CURL.NET???

  • Thread starter Thread starter Conrad F
  • Start date Start date
C

Conrad F

Hi,

If any Microsoft people are listening....

Are there any plans for the new web language called "Curl" to be
supported in .NET (ASP.NET)? I ask as Curl represents the first step
to true OO programming of web pages and not just running "scripts"
which have in-roads into service DLLs of other languages such as C#.
Surely MS will put Curl into .NET at some time since it is designed to
work with DirectX and OpenGL?

Thanks for any info available on this topic.

Conrad
 
I'm just curious about your statements about Curl. I've never heard of it,
but you say it represents the "first" step towards "true" OO programming of
web pages.

ASP.NET "is" true OO web development, so what does Curl offer that we don't
already have?

Thanks.
 
heheeh My thoughts exactly.. C# is OO with some extras.. I am confused as
well..
 
From a quick glance, CURL appears to be targeted as more of a replacement
for HTML\Javascript than a replacement for ASP.NET, although it can run as a
general purpose language. However I don't think it has as much of a chance
to make it as, say, an xml based language would. I know its one I wouldn't
use(tho, I won't use html or javascript directly either if I can avoid it,
web languages suck in general).
However, more interesting, since I'm not a major ASP.NET user, what kind of
markup generation support does asp.net 2.0 have? Will it allow targetting
other markups(outside of (x)html)?

I could see this language being converted to .NET, although I would not
expect Microsoft to be the company that does it. I don't see a current
project, a place for CURL fans to provide something to their community,
mayhaps?
 
Scott M. said:
I find it hard to believe the HTML/JavaScript will be replaced by anything.

As do I, although they are both among the worst languages\frameworks I've
ever seen, but that isn't really the important issue, I suppose. Standard
and commonplace doesn't mean it is a good thing, just the usual thing.
 
Hi again,

I think Daniel O'Connell echos more my viewpoint. Somebody in this
thread said "what's wrong with HTML/JavaScript?" etc.... well, quite
simply why use more than one language if one will do? And as for an
earlier OO comment... why put your OO in DLLS and access objects by
VBScript or JavaScript when you can have one inline webpage completely
written in an OO language? Why be stuck with the DOM (which is an
object model yes, but not very customizable or extensible as it would
be in true OO fashion... what about being able to make your own and be
independent of a pre-existing DOM?

C#, when being used in ASP.NET will still generate HTML *script* and
so you have C# mixed with HTML, possibly the odd bit of JavaScript and
java, VBScript, Perl, PHP if you so choose - and somebody said this
was true OO? I think not. Think about it... would a C++ desktop
programmer write code and shell out to several other language because
you cant do something in C++? The desktop app world has been around
longer and has a certain "language maturity" about it... the web
doesn't. We use more than one web language because no single language
exists for the web that is ideal in all web programming scenarios -
then comes CURL as I understand it.

Hmm... I kinda think that the day I write a web page top to bottom in
a language without redirecting out to some other language will be the
day I agree web programming has come of age. I should be able to write
code the same as I would for a desktop app and say "I want this via a
web browser so it is web based" and without doing anything it will
work. CURL represents this.

Currently, if I create a label control in C# for ASP.NET, somewhere
some HTML will be generated and interoperability between the two
languages built. CURL aims to do away with this - you write one page,
in one language, no shelling out to something else and it is truly OO.
Of course you need a runtime in the browser, or the browser wont be
able to load the page, but I think the idea of one true OO language,
no need to know "scripting" languages (save XML) is a good one and if
a browser just needs a runtime (as does .NET with the CLR and Java
with its runtime) then why not CURL?

Regards

Conrad
 
Conrad F said:
Hi again,

I think Daniel O'Connell echos more my viewpoint. Somebody in this
thread said "what's wrong with HTML/JavaScript?" etc.... well, quite
simply why use more than one language if one will do? And as for an
earlier OO comment... why put your OO in DLLS and access objects by
VBScript or JavaScript when you can have one inline webpage completely
written in an OO language? Why be stuck with the DOM (which is an
object model yes, but not very customizable or extensible as it would
be in true OO fashion... what about being able to make your own and be
independent of a pre-existing DOM?

It forms conceptual OO, to the point where the developer doesn't have to(or
shouldn't have to) deal with what the output is. HTML is an output when you
start to consider ASP and JSP, not so much a language, just as jpeg, PSD's
and bitmaps are outputs for photoshop. Ideally, with a well written system
CURL could be an optional target language, I'm not familiar enough with asp
and its rendering engine to know.
C#, when being used in ASP.NET will still generate HTML *script* and
so you have C# mixed with HTML, possibly the odd bit of JavaScript and
java, VBScript, Perl, PHP if you so choose - and somebody said this
was true OO? I think not. Think about it... would a C++ desktop
programmer write code and shell out to several other language because
you cant do something in C++? The desktop app world has been around
longer and has a certain "language maturity" about it... the web
doesn't. We use more than one web language because no single language
exists for the web that is ideal in all web programming scenarios -
then comes CURL as I understand it.

The web has alot of problems, HTTP and html may well be at the root of it,
but changing that simply will not be an easy task. W3C standardization is
probably the first step, any tech that doesn't have that(and doesn't come
from MS or IBM) may well not have much of a chance.
Hmm... I kinda think that the day I write a web page top to bottom in
a language without redirecting out to some other language will be the
day I agree web programming has come of age. I should be able to write
code the same as I would for a desktop app and say "I want this via a
web browser so it is web based" and without doing anything it will
work. CURL represents this.

Except whats the point in writing the same for a desktop app if you can't
use it as a desktop app? HTML is a markup language that effectivly describes
applictions, it isn't the best at it but it is what it does, CURL is another
representation of the ideal, not nessecerily the best one.
Currently, if I create a label control in C# for ASP.NET, somewhere
some HTML will be generated and interoperability between the two
languages built. CURL aims to do away with this - you write one page,
in one language, no shelling out to something else and it is truly OO.
Of course you need a runtime in the browser, or the browser wont be
able to load the page, but I think the idea of one true OO language,
no need to know "scripting" languages (save XML) is a good one and if
a browser just needs a runtime (as does .NET with the CLR and Java
with its runtime) then why not CURL?

Don't know, unless its considered a standard I doubt it has much of a
chance. Are there other competing languages or frameworks targetted at the
same goal?
 
Daniel,

I take on board some of what you state about HTML just being an output
language. However I would not agree it "descibes an application" since
it cannot describe object dependencies or even define a class unless
using custom DLLs and JavaScript (or something else that can), which
was much my point. HTML has always been just language to describe how
something is presented in a web page - it has little other ability.

A webpage cannot be just JavaScript or any other langauge other than
HTML, it must start with HTML...and we use background processors like
ASP and PHP to pre-process and add deterministic logic ability to
HTML. Whatever a web page extension may be...ultimately what is
EXECUTED by the browser is HTML as can be witnessed by viewing the
source of any web page you navigate. Another language has had to be
used to deterministically generate the appropriate HTML. It would be
good if this was not the case. The appearance is a web-application,
but the real application is the logic in the other languages being
used behind the scenes and not the HTML.

I am not the most familiar with ASP.NET either, just a C#/VB.NET
developer writing distributed .NET systems whereby there is some
overlap with respect to security and typical problems of accessing
data across global networks.

I don't know if any other languages are aimed at sorting out web
langauge problems apart from CURL.

I agree it would take some backing to get a change - indeed from the
W3C! In this case, CURL is under development from a team which is
headed-up by Tim Berners-Lee and we know what he is known as the
father of! If he sees Internet v2 using a full-blown OO language and
not the current "pick and mix" approach, then I am with him on that
one, and he carries weight with the W3C. :)

I think we should be able to use one language targeted at a browser
whereby the code (not a "script") is delivered to the browser and run
by just-in-time compilation through a runtime. How that looks in a
browser depends on many things and is determined by the logic the
browser processes and not some powerful application tier sitting
behind a row of web-servers in a central location. The client browsers
have powerful machines these days - when browsing that power is
underutilized and the problems of concurrency can stress web-servers
which must develop pages just-in-time by generating them through ASP
or PHP an the ilk. In the CURL approach, the deterministic logic would
reside at the browser end. Current PCs are more than capable of this
and it would free up the processing power in corporations for other
things than running ???,000 users web-applications on some machine in
a data center.

I for one, will be looking to learn Curl as soon as I can. There are a
few good books on the subject that were lurking around as soon as
early as the fall of 2001. I very much like the concept - you can
probably tell :) All I actually am interested in is if Microsoft are
going to make it part of their .NET strategy and allow to become part
of the VS.NET IDE - if not, I may end up having to learn someother
companies IDE - in the early stages of CURL this is a must anyway. But
VS.NET is most things to most PC developers and if Curl takes off, I
think MS will take it on board as part of their standing "embrace and
extend" policy.

Conrad
 
Conrad F said:
Daniel,

I take on board some of what you state about HTML just being an output
language. However I would not agree it "descibes an application" since
it cannot describe object dependencies or even define a class unless
using custom DLLs and JavaScript (or something else that can), which
was much my point. HTML has always been just language to describe how
something is presented in a web page - it has little other ability.
Sorry, I should have said It describes an application interface, when used
for devising applications. Javascript provides some functionality, but the
constant postback for processing seems silly to me.
A webpage cannot be just JavaScript or any other langauge other than
HTML, it must start with HTML...and we use background processors like
ASP and PHP to pre-process and add deterministic logic ability to
HTML. Whatever a web page extension may be...ultimately what is
EXECUTED by the browser is HTML as can be witnessed by viewing the
source of any web page you navigate. Another language has had to be
used to deterministically generate the appropriate HTML. It would be
good if this was not the case. The appearance is a web-application,
but the real application is the logic in the other languages being
used behind the scenes and not the HTML.

I am not the most familiar with ASP.NET either, just a C#/VB.NET
developer writing distributed .NET systems whereby there is some
overlap with respect to security and typical problems of accessing
data across global networks.

I don't know if any other languages are aimed at sorting out web
langauge problems apart from CURL.

I agree it would take some backing to get a change - indeed from the
W3C! In this case, CURL is under development from a team which is
headed-up by Tim Berners-Lee and we know what he is known as the
father of! If he sees Internet v2 using a full-blown OO language and
not the current "pick and mix" approach, then I am with him on that
one, and he carries weight with the W3C. :)

I think we should be able to use one language targeted at a browser
whereby the code (not a "script") is delivered to the browser and run
by just-in-time compilation through a runtime. How that looks in a
browser depends on many things and is determined by the logic the
browser processes and not some powerful application tier sitting
behind a row of web-servers in a central location. The client browsers
have powerful machines these days - when browsing that power is
underutilized and the problems of concurrency can stress web-servers
which must develop pages just-in-time by generating them through ASP
or PHP an the ilk. In the CURL approach, the deterministic logic would
reside at the browser end. Current PCs are more than capable of this
and it would free up the processing power in corporations for other
things than running ???,000 users web-applications on some machine in
a data center.

I for one, will be looking to learn Curl as soon as I can. There are a
few good books on the subject that were lurking around as soon as
early as the fall of 2001. I very much like the concept - you can
probably tell :) All I actually am interested in is if Microsoft are
going to make it part of their .NET strategy and allow to become part
of the VS.NET IDE - if not, I may end up having to learn someother
companies IDE - in the early stages of CURL this is a must anyway. But
VS.NET is most things to most PC developers and if Curl takes off, I
think MS will take it on board as part of their standing "embrace and
extend" policy.
It depends, CURL seems to be aimed pretty well against MS's current fat
client designs. I don't know if they will support a technology, atleast in
its infancy that could undermine their overall strategy, it would simply be
bad business practice.
Also, I'd MUCH rather write in C# instead of CURL, so given the choice I'd
produce a easy to distribute fat client(or even an ASP.NET page) before I
wrote anything in CURL, but thats me, and I really dislike the web(when
viewed as http & html) as a whole, ;).

However, to find out, post to the asp.net group and ask about the output
targetting possibilities of asp.net and asp.net 2, that would be your first
step into finding out if any support is planned\possible. I would suspect
CURL, like HTML, would be written more often as a generated language than by
hand. One problem I have with it is its not sgml derived and that means web
pages no are longer parsable with any existing parsing tool, let alone
xml(which everyone seems to be flocking too), there is also the issue of
IP(if the entire program is sent in source form to the browser, there is no
level of protection provided), others may exist and it is certainly not my
area of expertise. The asp.net people would know more about the reality and
the theory than I do, I'm quite sure.

Also, if it is possible to output to arbitrary markups, you could always
consider figuring out how to write a CURL outputter and help the community,
;).
 
Back
Top