Friday, May 29, 2009

more thoughts about Google Wave

I recently wrote:
It would be nice to be able to actually markup hypertext... but that still appears to be outside the range of feasibility.

Maybe it could be called TLM - Text Language Markup?

It turns out the given name is Wave, and is a dialect of XML still being tweaked by the wizards of Google and a few thousand developers who just saw it for the first time this week.  I spent the time to see the demo, and have been following up to see what others think. The gut level reaction seems to be one of  hope that this is something good, because it does seem to be a game changer.

The void being filled here is hard to describe, I've been trying for years... it all boils down to context. When you send an email and then reply, you're forced to use all sorts of mechanisms and tools to attempt to keep your train of thought, your conversational cache, your context.  Every step away from being able to just add a note, circle something, highlight or annotate text to draw attention makes conversation less efficient. Wave is going to provide a mechanism that does a much better job of preserving context. 

I expect Wave to succeed because of the good Kharma that Google has built up, along with their pledge to open-source most of it... which will greatly help adoption as a defacto standard.

The money quote for me was Tim O' Reilly's mention of the need for granularity when editing book manuscripts, which I feel vindicates some of my howling in the wilderness these past years. 

Google wave - Web 2.0 at last

Google has acidentally created the first good realization of web 2.0.  By providing a way to mark up hypertext, they are on the path to resolving one of my long term frustrations with HTML and Web 1.0.  It will now be possible to collaborate in fine grain, with minimum loss of context because of tools that lack the ability to point at a part of a document.

As Tim O'Reilly says:
Our experience with collaborative editing of book manuscripts at O'Reilly suggests that the amount and quality of participation goes up radically when comments can be interleaved at a paragraph level.
Colaboration and sharing of data is about to take off in new and very powerful directions. I highly recommend you take a few hours, watch the demo video, and dive in to see what the future is going to bring. It's very exciting.

Tuesday, May 19, 2009

More on why Google sucks

Google forces you to try to figure out keywords unique to the subjects you are looking for. This doesn't always work, and is especially hard to use for conversations that are non-technical and thus have the least amount of topic-specific jargon to latch on to. 

The next great leap will be to figure out the subjects of text, so that you can match on multiple areas of interest, and explore the intersections.

Why google sucks

Ideas are like Reese's peanut butter cups... they are all the synthesis of other concepts in new combinations. Creative energy is the exploration of new combinations to yield new and interesting combinations, some of which turn out to have value. Ideas are all a synthesis of *more than one* pre-existing concept. Thus any catalog of ideas necessarily will have all of the component ideas cataloged so that you can search for combinations of ideas...   a single heirarchy will fail miserably at this task. Tags with a folksonomy have a much, MUCH better chance of yielding positive results, and being a better tool.

Why virus scanners are doomed to fail

Intent is very important, for example... spam is email with an intention to push a product. Malware and trojan horses are similar in intent, but even more malicious. 

The user of a computer can not tell what the intent of the author is... thus it's necessary to provide a mechanism for limiting the scope and actions of that program, in an effort to help with that judgement.

The current group of operating systems do not provide a way to limit the scope and environment of a program prior to its execution, to the ordinary user. 

Virus scanners all assume that programs can be examined and found not to be of ill intent merely by checking them against some arbitrary lists. This can be seen to be a losing battle when viewed from this perspective.


Intent is very important. For example, if there is an auto accident with a tragic outcome, the driver may get the sympathy of the community if they judge the intent wasn't there... on the other hand, if malace aforethought is judged to be the intent, the driver may be executed by the state. 

The intent of an action has very little to do with the reactions of others... it is important to divorce your feelings from the situation and to be objective if you are attempting to judge the intent of others. A 2 year old girl has little intent beyond the next few minutes amusement or contentment, for example. It's very hard sometimes to keep this in mind, but a good parent will do so, even if it's after the fact.

Empowering IT

It is the IT departments mission to provide a stable and consistent set of tools for our users... empowering them should also be the focus here... providing the best tools and skillsets possible to maximize the reach of the users is our goal.

Empowering Virginia

We are limited by our language, vocabulary, and skill set in using them. I believe the "classical" education was designed to give the largest possible set of tools to the pupil, which is in stark contrast to the currently accepted system which provides a lowest common denominator.

It is a parents responsibility to provide their children with the best possible set of tools, to empower them as much as possible. It is THIS that should be the focus, not the idea of giving them a head start in a race. The head start can rapidly be overcome by someone who has better skills negotiating with others, and can communicate more efficiently.

It is with this realization that I'm going to refocus my energies as a Father to Virginia.

Monday, May 18, 2009

Expressing limits - can't be done yet

If someone came to you with a program which you knew might be evil... how do you express to the operating system that you wish to run it, with only the following access:

  No internet or other access?

You can't... your power to express this is non-existent. This forces you to have to trust each and every piece of code you run to not make system wide changes.... which is just plain stupid.

Virus scanners check the code against a list before running it... this list is never perfect, and there is a delay in adding new entries...

Thus security on a pc is not perfect. This will not change until there is a way to express the limits on a program prior to (and during) it's execution.

Operating systems came about as a means to share resources safely... they aren't done yet doing that job.

Tools - it's all about tools

Better tools... it's all about better tools. A good tool reliably allows you to extend your reach, and do things more efficiently, with more control.

The internet is still relatively young... the web is now 20 years old, and we're still figuring out the tools that can be built. Mashups are cool, but being able to program enough to put together something that can be optimized by others is a very good thing.

The complexity of putting a simple "Hello World" example in place using the LAMP (Linux, Apache, MySql, PHP/Python) precludes a lot of us from using it as a tool. Perhaps we need something a bit easier?

Tags - an oops

I started to get rid of tags from entries on my blog because I thought a shorter tag list would be more useful, but I've figured out that is not the case. There is an expressive force in tags that is lost when you cut out the free association aspects of it. You need to be able to make those serendipitious connections, this is where ideas come from.

I'm going to review all of my previous posts, and put tags as appropriate on them, as time permits. This will allow eaiser discovery of the good stuff, and make the barriers to it lower for someone who happens in.

The limits of power

What you can express limits your actions. If you can't communicate an idea clearly to others, who wish to help you actualize it, you are stuck with the resources you personally have control over.  Thus it's best to have a toolset that expands your capabilities.

Ideas are worthless, it's been said... and there are good arguments to support that. However, not all ideas are worthless, there are some small fraction that could do great things, but are trapped in isolation because the person who concieved of them can't express them.

The flaming of newbie questions in the Usenet/Linux culture limits the number of people who will adopt it as a tool, as well as limiting those who might otherwise pick up a new language, try out something new. Clearly this is broken, or at least immature.

Teaching tools to empower others gives them the option to exert their own work, and their own time, to try to give their ideas life... this parellizes the problem of weeding out the bad ones, and makes the possibility of a good idea coming out of a population of X people increase because they have a better set of tools.

Thanks to Daryll for putting up with my first run through of these ideas.

Thursday, May 14, 2009

Small improvements in the right places

I'm cleaning up the tags on this blog, getting rid of the most single instance items which are noise. I'm also clearly tagging all of the rants, just because I think I need to. These small changes should at a bit of value to this blog.

Wednesday, May 06, 2009

Blogging forces some unique constraints upon the author, these are mostly societal and expectations, not technical. The blogger is expected to write a short essay (french for "to try") on a regular basis. This stream of text is also expected to stay on a single topic, the best to feed the point and click sensibilities of the larged audience possible.

No more!

Blogging is broken. Here's why.

The short form doesn't allow for the necessary amount of exposition necessary to lay the groundwork for common understanding that is necessary to overcome the limitations of text and the vagaries of human language. There are always differences between the intended meanings of the author, and those inferred upon reading. It's necessary to supply more than one set of explanation to help the reader decide which of their multiple guesses at the intent of the author is closest. 

Conversation involves listening to the other partner, and giving feedback on the concepts discussed, to make sure your minds are as close to synchronized as possible. This helps then ensure clear communication of the intented subjects. The better the communication, the further the reach from the everyday is possible.

In order to be able to discuss deep subjects, specific vocabularies must be developed... specific metaphors, and examples. This is the reason doctors have their own medical speak, and programmers know what loops are.

The short form doesn't allow this exposition at length. It forces an artificial division of text... which reduces the amount of signal, and increases the proportion of text used to frame each piece... thus it's not best for very deep subjects.

The limits of what you can express and explain are the very limits of your ability to change the world. If nobody was able to explain the chain reaction, the atom bomb would have never made it out of the theory of one physicic and into reality.

Language is power.

Blogging can be tweeked, ever so slightly, to be more powerful. The limitations are not technical, they are societal, so we don't even need new tools... just new expectations.  Fortunately, expectations can be adjusted as a matter of intent.

Blogging tools allow keywords on pieces of exposition, to link things together. Blogging tools allow the linking of a longer explaination seemlessly into your text, to help shorted the text for the reader already in sync, while providing additional support for the reader who is not familiar with the topic, or has additional doubt about the intent of the author.

The need to stay on a single topic is an artifact of the pre-tagging days of blogging, and is hereby declared obsolete!  (HA)

Actually, if you give up the idea of having a stream of regular readers, and rather chance meetings of other minds, your point of view is shifted... and your emphasis should then be on allowing the best (easiest) discovery of the content you already have. Linking back to previous work related is one of the most powerful tools to accomplish this. The calendar based archive is appropriate for a single-topic blog, but the entire creative output of a person will necessitate new and better approaches.


Suggestions for improving blogging in general.

Don't waste time and emotional energy limiting yourself to a single topic, issue, etc...  go ahead and enjoy the freedom of free association, but please be sure to tag things, and nuture your own links, and links to others.

Revisit your older work, update, revise, correct it... much as if you were to produce a new edition of a book... if you expect people to read it, you should be willing to take the effort to fix mistakes that you've found, and to clarify it.   Mark your changes to help avoid confusion if they are significant.

Give up the idea of having to do something every day... just spend time on it when you like.  If you need to peek at your Google analytics, go ahead... but it's like dieting, there are ups and downs, and you have to keep focused on the big picture and not the numbers immediately in front of you.

Tag everything, prune the tags (get rid of tags only used once that you doubt you'll use again) to make your index smaller.

Ask others for feedback... give others feedback.  It's lonely here out on the small end of the long tail. Your time and attention is the most valuable thing you have to give. A little bit goes a long way.


In summary... blogging daily on a single topic is broken... blogging at length with good links back to relevant topics and a good index is better.

It's the same tool set, you just use it a bit better. You extend the reach of topics you can discusss, and thus extend the extent to which you can change the world.

Which is what we all want in the end... to leave it better than we found it.