seems like there might be two screen savers being started but I shall investigate that tomorrow. I have the same issue at both work and home so it is more likely to be related to Ubuntu + GNOME 3 or something about the way I set things up.

I  usually resolve this by logging into the console and here a neat trick for killing all our processes in one fell swoop.

Another thing I have been doing a bit more of recently is gaming which involves rebooting in Windows.

Both of the above leaves me with a restarted workspace. Starting up the applications pops them all into the same workspace. Chrome is especially a nightmare. I might have 135 open tabs but they are in about 6 windows spread across four workspaces.

It is annoying to have to distribute these things out each time.

After having done much research, I have not been able to find a clean automated solution.

There are two half solution that I have found however.

The first one is Devil’s Pie and for a graphical interface gdevilspie. According the website for Devil’s Pie, it is “A totally crack-ridden program for freaks and weirdos who want precise control over what windows do when they appear. If you want all XChat windows to be on desktop 3, in the lower-left, at 40% transparency, you can do it.”

Unfortunately, that is exactly what it is. If you pre-determine where you want your windows to be, you can use this very useful application. However, that is not quite what I want. I want the current configuration to be remember. Exactly like how Chromium remembers which tabs are in which order in which windows and their position on the workspace, but for multiple workspaces.

Unfortunately, I couldn’t find any way to save the current state.

There is however, another tool I found scouring the web.

libwnck-3-dev is what I installed on my Ubuntu box. There are two key commands here

This will list all the windows across all the workspaces. To get more information on a specific Window,

The XID is the number returned next to each window from the first command. The post that I  mentioned above has a nifty tool attached that saves the window positions and can also restore them using wnckprop.

However, it saves them based on the Window title. This of course doesn’t work for Chromium or such Windows that changes the title each time you change the tab.

However, if the save is the last command you run and the restore is the first command you run after opening up the windows, it can restore the windows into the correct workspaces.

With the idea of the dynamic workspaces in GNOME 3, you might have to initialise the workspaces first but it is better than spending five minutes after logging in each time re-arranging windows…

I installed synergy on both ends, with the linux one as the server

and the mac as the client

http://sourceforge.net/projects/synergykm/

and it worked.

There was just one very very annoying problem. The Ctrl key and Cmd keys were different. This really messed with my muscle memory. After some hunting around, I just had to update my .synergy.conf file in linux. Here is the relevant section

et voila. It now works a charm. I  have neglected the configuration of the synergykm and synergys but these can be figured out easily ;-)

Being a bit of a purist, I ran Debian{for very many years but found their stable releases lagging behind far too much. This was largely due to their perfectly understandable view of it being ready only when it is right.

For a while, I ran their unstable distribution called Sid{ based on the disturbed, hyperactive 10 year old boy in the film Toy Story{ The idea being that Sid breaks things, and it certainly did. While it taught me a heck of a lot about linux (and the terminal), my computer was broken on a very regular basis.

I switched down to the testing version and that helped ease the pain to a very large extent. I had always thought that Debian with a more regular and shorter release cycle would make a world of difference. Clearly, Canonical had the same idea.

Thus Ubuntu was born and it has grown from strength to strength. Its latest distribution of 9.10 codename karmic koala released October 2009, is a massive step forward.

Stephen{ the head of krayatec{was so impressed by the new release that he conducted an experiment. He asked three people in the office who are not tech savvy to install and try out the new release. My view was that it was still too early for such kind of an adoption. I felt that pushing people to try it out at this stage would damage the reputation of the user friendliness of Linux rather than help it.

I must admit that I was proven wrong.  All of the installations went smoothly and it was possible to log in and do the things that they wanted to do.

Do they now use Linux instead of Windows - No! There is still a learning curve and with tight deadlines and little time for re-learning how to navigate around a computer, it remains an experiment.

However, it does answer one question. Can a user who is not tech-savvy, pick up a CD/DVD of the latest version of Ubuntu and run with it? The answer, from this very tiny experiment is a resounding Yes!

As someone who is technically very demanding, I have very few complaints about the latest version. The only one is that it still looks largely the same as the previous version. Then, the release is a collection of tools, so this is understandable.

My three favourite things about the new release are:

• substantially faster bootup times
• Empathy (particular for people-nearby which works great in the office)
• Cloud (Client & Server)

In my opinion, this is the beginning of the end. Linux, has finally taken a leap that shows its potential to change everything… I would wish it luck, but it looks like it doesn’t need it ;-)

Why not download a live cd and try it out?

I needed host networking to be able to access network resources (Samba shares etc.) which does not work if the guest OS is on NAT :-(

Solving it was easy though… I assume Vista is installed as a guest with the guest additions and that your user account is a part of the vboxusers group.

On the linux host, first install bridge utils. I run Ubuntu, so it was as easy as:

Next, you need to set up the bridge; again, easy on Ubuntu:

add the following section to /etc/network/interfaces

Add the interfaces to VirtualBox

Within the VirtualBox Guest settings, choose Host Networking and fo the interface, choose br0

bring the interface up:

and start your guest os… et voila, it just works…

I am making some assumptions based on how I see twitter working. I have little information about how it is architected apart from some posts that suggests that it is running ruby on rails with MySQL?

Twitter is in the rare category where there is a very large number of data being added. There should be no updates (except to user information but there should be relatively very small amount of that). There is no need for transactionality. If I guess right, it should be a large amount of inserts and selects.

While a relational database is probably the only viable choice for the time being, I think that twitter can scale and perform better if all the extra bits of a relational database system was removed.

I love challenges like this. Technical ones are easier ;-)

If I didn’t have a lifetime job, I would prototype this in a bit more depth. Garry pointed me in the direction of Hadoop. Having had a quick look at it, it can take care of the infrastructure, clustering and massive horizontal scaling requirements.

Now for the data layer on top. How to store and retrieve the data. HBase is probably a good option but doing it manually should be fairly straightforward too.

From my limited understanding of twitter, there are two key pieces of functionality, the timelines and search.

The timelines can be solved by storing each tweet as a file within a directory structure. My tweets would go into

/w/o/r/d/s/o/n/s/a/n/d/<tweet-filename>

The filename would be <username>-<timestamp>

For the public timeline, you just have a similar folder structure, but with the timestamp, for example, the timestamp 1236158897 would go into the following structure as a symlink

/1/2/3/6/1/5/8/8/9/7/<username>

For search, pick up each word in the tweet and pop the tweet as a symlink into that folder. You could have a folder per word or follow the structure above.

/t/w/i/t/t/e/r/<username>-<timestamp> OR

twitter/<username>-<timestamp>

You would then have an application running on top with a distributed cache with an API to ease access into the data easier than direct file access. Running on Linux, the kernel will take care of the large part of the automatic caching and buffering as long as there is enough RAM on the box.

This can in theory be done without Hadoop in between and separating the directory structures across multiple servers but that can have complications of its own, especially with adding and removing boxes for scalability.

You are also likely to run into issues with the number of files / sub-directories limits but they can be solved by ‘archiving’ - multiple options for that too…

Thinking about this problem brought me back to the good old days of working on the search mechanism within megabus.com. We needed the site to deal with a large number of searches on limited hardware when the project was still classified as a pilot.

With some hard work and experimentation, we were able to reduce the search time to a tenth of the original time.

I’ll admit that I don’t know the details or the intricacies of the requirements that twitter has. I have probably over-simplified the problem but it was still fun to think about. If you can think of problems with this - let me know; I wanna turn them into opportunities ;-)

Two main reasons for this is that most software that comes on Linux is highly customisable to start off with. The second reason is that if you don’t like something, you can change it.

There is also the nice thing that most things that you think would be cool or useful in software is already available in some form since someone else thought so too, but before you did and has had the chance to spend some time building it.

I love this so much so that I have often put together a quick linux box for doing things that one could easily replace with an embedded device like a router. I have swayed between the two options based on how much I want simplicity vs flexibility.

One of my favourite responses to someone telling me that we need something that we don’t have is - “we’ll build one”… The software customisation / writing has turned into a metaphor that I apply across more and more things. You need a new table with custom bits - let’s build it. You need a classic car with all the modern gizmos - you know what - let’s just build it.

This has its pro’s and cons. For one, it feels like anything is possible. It also becomes very frustrating to work with limited, limiting, or closed source software (esp when you just want to fix a quick bug that really irks you). It also eats up all your time as you try and do all the things you want… just because you can…

Striking a balance is hard especially when a client asks if it is possible to do something very specific. The answer is of course yes and there is a question that goes with that response. At what value does it become cost effective and provide a good Return On Investment(ROI)

We are not necessarily controversial, just that we hold a view that is usually a little different from the mainstream views. It could be said that we bring the alternative to the mainstream.

But then, so did some world governments, bringing open source software into their work places, successfully or unsuccessfully in the last few years instead of Microsoft.

Someone recently suggested that we were anti-microsoft. I don’t think that is case. Microsoft has its place in a technology infrastructure. It is simply that its position is usually overrated or misplaced. As far as desktops for technically shy users are concerned, there is really no alternative but Microsoft Windows. I can hear the Mac users scream that Macs are also an alternative. Theoretically, yes but the fact is that they are too expensive for someone to dabble with it. This is precisely the reason that Microsoft Windows dominates the desktop market.

We support and use Linux. In fact, the majority of the desktops in the office run Linux (Ubuntu as it happens) but people who have a non-technical role use Windows. They could use Linux but Windows is better suited to their role.

This is not necessarily a cost-saving decision. Sure, we have saved thousands of pounds by sticking to Linux instead of using Windows but that is a co-incidence more than anything. In some ways, it is a testament to the skillset of the people who work at Kraya that they are comfortable with Linux. The mindset of Linux is in alignment with the mindset of a developer.

I used to develop in Windows and I often found myself fighting with Windows, whereas with Linux, it just fits. There are several reasons for this. One being that Linux forces you to understand what you (trying to ) do to a bit more depth instead of pretending its magically taken care of.

I am not, for one moment implying that developers who use or develop on the Windows platform is inferior or not as skilled. Simply that my experience was that the Windows platform made it easier to do things badly and more difficult to do things well.

Microsoft has done wonders in bringing technology to the masses and making it more accessible. However, there is still a massive barrier, even for people specifically in the technology sector to appreciate and use technologies which require a bit more experience or knowledge to use appropriately.

There are a couple of really good examples. PostgreSQL is a powerful outstanding database server that can easily compete with Microsoft SQL Server and Oracle. However, very few people know about it and even fewer use it.

MySQL on the other hand is also an open source database server but is much more widely used and accepted.

It surprises me when MySQL is used when PostgreSQL is, from a technical perspective better suited. MySQL is faster than PostgreSQL at the cost of poor transaction managment (at best). For any system where data integrity is even remotely important, PostgreSQL is a better choice. However, since there are better GUI tools for MySQL and since it is easier to get the hang of, it gets chosen.

This give technology and people in that sector a bad name. Every tool or software has its place, and should be used in an environment where its strengths are displayed, not its weaknesses. We have instances where we use multiple database servers within one project. PostgreSQL for all the data integrity sensitive areas and MySQL for the speed sensitive areas. Sometimes you want integrity and speed. In these cases, you have to make a choice based on which is more important or layer the databases to use the strengths of both.

Metaphorically speaking, MySQL is a hammer, and PostgreSQL is a sledgehammer. Would you use a sledgehammer to crack a nut, or a hammer to crack a slab of concrete?

Before someone jumps down my throat, I am not suggesting that PostgreSQL is better than MySQL or vice versa - just that they both have different goals, different strengths and weaknesses. They have spent a lot of effort to converge and strengthen their weaknesses but not matter the amount of convergence, their core goals are still different that they will never truly be able to remove their weaknesses without giving up some of their strengths as well. One tool cannot be both a hammer and a sledgehammer…

I was pleased to find out that this was also available for linux. Great... Lets try it out.

First stumbling block is the requirement to register on the website before I can download a trial. I am a firm believer of try before you buy. I should be able to register but it should be entirely my choice.

I am more comfortable with registering before buying or for the use of a free piece of software. However, registering for a trial always irritates me. This was also the case when I wanted to trial InDesign / Illustrator the other day.

After registering, there was the irritating wait for the email to arrive. Now, this is irritating. When I want something, I want it NOW. I hate waiting. Adobe did not make me wait for the confirmation email of registration before downloading the trials. There are two good reasons as to why this irritates me.

1. Email, as reliable as it is generally, can take time. In theory, this can be anywhere from a few seconds to hours. How about if my mail server is currently down. Or even more importantly, what if I have shut down my mail client so that it does not keep distracting me from something that I am trying to do. Opening up my mail client, I now want to find out about the other emails that are in my inbox and whether any of them require an action...
2. I have reluctantly provided details about myself. Confirming my email address before I am allowed to download a trial suggests that Tasktop does not trust me enough to just let me download the trial. The software has started off on the wrong foot. How much of an issue is it really if someone gave the wrong details before downloading a trial. Is it really that important that you are able to keep bugging them via email to buy the product?

I was curious enough to jump through the hoops to download the product. The first thing I noticed is that there is no 64bit for Linux :-(. More steps involved in installing this on my 64bit machine. So instead, I installed it one of my 32bit machines - save time.

Once the download completed, the steps on the website suggested that I needed to configure it (with ./configureTasktop.sh) and then run Tasktop. The configuration step required no input from the user and outputted nothing. I have to ask:

1. Why is the configuration step not integrated into Tasktop and configured to run once? Alternatively,
2. Why does the configuration step, not start Tasktop right after.
3. Even better: Make Tasktop a symlink to configureTasktop.sh, which then relinks that to the Tasktop Binary with the configureTasktop running Tasktop right after. This means that from the users perspective, they are always running the same command, and you save any cost associated with run once checks.

I finally got Tasktop to run and it asks me if I want to install the firefox addon to integrate with Tasktop. I want to see how it integrates, so I do. Of course, this is yet another step.

A restart later, I was ready to try out Tasktop - or was I? We use bugzilla to track tasks and I wanted to integrate that in similar to how I do it in Eclipse. This was also trickier than I expected.

I went into the partner connectors section which did not cover bugzilla, which I assumed meant that it came with Bugzilla integration by default. This is true but how the hell do I get there to configure it. It took me a little while to find the configuration section (there are no menus). Once I was there, I wanted to get back to the original layout which was tricky since the "close configuration" button was nicely hidden away up at the top right.

Once I had this working, I tried out the active/deactive mechanisms and this works just the same as in Eclipse. Except with the Firefox plugin, it adds in the links that you browse as part of your context - GREAT!

Add in a task to blog about it and went through writing half the document, then decided to de-activate it before I started working on something else. All the firefox tabs were closed - again, great...

The problem is that when you re-activate the context, it just clears the tabs in firefox and shows you the links you last had open. The page titles for the pages that I had open were the same for a few, so going through them trial and error to get to the blog post was tricky. More importantly, the cookie was already gone and I had to re-login. This might be a timeout issue with Wordpress so wont tag that against Tasktop.

I haven't tried linking folders / files yet but considering that with the above process taking me more time than I expected due to the sheer number of steps involved, I shall have to leave that to another day. In all honesty, it might never happen.

I do like the time logging feature of Tasktop as it tells me which tasks I spent my time on in different chart formats. This is great. However, I have a problem in that this is on an individual basis. I see nothing on here about how a team leader can link in Tasktop used by the team to calculate total time spent on a project / task. This is a necessary feature for a tool like this in the team environment.

It is possible that all of this is easier in a windows environment. Possibly because it was built on there, but more likely because Windows users are used to taking several steps to achieve something (what is it - 7 clicks to delete a file in Vista?)

Having ranted on for a while, dont get me wrong. I think that Tasktop is a fantastic concept and with a bunch of tweaking can be a very intuitive tool to use. However, at the stage that it is in, it does not do what I need it to do. It is actually more obtrusive than useful (e.g. by removing all my tabs from firefox when switching out of a context and not re-instating them on going back to the context).

Then, it is probably just because I simply expect too much... :-(

However, recently, I have been niggled by the amount of time it takes to switch perspectives on eclipse. It takes a good 4 seconds to switch between perspectives.There is also a noticeable lag when performing some operations.

To resolve this, I spent a lot of time looking at the linux real-time and low-latency patches. I had expected that running X11 applications remotely would not cause a bottleneck over a gigabit link. Turns out that I was wrong.

To test this, I ran a vnc server on the application server and found that switching perspectives on there was super fast.

To be able to resolve this, the first thing to do was to remove any latency put on the X->X communication by ssh.

We use gdm, so I had to enable to TCP on there first. Do this using the following config line in /etc/gdm/gdm.conf

Restart gdm

on the remote host, export DISPLAY

According to W3C, as of November 2008, IE(6/7) dominate 46.6% of the market with Firefox at 44.2%. Compare this to November 2007 when IE (5/6/7) dominated 56% of the market and Firefox only had 36.3%

It is interesting to note that between Nov 2007 and Nov 2008, Linux adoption (as far as internet browsing is concerned) went up a meagre .5% from 3.3% to 3.8%.

This means that a very large proportion of the firefox users are from the Windows Platform. Why is this impressive? There is technically no reason for a user on Windows to download Firefox. Windows comes with Internet Explorer, which should be adequate for all the internet browsing needs.

If 44.2% of all windows users went to the effort to download, install and use firefox instead of Internet Explorer which comes pre-installed, let me ask the question - if Windows came pre-installed with Firefox instead of Internet Explorer - how many would go to the effort of downloading and installing Internet Explorer.

While it is possible to install Internet Explorer on Linux, it might be a little unfair to answer this question based on the number of Internet Explorer's running off linux. People who run linux have proven to be biased against Microsoft anyway, so it would be a loaded statistic.

While I have no doubt in my mind that Firefox is better than Internet Explorer, I still don't feel that Firefox is perfect. It still feels far too bulky, with disproportionate memory usage and it is still not as fast as Safari in terms of page display.

Sure, the addons and themes functionality is great and useful. However, it would be nice if it was faster to load, faster to use and just felt more lightweight... like Safari does.…

Having said that, I am not going to switch to safari. I like the browser but it is still just not as good as firefox.

One of the points of open source software, should be to bring all the benefits of all the competing pieces of software into one but it just doesnt work like that. If Firefox had all the benefits of firefox as well as the benefits of Safari, I am sure the adoption rate would be far higher...

Lets take it one step at a time... I vote for firefox feeling a lot quicker and snappier for a wishlist... :-)

I could only ponder about the sheer stupidity of this teacher and wonder about the next generation of students brought up under this ignorance.

I grew up with Microsoft, with DOS 3 as my first Operating System and went through DOS 5, 6, Windows 3.1, 95, NT, 98, & ME.

I also played around with BeOS, and various versions of Mac.

I was then introduced to Linux turned into an open source zealot and wiped out my Windows installation in anger. Since then, while my primary operating system is Linux, I still have Windows running on my Laptop and have both Windows & Linux on my home computer.

I have since worked with Windows 2000, XP, 2003 & Vista. I love what Microsoft does with these products. They do innovative things, pick up features from other products that are useful and try to simplify things.

My Laptop came pre-installed with Windows and I never went to the effort of installing Linux and I use my home computer to play games, which (whether I like it or not) just handles games so much better.

As per the old joke, It is the software engineers job to make software as idiot proof as possible. It is the job of the universe to create bigger and bigger idiots. So far the universe is winning.

Linux & Open Source software (in general) takes a different approach to software. It should be easy to use and manage software but it also expects you to understand (or at least think about) what you are doing or trying to do.

Microsoft seems to be under the impression that this is not necessary. The user does not need to know what they are doing - they just need to know what is to happen. e.g.

Lets take a simple operation - deleting a file. Before Windows 95, this used to be a simple, difficult to undo operation. Windows 95 brings in the concept of the Recycle Bin (or Trash), a concept that was available on the Mac platform for quite some time.

After this point, you no longer delete a file on Windows - you move it to the Recycle Bin, which will delete them from the disk when the number of files in there exceeds the set capacity.

Now, from a users perspective, what they are doing is deleting a file - in fact, thats what the menu item says - Delete. But what happens is completely different. The file disappears from their folder. What they aimed to do - “make this file disappear” has happened. However, the file has not been deleted.

Windows has effectively lied to the user since it is “smarter”. If the user later discovers that they deleted the wrong file, it can be recovered easier. However, that is not the point.

Microsoft software, are in general rife with such miscommunications. I find this fairly insulting and this was one of the main reasons that I started using Linux.  If you ask it to delete a file - it deletes it. If you want to move something to recycle bin, it can do that too.

To go back to the original point, the ignorance shown by the teacher in this school is exactly the kind that Microsoft panders to. Microsoft allows (nay encourages)  its users to be as “simple” as possible and let Microsoft worry about the rest.

Don’t get me wrong. I think that Microsoft do a fantastic job in making software accessible and easy to use but it should also help educate it users on what they are doing and help them think about what they are trying to do. Don’t pretend or try to do their thinking for them. Thats their job.

“Give a man a fish; you have fed him for today. Teach a man to fish ; and you have fed him for a lifetime”