Salvatore Iovene

One lit­tle known new fea­ture of Maemo5 PR1.2 is the sup­port for cus­tom smi­lies in Con­ver­sa­tions (<bragging>which I implemented</bragging>).

I have made a smi­ley theme based on Tango icons, but it’s not yet in the repos­i­to­ries. Here’s what it looks like:

Tango smi­lies

You can install it by down­load­ing this file: conversations-tango-smilies_0.1_all.deb, and copy­ing it over to your N900.

Then open an X-Terminal, become root and do:

cd /home/user/MyDocs
dpkg -i conversations-tango-smilies_0.1_all.deb

Once it’s fin­ished, switch to Con­ver­sa­tion, go to Set­tings, and you will have a new but­ton, called Smi­leys theme. You can select your newly installed Tango theme.

Note: the theme will not work with SMSs and Skype chats. All the other accounts, and SMSs, will use the new theme.

After set­ting the new theme, go to an IM con­ver­sa­tion and see if the new icons are in the smi­ley picker. If not, a reboot should do.

Enjoy!

Before the last few years, I don’t recall any­one say­ing they were fans of some com­pany. I could imag­ine peo­ple say­ing “I like Ford cars” and “I pre­fer Ben&Jerry ice cream.”

But being fans? No, I’m pos­i­tive I’ve never heard or read any­thing like that.

Nowa­days I read and hear it on a daily basis. Peo­ple are fans of Google, Apple, Nokia. They lit­er­ally cheer for their favorite com­pany, watch­ing their direct (patent law­suits) and indi­rect (new prod­ucts com­pet­ing each other) bat­tles, and hop­ing that it wins in the end.

Face­book, being extremely unpop­u­lar these days despite me think­ing it’s no big­gie, is indeed risk­ing a damn lot just because of the cost of being uncool.

What does this tell? Is there a pre­cise con­nec­tion between the qual­ity of one’s prod­uct and its per­ceived cool­ness? Or is there a link between the mar­ket­ing and this cool­ness? Per­haps none, per­haps both. A com­pany goes build­ing an image, and it looks like each sin­gle tiny thing that hap­pens plays a part into gen­er­at­ing this pub­lic image.

Apple has seen a tremen­dous explo­sion of pop­u­lar­ity since the iPhone, and that’s helped a lot with build­ing an image. An image that’s really dif­fi­cult to erad­i­cate. Let’s be hon­est: the iPhone’s hard­ware is far from the com­peti­tors’ (see Nokia’s and HTC’s recent devices) and the soft­ware is barely catch­ing up (mul­ti­task­ing? It was about time.) Still, they’re going to con­tinue being cool, no mat­ter how many times they’ll mar­ket thinks like Cut & Paste like the best inven­tion since sliced bread. Why? You know it: because first impres­sions are hard to go.

So what has a com­pany to do, today, to be cool? Where is the line between meet­ing the expec­ta­tions for the sake of not dis­ap­point­ing the user base, and exceed­ing them with the risk of being ahead of the times?

I can iden­tify a few things to keep in mind.

Con­fi­dence. Whether we’re talk­ing about mar­ket­ing, or push­ing for a risky idea, hav­ing con­fi­dence mat­ters a lot. Deal­ing with large user bases comes with two big prob­lems: you can’t make every­one happy and you have to tell your user what they want while allow­ing them to feel that they’re get­ting what they thought they wanted. It is risky, but it’s the only way to lead. If your com­pany lacks the con­fi­dence to do that, it’s in trouble.

Hon­esty. Your users are smart. Maybe not all, but enough of them are smart enough to make noise. Don’t try to fool them. Per­haps you’ll sell a lit­tle more units in the short term, but your pub­lic image will deteriorate.

Integrity. Mak­ing mis­takes is OK, but it doesn’t work if nobody believes you. You need to have a clear set of val­ues, and stick to them.

All in all, it boils down to the DBAA prin­ci­ple, i.e. Don’t Be An Ass. Do that, and make sure you are able to give your users what they want and what they don’t know they want, and peo­ple will cheer for you.

This is really silly. And by silly, I mean even sil­lier than unlim­ited data plans capped at 5GB. Mr and Mrs Ver­i­zon have appar­ently decided that teth­er­ing your Android phone should cost you money.

Let me remind you what teth­er­ing means: your phone will cre­ate an ad-hoc WiFi net­work, and will share its 3G con­nec­tion with all the devices con­nected to that net­work. A typ­i­cal use-case might be pay­ing for a 3G con­nec­tion from your Android phone, and then con­nect­ing other phones, com­put­ers, tablets or what­ever to the Inter­net, shar­ing the same connection.

Pre­sum­ably, though, your con­nec­tion has a max­i­mum speed, and also a monthly data cap, so this means that whether you enjoy that con­nec­tion from one or from ten devices, the cost of oper­a­tion for Ver­i­zon will be exactly the same. Why are they charg­ing for this, then? Because they think they can, obvi­ously. To give you the impres­sion that they are indeed giv­ing you some­thing more, buy­ing teth­er­ing will actu­ally bump your unlim­ited 5GB con­nec­tion to some unlim­ited 10GB, which is dou­ble the amount of unlim­ited! Jokes apart, it looks to me as they’re sell­ing you some­thing you don’t want, masked as some­thing you want but should be free.

This is even more fool­ish than unlim­ited data plans capped at 5GB, because I see how “unlim­ited” is a mar­ket­ing word. Most peo­ple have no idea of how much 5GB is, and if that’ll be enough for them. But unlim­ited… that’s a dif­fer­ent deal, isn’t it? It sounds so much bet­ter. Again, most peo­ple will never encounter the cap, because the same peo­ple that don’t know what 5GB are, are unlikely to use them in a month, from their mobile device.

So you see, I kind of see the point of play­ing with words with “unlim­ited”, even though it’s not really fair. But charg­ing for teth­er­ing? I call that steal­ing; charg­ing the users more, when they’re really get­ting the same, and it’s cost­ing the oper­a­tor the same.

The Inter­net is now com­pletely sat­u­rated by peo­ple whin­ing about pri­vacy on Face­book, or lack thereof. I hon­estly fail to see, for instance, how its pri­vacy set­tings should be a bewil­der­ing tan­gle of options. Face­book has hun­dreds of mil­lions of users, so yes, it is going to be com­pli­cated to han­dle the pri­vacy set­tings. Besides the fact that it really isn’t. And so what if its pri­vacy state­ment has a larger word count than the Con­sti­tu­tion of the USA? Face­book has a hun­dred mil­lions users more than USA has cit­i­zens. One glove does not fit all hands, and they need their legalese.

I agree: upgrad­ing pri­vacy set­tings so that they are less secure than they used to is not the right way to go, but every­body is free to change them the best way they can. The more I observe the debate evolve, the more it looks like it’s founded on con­spir­acy the­o­rist delusions.

This isn’t sim­ply a mat­ter of “Face­book is the only game in town, so they can and will do what­ever they want”. Face­book comes for free, and it’s not really some­thing I would include amongst the must-have ser­vices, like fresh water. So you don’t need it, and don’t have to use it. If the sole fresh water sup­plier in your town started to pol­lute the water inten­tion­ally, say so that they could sell med­i­cine with the other com­pany they own, then you would have a problem.

Oth­er­wise, there’s noth­ing here. What? Rais­ing pub­lic aware­ness? I see. You can do that with­out try­ing to orga­nize Facebook-quitting days. It’s OK to talk about it, and indeed I’m doing it right now myself. But I’m just doing it in a more con­struc­tive way: check the safety of your pri­vacy set­tings by fol­low­ing the instruc­tions at http://www.reclaimprivacy.org/, and fix them to suit your needs.

Just don’t go blog­ging that Face­book is evil. You think you can just found a start-up and make a non-evil Face­book replace­ment? How long until you’re sink­ing in the bills from the host­ing com­pany, and start think­ing that “hey, maybe I should put ads on this thing”, and then, when the ads don’t pay enough, you start think­ing that “hey, per­haps I should tar­get the ads bet­ter: I need to share more infor­ma­tion about the users with the ad providers”. And so on.

Tweakr has been finally pro­moted to Maemo Extras, check it out at http://maemo.org/downloads/product/Maemo5/tweakr/.

You can now eas­ily install it from the App. man­ager. I’ll write up a new tuto­r­ial some­day soon.

Hi. I have released tweakr 0.0.16, which you should be able to install from Maemo Test­ing in a mat­ter of min­utes. A cou­ple of bugs fixed and the Silent pro­file hard­coded is what’s new. Update!

Hi, a quick update on Tweakr: it has now been pro­moted to extras-testing, since the reboot loop bug (thanks, hildon-home!) has been fixed.

Also a new fea­ture present: the Pro­file but­ton in the Sta­tus Menu gets replaced by Tweakr’s own but­ton (which looks iden­ti­cal), so that you won’t have get your Sta­tus Menu too crowded.

A new fea­ture in Tweakr allows you to extend the sound Pro­files to more than just Gen­eral and Silent.

Tweakr intro­duces the con­cept of Pro­file pre­set. Pre­sets are Pro­file set­tings which can be saved, deleted and assigned to the Gen­eral pro­file. This allows you to prac­ti­cally use as many pro­files as you want on the N900.

The nor­mal use-case is to first tune the Gen­eral pro­file for the new pre­set you want to save, then open Tweakr and choose the Save cur­rent Gen­eral pro­file to new pre­set but­ton. This will allow you to save the set­tings of the Gen­eral pro­file with a new name.

After that, you will find a new but­ton in your Sta­tus Menu, as shown in the fol­low­ing screenshot:

You can now select what­ever pre­set you have cre­ated before, and its set­tings will be applied to the Gen­eral profile.

If you’re not using Tweakr yet, go get it from its down­load page!

I have been work­ing on a lit­tle util­ity pack­age for Maemo 5, called Tweakr. It’s a Settings applet that lets you tweak lit­tle known set­tings that could oth­er­wise be changed only by edit­ing con­fig­u­ra­tion files by hand.

It has a plu­gin archi­tec­ture, so you could write your own plu­gin too. The ones I have so far are Desktop and Hardware keys.

Desk­top

This plu­gin allows you to edit the labels of the book­mark short­cuts you have on your desktop.

Hard­ware keys

You can con­fig­ure the behav­ior of the Power Key, and whether uncov­er­ing the cam­era lens unlocks the device.

More plu­g­ins and fea­ture com­ing soon!

Here’s some screenshots:

Dur­ing my cod­ing adven­tures, I have just found myself hav­ing to port an exist­ing Win32 appli­ca­tion to CMake. After writ­ing a mere 283 lines CMakeLists.txt file, and get­ting the appli­ca­tion suc­cess­fully com­pile, I fired it up to see if it worked, of course. I found it fail­ing when doing a WSAA­sync­S­e­lect, and fail­ing there didn’t seem to make any sense. So, to get to the point, I decided I’d just care­fully com­pare the compiler’s options gen­er­ated from CMake with the ones that were in the orig­i­nal Visual Stu­dio project file. After find­ing the dif­fer­ence, I decided I was gonna learn what those flags meant, so I googled for “visual c++ com­piler flags” and guess what? Noth­ing use­ful. The right string to google would be “visual c++ com­piler options”. But before get­ting to it, I had to unsuc­cess­fully go thru “visual stu­dio com­piler options”. You’d think that would cut it, right? Since “visual c++ com­piler options” did it. But, sur­pringly, “visual stu­dio com­piler flags” did actu­ally find what I wanted. Inter­est­ing combination.

I don’t really feel like blam­ing Google on this. Repro­duc­ing the same search pat­tern in the inter­nal Help func­tion of Microsoft Visual Stu­dio gave me the same success/no-success scheme. Besides, the whole point-and-click pro­ce­dure is a ter­ri­bly uncom­fort­able expe­ri­ence. Not only I have to get my hands off my key­board (inher­ently a waste of time), but also do I have to (mind: have to) wan­der my mouse around for a while to get to the information.

So, I won­dered, is there an alter­na­tive? I went to my rxvt ter­mi­nal emu­la­tor (which I run through Cyg­win, can’t really be both­ered with using the native Win­dows Ter­mi­nal), and tried to get some help from cl.exe. No luck. After some try-and-fail recur­sion, the least use­less thing I would find was cl.exe -help which would give me 106 (one hun­dred and six!) lines of doc­u­men­ta­tion. Wow, impres­sive, isn’t it? Com­par­ing, gcc’s man­ual page is 7820 lines, as for the lat­est stable.