--001a11c219ea98351004f82ab4d3
Content-Type: text/plain; charset=UTF-8
The reason you use python is because the development cycle can be so much
faster. Not because it doesn't have it's weaknesses. Where with rails, you
sort of have to do it their way or it becomes very painful.
Normally the logic part of web apps is pretty straightforward the hard part
is making everything talk together over the web that is difficult(database,
templating, user authentication, etc), and then there is JavaScript(which I
agree is horrible) if you want to do anything client side. Django makes
that part very easy, and flask would be more bare bones if you wanted to do
it more hands on. I would say at least read the django tutorial before you
write it off. It sounds like you understand code pretty well so I think you
could build what you are saying pretty quick there.
Then there is deploying it on some envoirnment to use it. Webfaction is
good for this or you good go platform as a service. But you should decide
where you are going to deploy before you build it so you can make sure of
their constraints or best practices they provide.
@max platform as a service is awesome which is why I mentioned GAE, but
there are many others, like heroku. But, you have to build something to go
on the platform first.
Cheers,
James
On Tuesday, April 29, 2014, max stalnaker
Post by max stalnakerI suppose i might get into a tech religion argument instead but I instead
You need to make a guess about future tech to build your world domination
app upon.
You are not yet a cool kid. Take a look at platform as a service There
is a fine chance that an operation like Hereko? Would handle all your
versioning issues and take care of the scaling from a few users to a few
billion with you just writing checks. Cost will run from zero at one end to
the GDP of several small countries at the other end. I rather like them in
particular because of ancient walk bys. This is all cloud stuff of course
and i am sort of puzzled about people talking about scaling to 10**9 users
and not starting with choices that imply that capability.
Try this: is there something in the non-existent mission vision that
implies a need to scale smoothly over 9 orders of magnitude with 5 9's
reliability? Is a vote error of 10 million votes on all world referendii
acceptable?
Max
Post by James HancockSorry I am on my mobile and hit send on accident.
Yeah, those little buttons are a pain ;-)
Post by James HancockYour technological ignorance is showing a little in your last post. Is
what
Post by James HancockI ment to say. You really wish we would have stayed with software
applications instead of moving to the serverside applications?
Definitely. Going a little off topic here, but the browser-based
architecture is incredibly limited, complicated, and brittle (meaning
it breaks with every little change, like version upgrades).
iPhone/Android is a little better because most things are apps and so
the developer has more control, but *ideally* versioning and
extensions would have been built into the OSs a long time ago so, for
example, instead of JavaScript which is a complete POS (and everyone
knows it which is why there are so many extensions and workarounds for
it), apps could have been built in any scripting language and the
necessary engines and libraries downloaded on the fly. This problem
could have been solved at several different junctures (e.g., when
Microsoft added automatically downloaded OS updates for Windows 95,
when HTML was originally defined to purposely exclude interactive
elements, or when Netscape chose to build in a weekend-hack scripting
language instead of defining an open architecture, etc.) but due to
shortsightedness and proprietary-think never has been. Like I said,
if it the web had been *designed* instead of evolving with the bizarre
set of constraints it has, I'd have a working voting system up and
running in days or weeks whereas now it's going to take me months.
As for "ignorance", I'll cop to that as, like I think I said, I burned
out and retired from software development about 10 years ago and so am
having to learn a lot about the current environment (and not liking it
much at all). But I've been developing software from microcode and
assembly languages through 3GLs and 4GLs up to AI languages since I
was a kid (40 years ago) so I think I have a pretty good awareness of
what works and what doesn't. And as for django/pyramid/flask or
anything based on Tcl or Python, talk about "old school"! I wouldn't
touch either of those languages with a 10 foot pole. Of course, it
doesn't help that the product I burned out developing was a major
competitor of those languages and yet still exists as a commercial
product (after enabling me to retire, unlike the the developers of
those other languages, who got the fame instead ;-)
Now that I've finished the main Rails tutorial I'm considerably less
enamored of it. The main concept that comes to my mind is "bag on the
side". The whole development environment is really just a collection
of hacks, kluges, preprocessors, and workarounds instead of a unified
architecture (which is why I had to spend the first few hours in the
tutorial *fixing* version incompatibilities in it!). It's really
designed for professional full-time developers who must deal with a
lot of different projects for different clients and where a project
may get passed around among several developers not all of whom will be
familiar with the overall design. Which unfortunately doesn't really
describe this project. I've written to my old crew to see if maybe
it might be faster and easier to customize the old product (which is
now open source) than punish myself by doing it the Rails way, but
using Rails is still the backup plan.
Post by James HancockAlso, I have a hard time believing this would eclipse Facebook someday.
In
Post by James Hancockthe US you vote 1 time ever 4 years and how much time do they actually
research and post opinions about it? Not very much.
Maybe my FB "friends" are different, but although I'd agree that stuff
like cat videos and pictures of food are still the main fare, bitching
about governments, corporations, etc. make up a significant percentage
of the posts. Imagine what it would be like if they actually believed
that posting would *change* things...
Post by James HancockHow much do they hang
out and talk with their friends? A lot lot more. They even watch cat
videos
Post by James Hancockmore. I think a good measure of this is day to day how much does the
average
Post by James Hancockperson talk politics? If you can answer that, I think you would get a
good
--001a11c219ea98351004f82ab4d3
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
The reason you use python is because the development cycle can be so much faster. Not because it doesn't have it's weaknesses.?Where with rails, you sort of have to do it their way or it becomes very painful.<div><br> </div><div>Normally the logic part of web apps is pretty straightforward the hard part is making everything talk together over the web that is difficult(database, templating, user authentication, etc), and then there is?JavaScript(which I agree is?horrible) if you want to do anything client side.?Django makes that part very easy, and flask would be more bare bones if you wanted to do it more hands on. I would say at least read the django tutorial before you write it off. It sounds<span></span> like you understand code pretty well so I think you could build what you are saying pretty quick there.</div> <div><br></div><div>Then there is deploying it on some envoirnment to use it. Webfaction is good for this or you good go platform as a service. But you should decide where you are going to deploy before you build it so you can make sure of their?constraints or best practices they provide.<br> <div><br></div><div>@max platform as a service is awesome which is why I mentioned GAE, but there are many others, like heroku.?But, you have to build something to go on the platform first.</div><div><br></div><div>Cheers,</div> <div>James<br><div><span></span><div> <br>On Tuesday, April 29, 2014, max stalnaker <<a href="javascript:_e(%7B%7D,'cvml','***@gmail.com');" target="_blank">***@gmail.com</a>> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> <p dir="ltr">I suppose i might get into a tech religion argument instead but I instead will say the obvious:</p> <p dir="ltr">You need to make a guess about future tech to build your world domination app upon.</p> <p dir="ltr">You are not yet a cool kid.� Take a look at platform as a service�� There is a fine chance that an operation like Hereko? Would handle all your versioning issues and take care of the scaling from a few users to a few billion with you just writing checks. Cost will run from zero at one end to the GDP of several small countries at the other end.� I rather like them in particular because of ancient walk bys.� This is all cloud stuff of course and i am sort of puzzled about people talking about scaling to 10**9 users and not starting with choices that imply that capability.� </p> <p dir="ltr">Try this:� is there something in the non-existent� mission vision that implies a need to scale smoothly over 9 orders of magnitude with 5� 9's reliability?� Is a vote error of 10 million votes on all world referendii acceptable?</p> <p dir="ltr">Max</p> <div>On Apr 28, 2014 7:47 PM, "Scott Raney" <<a>***@metacard.com</a>> wrote:<br type="attribution"><blockquote style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
On Mon, Apr 28, 2014 at 5:26 PM, James Hancock <<a>***@gmail.com</a>> wrote:<br>
> Sorry I am on my mobile and hit send on accident.<br>
<br>
Yeah, those little buttons are a pain ;-)<br>
<br>
> Your technological ignorance is showing a little in your last post. Is what<br>
> I ment to say. You really wish we would have stayed with software<br>
> applications instead of moving to the serverside applications?<br>
<br>
Definitely. Going a little off topic here, but the browser-based<br>
architecture is incredibly limited, complicated, and brittle (meaning<br>
it breaks with every little change, like version upgrades).<br>
iPhone/Android is a little better because most things are apps and so<br>
the developer has more control, but *ideally* versioning and<br>
extensions would have been built into the OSs a long time ago so, for<br>
example, instead of JavaScript which is a complete POS (and everyone<br>
knows it which is why there are so many extensions and workarounds for<br>
it), apps could have been built in any scripting language and the<br>
necessary engines and libraries downloaded on the fly. This problem<br>
could have been solved at several different junctures (e.g., when<br>
Microsoft added automatically downloaded OS updates for Windows 95,<br>
when HTML was originally defined to purposely exclude interactive<br>
elements, or when Netscape chose to build in a weekend-hack scripting<br>
language instead of defining an open architecture, etc.) but due to<br>
shortsightedness and proprietary-think never has been. Like I said,<br>
if it the web had been *designed* instead of evolving with the bizarre<br>
set of constraints it has, I'd have a working voting system up and<br>
running in days or weeks whereas now it's going to take me months.<br>
<br>
As for "ignorance", I'll cop to that as, like I think I said, I burned<br>
out and retired from software development about 10 years ago and so am<br>
having to learn a lot about the current environment (and not liking it<br>
much at all). But I've been developing software from microcode and<br>
assembly languages through 3GLs and 4GLs up to AI languages since I<br>
was a kid (40 years ago) so I think I have a pretty good awareness of<br>
what works and what doesn't. And as for django/pyramid/flask or<br>
anything based on Tcl or Python, talk about "old school"! I wouldn't<br>
touch either of those languages with a 10 foot pole. Of course, it<br>
doesn't help that the product I burned out developing was a major<br>
competitor of those languages and yet still exists as a commercial<br>
product (after enabling me to retire, unlike the the developers of<br>
those other languages, who got the fame instead ;-)<br>
<br>
Now that I've finished the main Rails tutorial I'm considerably less<br>
enamored of it. The main concept that comes to my mind is "bag on the<br>
side". The whole development environment is really just a collection<br>
of hacks, kluges, preprocessors, and workarounds instead of a unified<br>
architecture (which is why I had to spend the first few hours in the<br>
tutorial *fixing* version incompatibilities in it!). It's really<br>
designed for professional full-time developers who must deal with a<br>
lot of different projects for different clients and where a project<br>
may get passed around among several developers not all of whom will be<br>
familiar with the overall design. Which unfortunately doesn't really<br>
describe this project. I've written to my old crew to see if maybe<br>
it might be faster and easier to customize the old product (which is<br>
now open source) than punish myself by doing it the Rails way, but<br>
using Rails is still the backup plan.<br>
<br>
> Also, I have a hard time believing this would eclipse Facebook someday. In<br>
> the US you vote 1 time ever 4 years and how much time do they actually<br>
> research and post opinions about it? Not very much.<br>
<br>
Maybe my FB "friends" are different, but although I'd agree that stuff<br>
like cat videos and pictures of food are still the main fare, bitching<br>
about governments, corporations, etc. make up a significant percentage<br>
of the posts. Imagine what it would be like if they actually believed<br>
that posting would *change* things...<br>
<br>
> How much do they hang<br>
> out and talk with their friends? A lot lot more. They even watch cat videos<br>
> more. I think a good measure of this is day to day how much does the average<br>
> person talk politics? If you can answer that, I think you would get a good<br>
></blockquote></div></blockquote></div>
</div></div></div>
--001a11c219ea98351004f82ab4d3--