1. Easy video recording – let the user take videos with every webcam within seconds, then upload to YouTube or similar. A similar proposal (a simple video editor) is on the lifehacker.com five feature request list.
2. Stream capturing – saving streamed video data doesn’t work so easily with all those different streaming formats. For some, you need a RTSP stream catcher, then maybe a RTMP stream catcher and for some you seem to be able to use just mplayer. And then there are many cases where all fails. Technically, what can be played can also be saved. But then there is the big Flash Player obstacle – some Flash videos are well-protected. Gnash may help there.
3. PDF reader&editor – one tool that allows for reading PDFs, annotating them, publishing&sharing the comments, manipulating the PDF itself, adding additional layers, manipulating PDF metadata, etc. Just like the Adobe Acrobat Reader in its more expensive variant – but as open source tool with the ability to write plug-ins and integration into Gnome or KDE (or any) desktop. Okular is already on the right track! See also my article about editing PDF metadata.
4. Centralised instant messenger and (video)telephony – unite Pidgin, Skype and other Videochat and IM apps in one UI. Maybe put this together with microblogging tools, since people use their IM status messages like microblogging anyway. Skype has announced to open-source parts of their Linux client, so this is not totally out of reach. Open source alternatives to TweetDeck are also there, for example Gwibber. See also my article on microblogging and news.
5. Metadata in file browser – make the file explorer a metadata editor, paving the road for a semantic desktop. Even the Windows Explorer can do better than Nautilus for now! But then I haven’t tried KDE’s Dolphin for a while and this might be the right thing to do… See also my article on music metadata as well as my article on photo metadata.
6. Asset manager – even one step further, make the file browser a pluggable asset manager, that can take the shape of a photo collection manager, a scientific paper organiser or a website bookmark manager. So far I know only of commercial asset managers and haven’t yet investigated which one runs on Linux and might be useful for me. Do you have any recommendations?
7. Annotation everywhere – a note-taking application that can annotate every single file or item on the desktop. This way you can relate a specific email to a task, to a note, to a website, to an application and a specific file – thus documenting entire work-flow states for later continuation. Well, there is Tomboy for now. See also my article on note-taking.
8. Private browsing – create for Firefox or any other browser a CSS/Javascript security model that avoids CSS privacy hacks by not letting any information about how the HTML rendered leak into the Web. That would include creating an open source Flash plug-in that doesn’t publish all Font and SuperCookie information. See EFF’s PanoptiClick for more information about this. The SuperCookie issues can be softened with the BetterPrivacy Add-On for Firefox.
9. Easy emulation – integrate a Dalvik VM naturally into the desktop, so that it’s very easy to install&run Android apps from the applications menu. Maybe the project to implement a Dalvik VM in Java is the right way to do this. Of course, the same would be nice for Wine but I don’t consider this an option because Dalvik is open and Windows isn’t.
10. Synchronise data with external sources – I want to backup all configuration and some data files with a variety of places: external hard-disks and remote storage services in the Web (encryption is necessary here). Ubuntu One is already a big step forward but I really want to backup all configuration so I could crash my computer, buy a new one, hit the “reinstall the software that was there before” button and then everything is back to normal. This is (almost) technically possible. Another road is, that I want to backup the data stored elsewhere (Delicious bookmarks, Google Reader Shared news, Facebook comments, etc.) to my home computer so I’m not stuck with one provider forever (so I can quit Facebook some day). This seems to be impossible for now, but the problem lies in coding “adapters” that take data from one service and move it to the other one.

Am I the only one who wants these features? Are they that hard to implement? (Yes) Hey, for most of these features, I would pay some money (depending on how well it’s implemented). Oh well, and I admit that these features are not really Linux-related. It’s just that I use Ubuntu and would want to have solutions available on open platforms. I guess web-apps and Java- or .NET-based apps would be OK for me, too – but then look again at the wish-list and you’ll see that most features require desktop applications.

If you have suggestions for applications that solve one of those problems at least somehow a little bit, please leave a comment.

What is your favourite not-yet-there Desktop/Ubuntu/GNU/Linux feature?

The penguin image (Tux) is licensed from linux.org under a CC-BY-SA license

I’m not going to do this, but in this post I’ll present some of my favourites from the last months at MathOverflow (omitting the more subject-specific ones):

• Success stories: Kevin Buzzard asked for errata on Cassel-Fröhlich’s Algebraic Number Theory book and the London Maths Society (LMS) is going to reprint that book, including an erratum. See also the discussion on meta.MO if errata requests are appropriate for MO and the Errata Wiki, an attempt to provide a common place for these things.
• Success stories: Timothy Gowers asked if MO has led to mathematical breakthroughs, and the answers list some cases where it at least helped a lot.
• Tools: Ilya Nikokoshev asked for tools that help in organising research notes.
• Tools: Elisha Peterson asked for tools that help in organising papers (toread, tocite, etc.) and the answers were helpful for my discussion on how to manage papers.
• Tools: Anton Petrunun asked for tools that help in collaborative paper writing.
• ToRead: Ilya Nikokoshev asked for “a single paper everyone should read” and there are some nice suggestions, if you don’t have enough to read yet
• ToRead: David Hansen asked for papers that maximise the ratio importance:length and there are some suggestions which won’t take much time to read!
• Teaching: Gil Kalai asked for fundamental examples in different branches of mathematics. The thread seems to me to be a very useful source to look for motivation.
• Teaching: Michael Hoffman asked what the undergraduate curriculum is missing and there are various answers and interesting controversies.
• Teaching: Ryan Budney asked for success stories with curriculum reforms (at average research universities).
• Fun: Mathematics in the real world – Theo Johnson-Freyd explains how to calculate using a very long frictionless one-dimensional billiard table.
• Fun: Andrew Stacey asked how to respond to “I was never much good at maths at school” and there are lots of serious answers, like this one by Andrew Tuggle which I like most.
• Understanding: Reid Barton explains how spectral sequences generalise long exact sequences.
• Understanding: Andrew Critch explains a geometric picture of flat modules where “flat” really means “flat” in an intuitive sense.
• Understanding: Scott Carnahan asked for a motivation for (higher) algebraic K-Theory.
• Understanding: Javier gives a rough idea about the field with one element (F_un en français)
• Understanding: Yuhao Huang asked “Why is the decomposition theorem awesome?”, but there are not many helpful answers despite links to the de Cataldo and Migliorini article. Have a look at the “related questions” section in the sidebar. I put this question here only because of the decomposition theorem winter school in Freiburg, Germany this month (Feb 2010).

The software MO uses is very well suited for this kind of excerpt, since there are permalinks not only for questions but for answers, too.

So maybe someone who is already thinking about being a math blogger will adopt this idea and watch out for nice general-interest questions&answers on MO, to blog about them occasionally (I won’t).

a simplicial model category is just an enriched model category which is enriched over the monoidal model category of simplicial sets.

but details are also to be found below.

The simplicial model structure on simplicial sheaves on a topos

In Definition 1.2, for every small site , a model structure on is defined:

1. The weak equivalences are the stalkwise (pointwise) weak equivalences
2. The cofibrations are the monomorphisms
3. The fibrations are defined via the right lifting property with respect to acyclic cofibrations

Remark 1.3 is a technical subtlety. If you happen to have a conservative set of points of a topos , then weak equivalence of a morphism of sheaves on can be tested pointwise: , where denotes the weak equivalences in the standard model structure of simplicial sets. A conservative set of points is just a set of points that is a conservative family of functors, which is by definition, that the product functor is a conservative functor.
A functor is conservative if it reflects isomorphisms. That means, isomorphism implies isomorphism for each morphism .
This technical lemma is used later in the text, but the homotopy sheaves are not, so I guess you can forget the proof details when reading the text for the first time.

See also: conservative functor in nLab

Theorem 1.4 (the structure defined by is a model category structure) cites the result of Corollary 2.7 in Jardine: Simplicial Presheaves, in no. 47 J.Pure Applied Math, 1987 which is originally due to Joyal. Since the article is behind a paywall, I’ll give you a rough idea:

• (MC1), (MC2) and (MC3) are deduced from the model structure on simplicial sets.
• (MC4) relies on the fact that the morphism from a presheaf to its associated sheaf is a weak equivalence and then applying the axiom for with the global fibration and topological weak equivalence model structure. (MC4) for is proved with a trick that uses (MC5).
• (MC5) is essentially a small object argument.

The corresponding homotopy category of on is written .

See also: small object argument in nLab

Proper model categories

Remark 1.5 states that the model structure is a proper one. The proof is available in Jardine, J.F.: Stable homotopy theory of simplicial presheaves, in no. 39 Can. Math. J, 1987 which is available for free here.

A simplicial model category is proper if

• (P1) the pullback of a weak equivalence along a fibration is always a weak equivalence,
• (P2) the pushout of a weak equivalence along a cofibration is always a weak equivalence.

(P1) is proved for simplicial sets via fibrant replacement, such that one has a cartesian diagram up to weak equivalence, and then application of K. Brown’s coglueing lemma, which is Lemma 1 on page 428 of Brown, K.: Abstract Homotopy Theory and Generalized Sheaf Cohomology, in Vol. 186 Transactions of the American Mathematical Society, 1973 which you can download from the nLab for free.
(P2) is proved for simplicial sets in a dual fashion, using the fact that simplicial sets are always cofibrant and a dual of Brown’s coglueing lemma.

For simplicial presheaves on a topos, the proofs are similar. For (P1), fibrant replacement yields a cartesian diagram (up to weak equivalence) in which all objects are locally fibrant simplicial presheaves (which form a category of fibrant objects) and the coglueing argument can be applied. For simplicial sheaves, (P1) and (P2) follow since the associated sheaf morphism is a weak equivalence.

It should be mentioned that (P1) is also called right proper and similarly (P1) left proper.

See also: proper model category in nLab

Functorial fibrant replacements (1.6)

(MC5) demands in particular, that every morphism is functorially factorizable into a fibration after an acylic cofibration.
A resolution on a site (which carries a model structure) is defined to be a functor and a transformation such that for every simplicial sheaf , the object is fibrant and is an acyclic cofibration.
Indeed, if is a morphism, we can factorize it into an acyclic cofibration followed by a fibration. Rename the acyclic cofibration and the object , then is a fibration, thus fibrant. Voilà – since (MC5) demands this to be functorial, the functor/transformation conditions for a resolution are fulfilled.
It should be clear that this works the same way for cofibrant replacements, although we won’t need this here, since in the simplicial model structure we’re looking at on , all objects are cofibrant.

See also: Kan fibrant replacement in nLab

Simplicial model categories

For every two objects , we defined

If you already know about the “subtleties” in the definition of simplicial model categories (maybe from my article about simplicial model categories), skip the next two paragraphs.

A category is a simplicial model category if it is a model category that is enriched over simplicial sets, that satisfies the additional axioms (Quillen):

• (SM0): for all and all finite simplicial sets , and exist.
• (SM7): If is a cofibration and a fibration, then is a fibration of simplicial sets, which is trivial if either or is trivial. (The S denotes the simplicial mapping object of ).

(SM0) is also phrased “ is powered and copowered” and sometimes already included in the definition of an enriched model category (like I did in my article about simplicial model categories). (SM7) is also phrased “the copower functor is a left Quillen bifunctor” and sometimes already included in the definition of an enriched model category (like I did, again). So, if you take the “modern” definition of a model category enriched over a monoidal model category, those axioms are already included (I put them in here just because they will show up in the literature and also because you might not have read my article about the definition of simplicial model categories).

Lemma 1.10, different notions of equivalence are the same

For fibrant and a morphism, these three statements are equivalent:

1. is a simplicial homotopy equivalence,
2. is a weak equivalence,
3. is a weak equivalence.

The proof indication is mostly a list of references, so let’s have a more detailed look, which will then finish this posting.

• (2)=>(1)
  factorise the weak equivalence into a cofibration followed by an acyclic fibration . Then is a weak equivalence again (by 2-out-of-3). By an argument in Quillen’s Homotopical Algebra (Corollary 2.5), obtain a retraction of by the lift in the diagram
  
  and then get a simplicial homotopy from to by the lift in the diagram
  
  and now is a simplicial homotopy inverse of . To actually obtain a simplicial homotopy inverse of , we’re going to build a simplicial homotopy inverse of . For this, observe that all objects are cofibrant (since cofibrations are by definition just monomorphisms), and that the dual statement to what we just proved is that a trivial fibration between cofibrant objects is a simplicial homotopy equivalence.
  What is ? What is ? you might ask. The object is just the simplicial set , whose geometric realisation in looks like the interval , hence the name (and I used this notation here because it’s the same as in Quillen’s book). The object is the internal mapping object . If this remains unclear, you might want to read some introduction to enriched category theory.
• (1)=>(3)
  We will not try to construct a weak homotopy equivalence but a homotopy equivalence:
  Using the definition of for and , you’ll see the canonical isomorphism . Now take a simplicial homotopy inverse to the map and choose a simplicial homotopy between and . This yields a map which, composed with the canonical isomorphism above, is the homotopy between and we’re looking for. The other composition is handled the same way.
• (3)=>(2)
  From SGA4 6.8.2 we learn that every point of has an associated functor , where is the category of neighbourhoods (French: voisinages) of . A neighbourhood is a couple where and . The cofiltrant category of neighbourhoods of admits a small cofinite full subcategory, so by abstract nonsense the functor is a pro-object in . A pro-object is, by definition, just a functor from a small cofiltered category to (think of it as a diagram to form a projective limit, hence the name). Let’s write the pro-object , hiding the small cofinal full subcategory of in the indices.
  Now for a point , is a filtering colimit (=projective limit) of all , thus a filtering colimit of weak equivalences. We conclude that is itself a weak equivalence. Since this holds for every point, is a weak equivalence.

This is something not to miss if you’re interested in mathematics, and it might also be valuable if you’re taking a first course in complex analysis. Even after you’ve taken a course on complex analysis, you might enjoy the animation of the Hopf fibration (which I liked most).

Go straight to watching the videos in English.

via The Math Less Traveled (via Wadler’s Blog)

Searching for a mysterious “asdf jkl” bug is really difficult. I found about 3 different laptop keyboard problems similar to this one, although they always had problems with the “Enter” key, too. My “Enter” key worked fine. Even if those problems would have been more similar to mine, it wouldn’t have helped. Nobody who has had those strange keyboard issues has found a solution so far. First I looked for Ubuntu/Linux related problems, and looked through various log files, then I found out the keys didn’t work even on the BIOS level. Oh, it was horrible.

Then I found a solution.

The keys were dirty. Not like the yellow patina that you can see on old keyboards – I clean my keyboard regularly, so this doesn’t happen. Dust stuck deep inside, almost invisible. If it would be visible, I wouldn’t have looked for a software problem for an hour… I suppose the dust got wet from humidity in the air (sometimes I use my laptop in the kitchen).

How to repair a Lenovo Thinkpad T60 laptop keyboard that has some kind of strange some-keys-no-longer-working issue:

Turn off the computer and remove one key with your bare hands (gently). It’s most easy to begin with “g”, “h” or “b” because you can pull off the TrackPoint (the red mouse thing) first. Then you can remove all letters (or even more, at least remove the letters that don’t work).

Don’t forget your keyboard layout. It won’t be so easy to google the correct layout if you can’t type!

This is the place where you can find hidden dust. Removing this dust was crucial to get the keys working again (for me).

I wrote this, so that other Laptop users with the same problems can find the solution in google and don’t have to spend an hour of their life searching for non-existent bugs, like I did.

As I said, you don’t need to be interested in the Hodge Conjecture (although it’s one of the so-called millennium problems, and you would get a million $ for a proof or disproof). You’ll get a glimpse on some central ideas of geometry from this talk, including:

• Projective plane (where every two lines intersect in a unique point, but maybe a point at infinity)
• Cycles (certain “shapes” in spaces, like points, lines, circles)
• Complex numbers (you don’t need to remember i²=-1)
• Polynomial equations (like x²+y²=1)

At the end, it’s a bit fuzzy about integrals, differential forms and the actual Hodge Conjecture, but if you weren’t looking forward to learn anything about the Hodge Conjecture, this spares you about 30 minutes of the video.
If you’re really interested in the Hodge Conjecture, take any book about Algebraic Geometry and learn some of what’s in there. If you did this already – then what are you doing here?

When people ask me about the stuff I learn in mathematics, I try to explain something about homotopy/topology, circles defined by polynomial equations or projective space. This usually conveys the idea that mathematics is not all about number crunching or calculus. So this video is in the same spirit of explaining what mathematics is.

If you don’t want to know anything about geometry (maybe because you hate mathematics since school), then I recommend this video (10 minutes, but you can skip the last 5 minutes):
“The most important video you’ll ever see” [sic]

The words “cheese” and “wrap” are translated…

Let’s have a look at some of the best Web 1.0 math websites:

• The On-Line Encyclopedia of Integer Sequences – look up some short sequence of numbers to see in which patterns they fit.
• John Baez: This Week’s Finds in Mathematical Physics – John Baez has been writing his wonderful thoughts about mathematics, physics and the in-between for more years than I know what mathematics is. You can learn a lot from these notes. He has been posting it in sci.physics.research, sci.math.research, sci.physics and sci.math but now he also has an RSS feed, of course.
• The Mathematical Atlas – a hand-crafted tour through the various regions of mathematics, clustered along the AMS classification, spiced with many useful links. (I hope this will be relaunched as a community-based website one day. It deserves to survive).
• The Mathematics Genealogy Project – find out how half of all professors are descendants of Mersenne: 139335 mathematicians in the database, 61089 descendants of Mersenne. They have nice posters, too.

And now, before I sketch my vision of Math 2.0, for some of the best Web 2.0 math projects:

• The n-category café – a group blog about higher algebraic structures (especially n-categories) and physics. There are almost always interesting discussions going on.
• The Catsters – two mathematicians explain category theory (thus the name Catsters) in short, understandable snippets made exclusively for YouTube. Have you ever felt the need to learn what Monads are? String diagrams? Maybe you would be happy if someone would explain you limits and colimits. The Catsters do it, and they do it for free.
• The Open Problem Garden – a collectively maintained list of open problems in mathematics, ranked in difficulty. It’s still in an early phase of it’s life-time and somehow concentrated on problems with a combinatorial flavour, especially graph theory. Maybe you could enter your favourite open problem there?
• Wolfram Alpha – a mathematical data search and browse engine. You can look up statistics, perform comparisons and calculations and visualise this data. Very nice!
• Complexity Zoo – a website that collects computational complexity classes, with lots of helpful explanations and fact around them. At the moment of writing, they count over 480 complexity classes!
• Rigorous Trivialities – a group blog about algebraic geometry, with a huge series about “Algebraic Geometry from the Beginning” – which I recommend for it’s little intuitive text-snippets, where you can pick just what you need.
• The Secret Blogging Seminar – a group blog about algebraic geometry.
• The Unapologetic Mathematician – John Armstrong’s high-level educational math blog. You can pick some topic you want to learn and track back the links to the point where you’re on safe ground. This way, learning is much more efficient than using a linear book. Covers, for example, some category theory.
• What’s new – Terence Tao’s blog. He describes it with the words “Updates on my research and expository papers, discussion of open problems, and other maths-related topics”. Well said, worth a look!
• Timothy Gowers’s blog – currently obsessed with the PolyMath project (see below).

UPDATE: For a list less biased by my personal interests, see the thread “most helpful math resources on the web” on MathOverflow

Okay, now what is Math 2.0?

Math 2.0 is mathematics done collaboratively in genuine new ways over the internet.

This means, a website qualifies as Math 2.0 if it changed the way mathematicians collaborated.

However, it seems like the school education community, more focused on children, uses the term Math 2.0 as a buzzword for “learning math over the internet”.
From the Delicious tag Math2.0 you can see that the term is also used for math blogging.

Keep in mind that all this 1.0, 2.0 buzzword terminology is just tagging some websites. It’s not important, and as Tim Berners-Lee says, the web was always about communication from person to person, it’s nothing new.

My favourite Math 2.0 projects:

• The nLab – the wiki associated to the n-category café, an attempt to structure the discussions and facilitate re-use. This way, the nLab users build an expert encyclopaedia about their subject. Since it’s a subject with intense research going on, it’s more like their secret lab book than like the consensus-based Wikipedia. The rather inclusive viewpoint instead of the encyclopaedic exclusive viewpoint of Wikipedia has already created a very helpful collection of references. The nLab personal lab wikis have already shaped how people do their mathematical research, thus it truly qualifies for Math 2.0.
• MathOverflow – the mathematical question&answer web site, intended to be used by mathematicians (so, no homework questions on this site). Without MathOverflow you would have to know the right people. With MathOverflow you can just ask them.
• PolyMath – the first massively collaborative mathematics problem solving project. It was successful, so they’ve just recently started the next PolyMath project. Gowers and Nielsen have an article in Nature about PolyMath.
• The Tricki – a wiki of problem solving tricks. It’s somehow in the spirit of Polya’s book about mathematical problem solving, but much more practical, solution-centered in concrete situations. It’s something you couldn’t get with a book and it’s perpendicular to classical literature.

Usage of the terms “Web 2.0 math” and “Math 2.0″:

I have some ideas in my mind for a future Math 2.0 project, involving creative use of LaTeX, wikis and collaborative/social websites… but it will take another few months until the idea is ready to go public, and I still need to convince some collaborators to help me with the workload.

Where they talk about Web 1.0 and Web 2.0, the Web 3.0 is not far. Clearly, somebody must fill the buzzword Math 3.0 with some nonsense! Since this post is already long enough, I will speak about the semantic web, Web 3.0 and the great potential for mathematicians another time.

(And I’m really sorry that I didn’t list all good math blogs or other math projects, not even all my favourite ones, like this wonderful blog about motivic stuff. However, if I missed a popular one, I would be happy to hear about it in the comments.)

How to manage news?

It is vital to get at least some news. You need to know about political developments, to be informed when it’s time to cast your vote (or, if you’re not living in a democracy, when it’s time to protest). You need to know about developments in your work, so you can adapt and don’t risk losing your job because you’re too old-fashioned. You need to know about economy if you’re investing money. You need to stay informed about every project you want to participate in. Maybe you even need to know what pop-stars do, because if not, you have nothing to talk about with your friends.
How to cope with this information overload?

Disclaimer: if you’re using a feed reader/aggregator and/or know about Google Reader, scroll to the bottom of this post. There’s not much new for you in here.

Why information overload?
Imagine you were living in the newspaper times, without television, without internet. Then, you could read (parts of) the local newspaper every morning and that was it. No problem yet. Of course, it would have taken much longer than today, to transmit the latest information about some projects, developments in society, and so on. Since everyone has had the same conditions, it wasn’t a problem to be that uninformed.
Now imagine you were living in the television times, without internet. Then you could always get the latest “breaking news” and stock market information, and it was vital to have this information because everybody had access to this information, so you would risk a disadvantage when being uninformed. It was easy to select the information you needed, because you could just find out which channels are sending the information you want and then stick to these channels.
(Sorry for skipping the “radio” times.. here almost the same applies as for television).
What’s different with the internet? You’re no longer restricted to 100 channels sending information in a linear fashion. It’s a completely uncontrollable mess, just way to much to consume, even if you would try to do nothing else. Today, it’s not only vital to get information, it’s also absolutely unavoidable to select important information – for everyone.

Another problem that’s creeping in since the invention of radio and television is the restricted attention span: human brains are not capable of doing several things at the same time – it just looks like that because we’re able to switch very fast from one task to another. Switching tasks is a problem, sometimes. When you’re concentrating on something cognitively challenging, you will need about 5 minutes to get back to your level of concentration if you were interrupted. Another problem is, that our brain is relatively good at doing very different things in short alteration (like playing piano and drinking water and thinking about stock markets), but it’s not that good at doing very similar things in short alteration (like speaking one foreign language and then another). That means, watching TV while working does not really work for jobs that involve information processing (at least, you won’t be as good as you could).
The internet is even worse: many people have e-mail notifications, mobile phone ring tones, twitter notifications, etc., so they’re interrupted in their work very often. Since many jobs today involve information processing and getting the latest news is information processing, our brain has severe problems in doing this in short alteration.
An interesting text about the question how the internet changes the way we think can be found on the BackReaction blog.

Software can be used to control the news flood, if you use it wisely. For example, you can use an instant messenger like ICQ or Skype to let people contact you, but without an audible notification. Then you’ll only read their messages when you’re not in a state of concentration. The same applies to (mobile) phones: you should mute them for some time each day. A phone call would interrupt you immediately, while a short message is controllable by you – it’s you who decides when to digest the information.
Ask yourself: do you really need the latest latest news? Most likely it suffices to get the news some hours later, too. Instead of watching TV and hoping for not missing the breaking news, you can use feeds in the internet, so you get news immediately but are able to consume later. Feeds offer another problem solution: You don’t have to navigate to 50+ websites to know what’s going on about your favourite projects, to get your favourite opinions about politics, etc.. You just subscribe to the feed of each website and a software manages the information flow for you.

Wikipedia has a nice explanation what a web feed is. You can learn about feed aggregators/readers there, too. I’m using the Google Reader, which is a web-application, so you can use it from everywhere on any computer that has internet access. It offers collaborative features: you can select information entries to “like” them or to “share” them, if you want even with a personal comment. The software then displays on top of every entry how many people “liked” this entry, so you can use this as a guideline in the selection of important information. To “share” a news item means it’s added to a special feed, the shared items feed, which your friends or co-workers can subscribe to. This way, an organisation can effectively process the information that’s important for them and then distribute it to all members. Most aggregators offer ways to sort the various news items, so you could select some to read them later and delete some items unread.

(comic licensed from Oliver Widder under a Creative Commons Attribution-NoDerivs 2.0 License)

Feeds offer interesting ways to organise various kinds of information. If you’re a Wikipedia author, you can use feeds to watch thew pages you’ve written. If you’re using MathOverflow, you can use feeds to get the latest questions about your favourite mathematical topic (same applies to StackOverflow). Every blog in the internet has a news feed. You can avoid twitter, because each twitter account has an associated feed. Podcasts are nothing than feeds with sound- or videofiles in it.

Another software recommendation besides Google Reader is the Android software NewsRob (which works only on Android smartphones). NewsRob synchronises with your Google Reader account and is able to save the feed items (and even the websites that are behind the feed items) for later use. This is really helpful if you don’t have an internet-everywhere contract with your mobile phone provider and still want to read the news somewhere where you don’t get internet access (in a train, for example).

UPDATE: At the time I’m writing this, Google presents a new feature: Subscribe to websites that don’t even have feeds. This way you can monitor changes on any site that’s important to watch. Google creates a feed with the URL http://www.google.com/notificationservice/webchanges/webfeeds/12345.... so you can even share the feed. For those who don’t like being watched, you can exclude your site from the feature by standard methods.

I admit: this is one of the most absurd posts ever…

On searching “smell file format”, the results are really absurd, too.

But this way, I came across this strange patent for digital cameras that don’t only capture pictures or videos but smells, too. Now you would think of a real odour-capture-device, but this patent suggests just a user interface to set one odour out of a list of pre-defined odours. The purpose of such a technique is clear: you can later display the image along the odour via a dispenser. Yeah … everybody has been waiting for this.

Of course, an “odour” is just a collection of molecules, since the human nose is a detector for chemicals in the air. Maybe you have never thought about this, but fishes in deep sea orientate by odour instead of light. So, an odour file format would be almost the same as any molecule file format. Maybe one could get a good file compression by knowing which molecules can be detected by the human nose and which can’t. That would be something like the list of pre-defined odours list in the patent above.

I’m pretty sure that, if it doesn’t exist today, there will be an odour file format one day.

Then I searched (again, out of curiosity) for such a dictionary which lists all smells humans can detect, together with their molecular composition. I wasn’t that successful, but at least some nice results:

• Fantastic Flavours has some nice articles about some of the most interesting flavours, with many links pointing to more information. Surely interesting for anyone who likes cooking, wine, coffee, tea, fruits, …
• The LRI & Odour database is funny to browse, too. For example, you can look for “banana” and see that bananas contain some chemicals that are also contained in melons.
• The Flavornet lists many flavours (most flavours are odours). Very interesting: the odour classes (there is a Maillard class, for example. The Maillard reaction is the chemical reaction that makes roasted food so tasty). Beware: the Flavornet crashed my Firefox, I had to use Chrome to get it working.