Main Area and Open Discussion > General Software Discussion

Godin: the end of dumb software

<< < (4/6) > >>

iphigenie:
Understood. And be assured, you guys are far from clueless. Unfortunately, what you're asking for often cuts right to the heart of most of what's wrong with the current state of 'computer science.'

Marvin Minsky (one of the founding fathers of AI) once complained that there were far too many smart people working on the "easy problems" (like designing a 'better' word processor) while most of the real problems were largely being ignored. Truth is, many of the fundamental questions in computer science have yet to receive definitive answers. And for better or worse, most people 'outside the profession' are unaware of that.
-40hz (September 13, 2009, 10:50 PM)
--- End quote ---

According to my partner, even inside the profession many people have stopped asking the questions. You'd think we have found the answers, the right ways to abstract domain problems and write code that is correct and does what we expect every time...

But then a lot of it is over my head, I'm just married to someone who cares about these things and has been working on the theory of programming languages for 15 years now, with the firm belief that there are right ways to figure out - if anyone really is interested in problems like completeness and correctness and can talk for hours about individual languages, what they did good and where they went wrong etc. I'd love to put you in touch...

40hz:
I'm just married to someone who cares about these things and has been working on the theory of programming languages for 15 years now, with the firm belief that there are right ways to figure out - if anyone really is interested in problems like completeness and correctness and can talk for hours about individual languages, what they did good and where they went wrong etc. I'd love to put you in touch...
-iphigenie (September 15, 2009, 12:14 PM)
--- End quote ---

Sounds like a very interesting individual. I'd be happy to chat with him. But better yet, maybe you could convince him to join us here and start up a thread on the topic? I'd love to hear what someone who is actively involved in language research and design has to say. And I'm sure I'm not alone.

"Nothing improves your ability to solo better than playing with musicians who are better than you." 

Note: the above saying has been attributed to so many musicians that it would take a day to name them all. Feel free to insert your favorite. Odds are, he or she probably did say it at one time or another. >:D

 :)

BTW: Does he have any publications or a webpage?

iphigenie:
As to the question of the program figuring things out for you or not, I think the solution is

a) do figure things out for the user especially if they are outside the core goals of the user
b) make it transparent, easy to correct/undo, and easy to tweak

I'm with Mouser on the annoyance that are word processors when you are in writing mode - as a matter of fact I cannot write in any of the current word processors - all of them just volunteer formatting in the wrong way at the wrong time, and frankly when I am writing this improvised formatting gets in the way, it breaks the flow. But I think it is in part the right way for a word processor to do this, it is just that people who get bothered shouldn't use them for first drafts, we should use a text editor or a structural editor (actually, a smart word processor might have a modal component - in draft mode just let me dump stuff, in revision mode focus on corrections and merging versions, in polish mode help maintain a clean and consistent styling etc.)

But I would like smarter software in the sense that with a lot of software I still spend far too much time managing, and too little doing/thinking what really was my goal. Managing the management, and managing the process.

Let me give a few examples:

1. email
Most people keep email, but since most programs make it hard to find email, get an overview of a history, or drill down through email, we end up creating folders. And once a folder has too many emails in it so we once again cannot find things easily with a browse or a quick search, we tend to create more folders and move email around. Before we know it we spend more time moving filing and finding email than we do reading and writing it, or acting on whatever it is.

Then you use a tool like Opera's M2, Gmail, or a methodology that shows how to do virtual organization in outlook, or use an add on like xobni, and you realize how much time you save when you stop organizing the email

None of the tools mentioned above pulls the whole job off, each has some nice ideas that you wish the others had to be perfect...

2. launchers
We've all tried menu systems and the problem is that you have to keep maintaining them - adding new programs, removing old ones, categorizing - then you switch to the dynamic launchers (like FARR), or the self-building launcher (I use "task commander" on windows) and yes, the tool does some thinking for you, but you can tweak it, and you stop having to constantly organise things

For example task commander adds every program that is run to its launch list - whether they are run from the command line, started via file manager, or the start menu. I find it a lot less time consuming to go and tell it to hide the ones I dont want to launch again, than it was adding programs manually to a launcher. This is especially great if you reinstall windows and have a lot of tools which dont need an install and can just be run from their folder (for example you dont want to reinstall your games when they dont need it), with a traditional launcher you have to add them all, with task commander they just appear the first time you run them... easy :D

3. Information
To me the big problem with outliners and most notes/pim/information managers is exactly the manager bit - it assumes that what I want to do is manage my information. No, what I want to is to store, save, find again, and be able to USE my information. The management part just came about as a necessary evil in order to not lose or be overwhelmed, but managing the information is not what we want to do, its what we want the tool to do for us.

I used to manually file my photos, now I let a tool file them on a pattern (date based) for me, and use the tags etc. to find it again - far less time wasted. I want the same for my information, i dont want to have to manually put it in a tree or whatever, i want to capture it, and know it has been put somewhere half logical on a pattern I am aware of (perhaps date based). Then i want to be able to virtually retrieve it and organise it on a per-project basis - ideally after the tool has used semantic clustering so i dont have to do stupid basic housecleaning...

PS: I have to disagree with Seth - the default on the web is not to be smart, it is to be cool. Smart is wholly optional and often gets in the way of cool or popular

iphigenie:

Sounds like a very interesting individual. I'd be happy to chat with him. But better yet, maybe you could convince him to join us here and start up a thread on the topic? I'd love to hear what someone who is actively involved in language research and design has to say. And I'm sure I'm not alone.

BTW: Does he have any publications or a webpage?

-40hz (September 15, 2009, 12:44 PM)
--- End quote ---

I am trying to make him join here or start a blog etc. He's been off the radar for a while, writing lots of tidbits and building a provable consistent yet usable language - but it's all on our computers and backup...- it started as an example language for reasoning about and teaching certain concepts, but it got more ambitious over time...

Will keep trying :D

40hz:
PS: I have to disagree with Seth - the default on the web is not to be smart, it is to be cool. Smart is wholly optional and often gets in the way of cool or popular
-iphigenie (September 15, 2009, 01:03 PM)
--- End quote ---

@iphigenie - I'm so glad you said that. :Thmbsup:

I've bashed the concept of 'cool' so often that my social circle has pretty much stopped listening to me.  ;D

IMHO, what passes for cool (in the Web 2.0 sense) is seldom more than a gaggle of rehashed ideas decked out in designer clothes with optional software patent applied for.

Navigation

[0] Message Index

[#] Next page

[*] Previous page