Andriy Drozdyuk

USSR Animation

2011-12-08T13:48:00.001-08:00

Some links to some USSR cartoons:
1. Gil bil pes (about a dog - setting in Ukrainian village)
2. The Three from Prostokvashino
3. Hedgehog in the fog
4. Cinderella
5. Kolobok
6. Nu pogodi (similar to "tom and jerry")
7. Tale of tales (artistic)
8. Kids cartoon

How is convex polygon like our society?

2011-11-12T19:19:00.001-08:00

Convex polygons are really interesting shapes in mathematics. They are surprisingly simple to define but have many interesting mathematical properties. But they apply not only to mathematics, but also to our everyday lives.

What is "convex"?
Let us first define what a convex polygon is. We will look at 2D only, as it is simple to imagine. So, a shape is convex if for two points that we pick in it, a line segment between the points is also part of that shape. In the image below you can see some examples of convex and non-convex shapes. Note how for the non-convex ones we can always find two points that will have a line segment between them fall outside the given shape.

Convexity in our society

How does convexity show itself in our society? Simple. Any two members, that are accepted by society imply that any member "in-between" them is also accepted. For example, if society considers a person of height six feet two as tall, and a person of height six feet as tall - then any person with a height in-between is also considered as such - tall.
This is a rather silly example, as "tall" people are already defined as some range, say - six feet to seven feet. However, this is not as clear when it comes to views on different issues like politics or religion. Say that both a person A=(90% religious, 10% republican) and B=(10% religious, 90% republican) are accepted as members of society. Well, then by convexity it follows that any person in-between these beliefs is also accepted, for example person C=(50% religious, 50% republican).

It would seem absurd for society to accept "extremes" - but to reject the "in-betweens". Indeed, it is rather that society rejects the "extremes", and tends more towards the middle ground, making itself convex.

This can be extended to politics, where, for example, a government cannot consist of only 1 corrupt politician and 99 non-corrupt ones. By convexity, such a government must contain politicians that are all "shades" of corruptness - some completely corrupt, while others only "recreationally" so.
That is rather an amusing perspective!

Why Eclipse Sucks 3

2011-08-26T06:13:00.001-07:00

As we are moving away from the problems in operating systems, more and more emphasis is put on the functionality within the browser.

Most of our everyday person activities can be carried out inside the browser. As more and more applications move over to the web, in a sense there are less and less desktop applications that remain.

What ends up happening is that desktop applications get less and less attention. This results in poorer quality desktop applications - because the expertise in desktop development is decreased.

Less users are hammering on the code - to test out the kinks in the software. The error reports are not always submitted (as they are in web-applications). As more and features and "versions" are release - the desktop application tends to get clogged with bugs and failures.

Desktop applications have to evolve. And the perfect example of where the evolution has to head is web applications.

Some applications are leading the way. Take for example, google chrome. Their auto updating feature mimics exactly the way web-applications operate. By ensuring that most users are running the same version there is not need to support legacy software.

Take as another example - Steam client (client for a online games store). By using an integrate web-browser they rid themselves of the buggy and error prone GUIs that most cross-platform desktop applications come with. By evolving towards the web-application they are making it better for the user.

So why not?

Why not implement a "single version" auto-updating policy?

Why not implement an automatic-bug reporting system?

Why not borrow browser's rendering to render you GUIs?

Eclipse is one of many desktop applications in need of "evolution". In order to make it easier to use, easier to support and update. Will it manage to do this anytime soon? Or will an alternative arise and beat it to it? Or will the developers be always content with old "desktop" application. I think it will be a rather long wait we wait to see the result of these questions.

-Andriy Drozdyuk

Why Eclipse Sucks 2

2011-08-11T07:09:00.001-07:00

Again this error makes no sense, and what's more it leaves one
completely helpless as to what to do next.
How can all the developers in the world be using such a poor quality
software is beyond me!
How come it is fine for the "geek" to be of anti-windows culture
(Windows seldom gives me errors when updating itself nor does it ever
crash anymore) but it is completely fine to be using buggy and slow
tool for their coding needs?
Is there no end to poor quality software, that just "kind-of-works"
that we will accept? Why is there no standard among programmers of how
good something should be - before they themselves actually use it? Is
it because the developers are adventurous and don't mind "hacking"
things to work properly, or is it because that by using buggy software
they acquit themselves of writing "good" software too?
In other words, if I am using a "kinda-good-enough" IDE, it is
completely fine for me to write "kinda-good-enough" software.

Why Eclipse Sucks

2011-08-11T07:01:00.001-07:00

This should not be happening.
Violates some of the common sense criteria:
1) Does not indicate what the error is
2) Does not indicate how to fix it
3) Does not indicate where to go for help

Converting Between Time Zones by Hand

2010-12-02T20:56:00.000-08:00

How does one convert time between different time zones? Sounds like a simple question.
With the abundance of information and ready-made solutions it is sometimes hard to find the essence behind the solutions.
Try searching for "time zone conversion" in your favorite search engine, and all you get are ready made converters or programming language libraries that do the task for you.
But How do you actually do it? What is the process?

Luckily Wikipedia has an answer and it is a rather simple solution.

Solution
Let's say we want to convert from certain time2 to time1.

Clearly the following is true:
time1 - timeZone1Offset = time2 - timeZone2Offset (1)

since subtracting offset from the time will give you the time in UTC(or GMT), and we really want both of them to convert to the same standard time.
For example 8pm in GMT-1 is the same as 9pm in GMT-2:
20:00 - 1:00 = 19:00 = 21:00 - 2:00

In any case, you can solve (1) for either time1 or time2, e.g.:
time1 = time2 - timeZone2Offset + timeZone1Offset

which, if we rearrange some more comes out to be:
time1 = time2 + timeZone1Offset - timeZone2Offset

and if we set quantitiy:
zoneDiff = timeZone1Offset - timeZone2Offset

we get:
time1 = time2 + zoneDiff

Which is rather neat.

PS: Of course, you must check that each part (e.g. hour, minute, seconds etc...) of the final time are not negative. Because that means it is wrapping around, so you have to subtract it from the appropriate value. For example for negative hours, you subtract it from 24, for negative minutes you subtract it from 60 etc.

File dialog improvement

2010-11-13T13:42:00.000-08:00

Have you ever been confronted with a file changed dialog? Perhaps once, twice or maybe 10 times per day. If you have - you know how annoying it can get to refocus on a dialog instead of the task at hand.
If you use Eclipse as your programming IDE you will often see something like the following:

To me it looks like this:

What about making it more intuitive?

Good movie: Hereafter

2010-10-31T12:42:00.000-07:00

Don't dismiss it as another Matt Damon flick. It is not. In fact, if you like any of the Matt Damon movies - you probably will not like this one.
As anything directed by Clint Eastwood it is a gem. Expect slow paced, relaxing and thought provoking film. It is almost as if Eastwood is pondering the end of his own life. Eerie.

Eclipse website confusing as heck!

2010-10-06T13:26:00.000-07:00

How do you get from here:

To here:

In other words, how in the world can you get to the latest Stream Stable Release version 3.7M2a, from the Official Eclipse website? Horrible.

The only way I could find it, is by googling for "Eclipse m5" then going to the download link on some wierd website, and finding the eclipse download site for all releases: http://download.eclipse.org/eclipse/downloads/

XNA forums gets a D grade from Y-slow

2010-08-03T07:53:00.000-07:00

XNA is a great framework for writing games, and now recently, Windows Phone applications. Unfortunately with every good microsoft product, there is always a bad side. Just like with MSDN of the past, Creators XNA Forum looms over the XNA community.

The problems with the forum are numerous. Just navigating around is sluggish, and pressing the "back" button on the browser always causes a page reload. But I really wanted to find out what the problems with the forum are. So I installed Firebug + YSlow webpage profiling utility.

It looks like the whole thing is clearly written in ASP.NET. The dreaded __VIEWSTATE is there - storing the whole session information client side. This indicates that the site is not using the new ASP.NET MVC framework, but the web-forms version of ASP.NET.

Running YSlow on the general forum section results in grade D.

Some of the problems reported by YSlow are:

Make fewer HTTP requests

Grade: F
This page has 8 external Javascript scripts. Try combining them into one.
This page has 3 external stylesheets. Try combining them into one.
This page has 19 external background images. Try combining them with CSS sprites.

What this means is that a browser has to make 8+3+19=30 requests! What's more, is that there are 11 more content image requests, 1 content html and 1 favicon request - brining the total to 43 requests.

Add Expires headers

Grade: F
There are 38 static components without a far-future expiration date.

http://forums.xna.com/assets/css/main_Default_FF.css?...
http://forums.xna.com/.../xnacommunityserver_Default_FF.css?...
http://forums.xna.com/.../menu.css
http://forums.xna.com/assets/scripts/common.js
http://forums.xna.com/.../RadMenu.js
etc...

What this means is that every time you reload the page, or come back to it later, the browser will request the same resources it requested before. Things like background images and javascript files, that do not change - could be simply saved on the client side without the need to fetch the again and again.

Compress components with gzip

Grade: F

There are 9 plain text components that should be sent compressed

http://forums.xna.com/forums/32.aspx
http://forums.xna.com/assets/scripts/common.js
http://forums.xna.com/.../RadMenu.js
http://forums.xna.com/assets/scripts/shellHeader.js
http://forums.xna.com/WebResource.axd?...
http://forums.xna.com/WebResource.axd?...
http://forums.xna.com/WebResource.axd?...
http://forums.xna.com/WebResource.axd?...
http://forums.xna.com/Assets/Scripts/WebTrends.js

One can easily compress the data on the server side, and then decompress it on the client side. Even for small amounts of information it is often faster to compress and uncompress it, then to send it in plain text, because the internet transfer speed is much slower than the processor.

Minify JavaScript and CSS

Grade: F

There are 6 components that can be minified...

What minifying means is removing all unneeded spaces and comments from the code files. For example minifying the following css file used on the XNA forums:

http://forums.xna.com/assets/RadControls/Menu/Scripts/4_4_5/menu.css

we reduce it's size by 21%. While this is not much, when taken together over several files, it can add up to significant savings for the amount of data needed to transfer over the network. This reduces the wait time for the user and improves the web-site responsiveness.

Configure entity tags (ETags)

Grade: F

There are 38 components with misconfigured tags (ETags).

ETags are a way of saying that the resource did not change. For example, say that the browser visits the forum the first time and finds and downloads an image Cat.png. Then the user comes back to the forum the next week, and the browser has to download the Cat.png again, say because the user is looking at the same page.

Now if ETags are properly set on this image, the browser will simply send a request for "Cat.png" and recieve a reply code "304" which in HTTP terms means "Did not change". Thus the browser knows that the image is the same and it should use a locally saved copy it got a week before.

This mechanism complements the expires headers, to make for returning visits to the site much smoother, and makes the "back" button behave like users expect - snappy and quick.

A Grades

I only included the F grades in this post as they are the most interesting cases. This is not to say that the forum is a complete failure. The following features all got an A-grade:

Put JavaScript at bottom
Avoid CSS expressions
Reduce DNS lookups
Avoid URL redirects
Remove duplicate JavaScript and CSS
Make AJAX cacheable
Use GET for AJAX requests
Reduce the number of DOM elements
Avoid HTTP 404 (Not Found) error
Reduce cookie size.
Use cookie-free domains
Avoid AlphaImageLoader filter
Do not scale images in HTML
Make favicon small and cacheable

Conclusion

In the end it is not that important what "grade" tools like YSlow give a site. They are simply a means of looking at something that is already faulty. If the XNA forums site was not sluggish in the first place, I would not have thought of running YSlow on it (except maybe to see how fast it was!).

What is important is how users feel about the experience on the site. Unfortunately, it seems that most of the users who are Microsoft programmers are used to the sluggish experience of online resources and don't seem to mind it.

There is no excuse for making the user experience online so horrible. What is more sad is that the web division of Microsoft is not going to help them. Just look at their forums.

Gamers on Steam watching the world cup

2010-07-12T04:26:00.000-07:00

stackoverflow, stackexchange 2.0 oh my!

2010-05-01T22:07:00.000-07:00

The problem
It is quite amusing, how the experts-exchange killer stackoverflow.com is becoming something similar. Ever since StackOverflow started, people wanted to make their own sites for similar Q&A types of sites.

Various topics could benefit from the similarly formatted Q&A sites. Unfortunately the SO team has not really been very open to open-sourcing it's source code.

First they tried to make some bucks from selling it, thanks to the business mind of Joel Spolsky. However not all developers were eager to shell out $130/month for having their own little stackoverflow clone.

So then, SO team, figured that they will never make big money by just having 4 paid sites (as that's all they managed to sell I think). What to do, what to do?

This is where stackexchange 2.0 comes in. They together with a community (who are mostly developers) will "approve" the topics for the sites they think are worthy, and will host it themselves.

Advanges? Ad revenue. Disadvantages? You can't really start your own. Even if your idea is great it may get downvoted by the community. After all - they are a community of developers, not say mechanics.

The solution

Free Django based Open Source Q&A system or OSQA for short.

These guys are essentially re-building stack-overflow from the ground up!

Not only is this an exiting free alternative to the SO, it is also great to see it being built on top of Django, as opposed to starting from scratch, and encountering all the same problems that a framework like Django already solved.

But the project is still very young. So I encourage anyone out there, who is interested in Django, Python to check it out!

You can download the latest release, and contribute back to the community by posting questions, bugs or patches at Meta OSQA.

Google Chrome keeps crashing

2010-04-25T08:48:00.000-07:00

ironic

(adj) humorously sarcastic or mocking
(adj) characterized by Chrome browser, produced by Google, crashing on Google's own web email Gmail

Installing Python 2.5 in Fedora 12 for Google App Engine

2010-04-16T09:43:00.000-07:00

This stupid install process gets me every time! How do you run Google App Engine which currently requires Python2.5 on a machine that has Python2.6 installed? This also goes for anybody trying to use Django with App-Engine.

Everytime I try I get an error like this:
Traceback (most recent call last): File "/home/andriy/lib/google_appengine/dev_appserver.py", line 68, in run_file(__file__, globals()) File "/home/andriy/lib/google_appengine/dev_appserver.py", line 64, in run_file execfile(script_path, globals_) File "/home/andriy/lib/google_appengine/google/appengine/tools/dev_appserver_main.py", line 82, in from google.appengine.tools import appcfg File "/home/andriy/lib/google_appengine/google/appengine/tools/appcfg.py", line 59, in from google.appengine.tools import appengine_rpc File "/home/andriy/lib/google_appengine/google/appengine/tools/appengine_rpc.py", line 32, in https_handler = urllib2.HTTPSHandler AttributeError: 'module' object has no attribute 'HTTPSHandler'

Now some people say that you need to "compile python with ssl support". But it ain't so. There is actually nothing to configure in the python install to make it include ssl. Python finds the ssl libraries automatically if they are installed. The trick is you need the dev ssl libraries.

So here it is for my own reference:

Download the openssl-dev packages! Yes... you need them: sudo yum install openssl-dev*
Download python2.5
Configure it: ./configure
Build it: sudo make
Alt-install it to make it reside near your current python install: sudo make altinstall

Now download App-Engine and make sure to use python2.5 command with it.

Multiple monitors

2010-04-05T10:07:00.000-07:00

Having recently read the Three Monitors For Every User article at coding horror blog, I started writing a comment in reply but decided to turn it into a post.

Multiple monitors and me

Personally I get distracted by working with anything more than 1 screen.

I think anyone who ever tried working on a single 24" will not go back to 2x19", neither will you switch from single 30" to 3x20" screens.

There are a few big problems for me with a multi monitor setup:

Brightness - there is just too much light coming from the 2 or 3 monitors. Much more than I need and my eyes get tired, having nowhere to look for rest.
Interference - whenever I try to concentrate on a single screen, I am always conscious out of the corner of my eye of another screen on the side. It is always there and is very distracting (especially if it has content with bright/white colors). If there was a way to "turn-off" and turn-on the extra monitors at my whim - then perhaps it would be worth it.
Separation - there is no clean separation of "desktops" on multiple monitors. I cannot drag something to a different monitor and forget about it. My mouse always moves off my current screen and sometimes I am hunting for it by moving it around like I am playing quake 2.

In the end

I think multiple monitors are more of a "hip" thing to have, that many people think they need. I am better of having a single large screen with multiple software desktops like workspaces in Gnome (to bad windows 7 has no built-in virtual desktops).

Unfortunately the prices for large screens are still exponential in relation to their sizes. I am quite happy with my 24" but I would love to get my hands on a decent 30" - which unfortunately runs more than $1000 these days.

Canada Revenue Agency Fail

2010-03-28T16:00:00.001-07:00

Recently I tried to change my address through Canada Revenue Agency's website. Silly rabbit! Navigating to http://www.cra-arc.gc.ca/esrvc-srvce/tx/ndvdls/myccnt/menu-eng.html My mistake of course thinking that Chrome might be supported! Indeed. How about second most used browser on the planet - Firefox? Why of course not:I don’t know what kind of incompetent web-developers CRA hired, but even a 14-year kid could create a simple html website that allows users to enter some information WITHOUT java required.
I suspect that they might be relatives of the boss in the company.

Gmail chat

2010-03-15T13:09:00.000-07:00

My status message has a curious effect on me...

Canada and the saddest thing about Tax Software

2010-03-14T17:19:00.000-07:00

Every year I and everyone I know, about this time of year, start clamoring about preparing their tax returns. It is a painful process for many, mainly because it is both mathematically challenging and lengthy. Neither of which the average lazy joe is very interested in. What to do what do to? Well, most of the people much rather plunk down around $50 and let someone else do it for them.

However, with the advent of the tax-software, some are actually getting ready to do the taxes on their own. There are always concerns when using "software" to do something critical, like your personal finances that government will make sure to scrutinize you about. Things like:

Bugs in software
Sometimes bugs make the software frustrating to use! However this is getting better with more competition on the horizon.
Crashes
Paper taxes never "crash", unless you are not using a pencil and eraser.
Inaccurate calculations
What happens when the programmer who wrote the software, misplaces a decimal point?
Missing features
What happens when some tiny little piece of required functionality is missing? You may be tempted to overlook it and submit your tax as is - only to be caught "cheating" later on.

Also, if you are like me, prefer the hassle-less Online versions of tax software you are also exposing yourself to one more bonus problem:

Privacy!
Some company who has hired a bunch of out-of-college programmers to add up numbers is going to have ALL your financial history on file.

Well, well - that is quite a bad decision to make on the part of regular joe if he doesn't want any headaches! Maybe using paper forms is the best solution after all.

The Solution

Wait a minute! Did you say paper forms?

One thing always bothered me about those things. I mean have you ever though about:

How in the world does government process your paper tax forms?

Well, I would imagine the have some people. doing something like this:

Look at your scribbly handwriting
Record each line that you wrote - onto the computer
Recalculate ALL of your calculations

Do you see any redundancy there? I mean why would they take the things you already "wrote" and "re-write" them?

For me the most obvious elephant in the room solution is for the government to simply

Write their own tax software!

What would be the advantage of this approach? Well...

Accuracy
Who else is going to tell you what you missed to fill out - but the government?
Automatic calculations
Why should I use a silly calculator and make silly mistakes when computer is much more apt to do arithmetic!
Immediate fraud detection
If something doesn't fit - they can go after you immediately!
Helpful history
Part of the information can be automatically filled in for you.
Privacy
Well, just about as much privacy as you are getting by filing taxes by any other means. At least, the hope is, that an average hacker Joe won't gain access to it, or your information won't be sold to the highest bidder.

Why it's not happening

Well, well you ask. This is all great and such but why doesn't the goverment do it? Why don't they create a software package like this? Surely some things must be stopping them... For example:

It would be a huge and expensive project to undertake.
Really? Paying a team of developers for 1 year to develop it is more expensive than paying hundreds, if not thousands employees who scan the paper forms and enter the information into computers, for 10 years? I mean realistically this could have been done 10 years ago.
In the extreme case - just buy one of those existing companies that do it already!
It will be hard to use and not user friendly
Well, better make sure that it is user friendly. If it's not - I'm not voting for the party who made this project so bad.
JOBS - think of all the jobs that will be lost!!! Think of the little guy who does your taxes!
Well, sooner or later technology wins, and you have to adapt in some way. Also, I don't think that it's completely moral to make a killing of this type of business.

Conclusion

The personal tax software and businesses have grown into an industry over the last decade. However the government, the only reason for the taxation itself, has no interest in it whatsoever!

Things like online tax software, would be trivial to implement for the government. Online systems are definitely not new to governments. As was the case with parking tickets for example: it quickly realized that getting money from the population online is much more efficient than tying up phone lines or appeal courts.

When will the government implement a viable alternative to all the paid tax software out there? My guess is probably not any time soon. There seem to be some "hidden" factors holding this type of program back. What those are - we'll never know.

In the meantime, you are left with the usual choices: diy, hire a guy or do it online with a third party company.

Good luck!

Debugging production Django deployment

2010-03-06T21:41:00.000-08:00

Deploying Django is a process that can drive one bananas. There are a lot of things to setup to go from your development environment to the production. Aside from the regular hassles - there come special little buggers that can really make you mad. If you ever had problems with 500.html pages, url configurations or import errors - you know what I mean.

However it doesn't all have to be that bad. There are quite a few steps you can take early on to minimize the pain. If you do get into trouble - there are some things you can do to debug out of it.

Preventative Measures

Deploy early
If you want to avoid the hassle of debugging 100 things at once, deploy your project as soon as possible. By "as soon as possible" I mean right after you create it - when there are no models, views and funky url settings!
Commit often
You are of course using source control for your project, correct? By committing often you allow for greater granularity of your project. In a way you "isolate" bugs in a particular commit. This makes finding them later on particularly easy.
Deploy often
Same as #1 - the only way to make sure it works - is to put it out into the wild!
Separate production from development.
You should have something like this in your settings.py file:
if socket.gethostname() == 'server40.online.com':
DEBUG = False
else:
DEBUG = True
and feed all your other settings (i.e. database, media, etc) from this variable. This will make for painless deployment.
Use dynamic path names
Don't just hardcode "C:\Documents" or "/home/user/app" in your settings. Don't be lazy. This is how we do it:

import os
SITE_ROOT = os.path.dirname(os.path.realpath(__file__))

and then go crazy for all the other path veriables:
MEDIA_ROOT = os.path.join(SITE_ROOT, 'core/media')
TEMPLATE_DIRS = ( os.path.join(SITE_ROOT, 'core/templates') )
and so on...
Checkout source code into production
Don't just copy your code to production server with FTP or SCP! Install revision control like git or mercurial on the server, so you can hack away at the code without being afraid of messing things up, and more importantly - not loosing your important "tweaks" that you can only do in production.

Debugging the Deployed

Check your path and full-qualify the import names
Make sure you python path is correct. Sometimes the modules you can see in your development - are invisible in production.
If your project structure is something like this:
/project
---/app
------- views.py
and you are importing your views in your urls.py files like so:
from app import views
than you might get a wierd handler500 error that makes no sense at all.
Either reference your modules with full qualified names, i.e. "project.app.views" or put the path to your app directory into your path: "/home/user/blah/project/app".
Back up - test - roll forward
Sometimes you can't tell if the bug is from your latest update. Because production can keep some code in memory - one often fails to notice a bug until it is more than one update later.
Your first order of business should be to find the bug exactly. Figure out which "commit" added that bug into the system.
For this you must follow these steps:
a) Roll back to some point in the code where it worked
b) Restart your server, services etc... Make sure it works.
c) Roll forward to the newer revision
d) Restart. If everything still works - go to c).
e) If stuff is broken Roll back to some point before a).
Essentially you are doing a "binary" search on your commit history to isolate the bug. If you are using git - you can automate this process with git-bisect.
Once you find the offending commit - roll your code to the revision before it, and start by copying by hand the code from the offending revision.
Hopefully because you kept your commits small, it should not be too long before you find the bug!
90% it's urls and imports
Most likely you screwed up your imports or urls. Check them over.
The other 10% it's your 500.html and 404.html
Did you create your 500.html and 404.html pages in your templates directory? The production environment won't work without them.
User postgresql_psycopg2 and not postgresql
This is more of a tip than debugging technique, but one often confuses the two. The new psycopg is termed psycopg2 and so we must use postgresql_psycopg2 for our database engine setting.
Set DEBUG to True on production
Just below your logic which determines your DEBUG variable, overwrite it temporarily with true.
if blah..
DEBUG = True
else:
DEBUG = False
DEBUG = True # Overwrite temporarily

This is a real hack, but it can temporarily give you some great insight into the problem.

Conclusion

In the end if you try to keep the "production" on your mind from the start, your deployment should be not a problem. Because one does not deploy very often, one tends to forget all the mistakes that were made previously. This makes it really challenging to get it "right" every time. Some time is spent going over the error logs and figuring out the same problems. However we cannot solve this. It will always be like so.
The best thing you can do is prepare for this process ahead of time - and once there, have effective ways of dealing with it.

Websites that suck: Futureshop.ca

2010-02-28T13:37:00.000-08:00

Today I'll be starting a new series of posts called "websites-that-suck".
Now in order for a website to suck it has to simply be:

Non-functional
Confusing

The nominee of the day: Futureshop website!

1. Non-functional

Let's see. When you try and place an order online and check it's status you get:

Great! Now let's try to cancel and order:

Yes! At least I know that there is an error in English AND French. What would I do without the french translation? Probably the same thing as without it - stare at the stupid error and do nothing!

2. Confusing

Now we come to the second criteria: "confusability". When one opens Futureshop site we can see a clown trying to help you by encouraging to you "ask a question":

Every time I come to the site and I want to search for something, I always end up typing my search terms into the "Your question here" box. How many damn search boxes can there be???

Here Futureshop - let me fix it for you:

Conclusion

Futureshop website "sucks" because most of their effort is spent on "bling" and "blang" and not on content or ease of use. Granted - not everything is bad. The good features are: search that actually works if you figure out how to do it, helpful customer reviews (not sure how reliable or "unabiased" they are) and product specs right on the site (even though some specifications - especially for notebooks are completely wrong).

Now one can easily fix the layouts and most of the usability issues, however the funcionality with "ordering" products online is fundamentally flawed. If you are thinking of shopping online - don't. Not only is the ordering plagued by errors, but it also performs POST request at every step, making the use of the browser BACK button completely impossible.

In the end - it boggles one's mind how a "computer" shop's website can be so bad and suck so much.

Links

Futureshop made it to the CANADA'S 50 WORST WEBSITES

Comparing times in Ruby

2010-02-12T08:06:00.001-08:00

Did you know that Ruby stores time as the amount elapsed since 1969's New Year?
> irb
> Time.at(0)
Wed Dec 31 18:00:00 CST 1969

Recently I looked at an interesting problem in Ruby. It had to do with comparing two Time objects. Given two time objects, say time_1 and time_2 you want to see which one occurs at a later time.
The problem is that they can be of different dates, e.g.:

> time_1 # 6 PM on Dec 31
=> Wed Dec 31 18:00:00 CST 1969
> time_2 # 7 PM on Dec 30
=> Tue Dec 30 19:00:00 CST 1969

> (time_1 < time_2)
=> false

Solutions
There are many ways to go about this problem. One of which uses a Time.parse function to extract time from both time objects and equate the date. Something like this:

> norm_date = Date.today
> parsed1 = Time.parse("#{norm_date} #{time_1.strftime "%H:%M:%S"}")
> parsed2 = Time.parse("#{norm_date} #{time_2.strftime "%H:%M:%S"}")
> parsed1 < parsed2
=> true

However this solution resorts to string manipulation.
An alternative is to use a function Time.at which basically takes a number of seconds and returns a time object.
So we can do the following:

> conv1 = Time.at(time_1.hour * 60 * 60 + time_1.min * 60 + time_1.sec)
> conv2 = Time.at(time_2.hour * 60 * 60 + time_2.min * 60 + time_2.sec)
> (conv1 < conv 2)
=> true

It might seem like this does not make much difference, so I decided to run a little test.
Do each conversion 5000 times and see how long it takes.

Results

Here are the results:

> ruby time.rb
0.016277
2.704446

That is over 150 times difference in speed. Can you guess which method is which?

Conclusion
If comparing times in Ruby use Time.at function to normalize your times and than do a comparison. Strings are slow. In Ruby or any other language. If one can resort to integer arithmetic - it goes much faster and less error prone.
Also I learned that python does times in a more sensible manner than Ruby:

> python
>>> t1
datetime.time(18, 0)
>>> t2
datetime.time(19, 0)
>>> (t1 < t2)
True

P.S.
What is the equivalent of dir() python function in Ruby?

Source code

# Compare Time.at method to the Time.parse
# by converting two time objects and comparing them
require 'time'
require 'date'
t1 = Time.now + 12
t2 = t1 + 13

LOOPS = 5000
int_time = Time.now
(1..LOOPS).each do |i|
    # Standartize the times
    conv1 = Time.at(t1.hour * 60 * 60 + t1.min * 60 + t1.sec)
    conv2 = Time.at(t2.hour * 60 * 60 + t2.min * 60 + t2.sec)

    if conv1 < conv2 then
    end
end
int_end = Time.now
int_time = int_end - int_time
puts int_time

parsed_begin = Time.now
(1..LOOPS).each do |i|
    # Standardize the times
    parsed1 = Time.parse("#{Date.today} #{t1.strftime "%H:%M:%S"}")
    parsed2 = Time.parse("#{Date.today} #{t2.strftime "%H:%M:%S"}")

    if parsed1 < parsed2 then
    end
end
parsed_end = Time.now
parsed_time = parsed_end - parsed_begin
puts parsed_time

Python 2.6 - an orphan child

2010-02-11T06:28:00.000-08:00

The current dominating version of python is 2.5. Most of the frameworks and vendors have comfortably settled on 2.5. Some of the upcoming technologies will support Python 3 exclusively. That leaves our old friend Python 2.6 out in the cold.

The situation
Python 2.6 - it is really not worth the effort to support it. I mean if you start breaking your code and such, you might as well upgrade to Python 3.

The problem
Python 2.6 provides a lot of nice features over Python 2.5 ("except as" is my personal favorite). A lost of the libraries are standardized (e.g. json over simplejson) and improved in terms of performance.
However - it is not backward compatible to 2.5. The version difference of 0.1 makes it very easy to start using Python 2.6 for everything only to realize at the end that one has to regress to 2.5 for production.

The solution
Either stick with python 2.5 or stick with python 2.5.

Conclusion
It seems that in quest of progress, and by splitting their core language into three, Python community has went against one of it's core principles: "Simple is better than complex". After all if your application uses components all of which might require different versions of Python (e.g. django, apache, mysql) you suddenly find yourself in the position of the "System administrator" instead of a programmer. In this respect, even Java is easier!

What Firefox learned from windows: Restart!

2010-02-10T19:54:00.000-08:00

Designing Java If statement

2010-02-10T19:24:00.000-08:00

We want to have something that says "if condition is true then do some-thing, else do another-thing". How should we design it?"
Attempt #1:



if EXPRESSION STATEMENTS else STATEMENTS

Hm... this leaves us with the dangling else problem. How about:
Attempt #2:



if EXPRESSION { STATEMENTS } else { STATEMENTS }

Better. But I think if looks pretty lonely, let's add some parenthesis:
Attempt #3:



if (EXPRESSION) { STATEMENTS } else { STATEMENTS }

Perfect!

Web application browser compatibility

2010-02-10T19:04:00.000-08:00

Web application requirements

Firefox 2 or IE 7

Desktop application requirements

Windows Vista, SP2
.NET runtime 2.0
Visual C++ runtime
Ability to enjoy exotic user experiences