I can't resist

[Markus]

After wrestling with the urge a few days I'm going to give up and let
the group decide.

If people are interested, I'd like to present next time, and give a
response of sorts to Matt's Wheeler presentation. Not a rebuttal,
exactly, nor a concurrence, more a sort of orthogonal counterpoint. I'd
like to push a little harder on some of the areas that he glossed over
and explore a few of them in greater depth. If his presentation was the
"skipping off the top of the atmosphere" view I'd like to look closer at
the chemistry and thermal budgets of a few particular undersea vents on
what I believe (but can not prove) is the same planet.

-- Markus

[Phil Tomson]

I think you just signed yourself up Markus.

Phil

[CPSolver]

On Sep 26, 5:59 pm, Markus <Mar...@reality.com> wrote:
> If people are interested, I'd like to present next time, and give a
> response of sorts to Matt's Wheeler presentation.  Not a rebuttal,
> exactly, nor a concurrence, more a sort of orthogonal counterpoint.  I'd
> like to push a little harder on some of the areas that he glossed over
> and explore a few of them in greater depth.  If his presentation was the
> "skipping off the top of the atmosphere" view I'd like to look closer at
> the chemistry and thermal budgets of a few particular undersea vents on
> what I believe (but can not prove) is the same planet.
>
> -- Markus

I'm interested.

Matt's talk stretched my mind in a way I enjoyed. I'm thinking we
were experiencing something like what cavemen were thinking when they
invented the early grammatical elements of spoken words and
sentences. I can imagine a caveman/cavewoman saying "cave fire" and,
like Matt's software, attempting to decide if that means "The cave is
on fire and we should put out the fire" or "We should start a fire in
our cave" or "I'm cold".

I look forward to you/Markus entertainingly stretching my mind in any
direction that software can take us.

Richard

[Matt Youell]

On Sep 26, 10:24 pm, Phil Tomson <philtom...@gmail.com> wrote:
>
> I think you just signed yourself up Markus.

Indeed. I cower in curiosity. :)

--
-/matt/-
http://youell.com/matt

[Matt Youell]

On Sep 27, 10:57 am, CPSolver <p...@SolutionsCreative.com> wrote:
>
> Matt's talk stretched my mind in a way I enjoyed.  I'm thinking we

Excellent. And I'd hate to stretch anyone's mind in an unenjoyable
way! :)

> sentences.  I can imagine a caveman/cavewoman saying "cave fire" and,
> like Matt's software, attempting to decide if that means "The cave is
> on fire and we should put out the fire" or "We should start a fire in
> our cave" or "I'm cold".

Interesting, because I have been using a caveman metaphor to explain
why categories are the fundamental abstraction in Wheeler. My premise
is that naming things is fundamental to our understanding of
everything around us. Eventually we abstract names into categories.
And there you are.

Also, btw, I wanted to thank you for your question about whether
``"Hello, world!" print`` would work. That was something I had
considered several months ago, never tried, and subsequently
forgotten. I'll use that as my introductory code snippet in future
demos. I think it will grab attention better than my original. :)

--
-/matt/-
http://youell.com/matt

[Markus]

> Also, btw, I wanted to thank you for your question about whether
> ``"Hello, world!" print`` would work. That was something I had
> considered several months ago, never tried, and subsequently
> forgotten. I'll use that as my introductory code snippet in future
> demos. I think it will grab attention better than my original. :)

I don't know. The vanilla 'print "Hello, World!"' worked pretty well
because you were able to riff off it's unsurprising nature a bit. Then
when you reveal that it's not doing what we expect, you're doing it in
the face of our complacency. I'd say introduce the plain looking little
girl first, _then_ lead us to discover that she's an alien creature with
strange powers.

-- Markus

[Phil Tomson]

>
>

Hey Matt,

You had 3 books on the table as I recall: "if I were a set", and
old-school math book by M. Ward, and Generative Programming - you
explained why you had the first two, but not why you had the third.
Was that something that you just didn't get to?

Phil

[Eric Wilhelm]

# from Matt Youell
# on Sunday 27 September 2009 23:35:

>Interesting, because I have been using a caveman metaphor to explain
>why categories are the fundamental abstraction in Wheeler. My premise
>is that naming things is fundamental to our understanding of
>everything around us. Eventually we abstract names into categories.
>And there you are.

But remember that premature categorization is the root of all evil.

Though I think you might be avoiding that by allowing things (I
know, "there are no things") to belong to multiple categories.

--Eric
--
"I've often gotten the feeling that the only people who have learned
from computer assisted instruction are the authors."
--Ben Schneiderman
---------------------------------------------------
http://scratchcomputing.com
---------------------------------------------------

[Matt Youell]

On Sep 28, 8:55 am, Phil Tomson <philtom...@gmail.com> wrote:
>
> You had 3 books on the table as I recall: "if I were a set", and
> old-school math book by M. Ward, and Generative Programming - you
> explained why you had the first two, but not why you had the third.
> Was that something that you just didn't get to?

Yes, I should have explained that one. I brought "Generative
Programming" (Czarnecki & Eisenecker) as a possible point of
reference. Their main thesis seems to be that we can bring automation
to software development like Ford did with automobiles, and that DSLs
and generic programming will save us all. Sounded plausible whenever I
bought the book (early 2000s) but I'm far more skeptical these days.

While I wouldn't recommend the book based on it's thesis, the coverage
of many topics is good and I could recommend it based on that
coverage. There is a heavy C++ bias to the book, and I think that
limits their perspective. For instance, their chapter on generics is
an unwitting love song to duck typing. Still, smart people wrote a
smart book. It's academic but reads a little bit easier than a typical
academic paper.

What does this have to do with Wheeler? I've been reading/re-reading
this book quite a bit lately while working on Wheeler. There are some
topics such as Aspect Oriented programming that are explained well. In
leading up to AOP they discuss Subject Oriented Programming (http://
en.wikipedia.org/wiki/Subject-oriented_programming) which explores one
of the central problems that I'm trying to solve with Wheeler: How do
you share information across perspectives?

There were also some other areas where they had a detailed explanation
of concepts like dispatch, typing and polymorphism, where I could sort
of mentally check myself and compare where I thought Wheeler was with
an actual reference point in traditional programming.

I wouldn't run out and buy it, but if you can get it at the library it
might be worth a weekend of perusal. I'm still using my copy but I'll
probably have it up on the PDX Virtlib in a few weeks or so, (http://
virtlib.builtsoftware.com) then anyone can borrow it.

--
-/matt/-
http://youell.com/matt

[Matt Youell]

On Sep 28, 10:50 am, Eric Wilhelm <scratchcomput...@gmail.com> wrote:
>
> But remember that premature categorization is the root of all evil.

:)

> Though I think you might be avoiding that by allowing things (I
> know, "there are no things") to belong to multiple categories.

Spoon boy: Do not try and bend the spoon. That's impossible.
Instead... only try to realize the truth.
Neo: What truth?
Spoon boy: There is no spoon.
Neo: There is no spoon?
Spoon boy: Then you'll see, that it is not the spoon that bends, it is
only yourself.

[Phil Tomson]

On Mon, Sep 28, 2009 at 1:17 PM, Matt Youell <softb...@gmail.com> wrote:
>
> On Sep 28, 8:55 am, Phil Tomson <philtom...@gmail.com> wrote:
>>
>> You had 3 books on the table as I recall: "if I were a set", and
>> old-school math book by M. Ward, and Generative Programming - you
>> explained why you had the first two, but not why you had the third.
>> Was that something that you just didn't get to?
>
> Yes, I should have explained that one. I brought "Generative
> Programming" (Czarnecki & Eisenecker) as a possible point of
> reference. Their main thesis seems to be that we can bring automation
> to software development like Ford did with automobiles, and that DSLs
> and generic programming will save us all. Sounded plausible whenever I
> bought the book (early 2000s) but I'm far more skeptical these days.
>
> While I wouldn't recommend the book based on it's thesis, the coverage
> of many topics is good and I could recommend it based on that
> coverage. There is a heavy C++ bias to the book, and I think that
> limits their perspective. For instance, their chapter on generics is
> an unwitting love song to duck typing. Still, smart people wrote a
> smart book. It's academic but reads a little bit easier than a typical
> academic paper.

I had a similar impression of the book... maybe that's why I never got
very far into it. It seemed to me that the kind of thing they were
promoting is kind of difficult in C++ (which is why they needed
several hundred pages), but rather trivial in languages like Ruby
(with metaprogramming), Lisp (with macros) or even OCaml (with
MetaOCaml: http://www.metaocaml.org/ ). But yeah, if you wanted to do
that sort of thing in C++ then that book is for you.

Phil

[Matt Youell]

On Sep 28, 1:26 pm, Phil Tomson <philtom...@gmail.com> wrote:
>
> I had a similar impression of the book... maybe that's why I never got
> very far into it.  It seemed to me that the kind of thing they were
> promoting is kind of difficult in C++ (which is why they needed
> several hundred pages), but rather trivial in languages like Ruby
> (with metaprogramming), Lisp (with macros) or even OCaml (with

Yeah, Ruby would seem to be the endgame there. Or possibly Haskell if
they favored types over objects. I wonder how/if their POV has changed
in 10 years?

[Matt Youell]

On Sep 28, 1:51 pm, Matt Youell <softbuil...@gmail.com> wrote:
>
> Yeah, Ruby would seem to be the endgame there. Or possibly Haskell if
> they favored types over objects. I wonder how/if their POV has changed
> in 10 years?

Curious, I googled to find out and found a paper with a title I"m sure
Phil will like:

"DSL Implementation in MetaOCaml, Template Haskell, and C++."
^^^^^^^^^^^^

:)

[Phil Tomson]

Yeah, I have that paper around here someplace... I read it while back.

Phil

[Igal Koshevoy]

I'd travel to the Wild West, the Badlands of Beaverton, to hear this talk.

I'm also sad that I missed Matt's talk.

Lastly, please remember to publish events to Calagator in advance, they
help me and others remember that there are meetings coming up.

-igal

[Matt Youell]

On Sep 29, 3:29 am, Igal Koshevoy <i...@pragmaticraft.com> wrote:
>
> I'm also sad that I missed Matt's talk.

Sorry you missed it too! I'm sure Markus' talk will be more
provocative though, so don't miss that one.

> Lastly, please remember to publish events to Calagator in advance, they
> help me and others remember that there are meetings coming up.

Calagator needs a repeat button. There's no getting around it. Manual
adds are a PITA and (obviously) prone to human error. (Like totally
forgetting to post.)
As our envoy to Calagotoria, please make that happen. #kthxbye :)

--
-/matt/-
http://youell.com/matt

[mar...@reality.com]

If the argument against auto-repeats is that cruft would accumulate and no one would own removing it, how about a "pester me to schedule the next one" button that send the created an e-mail with a link that would pre-fill the form with what the auto-repeater would have done (including setting pester me) a week or so before the next meeting?


Sent via BlackBerry from T-Mobile

-----Original Message-----
From: Matt Youell <softb...@gmail.com>

Date: Tue, 29 Sep 2009 18:50:23
To: WestsideProggers<westside...@googlegroups.com>
Subject: [WestsideProggers] Re: I can't resist

[Igal Koshevoy]

Markus wrote:

From: Matt Youell:

On Sep 29, 3:29 am, Igal Koshevoy <i...@pragmaticraft.com> wrote:
    

Lastly, please remember to publish events to Calagator in advance, they
help me and others remember that there are meetings coming up.
      

Calagator needs a repeat button. There's no getting around it. Manual
adds are a PITA and (obviously) prone to human error. (Like totally
forgetting to post.)
As our envoy to Calagotoria, please make that happen. #kthxbye :)
    

If the argument against auto-repeats is that cruft would accumulate and no one would own removing it, how about a "pester me to schedule the next one" button that send the created an e-mail with a link that would pre-fill the form with what the auto-repeater would have done (including setting pester me) a week or so before the next meeting?

Markus & Matt, {I'm also CC'ing pdx-tech-calendar}

Thanks for the suggestions. I've been too swamped with other events to work on this, although I hope to have some free time after WhereCampPDX[1] over.

Providing a way to clone an existing event should be fairly easy. E.g., you search for and display an old WestsideProggers meeting, click a "clone" or "repeat" link (or something similar), would then get a "new event" form pre-filled with the old content but the date/time cleared. This should be easy to do and speed up the process of creating a new event similar to a past one. I created a ticket for implementing this earlier: http://code.google.com/p/calagator/issues/detail?id=399

Sending people reminder emails is a good idea, but is more technically complex because it requires more infrastructure changes. We'd need to implement user accounts (we can extract them from openconferenceware or pdxruby2), come up with a way to prevent this from sending spam, redo the export system which currently lets anyone download the entire raw database because we don't want to publish people's emails and other account details, and come up with some way for anonymous developers to download data with sanitized user accounts they can take over for local development. This isn't "hard" but would take a fair number of hours to get production ready, as well as requiring some emergency triage as edge cases break in production, and I want to work on other things first.

Anyway, if someone wants to get started on these or other tasks, feel free to do so and send us pull request or patch when you've made progress.

-igal

[1] WhereCampPDX, October 2-4, is a free unconference for geo-technology professionals and enthusiasts. Learn more at http://wherecamppdx.org/