Minddriven

Killerproject - the naked license agreement

2011-10-27T15:00:00.000+02:00

License agreements, terms of use, lines and lines of words that get unnoticed every day (who has time to read 60 pages of legalese ?) but are very important, expecially when things go wrong.

So, why not create a great online service for the management and analysis of this documentation?

The naked EULA

What should this hypothetical service do ?

It will store the localized versions of the license agreement of a lot of applications and services reported by users. It will extract the most important information, translating into summaries and / or infographics. It will highlight the risks and differences with EULAs of other similar services. It will highlight the differences between different versions of the EULA of the same service.

How to use it ?

Do you need have to sign up to a new online service? Go on TNEULA (The Naked EULA) and check the summary and the various 'traps' that are awaiting you, but without reading the entire blob of text.

Have you got yet another update of the iTunes terms of use ? Go to TNEULA and it will only shows what's new since the previous version.

Do you have some doubts about the conditions and responsibilities of Sony in the use of your sensitive data and credit card? Go on TNEULA and it will highlight this specific part.

What do you think? Do you like it ? I could definitely become a regular user. And even a user much more aware of the services it is subscribed to.

Who starts to write it ?

[this post is a lazy translation of the original one, in italian language]

Arduino Monster Case

2011-06-22T15:40:00.000+02:00

Directly from ‘the labs’, this is a small and weird off-topic experiment that will go on as long as my office is filled with Arduino boards.

The Arduino Happy Monster Case is a ‘smiley-monster-shaped’ case that can hold an Arduino UNO or Arduino DUEMILANOVE, and make it very attractive !

Happy Monster Case features:

Power supply and USB connector are easily accessible even when the board is in the case. I/O pins are not, so you’ve got to open the case in order to work with them.
The monster can stay up on it foots or lying.
Mouth and eyes can be left open, letting the board show, or closed by a plastic coloured sheet that gives a some colour.
The Happy Monster Case is available in kit together with assempling instructions and advices for alternative assembling (eg. are you mounting shields ?), and some additional spare part.

In the remote eventuality you’re going to like it, is available with Blomming. I hope it is a good start for other funny side projects.

Social thoughts on follow and defollow

2011-06-08T15:00:00.000+02:00

Off topic thoughts.

I’m not a social network enthusiast. I try to participate as fully as I can to those I like and, for the rest, just wander cautiously. At most I sign up, make experience, and then I cancel the account.

The thing that intrigues me the most, on social networks, it is the mechanism that connects users to each other. Technically this link is nothing particularly interesting, but entire book can be written about the emotional / cultural side of things.

Starting from the term used, which has an incredible weight.

There are social networks in which you are a friend, where you become a neighbour, where your are a contact and where, finally, you simply follow someone.

That's easy: follow.

Do you publish something I like or am curious with ? I follow your flow, so I can read something interesting.

Do you like my writing and public? You follow my flow, as long as you are interested ..

It sounds simple, but the interpersonal relationships on the Net get always complicated.Too often I see arguments and discussions about following and defollowing. Defollow used as annoyance, or follow used as a declaration of friendship.

It is me ? I have misunderstood ?

Just to clarify to 'those who follow me' the way I see it:

In my opinion, the social 'following' basically means 'looking in one direction'. The digital life is nice because it makes you do things that would not be able to do physically, like following 200 people at once.
‘Following' does not necessarily mean 'we are friendly and we're friends for life'. It means that 'what you write arouses my interest, I want it in my timeline at this time'.
The 'follow' and the 'defollow' are not forever. They are just a click away.
I can 'defollow' because even if your content is interesting, you are filling my timeline and I can not read others.
I can 'defollow' because I can’t follow too many people.
I can 'defollow' necause you don’t write anything.
I can 'defollow' because, finally, I realized that you write rubbish.

And that's all.

Which brings me to the final big question: "But it is more correct ‘unfollow’ or ‘defollow’ ?"

[this is a lazy translation of the original one, in Italian language]

Yes, this is Italy

2011-04-29T10:30:00.001+02:00

This is a official note from the website of the Italian Presidenza del Consiglio.

The note says:

With regard to what some news agencies reported incorrectly, the President Silvio Berlusconi was careful from making a prediction on the Scudetto with Milan also for obvious superstitious reasons.

This is Italy.

Hackers - it's your Research & Development division

2011-04-06T15:00:00.001+02:00

After reading this summary on how Sony is particularly inclined to conflict with the community of hackers, I link to this post on a success story with a totally opposite spirit.

And I wonder: when you release a product and there are fans willing to extend, add functionality, increase the hype and sales, how can you sue them ?

It is Research & Development, can't you see ?

Because, let's face it, an active and passionate community of hackers is a real department of research and development.

Sometimes much more efficient than the 'corporate' one. Even better of any Innovation Time Off, if I may dare ..

They are many, they are different. They have different mentalities, they think differently, they speak many languages. It is difficult to set up a department more diverse and thus more creative.
They are passionate. They work all hours. At night, with their own instruments, without pay. Do you think your employees will do the same ?
They are smart.Many of them are much more brilliant than many 'corporate' engineers. And it is a lively intelligence, creative, very suitable for research.
Low cost. They work for free. They buy the products. Sometimes adversely affect the licensing and copyright, of course. But it is a damage that pays for itself with increased sales.
They really use the product. Many companies claim the famous saying 'eat your own dog food'. But they don't do it. If they did, they made better products. Other companies do not even know what they are their own products. Hackers use them, really. They know them well. They love them. Who better than them can improve these products ?

In short, I would like to have products targeted by hackers. War ? Sue ? Are you kidding ? Cheers !

What do you think?

[this post is a lazy and cheap translation of the original one, in italian language]

Scalable web interfaces, how to set them up ?

2011-03-16T15:00:00.007+01:00

In a recent post by JP there was an interesting discussion about web interfaces, scalability and graceful degradation based on the type of device used.

By default this scalability is automatic and determined by the user agent and other features of the internet browser used to watch the pages.
But automatic recognition could easily choose a 'wrong' target (e.g. it can consider a tablet, that may be able to show a complete interface, as a smartphone just because it runs on Android).

Usually many websites give the possibility to choose between a mobile versus standard interface. Is it enough?

In a comment on the above post, thinking about the possibility of a choice of the user interface scaling level, I asked: is there a standard UI to choose the user interface details within a site?

1. Precise tuning

In addition to the classic choice of viewing mobile / standard, we can think of an interface used for tuning the level of detail of an application.

The target is for sure the videogame.

Well, we do not want this kind of stuff.

Surely there are games that provide a good 'user experience' even when you go to adjust the screen resolution and 3D rendering.

Is it better ?

Pages to separate the arguments, a list of options, no pop-up controls but quite readable choices that can be selected via left / right buttons.

Note the button 'Reset to Defaults', which allows the user to return to a normal situation after having messed up with all the settings.

Here is a very similar example.

The principles are the same. List of options, limited choices, easy to understand.

These interfaces allow a fairly precise tuning of the level of detail of a graphics application. But in some cases it is too much. A power user may find bread for their teeth, but the casual user is not able to make certain choices, and is lost.

Is it there a simpler alternative, although less precise?

2. Scaling the detail level

It is possible to simply leave the user the ability to scale the level of detail.

That is what the 'resource and file explorers' of all modern OSs do. It is just possible to determine the level of 'zoom' which shows the file.

This type of interface is certainly not detailed nor precise, but it is easy to understand.

In our case, the slider should move between 'light' (text-only mobile version) to 'Detailed' (full site).

Defect. A slider based on a single axis dedicated to the 'level of detail' does not allow us to change some other parameters like the type of input (touch, mouse or remote control).

3. Choose from a preset

In order to keep the choice of the device simple, yet quite different, why not think about to choice a 'preset' of settings dedicated to the most common client devices ?

The user can choose between 'laptop', 'smartphone', 'tablet' and 'TV / media center'. He may not only scale the level of graphic detail, but also choose the type of input. If possible, without going into an advanced setup.

The interface can be represented by simple icons. Each icon represents a set of settings dedicated to the specific device, and it might be changed later.

The ideal solution ?

Probably the best solution is a mixture of the above options:

Detailed setup (more control, more difficult).
Detail slider (less control, much easier).
Presets (less control, much easier).

It is not impossible to think of a combination of the 'difficult' interface with one of the two 'easier' solutions.

The real question is: is it worth investing in an advanced detail configuration ? Or the application must always choose automatically ?

What do you think is the best user experience ?

[this post is a lazy translation of the original one, in italian language]

The box of old projects

2011-03-04T15:00:00.001+01:00

I looked at the box of old projects.

Code snippets, ideas, applications started and never completed. My contacts and colleagues often hear me say: "I'm thinking of writing a tool for XXX" or "I have prepared a system to YYY." If after a few weeks there isn't news, my contacts and colleagues have learned that the new tool and system ended in the box of old projects.

I'm not ashamed to start more projects than those that come to a deployment. It is part of an innovation process that requires some victim.

But not always.

Today, for example, I found out an old persistence library that can be easily reused in an ongoing project. Other ideas and notes that seemed dead, I must admit, can still work if cleaned and updated.

My suggestion is to not throw away old ideas, old code, old unfinished projects.

Get it out every now and then. Try to remember the idea and the intuition that you had. Rediscover that old projects are better than what you thought. At the right time, they can come to aid.

Flint Lockwood, the nerdy inventor of 'Cloudy With a Chance of Meatballs', was able to resolve an apocalyptic disaster thanks to his old inventions that he had just thrown in the trashcan.

We can, too. We don't have spray shoes or flying machines. We don't have ratbirds, but in the past we did something interesting.

It would be a shame to let it die without a last look.

What do you think?

[this post is a lazy translation of the original post in italian language]

Git - irreverent quotes about the distributed source control system

2011-03-02T15:00:00.003+01:00

While re-reading old emails and chat logs, I enjoyed to collect comments, criticisms and 'phrases of the day ' about the use that my team makes of Git, the distributed source control system.

Here's a short list (certainly incomplete, some quotes are definitely lost):

"Git commands have been designed by a serial killer in a bad mood."
"Git - The source control system that makes you feel stupid."
"Git has the option --linus_will_review_this_code acting exactly as it sounds."
"Git has the option --divide_by_zero."
"Git is awesome, unless you develop code for Windows."
"Git does not need any user interface."
"Git has the option --and_fix_this_code."
"Only the Chosen can use Git to its fullest extent."
"Git has a command 'git raise --venture_capital'."

If you read these quotes you may think we hate this tool, but it's not true. We like Git. Really.

It 's just that we enjoy grumbles.

[this post is a blatant translation of the original one, in italian language]

do-Dots ... and fitness

2010-12-03T17:31:00.002+01:00

I will not go into details but I want to share this little new.

In recent days in upto9bits we worked on an interesting project in the field of fitness and biomedical.

Conceptually it is nothing complicated. Some exercise machines like treadmills and stationary bikes can be connected, controlled and monitored (along with exercises and physical condition of athletes) from a PC-based control software.

The good part of this project is the use of do-Dots to set up the system. This platform has enabled us to write Dots interfacing with fitness machines and take use of Scaddle as the default client graphical user interface.

Having written Dots for fitness equipment also means we can reuse them in other projects.

Other users can write solutions that involve stationary bikes and treadmills. We are already dreaming new applications dedicated to gyms, personal trainers and related topics. The manifacturers of these machinery, thanks to this opening, will probably sell more units.

It is exactly what we hoped to be able to do with this platform. Simplify, extend, multiply the solutions. And multiply the sales of hardware manufacturers.

And now, go !

[this post is a lazy translation of the original one, in italian language]

Italians are not like this ...

2010-11-17T14:42:00.000+01:00

Thank you, Newsweek, for the post titled "Bunga Bunga Nation". It is well written, complete, accurate and well documented.

But it is also important to point out that Italians are not represented by Berlusconi, that a lot of Italians have never voted him, and that a lot of Italians keep on going everyday despite of the actual government.

Qbo - how cute ..

2010-08-04T15:00:00.000+02:00

I don't know how I stumbled on the Qbo project website, but it took me so much.

Qbo is a cute little open source robot that offers everything you could expect from this kind of tool. Sensors, webcams, wifi, bluetooth, speech recognition.

I love it.

This is one of the devices we want to write Dots for. Imagine if you can easily connect it with your Home Automation system ("Qbo, shut off the lights in the sleeping room", "Ok!", click, click) ...

(it will also be a great gadget for demos)

The do-Dots project

2010-08-02T15:00:00.003+02:00

I'd like to introduce you to a new project I'm working on.

do-Dots is a framework dedicated to communication between devices.

It started after some discussions with a colleague regarding some home automation projects we developed in the last years. We were tired of writing a lot of communication layers and be compatible to dozens of different protocols in order to connect our hardware to different systems.

The result of these discussions is an application level, readable and simple ASCII protocol used as a common 'language' in a publish / subscribe network. Different components can connect to this network, adding features like user interface, logic, remote connection, gateway with other systems.

This 'kit' allows us to develop new custom automation projects faster and better.

The kit (or framework) will be soon available to customers and developers. We are working hard to make it more standard and readable, and to translate code and documentation.

In our vision, when you develop a new electronic gadget or automation system, you can write a Dot in order to connect it to a do-Dots network.

Writing a Dot is relatively simple. You don't start from scratch. You have examples, a guide, and tutorials. At the very best you only have to override some C# methods of a base class. Or modify and existing Dot.

When you've got your Dot, you can connect your device to a do-Dots network. And you immediately can:

Control your device via a desktop (Windows, Linux e Mac) or Web graphical user interface.
Write scripts to automate your device and other devices connected to the same network.
Have your device work with information shared in the network (eg. external sensor data).
Connect to a remote server, in order to backup data, mashup with web applications, and more.
Expand your network with Dots downloadable from the do-Dots library.

Obviously, you don't finish here.

do-Dots is not magic.

But is a good start. And solves some problems.

That's do-Dots. In the next months I will post some other information and examples, hoping it will be useful or inspirational.

Meemi.com during Italian match in FIFA World Cup 2010

2010-06-24T16:09:00.000+02:00

Meemi.com adjusts its graphics during the Italian matches in FIFA World Cup 2010.

Isn't it awesome ?

Meemi is growing ..

2010-05-07T18:50:00.003+02:00

This week @capobecchino, the author of Meemi.com, revealed some statistics of the recent growth of this microblogging platform.

In the first third of 2010, new accounts have significally overrun the total of 2008 and 2009 new accounts.
These data are impressive, especially considered that the very most of the accounts is from Italy, that this service isn't covered by mainstream media, and that competitors have very important names like Facebook, FriendFeed, Twitter and Buzz.

Still, Meemi has some very great points in the energy and openness of the authors, in features that are getting everyday more useful, and in a community of users that is really, really, open and conversable.

I think that no other social network can have discussions that start with a simple 'buongiorno' and last for over 100 days and 2000 answers.

I think no other social network is so useful to get help or to brainstorm ideas.

I think no other social network can end up in an 'avatar exchange day'.

I think no other social network has so many followers that spontaneously defend every member of the community trolled in its own blog.

If you are interested in joining Meemi, let me give you and advice. Don't assume it is Twitter. Don't post tweets.

Participate. Answer. Live it.

Be a meemer.

When your toddler grabs a digital camera ..

2010-03-17T16:04:00.000+01:00

When your toddler grabs a digital camera ... it may result in some pretty wallpapers.

Here they are. Use it at you own risk.

Good customers, Bad customers

2010-03-12T11:11:00.001+01:00

[silly translation from the original one, here, in Italian]

If you make a living writing software, on the long term you realize that the best customer is the one that, although does not know how to make software, at least understands what it means.

Immediately after is the customer who does not realize what it means, but at least is aware. And trusts you.

The worst, in this profession as in many other areas, is the one that does not understand anything but thinks he knows everything.

Meemi, an Italian Social Network

2010-03-04T15:09:00.002+01:00

[this is a silly translation from the original one, here, in italian]

This post is an answer to friends and colleagues that received my invitation to reach me in Meemi and don't know what Meemi really is.

I can describe Meemi, as a 'Twitter made in Italy, more complete, rich and structured'. But it would be a banalization.

If you look at features, Meemi is more like a micro blogging platform. It doesn't have the 140 characters limit, and allows to share formatted text, quotes, links, videos and pictures. It seems much similar to Pownce, maybe, but without the aridity and sense of strangeness that made me leave it.

In my opinion, when I look at web applications, I am sure that is not possible to compare them only on a feature basis, but mostly on the quality of the community of users involved. On this basis, Meemi is definitely superior to Twitter, Buzz and their clones. Even Facebook.

In Twitter, I try to be heard in a huge and cahotic cacophony of tiny messages, most of the times unreadable or obscure. And I have always the ugly sensation that no one really cares about what others are saying.

In Meemi, on the contrary, I can have a real discussion, interesting or futile. With people. I can share my thoughts with friends and 'friends of my friends', and I can meet people that share my ideas and interests. Everything in a simple and friendly application.

And that's why, friends and colleagues, I am inviting you in Meemi.

Reach me via this link !

The STAGE-A / STAGE-B theory

2009-03-06T23:57:00.003+01:00

Chatting with some colleagues in front of the coffee machine, I've heard an interesting theory about the cycle of recruitment in some companies, that can be summarized in a sequence of actions:

1. STAGE A:

The company hires a new employee. In order to save money and have more control of his work, the person is chosen without any particular experience or skills.

The work is being pursued, but the employee makes many mistakes and does not produce an exceptional quality work. He also requires considerable time from managers to be trained properly on any action to be taken, thus reducing productivity over the department.

This brings to ..

2. STAGE B:

The company hires an experienced person with skill and considerable motivation.

He is more expensive, but the company considers the new employee a good investment on the quality of work and productivity of the department.

The new employee works well. However, because of his experience, motivation and skill, he tries to improve the organization to achieve even more efficiency. This attitude can lead to a positive evolution of the entire organization or a friction and annoyance if the leadership is not keen to see a devolution of control to the experienced staff.
In the second case, the employee hired in Stage B loses confidence in the organization and starts to look around until the inevitable dismissal.

At this point the company gets back to ..

3. STAGE A:

The leadership, given the negative experience of the employee expert / skilled / motivated again decided to take a low profile guy. In this way, they think to gain control of its activities.

... at this point the cycle is repeated indefinitely, with the turnover of this company constantly focused on profiles of Stage B.

After a certain amount of time and the succession of stages, the company will lose these figures only to find themselves with employees hired in Stage A ...

I do not believe that this theory is applicable to all real companies, but it certainly is interesting and curious.

What do you think ?

Do you recognize this sequence in the history of recruitment of your company?

[Translated from MinddrivenIT]

Auto completion on iPhone - isn't it strange ?

2008-11-27T14:44:00.002+01:00

If you ever wrote something on an iPhone (SMS, notes or mail messages), you'll probably know that this phone uses an auto completion engine completely different from the T9 system available in other cell phones. This system improves the speed of your writing, but not always the quality.

How does it work ?

While digiting a word, the editor searches if it is the start of, or is similar to, any word available in an internal archive. If found, the editor shows a small popup above the word with a proposal. If you don't like the word proposed by the auto completion, you just have to touch the popup and it disappears. Elsewere, when you hit 'space' or 'enter' or another punctuation character, the proposed word is automatically substituted to the one you entered.

Is there something wrong ?

In a standard auto completion system, the proposed word must be confirmed by the user in order to be accepted. When the user looks at the proposal and finds that the word is right, he can choose to stop digiting the word and speed up the writing. On an iPhone, the proposed word must be refused, or it is accepted by default.

This behaviour is everything but intuitive. It works well only if the auto completion guesses perfectly all the words you are writing. When it doesn't, you have to keep constantly an eye on the popup because you need to close proposed words that don't fit. If you miss a word, you have to delete it and rewrite again.

An example. If I digit "My pc is very good" without looking at popups, the message becomes "Mg of os veri golf" (language setting on 'italian'. I hope that english language works better ...).

Great, it isn't ?

This approach, the need to refuse proposals, forces the user for extra activity, and does not allow to concentrate on the real task: writing a message.

In conclusion, the approach of the iPhone is for sure original and interesting, but thinking different not always brings up the best result !

[Translated from USERInterfaccia]

Poll results 3 - What music do you listen at work ?

2008-10-03T23:11:00.004+02:00

here it is ..

Team working .. and the coffee machine

2008-08-21T14:46:00.003+02:00

You know. We italians don't take work breaks around the water cooler but in front of a coffee machine.

Always.

The coffee machine zone is ideal to reduce stress, distract, reorder our thoughts. In front of the coffee machine we chat, relax, rest eyes and exchange useful information about the project in an attitude that is more informal of office meetings. With a coffee, the social hierarchy of the organization flattens out and shuffles. Here you can see the manager joking with a junior employee.

So, don't underestimate the importance of the coffee machine zone (or the water cooler zone), and above all don't prevent your team to make use of it.

If you want to take the maximum advantage of the coffee machine, always remember to keep it in an area that is isolated from the office. Is should be out of eye and ear contact, so that the people having a break can not distract who is still (trying to) work.

Still, don't isolate this area too much, or you'll give lazy people an opportunity to shirk.

The coffee machine area is also suitable for communication that is not strictly verbal. A bulletin board with fliers or announcements fits very well in this area. So, why don't celebrate your team's small successes by displaying a chart of resolved bugs (obviously in positive trend) or an anthusiastic feedback from one of your customers ?

[blog post traslated in english from Minddrivenit]

Great article on Toddler Driven Development

2008-07-30T12:53:00.001+02:00

You will probably already read it, but from A List Apart comes a great and enjoyable article by Natalie Jost about working at home with toddlers around you ...

How familiar it looks ...

An interesting article on User Stories estimation techniques

2008-07-14T09:30:00.004+02:00

I've read with interest an article on InfoQ by Jay Fields covering some tips on estimation.

In particular, some points raised my interest:

1. The concept of throwing an estimation, in which every developer should give his own estimation together with others so that no one is influenced by others. This concept is put in practice with a sort of rock/paper/scissor style game that it looks really, really amusing. I must try it.

2. The concept of estimations that get stale. It is very important for me, and I also written about it in a previous post, describing my usual practice to add an expiration date to every estimate. As time passes, estimates are less and less realistic until they are completely useless and meaningless. In this article, estimates get stale after 12 weeks, and it seems a good time limit for an average project.

3. The overall focus on people involvement. This concept applies in small rules like: no laptops; small team size (no more than 6); business people involved; everybody expected to estimate and read user stories. This is very positive, because estimation meeting do not have to be focused only on one domain expert, a team manager and a leader developer with all other people IMessaging around or just watching outside the window. The estimation, like development, should be the result of a whole team effort.

Please, hide what can't be used

2008-07-02T10:58:00.001+02:00

Joel Spolsky suggests not to disable or hide menu items when they can't be used. Instead, says, you should leave them active and tell why they can't be used.

I can agree in not leaving menu items grayed. They are boring and they distract the user from what is really active in the UI.

But, please, don't give other annoying messages to the user. He already needs to complete his own work.

Instead, hide menu items that are out of context, and the user will probably already know why.

[Don't hide or disable menu items via USERInterfaccia]

Poll results 2 - Lazy people win !

2008-07-01T14:49:00.003+02:00

Ok, this is the final result of the second poll:

... and I think it means that the readers of this blog are either rockstars or lazy people. I think I should find a deep meaning in this results but, well, don't take it too seriously ...

Starting a new side project on Heroku

2008-06-18T17:33:00.001+02:00

As I wrote in the previous post, I'm really attracted by Rails and I think I will start a small side project based on it.

Maybe an open source project, or at least something that can be used by other people. At the beginning I was thinking of a sort of project + budget + contact management tool, but I think it may be better to start with something simpler.

I signed up an account on Heroku: if you don't know what is it I can describe it as a Rails hosting service that integrates management, console, configuration and even the source code editor. Very neat !

Since I can invite some collaborators, if someone of you readers wants to give ideas, test things or even write some Ruby code, I can invite you to the project. You just have to ask.

Programming skills (and side projects)

2008-06-17T17:06:00.001+02:00

I think it is time to improve my programming skills in a different direction.

In the last year I worked full time on a project as a C++ (MFC) developer. In order to keep exercised with .NET Framework, I used to work on a small ASP.NET/C# project in my spare time.

Now I changed clients, and I'm working full time on a new big and complex project with .NET Framework and C#.

Wait, I'm not very sure I need to keep exercised in the same tecnology !

So, why don't move on another side project ? In the last year, I sporadically wrote some small tools in Ruby, and I was very happy with it, with its platform independence, with its design and flexibility.
Why don't improve my skills with Ruby, and in particular with Ruby on Rails ? They are cool, they are cutting edge, and I can boast of them in my blog !

I'm planning to write something in Rails: basically, I can start from something like that old, never refactored, and buggy Jirino project management tool (that is, a simple manager of bugs/features/documents that renders like a wiki and allows to assign, plan, tag, comment and move stuff around the team members) and improve it into something more flexible and extended, with features more similar to a contact and budget manager, with some standard format support like those recommended by microformats or data portability ?

Any cooler idea ?

Do you think it is the right language in the right time ?

Interesting links

2008-06-17T11:26:00.001+02:00

Some links that in my opinion are a must read:

Smashing Magazine

An inexhaustible source for graphics, design inspiration, icons, templates, fonts, wallpapers, tutorials and whatever you can ask for. All the content have really high quality, so you never waste time when you read it.

The Big Picture

A photographic blog of great, stunning, quality. I love it.

Freelance Switch

Resources and articles for freelancers, it seems expecially dedicated to web designers but contains suggestions and tips that can be useful for every professional.

Visual Complexity

Working examples of graphic visualization generated from complex network and data. The result is a large set of very interesting graphs and animations that can add inspiration and ideas.

OT - Shuriken tacks

2008-06-03T17:21:00.001+02:00

I think I want those “ninja tacks” ... they would be very catchy on a planning or a project status board ..

The right technology at the right time

2008-05-27T15:04:00.003+02:00

A colleague of mine, watching some post-it notes placed beside my monitor, said that no real programmer will ever use these. A real programmer would have used (or written) some really and cool high-tech software tool to keep those notes.

He was joking (I think), but I am sure that he is also wrong. Real programmers can actually use post-it notes when necessary.

It is just a matter of context, not a matter of technology.

When high-tech solution is needed, it is time to use it. But when high-tech solution comes to a cost, and that cost is not worth, I think that everyone will choose for a low-tech solution. Programmers included.

And you ? Do you prefer to use the right technology at the right time ? Or do you always cry for the cutting edge technology around here ?

And what is, actually, low-tech and high-tech ?

The fastest ToDo List is a ToDo Album ...

2008-05-15T22:51:00.008+02:00

I have to face it. Whenever I need it, it is easier to find the nearest digital camera than pen and paper.

With this in mind, I have tried to change my approach to the classical ToDo list technique. Why don't replace text notes with photos ?

That's my recipe:

1. When I see something that needs to be fixed (something to buy, to repair, to clean, to move), I take a snapshot of it. It is simple and it needs one second: the photo doesn't need to be a masterpiece, I just need that the subject is recognizable.

2. I keep photos stored in the camera (iPhone): I have not tried to upload or upload them to Flickr (or similar) because to me it seems too much complicated and slow. I just keep the iPhone with myself.

2. When I exit home, I get into the office, I exit the office or I enter home, I check the ToDo Album. If I find something that I can do now, I fix it.

3. When I fix something, I remove the photo.

That's all, and that's incredibly simple and intuitive. I want to point out that it is not the solution for all situations, but it is very fast. Believe me. The concept is that if I can take a fast snaphot, nothing else is required: I will easily associate the subject of the photo with the task to accomplish.
If you want to try, my suggestion is also to avoid to write notes, use uploaders, share images in the Internet or download stock images. If you try to do this kind of stuff, you will probably be slower than taking notes with pen and paper.

I have done a quick research, and found I am not the only one to use this technique and, for sure, not the first :)

Tell me if you tried this, and if it was successful.

(bonus - here are some examples)

.. coffe is finished ...

.. toothpaste is finished ...

.. my car needs new tyres ...

Jirino 0.6.1 is available for download

2008-05-12T22:53:00.005+02:00

After a while, it is available a new version of Jirino. I am sorry that adopting users need to wait so much time for newer releases but I've been very busy the last months, and I really, really did not have time to spend on this project.

Anyway, a little new version is here. What you can find:

. Multiple languages: Italian and English.
Jirino was already in english, but in order to ease the use for italian language users, I've added the multiple language support and this very first translation. Beware that not all the text is translated and that the language is automatically switched depending on your browser (or operating system) regional settings.

(do you want you language translated ? Just ask for it !)

. Item history and attachments are hidden and can be expanded.
A small layout trick that can make the Item page a little less messy.

. Item history in RSS and TXT format.
You can access the whole history of an item in RSS and raw text. Fine if you use feed readers or tools that need to watch the evolution of one or more items.

. Preview of the item body.
A small button allows to preview the text rendered with the wiki-like syntax.

. New tags to link attachments inside the document.
Brand new wiki tags:

{{{imgs:IMAGE}}} displays and attached image inside the document, in a thumbnail format.
{{{imgt:IMAGE}}} displays the text of the attached image, and a click will open the image itself.

Remember that new image attachments are automatically linked in the item body.

Some other (good) blogs from the web

2008-05-11T16:28:00.004+02:00

Other good reads from the web. You probably already know all these links.

Or not ?

456 Berea Street

Very good read if you are interested in Web design and accessibility. Posts are not frequent, but are always worth a look.

Signal vs Noise

The blog of 37 signals. It contains some news about their products, their design process and also very interesting links from the web.

43 folders

A blog about getting things done. Full of interesting stuff on organization, productivity and general clutter removal.

ReadWriteWeb

A good alternative of TechCrunch, if you like blogs talking about startups, investments and what's new in the Internet industry.

Geeks are Sexy

Technical news, toys and gadgets and funny things all around us.

Faces in Places

Ok, there is almost nothing to read, but it is sooo funny...

IKEA Hacker

Smart projects and alternative ideas that can be made with IKEA products.

The product guy

Nice ideas and interesting considerations on startups, web, user experience and modular innovation.

Let the skater skate

2008-03-31T18:35:00.005+02:00

I think that many (although not the most) of the problems in software projects come from the bad assumption that everyone can write software and everyone can do software related tasks.

This bad assumption means that a lot of software teams have to fight, during their project lifetime, with a number of external interferences. Have you ever seen managers or customers, or both, pushing for changes in the source code, the use of a particular (yet not essential) tool or technology, different estimates, or different design and implementation from the one choosen by the software team ? I have seen it a lot of times. And all the times it was very bad.

Where it all come from ? My opinion is that, from the perspective of external people, writing software seems easy.

Watching a good software development team at work it is like watching an expert figure skater. Jumps, twists, lifts and figures seem all very easy when you watch an expert. So easy that you may be tempted to try by yourself. But, when you try, you will probably fall at the very first jump. And if you try tied together with an expert skater, it is very probable that you both will fall !

Software people have made a lot of work, in the years, to ease the software development process: they invented new languages, practices, tools and instruments that make the process easier than it is. And it is easier than in the past.

And it looks easier. Really.

But even if some languages and tools are so simple that everyone can learn the basics of programming and software writing, there is a dramatical huge difference between software written by professional developers and software written by everyone. And this huge difference can make every project to fail.

So, please, enjoy the skater work, but let the skater skate alone.

Workaholics, productive folks or zombies ?

2008-03-13T22:52:00.003+01:00

Some days ago a post from Jason Calacanis created a lot of discussions and critics. The hot phrase (that was tamed later) was the following:

Fire people who ~~are not workaholics.~~ don't love their work... come on folks, this is startup life, ~~it's not a game~~. don't work at a startup if you're not into it--go work at the post office or stabucks if you're not into it ~~you want balance in your life. For realz.~~

Calacanis was really criticised for it. Since in my small blogger experience I was criticised, too, when I proposed to fire very unmotivated people (a.k.a. zombies), I think I can say something about that.

There are just 3 small things I want to point out:

1. There is a huge difference between Workaholics and productive people. I think Calacanis was thinking the latter while writing the former. You see, productive people are great because they can do a lot of job and still have a life out of the office.

2. Startups need productive people, not workaholics. The reason is that startups need fast results. And results are achieved by people focused on their tasks, not on the job by itself.

3. Startups, also, don't need to fire people. They need, on the contrary, to hire the right people. If they have to fire people, they are going to fail.

2008-01-22T23:10:00.000+01:00

Before getting used to the standards of the project I'm currently working in (that is, methods of 3000+ lines full of obsolete comments and 2-letters-variables), I was quite rigid in what I supposed to be my standard in the size of methods.

I can summarize my standard with the following numbers:

Less than 10 lines – Small Size Method (small is good!)
Between 10 and 100 lines – Medium Size Method
Between 100 and 500 lines – Big Size Method
Above 500 lines – Pachiderma Juggernaut Size Method (refactoring needed !)

The measure is based on C++ code, without comments. Useless to say, I try to keep every method at small or medium size. Few, rare methods can reach the big size, but I usually avoid it if I can.

And you ? What are your standards ? Do you think my methods contain too much or too few code ?

Source code rottens ?

2008-01-10T22:35:00.001+01:00

Yes, I know. Old habits die hard.

I'm always been used to do frequent commits and merges with Subversion. Many commits per day, with small, isolated changes of source code.

I like it. It seems that the project is continuously reshuffled. Alive.

Therefore, now that in the project I'm working on the standard is to commit rarely with big portions of changes, I have always the impression that my changes 'rotten' until they are out of the project trunk.

Yes, I can update my local folder frequently, but I know it is not the same thing. When I commit a change, I'm also putting the changes under the hands of the other developers. They will look at it, they will compile it, they will run it together with their own changes.

Until my brand new code is in local, it is isolated from the Real World. It doesn't join the battle. It lives in a small, separated suburb.

And rottens.

Bleeding fingers, and programmer’s forma mentis

2007-12-11T22:01:00.000+01:00

Family man life. After both me and my wife cut our fingers with the (extremely sharp) twin blades of our herb chopper, I decided it was time to make a simple blade protection with paperboard.

What I did is documented in the following photos. As you can see, the result is not a piece of art, but at least it works. Fingers are now safe.

What makes me thinking is the fact that my wife admitted she couldn’t even think she could make such an object, and while I was working on it she could not figure out what I was actually doing !

In my head, on the contrary, it was everything very clear. So clear that I did not need for measures, plans or even pictures on a sketchbook. Even the small triangles I cut on the sides of the strip (needed to curve the paper around the blades) were extremely natural to make.

Now, this is my thought. This object is not art and it is definitely not an example of high engineering. But it works and was extremely simple to build. This is not the creation of a genius but, still, not all the people have the ability to create objects this way.

I think it is a skill that almost all programmers have. A sort of ‘forma mentis’.

We probably share it with carpenters, bricklayers or designers, whereas interpreters (like my wife), writers, teachers, actors or speakers probably lack of. It is the ability to figure out something that does not exist yet, and the ability to understand, in detail, how to make it real.

This is actually what we do every day.

Do you think I’m right ?

SVG emoticons, free

2007-11-27T22:30:00.003+01:00

Even if I promised myself I wouldn’t bother readers about one of my favourites side projects, Jirino, I’d like to share this stuff with you.

The following SVG file is the clipart for Jirino emoticons: I’m quite satisfied of it because, even if I am not an artist or a designer, I think I captured the balancing between fun and flat graphics that I’d like to put on this tool.

Now I want to share it with you. Feel free to use it, to change it and to distribute it. Since it is incomplete (there is still a lot of common smileys to draw), I will appreciate very much if you return an updated file through the comments of this post.

Enjoy !

Note: you should see the image only if your browser supports SVG format. Anyway, you can still download the file here.

Poll Results: your preferred programming language ?

2007-11-21T00:03:00.001+01:00

End of the first poll !

With just 31 votes, it does not claim to have iper-realistic-and-important results, but still surprised me. Even though the winner of the programming language contest was Ruby (32%), the victory wasn't easy: C++ has been a bold opponent (29%) that was neck to neck until the end.

What to say ? Ruby is portable, simple, flexible, trendy and welcome by the agile community.
C++ is portable, mature, extemely diffused and very powerful. I can say I prefer Ruby, but certainly I have more experience with C++, so I'm happy with both.

And now the next poll: what to vote for ? The next U.S. presidential candidate ? Nooo, I think you should vote for your favourite ... hex magic number !

.. have fun ..

Cool web apps from the TechCrunch40 conference

2007-09-27T23:48:00.000+02:00

What impressed me more from the TC40 conference ? In my opinion, Ponoko is the most intriguing. Think of something like Lulu.com but applied to wood and laser cutting manifacturing instead of books. You design your stuff, choose the materials and cost, and publish your design. Then you can sell your objects online via a virtual showroom, or you can sell or distribute only the design. Obviously designs are not strictly limited to tables or shelves: you can design whatever you want. I think I’m going to install a CAD: there is a couple of objects I always thought of ...

CrowdSpirit is the other web thing that impressed me. It is a sort of “open source and social web for hardware projects”. It all starts with an idea posted on the page: if it is interesting, other people will follow it and refine the project. When the design is good enough, some manifacturers can make a prototype. People involved in the project, then, can act as beta testers and test the prototypes. When the object can actually be distributed, all people involved in the project can earn money in proportion to their contribution. Smart ! (let’s wait for the first real projects – by now, it seems that the most popular if electronic wall calendar)

Source code and lullabies - (Infant Driven Development)

2007-09-18T21:09:00.000+02:00

I am a brand new dad, so in these days I have very little time to spend in programming on side projects. However (strange but true) I have been able to start again the development of one new version of Jirino, my bugtracking and project management tool.

How I have been able to find the time for this project ? Finding out practices, tips and techniques that helped me in this state of emergency !

What tips ? Here is a small list:

Be Portable – all the development environment you need must be installed in a notebook (or some similar portable device). If you need servers, install them also on the development PC. Use the local server when you can’t be connected to the network. Use wireless network, not cables ! Your network will be accessible from every place within the signal range. So you can work, say, in the kitchen, in the bathroom, in bed ..

Fast Start – Everything needed to develop must start and be available within 30 seconds from the power up of your notebook. There should not be manual confirmation or manual tool selection.

Forget The Big Thing – the target for each working session must be resized to very small goals – don’t even think to complete a feature during a session. Try to complete only that 10 lines of code. Test Driven Development helps ! It allows you to focus on very small targets (the single test has to run with success) that -luckily- most of the times coincide with real features.

Forget Time Management – Be prepared to work every time you can. Try to make plans on target results (eg. ‘complete the text search feature’), not on daily tasks. Forget time management practices, from tdP to time tracking. If you still want to track working time, use a small timer that can count seconds or minutes, not hours. Make sure that it is possible to start and stop it automatically or with just a click: if it takes too much time to manage, forget to use the timer. Use your precious time to write code !

Shortcuts Help ! – Prepare keyboard shortcuts for everything you need. The shortcuts should be selected with a single hand… and tools like Launchy or Colibri help a LOT !

Your Tools Help ! - Every single tool that can be used to avoid manual work is important. Try to write your own tools or macros when you see that you spend too much time in repeating tasks.

Every Moment Is Good For Development or Design – Even moments that usually are not. Depending on your concentration capabilities, you can develop during breakfast or plan / design during driving. In the latter case, you can try to record voice notes instead of writing / drawing .. but be careful !

Choose The Simplest Technology (If You Can Choose) – Sometimes there is no requirement for a particular technology or language. In this case, choose the simpler one. Prefer Ruby over java. Prefer Rails over ASP.NET. Prefer small IDEs over huge-with-every-feature-you-can-think behemoths. Prefer ASCII readable formats.
If you are in doubt, however, choose the technology you know better: complex technologies in which you are expert are easier and take less time than easy technologies you still don’t understand completely.

Keep yourself disconnected – Surfing the web, reading blogs and feeds, instant messaging and socializing on the network takes a lot of time. If you want to take maximum advantage of your programming time you should consider to keep yourself disconnected from the Internet or at least force yourself to close all the Internet tools not strictly related to software development. If you want to keep some time for this kind of activities, be sure to connect the Internet only after you have completed the development tasks.

Motivation – Love you project ! No best practice or agile technique can help to succeed as simple motivation. Be organized, be methodic and you will fail if you don’t like what you do. Be motivated, care about the project on which you are working, and you will probably have success no matter if you are very well organized and disciplined or not.

Conclusions

Take it as it is. This post is just a list of small tips that worked for me, my project and my family. Please don’t get it too seriously and –please- don’t flame: I guarantee that I’m not going to create a new methodology or to sell you management books on IDD !!

Five.Posts.Past

2007-09-04T22:57:00.001+02:00

Francesco was tagged by Ilker and tagged me again. This time the post theme is: share 5 old posts that have some value (for you).

These are mine:

Pair programming with the customer (because customers are great, sometimes)

Estimates are like milk... (because now is not tomorrow)

Too many CAPTCHAs are a serious usability issue (because too often I'm said not to be human !)

Zombie developers (because it was the first post translated in other languages)

Which Language ? (because it was the very first post)

... mmm, well, who can I tag now ? Flavio, Gianchy, Dan, Paolo and Catepol.

Some (good) blogs I follow

2007-08-29T22:14:00.001+02:00

Well, tonight I’m very tired and I don’t want to write about programming issues or geek stuff, so I will make an un-ordered list (everybody needs a list) of the blogs I like the most.

(That means the blogs in english language I regularly read)

IQ Blog

Usability, usability, usability. Even not strictly related to technology or computers. Nice to read and sometimes very clever.

TheThinkingBlog

The interesting work from a cute girl, it contains lots of news and facts. Nice layout, a worth reading. As the title says, it keeps you thinking with the variety of contents.

Destraynor

Usability, technology and even programming stuff. Smart guy, always fair and clear.

Mapelli.info - Blogging tips and social trends

Lots of stuff (both technical and non-technical) about the art of blogging. Useful tips and good questions that start interesting discussions.

NOTCOT

A network of design sites and nice stuff. Is it a blog ? I don't know. But is nice to read for new, beautiful, things.

LifeHacker

The best source for hacks and tips of every argument, not only technical stuff but even lifestyle, self improvement and organization, and so on.

Jeff’s Junk

A lot of .NET stuff, well written. Useful.

Mike-O-Matic

Smart things on software development, open source and web.

Rands in repose

Tech life, writing, management. Exceptionally well written (the author has an unique style I'm not very good to describe).

Creating passionate users

One of the best blogs on user experience, authored by Kathy Sierra and Dan Russel. Currently suspended (after events you know for sure), but I hope they will be back soon.

Here they are, my top 10. I usually follow more blogs, both in english and italian, but a complete list would be very very very long. By the way, compiling this list has been a great opportunity to read not only the posts, but also the profiles of the people behind the blogs ..

.. and you ? Are you a reader of a blog that is really good ? Please, let me know ..

Jirino 0.6 is coming

2007-08-24T23:51:00.000+02:00

What a pity. I’ve been told that, due to management decision, the Jirino free bugtracking tool is no longer used in the first team that adopted it. Bad news. But not so bad.

First, I’ve been told that users liked Jirino, and the lack of features or robustness was not the cause of the dismissal.

Second, there are still three other ‘official’ projects using Jirino.

Third, I have to admit that it is time to close a new release with a lot of new, interesting things.

The 0.6 release of Jirino will (probably) contain these features and changes :

. Bugs ? No, items: Jirino is going to manage not only bugs, but also generic documents or tasks. In this way, the whole tool will be more similar to a project management tool than a simple bug tracker. You can use Jirino as a todo list manager, as a Wiky-like tool, or even as a very simple blogging platform.

. Tags: a simple tagging system will replace ‘modules’ and will allow to organize and group items better than before. With more than one tag for each item it is possible to group items by milestone, by customer, by release, by component or everything you can think of.

. User accounts: administrators can add, remove and set user permissions. Users can change their password and settings. Users will also remember the status of the last session.

. Wiki parser: the Wiky syntax parser will be extended so that it will be possible to format text as source, to insert separators, and to enter macros like ‘{{255}}’ rendered as links to items.

. Printing: Style sheets dedicated to printouts allow to make well looking (and usable) reports, without logos and menus.

. Database: for the very first time, Jirino uses a database layout different from the BugZ one. Forgive me, Maxime, but it was necessary to change some tables in order to manage the new features. Are you using 0.5.2beta ? No problem, a small tool will allow to convert an existing old database into a new one.

. Nessence: this is not visible to users, but now Jirino has simplified its own data layer by using the Nessence library, a very simple persistance module.

If you are interested in the next version of Jirino, please stick around. By the way, I think it's time to change the name of this tool: the current one seems related to a Java project instead to a .Net one. Have you got some idea ?

Related posts:

Tails, effortless bugtracking ?

Jirino is online

More kitchen timers ...

2007-08-18T09:35:00.001+02:00

Kitchen timers are cool: Tannaz Sassooni is another girl that is using a timer to break her working time in order to improve his own productivity. She splits her activities in 'chunks' of 15 minutes, but the whole concept is very similar to the Pomodoro Technique, by Francesco Cirillo.

Personally, I'm used to split my time in 25/30 minutes of 'working time' chunks and 5 minutes of 'free time' chunks. I think it fits better my capacity to keep the focus on tasks like software development and writing.

If you are interested in timing techniques or event if you are already practicing the Pomodoro Technique, the 'Fifteen Minutes of ...' post is worth a read: it also contains links to useful widgets and websites.

[Update - Tannaz is a girl ! Sorry !]

Software, application and tools I hate the most …

2007-07-19T23:40:00.000+02:00

(maybe for their bugs and bad design, maybe for my inability in using them)

SandCastle – What I thought it would be a very simple task to achieve (converting XML documentation into CHM files) was something like an odissey. SandCastle never worked on my PC.
I tried tutorials, samples and helping tools but I wasn’t able to see it succeed. The ridicolous thing is that I suppose the hardest part of the work should be the generation of the XML files from source code comments. This task was already accomplished by Visual Studio, so what else is needed ? To convert these files !
Now I am using a custom (simple and straightforward) converter, and I quickly uninstalled this complex tool.

MS Outlook – Complex, big and buggy (never let the database reach the size of 2gb !!) I used it a lot but never felt safe or confortable. Now I'm using Thunderbird. It does less things but it does it better.

SuSE Linux 10 – I installed it when I was planning to make some exploration projects with Mono. After some weeks, I decided that was not worth.
The official release on SuSE 10, on my PC, was less stable than Vista BETA2 ! It was frequently hanging, did not connect to the Internet via ADSL modem (only via Router), and failed even in very simple features like mouse wheel management … it made me very upset and less confident with the Linux robustness … fortunately some months later (the whole PC was formatted) I tried a Linux distribution called Ubuntu …

InstallShield DevStudio 9 – One of the most powerful installation frameworks but even one of the most chaotic and complex IDE I ever used ! Catchy graphics and layouts seem to be there only to hide a lot of illogical links and unintuitive settings ..

Painbrush – One of the most distributed tools for image processing can’t resize, crop, zoom properly or manage a decent palette. Why still distributing an application that lacks of such commonly used features ? I think pbrush is the same application I was using 10 (or more) years ago..
But in the meanwhile images are getting bigger, with more colors, and it is very common to copy them, change and transfer between tools.

The Office Assistant & The Office Toolbar – The only reason I choose a custom Office installation is to uncheck these components. I think they are two of the most useless and boring pieces of software never produced by Microsoft.

(in the same men-hours couldn’t they add features to Paintbrush ?)

And you ? Do you have tools and software you hate ? no matter how good it is ?

Take more breaks !

2007-06-24T11:12:00.000+02:00

I recently read this blog post from 37 signals saying that they are breaking their development time in 4 hours chunks. After a chunk you take air, think something different and when you restart you can re-evaluate your work, redesign or take a different approach if you were stuck.

This is great: taking a pause is always good during the development. The more breaks you take, the less wrong decisions you take, and you are always fresh and confident with your work.

But why don't we try to make better ? We can try to pause more frequently, not just every 4 hours but every half an hour..

.. the Pomodoro Technique, by Francesco Cirillo, is an italian agile practice based on the use of a chicken timer to set the rithm of your development time. The basic rules are simple: you get a kitchen timer (a tomato shaped one is always better); you set 25 minutes; you work without interruptions during the 25 minutes; when the timer rings, you set other 5 minutes in which you take a pause (read mail, have stretching, google around); when the timer rings again, you can start another pomodoro with more concentration and energy.

I tried it for several months, and I can confirm that this practice is not only funny, but it really works. You can try, too: the paper is now translated in english under the creative common license. Have fun with it !

... and my workspace, too

2007-05-29T20:19:00.001+02:00

Ouch ! I've been tagged by Francesco, so I'm going to expose my workspace, as suggested by TimeForBlogging. (click to enlarge)

Noticeable things:

1. No double monitor, although I love it.

2. Lots if IKEA stuff, including the desk, drawers, the Stokke-like chair, trash bin and shelves.

3. Lots of mobile and cordless phones.

4. Food. Healthy food. This time you can see cherries of 'Vignola' at the right of the laptop.

5. Old vinyl record player (above the two desktops) for my old 33rpm records.

.. and these are some of the workspaces I like the most :

David Pitlyuk (cool)

Eleanor (cute)

Oracolo (great integration with the blog's style !)

.. and this !! (yes, I'm going to be a daddy too)

Pair Programming with the customer

2007-04-12T10:15:00.000+02:00

Recently I had a development session of a software component directly in the customer’s office. We were sitting at the same desk, notebooks placed side by side, and I started writing code. The presence of the customer sitting beside allowed me to quickly ask for explanations and information on the structure of the framework and the hardware with which the component was going to be integrated.

At the same time, the customer was able to give me further indications, to check if the component met the requirements, and even to change some non crucial requirements.

At the end of the session (4 hours) the component was working, tested, perfectly ready to be integrated in the customer’s project, and we were satisfied of the results.

Afterwards, I realised an aspect apparently trivial: I was doing Pair Programming directly with the customer.

If you don’t know it (that’s difficult), Pair Programming is one of the practices adopted in the XP (Extreme Programming) methodology. It means having two programmers writing code on the same computer. Usually one writes code and focuses on syntax while the other is the co-pilot, checks for errors and gives high level design suggestions. Roles (and keyboard) switch frequently in a working session.

Pair Programming is based on the principle that two developers side by side are more productive and make less errors that the same developers working separated. In its core principle, pairing is based on participating and the completion of the skills of both the programmers. That’s why, for example, it is recommended to put a senior aside a novice.

So, excluding things like ‘switching of roles’ or ‘shared computer’, I think I had a real Pair Programming session with the customer. The complement of skills (I, developer, know more about the programming language and algorithms, you customer know more about the framework, the domain and the requirements) leads to a real conversation that increases the productivity in proportion with the contribution of each one.

The result was excellent, and in conclusion I just want to add some little considerations learned from this experience:

A) Pair Programming do not need to be intended in the ‘classic’ mode

It is not necessary to have two developers of the same team, it is not necessary to have a single PC, and it is not necessary to switch roles. On the contrary it is necessary to keep the conversation alive and that both contribute to the target (source code, documentation, a book or a design for a desk).

B) The “Pair Programming” label is not necessary

This is valid particularly in environments and cultures in which agile practices are looked with skepticism (I don’t know other countries, but at least in Italy it is very common). Try to say to a manager that you will make two programmers work at the same lines of code and that manager will be determined to fire you. Work together with a colleague or a customer with the right spirit of conversation and participation but without actually show off the ‘pair programming’ flag and you will still get excellent results without the need of evangelizing others.

Finally, when good results have already been achieved, it is simpler to say that pair programming works ..

Blogged with Flock

Estimates are like milk ...

2007-03-20T17:40:00.001+01:00

.. and after a while they turn bad.

This is not strange but I was really suprised when, some days ago, some management people complained that I changed a seven-months-old estimate of a feature never implemented in the main project.

Well, in my mind it was very clear why the estimate changed, but it was not as much clear for other people. Fortunately, a brief explanation made all people happy without the need of rude and long and boring negotiations. But this little misunderstanding left me with a sense of "something should be done better" ...

Let me explain some details. The project in which I make estimations is a long project whose lifetime is expressed in years. Macro-features are regularly implemented in iterations of 3..6 months. Micro-features and bugfixes are implemented continuously. Estimates usually measure Size, Features and Cost. These estimates don't produce Schedule and Effort, which are usually negotiated by other people.

The Cost of each feature 'officially' depends on Scope and Complexity. That's basically the reason why management people believed that if the feature is left unchanged, the estimation is therefore unchanged.

But there are a lot of 'unofficial' variables that can change the Cost of one feature in the time. Here are some of these variables :

Project Complexity - The project in which we add the feature can get simpler or more complicated in the design, depending on what has changed during the previous iterations. If a lot of good refactoring was made, it is probable that the design has gone simpler and the project 'accepts' the new feature joyfully. On the contrary, an iteration in which a lot of low-quality code was added, or in which massive altering features changed the overall design can make the new feature harder to add.

A wave of features - When a lot of similar features are developed in a single iteration, they are likely to cost less than each feature added in a single iteration. This side effect comes from a natural development thrust that cuts the 'startup' time for each change. If the developer has to insert a bold-magenta-border (?) in twentyfive forms in a day, it is probable that he will take less that adding the same bold-magenta-border in one form once a month..

Team Skills - Due to turnover or team changes, the team itself can be more or less skilled. In this case estimation can change to reflect the capabilities of the team, not in terms of effort but in terms of project knowledge and skills.

There are obviously a lot of other variables that change one estimation after some time, and the related considerations can vary a lot. For example, it is quite obvious that some of these variables (team skill, complexity of code) should be considered at technical level but they don't have to burden a customer.

What is important, in my opinion, is the consciousness that an estimation is the result of a process based on many variables in a precise moment. When this moment passes, the estimation should change. In an ideal world (or maybe in some months) there should be a tool that recalculates all estimations as the project evolves, for every single line of code. In such a scenario the management people could watch the development costs just like we watch stock exchange, so that they can choose the right moment to insert every single feature ("Bob, the 'tags' feature cost has just falled to 10 man hours - let's put it in the next iteration before it raises again !").

Well, it will be very nice ... but in the meanwhile I will add an 'expiration date' to all my estimates, just like milk !

technorati tags:estimates, features, software

Blogged with Flock

That's a cool way to manage source control

2007-03-10T10:32:00.001+01:00

Springloops is a "unique source code management tool focused on web development teams".

Seems like a kind of SubVersion hosting solution (examples show up TortoiseSVN !) expecially designed for web projects. Provides an elegant interface and manages deployment to multiple web servers (via FTP) and integration with Basecamp.

By now, at least during beta phase, all services are free.

I don't know if Springloops is really that good but I'd like to point out how much is elegant, simple and well designed their website: simple but cool graphics, no animations and scripts and flashies, and fine layout.

Well done !

technorati tags:tools, web, subversion, springloops, basecamp, hosting

Blogged with Flock

The zombie developer saga - when a zombie is fired

2006-12-20T08:07:00.000+01:00

In one of my previous posts, I made some observations on a kind of unmotivated developers that I simply call zombie developers.

Well, two weeks ago, one of these kind of developers was fired in a company in which I work. It was not a pleasant thing, believe me, expecially under christmas.

BUT, WHAT HAPPENED ?

I think that something bad has been grown up during months: a developer starts to get lesser and lesser motivated and feels very frustrated about his job, his professional results, the relationships with the boss and other teammates.
On the other side, the management of a small software company gets lesser and lesser results and feels frustrated about project results and the relationships with one the employees.
When a break point is reached, something must happen.

WHAT WAS WRONG WITH THE COMPANY ?

They monitored the results of this developer and, when something wrong was very obvious, they tried some fixes:

1. Physical shift: the company believed that in some way, the physical placement of the developer was not good. Too much far from the team, too much hidden from eye-to-eye contact. So he has been moved to a more central desktop, but the change was not useful.

2. Another chance: different kind of tasks were assigned to the developer, hoping to find an area (GUI programming, non GUI programming, different languages) in which he was more confident. It didn't work: results were constantly worse.

3. Explicit warnings: the developer was not fired without any sign of what was happening. He was warned by the boss, and asked to change his behavior.

The worst error, I think, was the fact that the company has taken a lot of time to detect a real problem with this developer. It seemed that some bad results on a project were due to the difficult of the project itself, not on the developer motivation. When the developer was moved to another, simpler project, the new team leader immediately complained about the bad results of the zombie, and this time the problem was clear.

WHAT WAS WRONG WITH THE DEVELOPER ?

1. First (and lethal) error: when the boss warns you, you have to seriously choose what you want to do. You can (professionally) do self-criticism and change your efforts on your job or you can (professionally) answer to the boss critics and try to defend or explain your position.
It is not acceptable to get the critics from the boss and continue to do what you were doing last day. You can't be found playing Solitaire or Minesweeper after you have been explicitly warned, unless you really want to be fired. Period.

2. Second error: the lack of communication with the entire team. This is a development team brought together, where people help each other. It is a natural evolution due to the stressfull projects in which they work. If the zombie had a real communication with the rest of the team, he would probably have received some help from them. Instead, the approach of the developer was to isolate himself, and to blame or give annoyance to other teammates.
Such an error can have dramatic effects: developers are not expected to be lonely cowboys riding in the sunset of a dusty desert. Developers are expected to be social and to work together on a target. It is not a matter of beying shy or boring. You can be shy or boring, but if you want to be a good developer, you need to learn to work in a team.

WHY WE TURN INTO ZOMBIES ?

If you watch the comments on the Zombie Developers post, you will find several descriptions on ways to produce zombies: projects got cancelled, developers moved around (as you can do with furniture), excess of bureaucracy and creativity killing.
Most of the times, zombies are just created by simple lack of respect for the employee.
I think that each company has the responsibility to create and maintain an healthy working environment. Even a company that do not believe in motivation needs to do something, because lack of results or the excess of turnover can be a serious problem in productivity. A software company, too, depends heavily on developer's creativity, healthness and motivation.

A team of zombies can't produce profit.

WHAT SAVES ME FROM THE ZOMBIE VIRUS ?

I am talking about zombies, but I was, sometimes, very near to turn into a zombie too. Some projects really suck, and sometimes companies use programmers really as objects, not human beings.

Still, I love this job and I always try new ways to motivate myself, instead of waiting the company to do this for me.

Much bloggers (The Retrospector, Angry, HBS, TechRepublic) have posted rich lists of suggestions to motivate geeks (or at least to avoid to un-motivate them): most of them are really talking about ways to avoid turning development teams into zombie flocks.

The most of these posts is simply great, but I don't like when they suggests futile stuff like gadgets, food or things like this. It does not seem to be realistic. I am a professional and I don't expect that the office in which I work is changed into a theme park or a bistro. I am just expected to produce results. The company is just expected to pay me, and to put me in the right conditions to do the job (with the right tools, a right environment and the information I need).

I don't want geek gadgets or snacks.

Well, some small features in a project give me more motivation than usual:

1. When I see my name in the project credits or in the project web page I am more motivated because I want that my name is only on successfull projects.

2. When I can propose or discuss solutions, instead of receiving closed designs choosen by others, I am more motivated because I can put my experience into the project.

3. When I can talk with other developers, and discuss things instead of having sterile 'ball passages', I am motivated because I can contribute to the evolution of a better team.

4. When the development team is a real entity, and has direct communication with the outside world, I am motivated because this is a team to work with.

That's all. If you have some comments on motivation, zombies and development I will be very glad to read and discuss them. Today I am a little tired and not very clear, so I recognize that the english quality of this post is very poor. I apologize about it.

See you (unless you are turned into a green zombie - mmmmgrhh)

Ooops, I will not experiment XNA, by now ...

2006-12-12T15:38:00.000+01:00

I'm not a game developer, but a lot of projects in which I worked on required graphics, sounds and multimedia effects with results very similar to those that can be seen in a common video game.

So, when Microsoft announced the XNA Developer Center, I was really excited. Finally there was the availibility of a non-expensive platform in which to program multimedia applications compatible with PCs and Xbox 360 (which is an interesting hardware even for applications that are not games).

I was planning to buy an Xbox for test and development when I opened the FAQ page for the XNA Developer Center. That's what I discovered:

Q: How can I share my Xbox 360 game built with XNA Game Studio Express with other Xbox 360 users?
A: To share your Xbox 360 game with friends, four requirements must be met:
. The individual you are planning to share the game with must be logged in to Xbox Live and have an active subscription to the XNA Creators Club
. The receiving user must have downloaded the XNA Framework runtime environment for the Xbox 360
. The receiving user must have XNA Game Studio Express installed on their own development PC
. The game project, including all source and content assets, must be shared with the receiving user. The receiving user then compiles and deploys the game to their Xbox 360. We are actively working on other ways to allow you to more easily distribute your games and are very excited about the possibilities this will open up for independent game development.

Oops, not very simple. There are really too many conditions to meet, by now, for such an application to be distributed. I think that only groups of entusiast friends can be the target for such a product..

Q: Can I store my XNA Game Studio Express game on my memory card or CD/DVD and share it with a friend?
A: No. Games developed using XNA Game Studio Express cannot be shared through a memory card or CD/DVD at this time.

Ooops, no CD/DVD distribution !

Q: Can I create non-gaming applications (such as a Media Center/Player) with XNA Game Studio Express?
A: On Windows this is possible, but the initial release on Xbox 360 is tuned to writing games. This is an area we are actively looking to the community for feedback on the types of applications they want to write for their Xbox 360.

Ooooops, only games ... !!!

These answers are the main reason why I will not experiment XNA by now. Still, I am quite optimistic, since I believe that most of the lacks in the distribution of Xbox applications should be due to techichal scheduling and not to marketing decisions. So, I think I'll wait the 'professional' version, and then check what can I do with such this kind of platform.

MindLink 0.0.5.1 Beta

2006-12-11T08:09:00.000+01:00

MindLink is a small .NET 2.0 tool written in C#.

It is a small experiment that was initially designed as a .NET clone of the great SlickRun utility, but evolved rapidly in something different.

How does it work ? MindLink stays hidden in the tray area for most of the time. When the user presses a shortcut (Win+Q), a small popup window asks for a 'quick word'. The quick word is then linked to something to execute, like an URL, an executable or a document to open.

Sounds really simple, no ?

MindLink is available via this link and is completely free.

Too many CAPTCHAs are a serious usability issue

2006-12-05T09:53:00.000+01:00

One of the best things in modern Web applications is the possibility to post feedback, comments, images or other resources on the fly, without the need to register to an online service that probably we will not use later. I think it brings more 'weight' to a lot of web sites. Just think how many intelligent, funny or smart comments were never sent before just because the author didn't have time (or didn't want) to register to the web site.

On the other hand, anonymous comment submittal must face an huge threat: spamming is always out there and filters need to protect servers and applications from a flood of unneeded posts.

The very common technique used today to filter unwanted posts, or at least to filter the automated ones, seems to be the CAPTCHA. Most common CAPTCHAs prevent bot access to the site content by asking the user to recognize some characters displayed in such a way that makes difficult to read from an OCR.

Yes, really difficult. So that most of the times Ugly Letters (a.k.a. stretch-that-letters-until-they-seem-insects-squashed-on-a-dirty-wall) can't be read even by people with a perfect sight ...

The most bad problem, expecially now that CAPTCHAs are so common, is that human recognition with CAPTCHAs don't let access web content to a lot of people with problems of sight. The worse part of it is that when the test fails, the user is labelled as 'not being human'.

This is not only discriminatory, but also offending !

Well, I did some research, and I am happy that what I'm writing is not a new issue. Just follow the W3C paper on this usability issue to read something on valid projects and alternatives to CAPTCHAs.

And now the YAWSI (Yet another Web Startup Idea).
What I'd like to see ? Well, there is a full plenty of Web startup companies around developing calendars, media sharing and tags. But, please, some smart developer doesn't want to start a service that makes available reliable, accessible and nice antispam filters ?

I think it's time for such an online service.

Trakkkit - a tracking tool that looks like the Reddit.com

2006-09-14T16:52:00.000+02:00

Some time ago, in order to achieve a a better project management for me and my colleagues, I was trying to design a powerful but incredibly complicated tool for planning and tracking.
After some hours of brainstorming, and some bewilderment, I asked myself if I really needed so much functions. The answer was negative: all I needed was a tiny ToDo web list to be used also as a tracking utility. To ease the work, and have a cooler tool, it would be very nice to 'borrow' the user interface concept from some already existing Internet service.

This is were Trakkkit came from.

Trakkkit is a very simple ASP.NET application written in C# able to keep lists of tasks and to track the time worked in an easy way. This tools manages (obviously) multiple users so that it is possible to work in team through the company network. The user interface is very similar to Reddit.com, even if objects and operations are quite different.

Some secondary features include simple statistics and reports, some Wiki-like formatting in the description of the tasks, and the possibility to submit messages visible to all the users. Other features in development include some tag support, and mail notification (by the way, the development tasks of the Trakkkit project are managed with Trakkkit itself).

The project is unmistakably in a rough and BETA stage and surely it needs some refactoring to the source code. Still, it works. So if you want you can download the latest release from the dedicated homepage (it's free).

All suggestions, bug reports are well accepted. I'm also looking for some goodwill programmer who wants to write some additional feature or enhancement ...

PocketPC tools I can't go without

2006-09-12T17:36:00.000+02:00

I am not a power PocketPC user, and I think I am not using my PDA very much in depth. Still, I use it for a lot of tasks: reading, writing, taking appointments, GPS-ing, calculating, playing music and videos, and so on..

I used a LOT of tools, freeware, payed and trial. Many of them are installed, used for a while and then removed. But when I (frequently) reinstall all my PDA sofware, there is a limited set of tools I can't go without. Here they are:

1. Windows Media Player

Luckily, I don't need to install it because it comes with Windows for PocketPC. It may not be the best player for MP3s and MPEGs, but it is very easy to use and accepts a lot of file formats. I have a LOT of (legal) music CDs converted by Windows Media and then copied on a SD card and I am very happy with it.

2. Adobe Reader for PocketPC

This is definitely NOT a good software, but is very useful because I like to port a lot of PDF files (books, articles, novels, manuals, etc) and read them when I need or have time to. The very useful feature of using PDF files on the PocketPC is that PDF files can be 'reflowed' to fit a very different screen, and it works quite well. Also, Adobe Reader remebers the last page of each file even when restarted and it is quite intuitive.

3. TranCreative Magic Button

This tool is a sort of program manager that is placed in the upper bar of the screen. It is smart and simple and allows to see which applications are running, and to decide which application should be kept in memory and which one will be really closed.

4. InsertCredit Battery Bar

This free application displays in as a line of the Today screen. It shows just the battery charge, but it is really useful.

And you ? Which tools are your favourites and need to be on your PDA ?

Vacation, at last !

2006-08-03T11:40:00.000+02:00

I'm going to vacation, in Calabria (5..27 august) and in the Ponza island (2..10 september), so I think I won't make posts for a while.

See you !

Very trivial GUI design error on PocketPC

2006-06-01T15:05:00.000+02:00

When I developed my first applications for the PocketPC I made a lot of design errors resulting in very hard-to-use tools.

This is one example. Watch the following image:

Do you find the mistake ? It should be obvious for expert developers, but is not very easy to find. So, try to imagine to be the user. Go on the lower edit control and change the selected text with the keyboard. The result will be like the following image:

Ouch ! Where the £%&$ is my edit control ? You will probably have something to teach to that stupid programmer ...

It just happened that the control was placed in the area that is normally reserved for in-screen keyboards, so that when the user edits some text, the control is hidden. This is not a real bug: the text can still be changed and nothing crashes. Though, the user experience is not the best, and probably this dialog will not be candidated to usability design awards. This is bad user interface design. Things work, but the user is not happy.

So, this is the first trivial hint on PocketPC GUI design that I want to share :

- Don't place edit controls in the keyboard area. -

In the following post(s) I will describe other hints, some very basic and others more subtle. I would have liked to read for these hints when I made my first mistakes ..

User interface design is particularly important if you plan to port an application from the PC platform to the PocketPC.
A lot of people will say: - Well, if your code is standard and well designed, you just have to build it with the a new compiler, fix some bugs, stretch some dialogues and you are done. -
Don't believe them. You will probably be able to compile, fix and run a lot of code, but in my opinion most of your time will be spent in redesigning an existing GUI.

Bye.

AJAX - is it a design pattern ?

2006-05-10T20:05:00.000+02:00

Almost everybody today seems to have something to tell you about Ajax, Web 2.0 and all that cool web stuff ... me too !

The first time I introduced myself into the gorgeous world of Ajax, it seemed to me that it was a nice technology to install, link, run and use to deploy cool web apps. Well, it was not really this way.

First, what is Ajax ? Billions of blogs give you this definition:

- The dumb will call it a Library (I don't want to explain why it is wrong).

- The dumber will call it a language (I can prove it, I've read this on a magazine !!).

- The smarter will call it a group of several technologies (which is better, because it makes use of existing technologies to solve a common problem).

- The even smarter calls it a Web Development Technique - that's Wikipedia definition - (and I quite agree with this because basically you can make Ajax with a lot of languages, variations, platforms).

... Im my humble, humble opinion, and if I am not too drunk, my personal definition of AJAX is a Web Development Design Pattern.

Why a design pattern ? Let's get back to Gamma. He quotes that "Each pattern describes a problem which occurs over and over again in our environment, and then describes the core of the solution to that problem, in such a way that you can use this solution a million times over, without ever doing it the same way twice".

This reminds me of web common problems, and web common solutions. Common solutions can be encoded into patterns because patterns are descriptions of groups of objects designed to solve a general design problem in a particular context.

What is my general problem ? I need to make changes on this page without having to re-read it from the web server. So, can I try to describe Ajax as a design pattern ? Sure, following the GOF general pattern template.

Pattern Name
Because it is Asynchronous Javascript and XML, we can call it AJAX.
like a soccer team ? nice !

Intent
The intent of AJAX is to make web pages more reacting by exchanging small chunks of information with the web server, so that the entire web page does not need to be reloaded each time it needs to show new information to the user. In this way, we want to improve the Web user's experience.

Motivation
The old, boring and hopelessly one-dot-zero web application does this: the user acts on the page, the page sends an HTTP request back to a web server. The server processes something and then returns another HTML page to the client. It’s good for hypertext pages, but now the web needs something else. Web designers need to fill the gap between web application and desktop applications. So they need responsiveness, flexibility and reactivity of the whole application, and not just relying on browser functions.

Participants
Web browser. Ajax engine. JavaScripts. Web server.

Collaborations
When is loading a webpage, the internet browser loads also an Ajax engine. The Ajax engine is responsible for both applying changes to the page the user sees and communicating with the web server for data exchange. The Ajax engine makes the communication asynchronous, so that the user does not need to wait the page to refresh before getting user interface feedback. That way, the user doesn't have to wait for flickering or waiting pages, but has always an application which is sensible to the user input.

Consequences
. "Interactivity" - the user has a more responsive service. This appears to be a Pros.

. "Portability" - not all the browsers support XMLHttpRequest, or the technologies involved in AJAX. This appears to be in the Cons.

. "The Back issue" - some criticizers are particularly sensible to the fact that the standard Back browser button loose his meaning in an AJAX application.

. "Asyncronicity" - There is lot of work to do when developing asynchronous code.

. "Javascript enabled" - Not all the users enable their browsers for Javascript.

. "Accessibility" - not all AJAX applications are still accessible to everyone, for example for the browsers for blind persons.

Implementation
At a very basic level, on the client side, use XMLHttpRequest to ask asynchronously for XML data packets from the server. When the data arrives, parse it and use the data to change your page on the client side. Access to page elements by using the DOM and simplify things with CSS. Do it with Javascript which is the most common script language you can run on the client side.

Known Uses
Almost the 90% of the Web 2.0 services.

So that's the AJAX design pattern. Do you think I was irreparably drunk ? Maybe. I am not the only one that uses this definition, and for sure I am not the first. Do you agree with me ? It doesn't matter. Why ? Because the most important question has to be:

- Does it work ? -

If the answer is - Yes - we don't really need to find the perfect definition, or to determine whether AJAX is a technology, a DLL, a design pattern or a technique. If it works, whe should make use of it to solve our design problems ... and to write the coolest web apps we can. (that's the hard game)

Tails, effortless bugtracking ?

2006-05-09T14:34:00.000+02:00

Even if I like to write my own bugtracking tools, I can't resist the need to try and search for new and different approaches to the bugtracking problem.

I'm especially interested in finding new ways to manage all these reports and descriptions, possibly focusing on the process of bug fixing instead of report maintenance, assignment and scheduling ...

Recently, I followed up a link to Tails, that seem to promise a lot of interesting stuff.

It is still in beta and is not possible, by now, to try it. But screenshots and the idea are very interesting.

Zombie developers

2006-04-19T14:59:00.001+02:00

In some rare occasions I had the misfortune of dealing with a weird category of developer: the zombie ...

This kind of guy can be identified during meetings with a good chance of success. If you want to find some zombie developers inside your team, try to announce conflicting fake news like these:

A - Our application won the 'World's Best Tool' award'.

B - The number of bugs in our application is over the population of China.

C - We received congratulations from the 2005 IT Nobel prize.

D - We are all going to be fired. Now.

... and so on.

What happens ? No matters if you are announcing extremely good or bad news, you'll find that the zombie developer will always keep the same face look. While the rest of the team will probably pass from Massive Hair Tearing to Extreme Cheering, the zombie developer will keep himself highly unemotional.

Why ?

WHO IS THE ZOMBIE ?

Basically, the zombie developer is a figure that, whilst working in your team, may be fine in every other place.
He has no interest in producing good quality code. Maybe he has no interest in producing code at all.
Independently from his skills (that do not get better in time), the zombie is not interested in improving his position, his career and in getting professional success.
Sometimes he is just too shy (there is a lot of geeks very shy, it should be the trite cliche of the nerd that marries his computer) and simply he is not good at expressing his own ideas and likings. Sometimes, he is just right with it.

WHAT DAMAGE CAN INFLICT ?

1 - PRODUCTIVITY

Here is another nice experiment you can use to identify zombies. During the meeting, put into planning a trivial task (change the color of the 'Apply' button) and evaluate it with an excessive amount of hours.
At the end of the iteration, check the completed tasks:

- The 'normal' developer has completed the task in less time, and he completed other tasks, too.

- The 'lazy' developer has completed the task in less time, but did take at least the same time for fixing the bugs he made.

- The 'Zombie' developer has completed the task *exactly* in the time evaluated.

Why ? You know that the zombie is not provided with his own intellect. So he has identified the evaluated time as a requirement for the task to be completed, and he takes that exact time to complete because you asked for.

-Nice !- you may think -I just need to assign him long tasks evaluated with few hours !-

Wrong. It doesn't work. Sorry.

If you do this, the zombie will accept the task without complains, and at the end of the iteration he will have not completed the task.
When asked for the reason, the zombie will justify that - Err... mmm... it needs more time -.

2 - (UN)DEAD WEIGHT

One single zombie developer can slow down the overall team progress and decrease the project quality, depending on how much tasks are assigned to him. The zombie developer adsorbs working time of managers, team leaders, testers and even other developers.

- He consumes managers and leaders time because they have to do extra job to find out the 'right' tasks to assign to him (generally not vital tasks) in order to limit the damage on the project.

- He consumes testers time because he generally does not produce good quality code, so that testers loose precious time in detecting trivial, ordinary, apparent bugs instead of finding more hidden an dangerous faults.

- He consumes other developers time because when he doesn't finish his coding and the software needs to be shipped, usually the same code needs to be completed by another programmer. Also, the zombie code frequently needs reviews and rewriting by other programmers because of its very low quality.

3 - TEAM MORALE

Remember that zombi-ism is contagious: when a zombie developer bites the normal one, he will transmit the zombie virus ...
So if you have even a single zombie developer in your team, it is very likely that you will count two or more zombies in a matter of some iterations.

That's because of the risk that a zombie is rewarded for being a zombie.

Why ? Because a developer that takes two weeks for changing the colour of a button will probably be assigned for the most easy, less critical, very trivial tasks in all the project. So he will complete all of his tasks, and he will never manage panic-bugs or extra working time, and does not have to explain anything in the meetings.

And he takes the same salary of more stressed developers.

So, unless you have very mature people in your team, it is likely that someone will think that being a zombie is not that bad ..

HOW TO CURE HIM ?

A zombie developer is extremely harmful, so you have to cure him or kill him before he does too damage to the team. So do it now, and do it fast.

1 - GIVE HIM COOKIES

First try. Let's give him a chance.

Talk with him, say the zombie that you will assign him a very important task to complete.
Keep him very controlled and, if he does a good job, reward him with more glamorous tasks.
Glamorous tasks are the cookies you waggle in front of him. The tip is to find something that he-likes-very-much. There is always a task really amazing for someone, you just have to find the right one. Some people like to do graphics and icons, some others like to find code snippets in the internet, some like to write web pages.

If you find the right cookie, it's just a matter of time. Gradually assign the zombie more and more challenging tasks and, attracted by the cookies, the zombie can even heal himself.

2 - THROW HIM IN THE SCRUM

Shock effect.

Give him something difficult to complete. Something that demands time, brain, skills and carisma. If you can afford it, put the zombie as the leader of a sub-project, and give him a junior developer to follow. Give the zombie the opportunity to design components. Put him in a position in which he is expected to publish news and documents, or articles. Possibly with direct feedback by other users.

The tip, here, is to put the zombie under-the-spotlight. Zombie developers tend to keep themself hidden, because they don't want responsibilities. So, responsibilities should find them.

This strategy is very risky, so be prepared for failure. Try to have a 'B-Plan' in the pocket.

3 - SUPPRESS HIM

If everything else fails, kill the zombie.

It seems that a shot in the head has good results, but is less gory to reposition him to another project, or even dismissing him. If you can't do it, talk with the boss or with someone who can do it. If you can't, the last choice is to isolate the zombie from the rest of the team.

Remember that the project is more important than the people, and the team is more important than the single component. If a developer is not likely to work in a team, the team should refuse him.

If you feel too cruel, talk with the zombie and explain him the problem in the most honest way you can. If he understands your position, it is a good result. If he doesn't, no hesitation or misgiving: do what is the best for the team.

Paint.Net

2006-04-06T14:07:00.000+02:00

Paint.Net is a nice Open Source paint application. It supports several file types, layers, special effects and does all you need without too much problems. It is also fast and has a cool GUI.

By now, it is my favourite tool for bitmap editing (InkScape is the one for vector drawing), because I was very eager to uninstall Gimp (Gimp is powerful, but I hate all that popup windows and how it is definitely clumsy in everything)

Oh ! he thinked ahead ...

2006-04-04T20:28:00.000+02:00

This great post by Matt Magain jokes around Web 2.0 apps:

http://ww w.sitepoint.com/article/application-fluff-hot-air

Matt speculates on the possibility of the promotion and selling of a new Web 2.0 app without actually developing it.

It's scary, because some of the foolproof strategies proposed are realistic. I really think you can sell a new web app that still does not exist.

You just have to:
1. Find a nice idea.
2. Insert some buzzwords in the concept (AJAX is mandatory).
3. Find a trendy name and design (remember that also rounded corners are mandatory).
4. Set up a blog, some fake reviews and feed.
5. Accepts members only by invitation.
6. Always keep the BETA label.

Well, I actually planned to do something similar, but now that this technique is revealed I think I have to write some real code ... uff ...

Nice try, but not very friendly ...

2006-03-29T10:58:00.000+02:00

Sometimes the attempt to create a user friendly application yelds to weird results ...

Skobee is a web site that solves a very specific need: managing meetings.
It has a very clear and smart user interface with readable messages, large buttons and a lot of clever solutions in design.

But ...

1. It is not possible to enter ZIP codes which are not in the U.S.

Well, are you developing an Internet application ? Don't you know that Internet apps market is The World ?
Ok, let's suppose you still want to use the Zip Code: do you really need ? You want this information for a better search of sites for meetings. Good. So, don't we have Web APIs that allow to place everyone in a world wide map ? It is possible to use Zip codes but also addresses and coordinates ...

2. Entering the time zone

Watch the following screenshot. Skobee is asking for my time-zone. Let's forget that they already asked me a precise location with the Zip code.
To select the time zone, they provide a simple drop down list with some common time zones, with a nice formatting so that they are readable and clear, easy to search, and don't take 800 pixels of width.
But ... were is Rome ?

I'm placed in Bologna and I usually search for Rome as my time zone ... So I start to think what other european city is in my time zone. London for sure is in GMT0 and I'm in GMT+1 ... Moscow seems too far. I try Amsterdam (which, in the drop down list, is placed before London) but I need to check my Windows clock settings just to be sure ...

Well, it is Amsterdam.

Thank you Skobee because now I know that Amsterdam is in my same time zone. I never realized it.
But, is it user friendly ? If the GUI cuts some information to improve readability, why don't add some items to the drop down list ? When I usually set the time zone, I search for 'Rome' and I search for 'GMT+1', because I know these information. Why cutting it ?

Skobee has made really nice usability solutions in this site. By now, it is wrecking it with little but annoying defects. I'm sure they will fix it at last (at the moment they allow to register without inserting a zip code) but in the meanwhile it is a nice example of small things to avoid when designing a user friendly application.

RallyPoint - an online collaboration tool (or just another online word processor)

2006-03-28T14:26:00.000+02:00

Another cool web app is RallyPoint. It is an almost-complete online word processor that allows simple collaborative tasks.

You can manage a group of pages (documents) and a group of users that can edit these pages. You can create and use document templates to avoid starting from scratch when you add pages. You can add documents, select which users can view or edit it, and start to manage these documents with your team of writers.

Nice features include the possibility to insert comments (other users are notified via email when a new comment is posted) and to lock the page while editing. It is also interesting the clipboard compatibility with MS Word format. Versioning is supported, so if you write some crappy sentences you can eventually go back ...

So, how to use it ? It may be useful to edit specifications or articles within a team of collaborators. It may be also used as a 'more complete' Wiki site, so that it is possible to share information. Just want to use it as a single user word processor ? It does not need to be installed, and the document is not stored on your PC, so you can edit your novel on any PC connected to the Internet, no matter if you are at work, if you are traveling or if you are just sitting at an Internet Cafè ...

http://www.rallypointhq.com/

Cool Web apps

2006-03-08T11:31:00.000+01:00

If you want to play with some cool new apps, here are some links:

http://preview.local.live.com

This technology preview website allows to search for addresses in some US cities, and to drive (virtually) to the destination with a double pane view made of a map (satellite photos or street maps) and the interior of your (virtual) car.
By now, this is not a great demo, but the idea is funny.

http://www.live.com/

This is a great page, still in beta. You can customize the page space as you want. You can subscribe to RSS feeds, watch for news, install gadgets (I really like the weather monitor) and manage email. The page is very dynamic so you can add, remove, reposition subwindows on the fly. Fine, clear and useful. A good example of Ajax at work.

Web 2.0 - in Italy ?

2006-03-02T09:02:00.000+01:00

"Web 2.0" is for sure one of the most hyped terms of these days.
Though, in Italy is not very well known yet. By now, seems a matter of gurus and well informed guys, so you hear a handufl of discussions not very tangible ("Web 2.0 ? Will it be compatible with my web site ?") and not much more.

But if you are used to follow Internet news and technical blogs, they all talk about Web 2.0 ...

So, why are we still ignoring it ? I think is for two basic reasons:

1. In Italy, we follow the States technical trends with a lag that ranges from some months to some years. I think it is because the IT industry here is not the same as the fashion or the sports cars industry. In the IT world we do not create anything, but just follow trends. In addition, italian IT industry is neither for a global market nor for common use: the most software products in Italy are built for other italian companies (above all, banks and insurance companies).

2. In Italy, Internet is not recognized to be a real media, a real source of information. Internet is a matter of hackers and music downloaders. Not for the news. So until the Television will talk about Web 2.0, it will not exist.

So, I think there is a good opportunity for a lot of smart and small software companies. Start producing Web 2.0 italian products before the mass of the big business will hurl into this market (and they will). Fortunately, in Italy it is not too late.

By the way, segnalo.com seems to be an italian Web 2.0 product. Is it the first ?

Jirino is online

2006-02-21T15:40:00.000+01:00

In the 9-developers team I'm working with, we always need a very fast but simple method to collect and manage bugs.

Fortunately, external bugs are managed with the very complete tool Atlassian Jira. Still, we can't use this tool for internal bug reporting: I think it is better because Jira it is very big, full of options, filters, users and it is too slow to use it inside a development team.

So, what to use ? In our first iterations, we accessed directly to an MDB file: this system was abandoned very quickly because not all the developers and testers were skilled with MDBs, and the results were ugly. Developers needed to manage bugs, not to learn how to hack Access databases ..

For a while, we used a simple Excel sheet. It was very simple to add new bug descriptions but we immediately had a lot of problems due to file sharing and management of the bug status (developers sometimes are very lazy and if you don't go and hit their heads, they often don't mind to update the bug status, so it is very common to read for bugs without assignee, or without status, and so on).

After some weeks, one of our developers has reported that on a previous project the BugZ application was used with success. BugZ is a very very simple bug tracking tool developed in ASP by Maxime Labelle. It is very nice, so we immediately transferred our database and started using it.

We are still very happy with BugZ: while it is designed for very small teams, it manages well our need to keep 12 people very up to date on the project status.

Alas, when you strongly use a tool, you may find its limitations. In the BugZ case, we had some small bugs we were not able to workaround. For instance, we were not able to make it work with our mail server, so we lost the very useful feature of mail notification. Also, we had some problems with the date format used into BugZ that swapped the day and the month digits producing some strange results in the ordering of bugs. Again, we had the need to make it run on the 'official' web server, that runs Apache, and ASP does not run with Apache ...

Well, not real bugs or serious flaws, but annoying issues that do not allow your team to work as they want.

So, after a very short code review I realized that could be very simple to rewrite BugZ in ASP.Net ... and the result is Jirino.

Jirino is basically the same application of BugZ, but written in C#. I don't think it is particularly better than BugZ (well, it is probably less complete), but it does what my team expects to do, and that's sufficient.

Jirino is also growing up. New features were added, and now it has it's own homepage at http://www.minddriven.it/jirino/

If you are interested in a simple bugtracking tool in written in C#, Jirino is worth a try.

Dont Click it !

2006-02-01T10:18:00.000+01:00

Who is interested (like me) in user interface design, can find on (dontclick.it) link a number of *really* interesting GUI ideas.
The theme of this site is to completely cut out the 'click' action (well, at least ONE click has to be done to enter the site, but the authors ensure it is the last one). The result is great and the entire site exploration is really worthwhile. So give it a try.

By the way, as a developer on the PocketPC platform I like to remember that, on PDAs user interfaces, the click (or the pick) is a normal operation and it is really the most intuitive action you can do on a PDA device (with stylus or fingers). On the contrary, other actions very easy to do on a PC (drag&drop, right click, etc.) are more difficult to simulate on a PDA.

So, if someone wants to make a user interface study similar to dontclick.it, here is the idea: study the PDA interface, and call your site onlypick.it ...

Which Language ?

2006-01-11T17:50:00.000+01:00

In the last post I was wondering what language should be used in this blog. The answer is: both English and Italian.

I will use italian when the audience is not -probably- huge, also when I use some slang or -example- when the humor is tipically italian and can't be understood in other countries.

All other posts will be written in english. Well, MY english .. not Oxford.

Some popular posts could also be written in both languages, too. The post title will always be written in the post language, so it will be easy to distinguish between IT posts and EN posts.

Hope you will enjoy ..