Should We Test It a Bit More?

[Editor’s note: Do read the comments of this posting about]

Always the big question; how much should we test it? When is it ready?

When should we launch the product? Should we launch alpha or beta (private or public)? Is it ok to launch so called perpetual beta (take Gmail for example)?

Pretty good questions.

Lately I’ve read that one should always launch early (see Startups in 13 Sentences for one example).

Well, today I received an e-mail from Mtv3 inviting me to beta-test their new real estate service ( I can’t actually remember if I signed for invite – might have done that.

Anyway, used the link provided in the e-mail. Whee, HTTP-basic authentication set to protect the pages! Professional? Not so, IMHO. Typed in the general user name and password combo (was there an ad agency who came up with these?) – what follows?

Screen capture below:

MTV3's new real estate service, as seen with FF 3.0.6, newest Flash installed (and working), running on WinXP

MTV3's new real estate service, as seen with FF 3.0.6, newest Flash installed (and working), running on WinXP

Pretty nice, isn’t it?

Plain blue, small tick mark from Adblock showing that there’s some Flash present.

A quick peek at code;  almost valid HTML, awful conditional comments used for IE and yeah, that’s about it. Flash and Ajax, which don’t seem to work with my sturdy old computer.


Accessibility? Can’t even evaluate it as I don’t see the content. Oh wait, I can, can I?

Yes I can: accessibility is not that good.

Lessons learned so far:

  • In order to use the site I have to have Flash 10 (latest version) installed.
  • Beta testing is a fine way to catch some bugs – as we see here;)

Small test with IE7 – no surprise – it works. (Screen grab below)
Pretty nice, though the slides are sluggish and so on.

jokakoti viewed with IE7

jokakoti viewed with IE7

Lessons learned so far?

  • not all users have the latest plugins installed
  • not all users have up-to-date computers (I’d reckon this works like a charm with my laptop)

Last but not the least, a small test with my favourite mobile browser, Opera Mini, using their wonderful Simulator/Demo. (Screen grab below, screen states that one has to have Flash and JS support).

jokakoti viewed Operamini simulator

jokakoti viewed Operamini simulator


  • Developing cross-browser, accessible websites is not easy
  • One can’t rely on users having newest versions nor fast computers
  • Beta testing is worth one’s while.

Helsinki, UN and Certain Surveys.

United Nations (UN) just released their survey about municipal websites, world wide. Actual title was “Digital Governance in Municipalities Worldwide (2007) ~ A Longitudinal Assessment of Municipal Websites Throughout the World” (PDF-link!).

What was the survey about? (Bolding by yours truly, quotes from the summary part of the survey) In the survey they “evaluated the websites of municipalities in terms of digital governance and ranked them on a global scale”. Cities and nations were selected as follows: “The top 100 most wired nations were identified using data from the International Telecommunication Union (ITU)…”.  and “The largest city, by population in each of these 100 countries was then selected for the study and used as a  surrogate for all cities in the respective country…”

To summarize: top 100 most wired nations were under inspection, of which biggest cities’ websites were evaluated. Helsinki is the most crowded city in Finland, so Helsinki’s website ( was evaluated.

Five components were evaluated:

  1. Privacy/Security;
  2. Usability;
  3. Content;
  4. Services; and
  5. Citizen Participation

I’ll leave other aspects alone and concentrate on the second component, usability. On the survey they examined “three types of websites: traditional web pages, forms, and search tools”. Sounds like a good selection to me, although in my opinion those are just parts of one website. What was then, in fact, evaluated? Branding, structure, consistency (colours, navigational elements etc.), requirements stated (clearly), availability of sitemaps and alternative versions for documents, basic forms’ usability, search tools and so on. Quite good a selection (imho), again.

In the content-component they also evaluated the “Bobby compliance” (sic!) or access to web site via a TDD Phone Service. (Side notes: Bobby was originally “a free online tool provided by the Centre for Applied Special Technology (CAST) used to validate websites for WAI and Section 508 compliance” (Wikipedia). At 2005 it became “Watchfire WebXACT” -tool (still free to use online) and was finally integrated by IBM at 2008 and is nowadays part of IBM’s Rational Policy Tester Accessibility Edition and therefore not available for free use.)

In the usability section Helsinki was ranked fourth overall and in the content section sixth overall. Not too bad, eh? About Helsinki was said: “Helsinki increased in its overall score and its ranking
significantly from those in 2005 and 2003. Helsinki was ranked 35th with a score of 34.62 in 2005 and was not ranked in the top five cities in any of the categories.” Now Helsinki is among top ten in four out of five categories (privacy, citizen participation, usability and content – if I’m not wrong).

Hooray for Helsinki and hooray for Finland, as the survey quite clearly states that “(The largest city, by population … was … ) used as a  surrogate for all cities in the respective country“. Points for Finland, not for Helsinki, I’d say, although finnish media wanted to state so (for example Helsingin Sanomat says that “Helsinki’s Website is the third best in the world” /  (HS: Helsingin verkkosivut ovat kolmanneksi parhaat maailmassa).To claim that Helsinki’s site is the 3rd best in the world is at best misleading and at worst just plain wrong.

Anyway, pretty nice results.

Why am I not pleased?

Helsinki’s site is in my opinion not as good as all this hype could lead us to believe. Good, yes, they’ve done a lot of work, but still! The URI structure is awful (take as an example), layout is table-based, front page has a lot of HTML errors and so on. Quite minor things, have to give you that.

The thing I’m worried about at the moment is that the given UN Survey will now be used as a way to measure basic quality of web sites in Finland (“Survey says that is the third best in the world!”) and the not-so-nicely-implemented features will be forgotten (at and other sites, too). Quality of finnish sites has improved but is not that good when accessibility and usability are taken into account. Even technical quality can be questioned in many cases.

To sum things up: Helsinki (or actually Finland) scored pretty well on municipal websites survey. Overall quality is ok, but in my opinion not as superior as the media tends to say. we still have a lot to do!

Pretty URLs (or Aamulehti redesign)

My almost weekly series of (failed or bad) finnish redesigns continues. This time my local newspaper, Aamulehti, decided to jump in the social train, big time.

Social software and Web 2.0, yeah, great things. But – should be applied carefully!

First things first, I use Firefox as my browser. I also have the marvellous adblock -plugin installed and basically all the finnish ad servers blocked.

Anyway, pretty URLs¨(Uniform Resource Locators). What does it actually mean? URL is “in normal” use the address for certain web page. (I’ll use URI, Uniform Resource Identifier, in this posting as almost the same as URL). For example URL for this page is:…lehti-redesignpretty-urls-or-aamulehti-redesign/
, where http is the used protocol, is the domain name and stuff after it represents the structure of the site (used to be folder structure, nowadays is something completely different).

Anyway, pretty URLs? Human readable URLs?

The above example is pretty clear, domain is, then comes the year, month and the title of this posting. Pretty pretty? (Also nice for search engines!)

Sidenote: do read the “Cool URIs don’t change!

Ok, back to Aamulehti.

Why do I think the redesign sucks?

  • URIs are not that pretty. For example the front page’s address (cool URI is is:
  • Front page is full of ads (except that it isn’t, thanks to Adblock) or empty space (using Adblock)
  • Site is very, very heavy with all the AJAX goodness
  • Lue lisää (Read more) -links used heavily on every page – WTF? Basic thing: don’t use phrases such as “read more” in links. Is bad for screen readers, is bad for search engines is bad for (list continues)
  • News on the front page are cropped automatically, which results in awkward sentences and so on.
  • Redesign caused the pages to crash (hey yeah, lot of requests? Get used to it!) for hours.

Bah. I don’t seem to have the energy to rant anymore, I give up. (Could go on and on about the blogs and everything).

There are some good and nifty features too, at But as the Site is down, no report on them. Perhaps later.

More redesigns that suck:www. (the city of Helsinki) reports as follows:

15.06.08 pages were transferred on a new technical platform

The City of Helsinki web pages were transferred on a new technical platform during week 25. All the URL addresses of the pages changed.

Did I say something about cool URIs earlier?

Could someone please, please, give me a great redesign? A finnish one?


Pretty please?

(Small editorial comment: I’ll add some screen grabs if the site works for an hour or two in a row…)

Tampere United Redesign

Front Page of Tampere United\'s redesigned site.

Tampere United, my local football (soccer, natürlich!) team, just released a redesign of their site.

My first opinion: WTF?

Flash, Javascript, font size even smaller than it used to be, almost everything displayed as images (and the list goes on and on and on).

77 validation errors (according to W3C’s Validator). And yes, although valid HTML is not the key to happiness, it is certainly a measure of quality.Tampere United - front page without flash

News -section at front page done with flash (sic!). WTF?

Antti Pohja’s so-called Blog: no RSS, no proper archives, no commenting. Nice stories, though, no complaints to Ato!

Tested with Opera Mini Simulator (latest version):

  • with mobile browsing off (no mobile view):
    Looks quite nice. News don’t work. Menu doesn’t work. Sponsor -list doesn’t work. Images look terrible (at lo-res).
  • mobile browsing on (mobile view):
    Fucking awful (imho)! (Menu comes first, news don’t work at front page, ad’s are really, really big and so on).

Visually the redesign is quite ok. Huge player up front is a nice touch, visually speaking. Colours go well with the team colours (blue and green).

Bearing accessibility and usability in mind the redesign is a total failure:

  • Menu system is based on drop-down menus (CSS-based, I’ll give you that!), but still really hard to use for people with disabilities or for the elderly.
  • Layout is fixed (try to enlarge the text, see what happens! How about lower resolutions, likeTampere United Front page without images 800×600?)
  • Reading order (linearized) is not logical.
  • Alternative texts for images and other content are not that informative (like the picture with team’s goalie, Mikko Kaven: Icon 01 Kaven) or seem to be missing completely
  • No images? Almost no content at Front Page! (Screen capture to the right)


I don’t like this redesign. There’s nothing new for me and the old features are even worse than they were. But, as I already am a fan of TamU, perhaps it doesn’t matter.

Perhaps time will tell if I grow to like this one, too. Like I did with the older ones.

There seems to be some new features, though. One is videos from games attached to reports (example), with interviews. Too bad this isn’t, technically speaking, not that good. Flash video, yes. Why, oh why does it start automatically?
Overall quality of videos is good, have to give you that!

Saved: Saavutettavuutta arvioimassa: tapaus Avustajaportti.

(Original at Hypermediaa ja elämää. Copied 28.4. Topic in english: “evaluating accessibility: case Avustajaportti)

Pidin Tampereen teknillisen yliopiston opintojaksolla Saavutettavuus vierailuluennon 20.11 aiheesta saavutettavuuden arviointi.

Luento perustui Invalidiliiton Avustajaportti -palvelun osana diplomityötäni suoritettuun arviointiin.


  • Määritelmiä saavutettavuudesta
  • Arviointi yleisellä tasolla
  • Arviointi W3C:n mukaan
  • Tapaus Avustajaportti
    • yleisesittely
    • saavutettavuus
    • saavutettavuuden arviointi käytännössä

SlideShare | View | Upload your own

Saved: Saavutettavuustason todentaminen, automaattitarkistimet ja luotettavuus.

(Original posting at “hypermediaan ja elämää”. Copied here at 28.4.2008. Topic in english: “conforming to an accessibility level, automated testing and reliability”)

Tämä “artikkeli” käsittelee automaattitarkistinten luotettavuutta ja käyttökelpoisuutta allekirjoittaneen lopputyön pohjalta (2006, TTY/Hypermedialaboratorio, valikoituja lainauksia myöhemmin tekstissä). Lopputuloksena työssä päädyttiin käytännössä siihen, että automaattitarkistin on hyvä työväline, mutta ei kelpaa sellaisenaan saavutettavuustason määrittelemiseen tai saavutettavuuden kokonaisvaltaiseen arviointiin.

Allekirjoittaneella on ollut jo pitkään mielessä tehdä testisivu havainnollistamaan automaattitarkistnten keskeisiä ongelmia TTY:n opiskelijoille ja muulle yhteisölle.

TTY:n Hypermedialaboratorion verkkopalvelun käyttökelpoisuus ja arviointi -opintojakso tarjosi mainion tilaisuuden, joten olkaa hyvä:

Kommentit, kehitysehdotukset ja muut vastaavat otetaan mielellään vastaan kommentteihin tai sähköpostilla (

(Otteita TTY:n Hypermedialaboratorioon 2006 tehdystä diplomityöstä “Saavutettavuus mukautuvan verkkopalvelun suunnittelussa ja toteutuksessa”, ss. 34-35, Ilkka Kaikuvuo, saatavana pdf-muodossa)

(Saavutettavuuden) Koneellisessa tarkistamisessa valittu sivusto tai sen osoite annetaan useimmiten verkosta löytyvälle tarkistimelle, joka tarkistaa sivuston tiettyihin periaatteisiin pohjautuen. Tarkistimet voidaan jakaa pääsääntöisesti tiettyä tehtävää varten tehtyihin tarkistimiin (kuten W3C:n HTML- ja CSS –validaattorit) ja tiettyä listaa tai muuta sellaista pohjanaan käyttäviin laajemman tarkistuksen suorittaviin tarkistimiin, kuten WebXACT [Watchfire 2003] tai Cynthia Says [HiSoftware 2003]. [Thatcher, Waddell et al. 2002 s. 165, Treviranus, McCathieNevile et al. 2000]

Laajemman tarkistuksen suorittavat palvelut ottavat usein kantaa myös verkkopalvelun muuhun laadukkuuteen ja käytettävyyteen, tiettyyn tehtävään suunnitellut tarkistimet määritelmällisesti suppeampaan ja rajattuun alueeseen.

Saavutettavuuden työkalujen käytöllä eli ns. koneellisella tarkistamisella voidaan todentaa automaattisesti tarkastettavissa olevat kohdat tarkistuslistoista ja kiinnittää tarkistajan huomio sellaisiin kohtiin, joita itse työkalu ei pysty tarkistamaan. Automaattitarkistimet ovat hyvä välineitä esimerkiksi sen tarkistamisessa, onko kaikissa kuvissa pakollinen vaihtoehtoisteksti, mutta tekstin järkevyyttä ne eivät käytännössä osaa tarkistaa. [Abou-Zahra 2005a, Clark 2002, Thatcher, Waddell et al. 2002]

Merkittävä osa esimerkiksi WCAG -suosituksen tarkistuslistan kohdista on tulkinnanvaraisia, jolloin koneelliseen tarkistukseen tulee myös aina liittyä ihmisen tekemä arvio kohtien toteutumisesta. [Abou-Zahra 2005a, Thatcher, Waddell et al. 2002 s. 164] Saavutettavuudessa ei myöskään ole kyse pelkistä teknisistä seikoista, vaan todellisten käyttäjien kokemus tulee myös ottaa huomioon. Useimmiten verkosta löytyvä automaattitarkistin ei pysty ottamaan huomioon ihmisen subjektiivista kokemusta sivuston käyttötuntumasta tai siihen liittyvistä käytännön ongelmista. [Kelly, Sloan et al. 2005 s. 50]

Pelkän automaattitarkistimen käyttö voi myös johtaa virheellisiin havaintoihin sivuston saavutettavuudesta. Tarkistin voi esimerkiksi väittää sivuston olevan tiettyä W3C:n saavutettavuustasoa, vaikka ihmisen tarkastaessa sivustoa selviä virheitä saattaisi olla runsaastikin. Toisaalta tarkistin saattaa ilmoittaa virheistä saavutettavuuden kannalta toimivalla sivulla. Eräänä esimerkkinä voidaan mainita englantilaisen yrityksen SiteMorsen tekemät virheellisiin lopputuloksiin johtaneet analyysit saavutettavuustyössä vaikuttavien tahojen sivustoista (vertaa [Isolani 2005]).

Automaattitarkistimet ja saavutettavuustyökalut eivät yksin pysty määrittämään sivuston saavutettavuustasoa vaan määrittelyyn tarvitaan aina myös ihmisen harkintaa. [Abou-Zahra 2005a, Kelly, Sloan et al. 2005, Thatcher, Waddell et al. 2002 s. 314]

(Lähdeviittaukset voi halutessaan tarkistaa alkuperäisestä teoksesta.)