So you bought this brand new HTC Hero and you tell everyone it’s on a par with the iPhone 3GS and its great browser? I mean, both are very recent Webkit-implementations aren’t they? Safari Mobile on iPhone OS3 is based on AppleWebKit/528.18, Chrome Mobile (or don’t they call it that?) for Android 1.5 on AppleWebKit/528.5+, and between 528.5+ and 528.18 there can only be minor differences? So HTML5-goodies (such as geolocation, localstorage and app cache) which Google is actively promoting, will work out of the box, just like on that dreaded iPhone 3GS, won’t they?
Sorry to bust your bubble, but Google seems to have decided otherwise; there’s no navigator.geolocation, no localstorage and no app cache on my HTC Hero (which is running Android 1.5 aka cupcake). You can access similar functionalities by calling the built-in Gears plugin, but mobile web-developers can’t assume that these HTML5-draft-specs are available on all modern high-end mobile handsets at all. Hell, even “big” Chrome 3.0.195.10 (which is based on Webkit 532!) does not seem to support these killer-features. Must be that Google is secretly pushing for Gears to become the default “rich internet enabler” instead of HTML5?
Month: August 2009
HTC-dude, where’s my keyboard?
In spite of my almost religious view on the ideal mobile phone, I bought a HTC Hero last week.
But let there be no doubt; it’s a great handset! My Hero sports a beautiful touch-screen, a nice -albeit young- Linux-based OS and a top notch webkit-based browser (with Adobe Flash 10, a first for a mobile device). The price is considerably lower then that of an iPhone and the platform is very open (esp. if you compare it to the golden cage Apple created for its ecosystem). I’ve installed several free apps from the Android Market and downloaded and installed a great AR-application from outside the Market without having to jailbreak anything (more on Android-apps in a later post).
But there’s one thing I really miss on my fancy device; a physical keyboard. Because as ancient as my Nokia e61i might have been, I really was more productive (as in “writing mails”) on it thanks to the (small) physical QWERTY-keyboard it sported. And while friends and colleagues assure me that I’ll get used to the virtual keyboard, and I’m sure things will indeed get better, we should not kid ourselves; nothing beats a real keyboard. Ever! So let the quest for a small compatible bluetooth keyboard begin!
HTTP upload MIME type hell
If anyone, like some colleagues of mine, runs into problems when uploading a CSV-file (or any other filetype for that matter) and checking mime-type (content-type), the following info might help.
Deciding the correct mime-type when doing an HTTP Upload is the browser’s responsibility, which produces … well, very mixed results. I made a little test-script and experimented with the same csv-file (with both .txt and .csv as suffix). See the results for yourself:
1. ubuntu 9.04, firefox 3.5.2:
.txt = text/plain
.csv = text/csv
2. work win xp sp2, windows firefox 3.5.2:
.txt = text/plain
.csv = application/vnd.ms-excel
3. work win xp sp2, windows ie6
.txt = text/plain
.csv = application/vnd.ms-excel
4. work win xp sp2, google chrome 3.0
.txt = text/plain
.csv = application/vnd.ms-excel
5. home win xp sp3, internet explorer 8
.txt = text/plain
.csv = application/octet-stream
6. home win xp sp3, firefox 3.5.2
.txt = text/plain
.csv = application/octet-stream
.xls = application/vnd.ms-excel
7. mac os x 10.5, Firefox 3.5.2:
.csv = text/plain
.txt = text/plain
8. mac os x 10.5, Safari 4.0.3:
.csv = application/octet-stream
.txt = text/plain
So as you can’t depend on browsers being strict while sending, you’ll have to be lenient while receiving, won’t you?
As found on the web (August 21st)
 |  |
 | Published WebTech news (august 2009)
|
 | |
| |
| |
| Shared Unix uniqueness |
| |
| |
 | Frank zoekt iemand die snel een testje kan doen op een mac. |
| |
| Frank mist een toetsenbord op z’n HTC Hero. |
WebTech news (august 2009)
I’ve started sending out a (maybe bi-)monthly “web technology news” mail at work, I’ll repost these here (excluding items/ remarks that are specific to my employee off course) hoping someone actually finds this useful.
1. The web wants to kill Internet Explorer 6, MS won’t
- msie6 (° 2001) is causing web developers headaches (it requires a lot of specific, non-standard code). some high-profile sites (such as youtube and digg) have announced they’ll stop supporting msie6 (cfr. http://www.wait-till-i.com/2009/07/14/did-digg-and-youtube-just-spell-the-end-of-internet-explorer-6/). web developers are mounting actions to urge users and IT-departments to upgrade (cfr. http://hey-it.com/, http://www.ie6nomore.com/, http://ie6update.com/ and others).
- microsoft confirmed however that they will continue to support msie6, according to the bbc even until 2014 (cfr. http://blogs.msdn.com/ie/archive/2009/08/10/engineering-pov.aspx and http://news.bbc.co.uk/2/hi/technology/8196242.stm)
2. HTML5: the web is growing up (but not fast enough, see previous paragraph)
- The same Microsoft (finally) joined discussions on the specifications for HTML5 (a new version of the language used to create webpages). It might be some time before Internet Explorer has the same level of support for the (draft) spec as Safari, Chrome, Opera and Firefox, but hey, talking is an important first step, no? (cfr. http://feedproxy.google.com/~r/webware/~3/iXVRC6Myrus/8301-1001_3-10305822-92.html)
- Google bought On2, a company specialized in video encoding formats and owner of a number of patents of the VP6-codec which is used in Flash Video (flv). Analists suspect Google will open source the codec so it can be used in html5, which supports video without Flash (cfr. http://news.bbc.co.uk/go/rss/-/2/hi/technology/8185910.stm). This could solve (or worsen) discussions in html5-land about what codecs are to be supported (cfr. http://arstechnica.com/open-source/news/2009/07/decoding-the-html-5-video-codec-debate.ars).
- 2 great articles about why html5 is important (and esp. for the mobile web) can be found here: http://feedproxy.google.com/~r/cloudfour/~3/-PtLhQmkuhw/ and http://www.smashingmagazine.com/2009/07/16/html5-and-the-future-of-the-web/
3. Misc
- According to research by an American company, interest in iPhone application-development is still very strong. More surprising might be the take-up of development of mobile applications for Google Android-base devices (cfr. http://gigaom.com/2009/08/10/the-battle-of-the-apps-google-vs-apple/). Looks more and more like Android is a serious iPhone-challenger (lots of Android-based phones are expected to arrive in Q4).
- Jakob Nielsen, the usability-pope, wrote a great article about usability of the web on mobiles here: http://www.useit.com/alertbox/mobile-usability.html, a must-read!
- Google published the specs and a reference implementation of “pubsubhubbub”, a light-weight semi-realtime communication mechanism for web-applications which is considered as one of the buildingblocks of what will become the “realtime web” (cfr. http://feeds.dashes.com/~r/AnilDash/~3/8TkbODhCEu0/the-pushbutton-web-realtime-becomes-real.html and http://code.google.com/p/pubsubhubbub/)
My blog laughs in your Facebook
Gisteren bij Peter Decroubele lekker ouderwets gereageerd op zijn tekst over hoe blogs aan populariteit lijken in te boeten ten voordele van Facebook en Twitter. En Peter linkt daarbij ook lekker ouderwets door een blogpost van Bruno Peeters over hetzelfde onderwerp. Al dat bloggrn, linken en reageren ondergraaft mijn hieronder hernomen (en lichtjes geredigeerde) reactie misschien enigszins, maar uitzondering en regel en diens meer zeker?
Met de opkomst van Facebook en Twitter is het belang van blogging als sociale netwerktool sterk verminderd. Statusberichtjes tussen de soep en de patatten laten zich nu eenmaal makkelijker schrijven dan regelmatige, min of meer vlot leesbare blogposts.
Ook het aantal reacties (en trackbacks en linken) lijkt overigens af te nemen, ten voordele van eenvoudiger (short-)URL’s, retweets, twitter-replies, facebook-comments en andere “vind ik leuk”-s. Blogs volgen, erop reageren en andere comments tracken is door het decentrale karakter van weblogs en door de beperkingen van feedreaders immers veel minder makkelijk. Ik krijg op Facebook dan ook gemiddeld meer respons op mijn daar automatisch geïmporteerde schrijfsels dan op m’n blog zelf (alhoewel dat ook van het onderwerp afhangt).
Dat alles betekent volgens mij overigens helemaal niet dat bloggen zal verdwijnen. maar ik denk dat het wel (terug?) meer maxi-dagboek en mini-journalistiek zal worden, zonder de “social” hype en zonder het incrowd-sfeertje (dat op Twitter een nieuwe thuis heeft gevonden). En al bij al is dat misschien toch niet zon slechte evolutie?
As found on the web (August 14th)
| | Liked 2 videos.   |
| | |
| | |
| | Liked 3 videos.    |
| | |
| | Shared Superlative |
| |  |
| | |
| | Shared Engineering POV: IE6 |
| | Shared Microsoft: IE6 Cannot Die |
| | Shared Litrik De Roy: HTC Hero |
| | |
| | |
| | |
| | Published Dochterken loves bling
|
| | |
| | Shared Become a Test Pilot! |
| | Liked 2 videos.   |
| |
|
| | |
| | Published My blog laughs in your Facebook
|
| | Liked New Slang  |
Warning: your computer might be infected!
Warning: your computer might be infected while surfing the web! Because these days web criminals aren’t just attacking government or corporate servers any more, but your browser as well. That way they can steal/ manipulate your data or install other malware. Most of these attacks happen while you are surfing and all the anti-virus software and spyware-scanners in the world will not fully protect your system and data (as Clopin found out while cleaning multiple PC’s of his family). The good news? There are only two big threats: malicious javascript and your own utter stupidity. And both can easily be countered.
Although stupidity is normal, you should try to live by one simple rule: don’t download software. Just don’t. And if you really must, only install if it comes recommended by at least one trusted source such as a computer-literate friend or a high-profile tech website. But don’t download software because a flashing red text on some obscure website tells you your computer is “infected”. Don’t “install a new version of Flash Player” to see that free porn movie. Don’t download, don’t install. Don’t!
Protecting against evil javascript (and malicious html, css, flash, java, …) is another issue. Attacks such as XSS, XSRF or clickjacking are barely visible. You’ll be happily surfing, clicking hyperlinks left and right, logging into your favorite web-apps and before you know it your data has been stolen or tampered, a password was reset or someone gained access to one of your online accounts. Entirely disabling javascript is not an option (you need it for most of modern web-applications to run), but if you’re a security-conscious Firefox-user there’s an easy solution; Noscript.
Noscript is a Firefox-addon that simply blocks all code (Javascript, Flash, Java and Silverlight) from being executed, protecting you by default against almost all types of browser-based attack (“almost”, as I’ve seen a nice proof-of-concept of a history-stealing web-page that only uses CSS-trickery, which Noscript can’t block). If you’re on a site you trust, you simply tell Noscript to temporary or permanently allow javascript -and other code- for that site and you’re back in web2.0-land. It may take some getting used to (a.o. to build a good permanent whitelist to allow your default sites to function), but it’s a great tool that can even double as a flashblock and (to some extend) adblock plus replacement!
So, to summarize; don’t install software and install Noscript and all will (probably) be well.
As found on the web (August 7th)
| | Liked 2 videos.   |
| | |
| | |
| | |
| | |
| | |
| | |
| | Published Voila le tube d’été!
|
| | |
| |