Mostrando entradas con la etiqueta google. Mostrar todas las entradas
Mostrando entradas con la etiqueta google. Mostrar todas las entradas

2010/05/18

OpenID

I'm a happy new user of http://myopenid.com.
  • Registration was very simple
  • The OpenID urls you get are nice (yourchoice.myopenid.com)
  • It supports multiple personas, i.e. during the signon process, you can choose differents sets of personal information to share with the requesting site
  • You get a detailed account activity report
Out of laziness, I've been resisting for years to get a proper OpenID account. Sometimes I've been using the one provided with blogger.com; others, my Flickr photostream url. And then sometimes I've signed up to places using my google or twitter accounts. In fact, a site requesting access to my list of contacts during registration with my google account has been the trigger to take care of this.

Two html lines have allowed me to use a vanity url (http://els.verg.es/xavier) and to avoid getting married to myopenid.com

2009/11/03

Analyze your traffic without you

Even when you have like a dozen visits a day, checking google analytics now and then is nice (hey! a vsitor from Belize!). However, it is nice to be able to tell what traffic comes from yourself and what comes from your two readers (if any). A bookmarklet can help you to enable a filter that will allow analytics to remove you from its stats, or even to track your own usage of your site.

Update: bookmarklet magic moved from the defunct googlepages to amazon s3 to http://code.verg.es/sharebookmarklet.html

2008/08/05

Gmail will automatically invite some contacts

Gmail will automatically invite some contacts

If there are other Gmail users whom you frequently email, you'll be able to chat and see each other online without having to send an invitation.
This wasn't the default setting some time ago, now it is, and I think it shouldn't be. To make it worse, disabling this behaviour is not reliable. My privacy is not happy.

2008/05/19

Notes to self (voice to gmail)

I often use my phone as a voice recorder to keep track of things that I don't want to forget while on the go. It has proven to be more reliable than my unreadable hand written notes. As an additional bonus, my kids love the idea and they often want to have their say in the recording, so processing the voice notes is often more fun than just listening to my odd ideas.

1955 portable recorder
However, the phone's interface to listen to them is quite crappy, and checking if I have some unprocessed note has not become part of my habits. They really belong in my mail inbox, so I now have setup things to send them to my gmail account (after an unsuccessful attempt of using ShoZu, that's already gone from my phone).

Teaching my Sony Ericsson k610i to contact gmail has not been obvious, since the net seems to be full of uninformed advice about this. As correctly reported here, the problem is that the phone is missing a root certificate required by to verify smpt.gmail.com. To fix it,

  1. Download a zip with a bunch of root certificates from http://www.thawte.com/roots/ (no need to answer to the form that you get there)
  2. Extract the missing certificate,
    thawte-roots.zip\Thawte Primary Root CA\Thawte_Primary_Root_CA.cer
  3. Send it to the phone using bluetooth, so that the phone recognizes it and prompts you to install it.
I only use the native email interface to be able to quickly send files from the phone, since http://www.google.com/mobile/mail feels more convenient.

I also had to upgrade RealPlayer to be able to listen the AMR format used by the phone. And fix the wrong .3gp extension associated to mime type audio/amr in HKEY_CLASSES_ROOT\MIME\Database\Content Type\audio/AMR

2007/10/21

Gmail spam filter sick?

In the past days I've been getting quite a few spam messages in my inbox. Not smartly disguised spam, but old plain lottery winning notifications and anatomy enlargement tips. Please, spam filter, get well soon: I miss you.

2007/05/27

Bookmarking things within gmail

Restoring a Firefox session accidentally told me that, when you open messages/chats in a new window, you get an url that you can bookmark/copy to another file. Nice.

Exploring things a bit further showed that gmail uses frames, so you can use the url of the inner frame to bookmark contacts, labels, filters, searches or any other thing that does not have the "New Window" link. In Firefox, use the popup menu, and "This Frame > Open frame in new tab".

Very handy for those of us that run our lifes in tiddlywikis.

Of course, Google is perfectly free to change these misterious urls not likely intended for public consumption...

2007/04/27

Evangelizing and toilets

One of Google's little secrets that has helped us to inspire our developers to write well-tested code: we write flyers about (testing techniques) and then regularly plaster the bathrooms all over Google with each episode, almost 500 stalls worldwide.

We've received a lot of feedback about it. Some favorable ("This is great because I'm always forgetting to bring my copy of Linux Nerd 2000 to the bathroom!") and some not ("I'm trying to use the bathroom, can you folks please just LEAVE ME ALONE?")
Stumbled upon this evangelizing technique in Introducing "Testing on the Toilet", while googling trying to decide which JavaScript unit testing framework use (yeah, shame on me, I've been doing lots of JavaScript lately without doing Test Driven Development).

They not only use a surprising evangelizing technique, but also have a great banner and motto for their blog: "Debugging sucks. Testing Rocks."

2007/01/09

My browser, my rules, my chrome

Summary: there are times when Firefox security gets in your way. Having a directory where you can place xul, html and scripts to run with chrome privileges enables you to shoot yourself in the foot and to create nice and handy hacks.

The itch that called for scratching: I wanted to hack something to let me upload several files at once to googlepages [1]. The problem is that, for very good reasons, you can not use JavaScript to change the value of input fields of type file: you usualy don't want scripts reading files in your disk. So, a bookmarklet, that to the browser looks as if it were part of the page, wouldn't cut it. Greasemonkey userscripts do run sandboxed, but not under chrome privilege [2], so I did not try that. I tried to give the page UniversalFileRead privilege [3][4], but failed.

A chrome folder. From Under chrome's influence, slightly modified,


In the browser's installation directory, create a new file chrome/mychrome.manifest and put a single line into it:
content mychrome file:///D:/whatever/mychrome/
Note that the trailing slash is significant. Then create the file D:\whatever\mychrome\hello.txt and add some text:
hello world!
After a complete restart of your browser, you can open that file using chrome://mychrome/content/hello.txt.
You need to keep in mind that chrome is catched, so you may not see the changes you make immediately. Setting nglayout.debug.disable_xul_cache to true [5] or the Extension Developer's Extension ability to reload chrome without restarting the browser are helpful here.

Future plans:
  • the bulk upload to googlepages is working nicely; needs some docu before uploading it.
  • I want to have a browser window with chrome privileges so that I can have more freedom of what I can do with some file: urls. Specifically
    • have known tiddlywikis granted UniversalXPConnect privilege without having to grant it to everything from file: and without having to mess with the capability.principal.codebase annoyance.
    • have local files use a script loaded from the web, or have a remote file load a script from my disk, so that I don't have to upload files while writing/debugging javascript.
Footnotes:
[1] googlepages: yeah, I know, I'm cheap!
[2] Mark Pilgrim's slides on sandboxing greasemonkey have a wonderful "how to become an expert" detour. Don't miss it.
And, just in case you are a powerpoint author, consider S5 (a Simple Standards-Based Slide Show System), the tool used for Mark's slides.
[3] UniversalFileRead: I added to my user.js
user_pref("capability.principal.codebase.googlepages1.granted", "UniversalFileRead");
user_pref("capability.principal.codebase.googlepages1.id", "http://pages.google.com/");
but still got a security exception.
[4] Highlighted link thanks to http://citebite.com.
[5] nglayout.debug.disable_xul_cache: I haven't tried it. As I write this,I realize that my user.js sets it to true while my prefs.js ignores it and keeps it to false. I'm clueless here. Update: A ";" was missing at the end of its user_pref line. Life is much better with this setting enabled!

2006/12/26

Google Tech Taks (engEDU videos)

You *have to* keep an eye on Google's TechTalks. I have not seen an obvious link, so here it is a search and a feed on that search (better, since it provides longer descriptions). They are mostly about computer science, but the range of topics is quite broad.

It is great that they are choosing to publish all these terrific talks they host.

Update: from http://video.google.com/googleplex.html

Google TechTalks are designed to disseminate a wide spectrum of views on topics ranging from Current Affairs, Science, Engineering, Humanities, Business, Law, Entertainment, Medicine, and the Arts.
I also learned about the Authors@Google series:
Authors@Google is a speaker series where thought-provoking, Zeitgeist-making, trend-setting authors come to the Googleplex to read from their works and share their thoughts with us. The following authors have agreed to release their talks to the world on Google Video.

2006/12/23

Irresponsabilidad corporativa

No escribo sobre plantar árboles, o no contaminar, o integrarse en las comunidades locales o... Mi tema es mucho más prosaico: no está bien que Google pida a los usuarios de Google Analytics que tecleen su identificador de usuario y contraseña en una página sin https. No basta con que las credenciales viajen seguras porque están utilizando un iframe con https para transmitirlas.

Nadie debería teclear su contraseña sin ver un candadito en su navegador.
Nadie, y, todavía menos Google, debería pedir a los usuarios que lo hagan.

2006/11/18

Unofficial PlanetIBM Search

Introducing... the Unofficial PlanetIBM Search, based on Google's Custom Search Engines and on the unofficial list of IBM bloggers maintained by Elias Torres. Search box in this blogs side bar.

I've set everything by hand now, but it does not look very hard to script it now that I've learned a bit about Google's CSEs. Lots of things pending: write the scripts to automate the maintenance, get a semi-decent logo, give more weight to the official unofficial IBM voice (http://www.ibm.com/developerworks/blogs/, executives...), make visible what sites are searched and with what weigth (as now, everything in Elias list, with the same weight), check if I can get Google to be nice and avoid inserting ads from our competition in the results page...

And, why? Because I think that it is a good thing to be able to search what IBMers have been writing outside the firewall. (And because the procastinator in me thought that this was far more interesting than completing two pending assignments that I have from two friends... Cris, Rodri, ahora me pongo, de verdad...).

It is often said that it is better to ask forgiveness than permission. (Wait! Don't let my kids know that I agree with this!). So, without being 100% sure about this being the right thing to do, here I go.

Update. Re: reasons, permission and forgiveness. The main reason for doing this is shaking the tree and hoping that this can help to get in place sooner an official search, using IBM technology, a list of bloggers maintained officially, rss available on searches, no adds, nice integration with ibm.com... Based on what I experience in the IBM intranet, putting in place a proper solution should not be hard at all.

2006/11/17

Mozbot: google con esto y sin lo otro

Un googlefight.com tonto me ha llevado a tropezar con mozbot.com, que añade y quita cosas a los resultados de una búsqueda de google. A veces es interesante excluir algún dominio de los resultados de la búsqueda, y mozbot lo hace muy simple. Otra cosa práctica es que, para cada resultado, puedes acceder a su historia en la máquina del tiempo (archive.org) con un par de clics.