tools for programming applets

[Andrew Thompson]

..95% of the people running your code are running microsoft browsers and
microsoft java implementations. ..

The last Microsoft 'Java' implementation I am aware of was the MSVM
that claimed to support 1.1.4. Sun sued them over it. Sun one.

That was all a *long* time ago. Which millennium are you trapped in?

 

[Lawrence D'Oliveiro]

jQuery ... does a lot of things wrong ...

So put your coding skills where your mouth is, and offer up something
better.

If you go to comp.lang.javascript you'll get quite negative opinion
on the usefulness of the jQuery library from the experts.

It still offers functionality in places where Java doesn’t even have places.

 

[Alessio Stalla]

In message


1TB drives can be had for pocket change these days.

What was the problem, again?

I'm not talking about disk space costs, duh! I'm talking about
complexity especially, and also time, bandwidth... you can't duplicate
a big DB at a snap of your fingers, and certainly I wouldn't do it
*just to test one applet*. Would you duplicate an arbitrarily big web
application to test the new version of a Flash app? Is simplicity not
a value anymore?

 

[Lawrence D'Oliveiro]

In message

I'm not talking about disk space costs, duh! I'm talking about
complexity especially ...

Rsync can replicate an entire file structure, no matter how complex.

, and also time, bandwidth... you can't duplicate a big DB at a snap of
your fingers ...

Considering how many days this thread has seen such pointless arguments back
and forth, you would have finished copying that terabyte database by now.

 

[Lew]

In message


Rsync can replicate an entire file structure, no matter how complex.


Considering how many days this thread has seen such pointless arguments back
and forth, you would have finished copying that terabyte database by now.

[APPLAUSE]

Lawrence, your reasoning is sound and your points are correct on this matter.

 

[Alessio Stalla]

In message
<884c9970-4a08-4606-8aff-a4b47e945...@q32g2000yqn.googlegroups.com>, Alessio
Stalla wrote:

Rsync can replicate an entire file structure, no matter how complex.

Considering how many days this thread has seen such pointless argumentsback
and forth, you would have finished copying that terabyte database by now.

[APPLAUSE]

Lawrence, your reasoning is sound and your points are correct on this matter.

Ok, I give up. Maybe I simply see things differently from the majority
of Java developers. I just cannot imagine to always have to clone a
server in order to test a client. Do you really think that e.g. Google
cloned the entire Youtube to develop their HTML5 player?

 

[Lew]

Ok, I give up.
Appropriate.

Maybe I simply see things differently from the majority of Java developers. I just cannot imagine to always have to clone a
server in order to test a client. Do you really think that e.g. Google
cloned the entire Youtube [sic] to develop their HTML5 player?

Irrelevant and false analogy. I conclude that you haven't a valid point if
you have to reach like that. You don't get to reframe the point in such a
ludicrous straw-man fashion. Well, you have a right to try, but it doesn't
change the truth that you are trying to resist.

No one suggested that cloning YouTube is necessary to test a browser. We did
suggest, correctly, that the applet security mechanism is designed to prevent
exactly what the OP requested, that there were good reasons for that, that
there were better ways to develop an applet that didn't have a problem with
the security restrictions, and that the overhead of setting up a single
development workstation with the resources to simulate the OP's production box
(a single node, a fact you overlooked in your eagerness to proffer useless
rhetoric) was minimal. All those things are true.

It is also true that Google does test even large-scale applications on a
separate infrastructure from its production servers for exactly the reasons
"the majority of Java developers", who actually know what we're talking about.
There's a reason why so many people agree with the truth, Alessio Stalla.
That makes your not-so-Ciceronian rhetoric work against the point you claim
you want to make and in favor of that made by the intelligent majority,
Alessio Stalla.

 

[Martin Gregorie]

I'm not talking about disk space costs, duh! I'm talking about
complexity especially, and also time, bandwidth... you can't duplicate a
big DB at a snap of your fingers, and certainly I wouldn't do it *just
to test one applet*. Would you duplicate an arbitrarily big web
application to test the new version of a Flash app? Is simplicity not a
value anymore?

Any self-respecting project will have a test version of the server on a
development box. This will be a fully functional, cutting edge version of
the server by definition and will usually only have a small database
since most developers generally don't put more data in it than they need
to support their test cases.

This is the server that should be replicated for client development, not
the production one. As a bonus, somebody has probably already needed to
replicate it in order to develop the existing client programs, so it may
also be easier to install on another development machine.

 

[Alessio Stalla]

Any self-respecting project will have a test version of the server on a
development box. This will be a fully functional, cutting edge version of
the server by definition and will usually only have a small database
since most developers generally don't put more data in it than they need
to support their test cases.

This is the server that should be replicated for client development, not
the production one. As a bonus, somebody has probably already needed to
replicate it in order to develop the existing client programs, so it may
also be easier to install on another development machine.

Oh, some sanity. I agree. Notice that this is different from the
advice that was given to the OP. Also, since the OP clearly does not
have such a development version of the server available (which is not
necessarily his fault), and he hasn't many resources since it's not
working for profit, "create a development server first" is not a good
answer, in my opinion, when easier alternatives exist (which are not
necessarily advisable in other scenarios, of course). I'm not
advocating against having a development version of the server! I
merely think that it's not an all-or-nothing situation.

Alessio

 

[Martin Gregorie]

Oh, some sanity. I agree. Notice that this is different from the advice
that was given to the OP. Also, since the OP clearly does not have such
a development version of the server available

How do you work that out? There may well be one, since I would hope that
any net-facing outfit would not be stupid enough to do development on or
against the live server. A development server may even be run on the live
server hardware if the owning organisation is really strapped for cash,
but they'd have to be really poor to be reduced to doing that: the box I
run my internal Apache, PostgresQL and mail servers on, a used IBM
NetVista, cost all of £100 back in 2004 and by now would be worth less
than the cost of hauling it away.

However, its not clear whether the OP has asked the right questions to
find out if a clonable development server exists and to get hold of a
copy if it does. All he has said shows a total unwillingness to even
consider doing so.

 

[Andreas Leitgeb]

I was looking to do some quick java development of applets. Here's my
situation:
1. I have a static server (ie: that I cannot touch) which serves my
client data (and applets).
2. a bare-bones client programming setup (vim and java compiler)
What I was hoping to do, therefore, is hijack the applets that are
coming from the server, and replace them with my own, compiled ones,
and hook the browser in such a way that when the applet is asked for,
my applet fires instead (hopefully in debugging mode) using the data
from the server as input.

If the static server is *not* using https, you could create some proxy
(transparent or not) that would intercept the requests for the applet-code
but pass through all the rest.

It seemed to me on overreading the thread, that this type of operation
was not intended by design, so it's *likely* (but not necessarily) so,
that no easier way exists.

There *may* even exist ways for an https-site, but I won't delve into
that. I might perhaps want to visit the USA some time in future and not
get arrested for some DMCA-BS applied to a triviality. It's just, that
if some web-developer stumbles over this thread, he should under no
circumstances come to believe, that "client side security" with applets
was anything else than utter insecurity.

 

[Michael Wojcik]

Why don't you run a local proxy (apache comes to mind but even
Tomcat/Jetty would do) for the site with some special rules to serve the
applet from where you have it and to replace the sites DNS with its
actual IP. Then map the DNS to localhost in /etc/hosts or
/Windows/System32/drivers/etc/hosts. Has worked fine with almost any
site for me in the past.

Was this really the only post in this interminable thread to suggest a
proxy?

To the OP: Look at your problem this way. You want a server you
control, so you can test your applet. You want the applet to be able
to request resources from that server. You don't want to duplicate the
production server.

So set up your local server to serve the applet from a local resource,
and to proxy everything else from the production server. A server can
provide some content without being the origin server for that content.
(HTTP specifies a number of proxying modes to solve problems like
this; see RFC 2616.)

The applet needs non-HTTP resources from the production server? Proxy
those as well. There are any number of generic TCP proxy utilities
floating around. I often use the TcpTunnel utility from the old Apache
Java SOAP toolkit, mostly because I've had it lying around for years
and there's been no reason to find or write another one, for this purpose.

Personally, I'd go ahead and duplicate the origin server, or a subset
of it, because that's usually the best option, for all the reasons
that have already been mentioned in the thread. But proxying is a
viable alternative.

 

[Stanimir Stamenkov]

Thu, 26 May 2011 11:35:55 +1200, /Lawrence D'Oliveiro/:

So put your coding skills where your mouth is, and offer up something
better.


It still offers functionality in places where Java doesn’t even have places.

You're obviously making silly remarks and no arguments. You've
stripped and haven't replied to my question "Which languages have a
jQuery equivalent?". I start to believe you have no clue what
you're talking about, really.

 

[Lawrence D'Oliveiro]

In message

... and he hasn't many resources since it's not working for profit ...

How “many resources†do you need? As I pointed out, even a terabyte database
doesn’t need expensive hardware to hold it these days.

 

[Lawrence D'Oliveiro]

Any self-respecting project will have a test version of the server on a
development box. This will be a fully functional, cutting edge version of
the server by definition and will usually only have a small database
since most developers generally don't put more data in it than they need
to support their test cases.

True, up to a point. You’ll probably need a full-size database to do
performance testing, that kind of thing.

And the most basic thing: testing what happens when you run out of space.

 

[Lawrence D'Oliveiro]

In message

Which millennium are you trapped in?

Funny, in this noisegroup, I get that feeling all the time.

 

[Lawrence D'Oliveiro]

In message

Sheesh. I swear that developers (myself included) have been in this
industry SO LONG that they take for granted the bulky, resource
intensive, extra-life-support-systems-required effort that programming
has become.

Who said it did?

 

[Lawrence D'Oliveiro]

Here's another way of looking at it: assume that readers only have the
one post to examine.

What a stupid assumption.

 

[Lawrence D'Oliveiro]

Thu, 26 May 2011 11:35:55 +1200, /Lawrence D'Oliveiro/:


You're obviously making silly remarks and no arguments. You've
stripped and haven't replied to my question "Which languages have a
jQuery equivalent?".

If you insist: “certainly not Javaâ€.

Own goal, anyone?

 

[Lew]

If you insist: “certainly not Javaâ€.

Own goal, anyone?

Once again, silly remark and no argument. The question is what languages
BESIDES Java DO have such a thing. The point is that the other languages
don't either, so there is no advantage to any language on that score. You
know that, you just don't have an answer to the valid point, so you engage in
your stupid nonsense, Lawrence.

Point goes to Stanimir, Lawrence. Next time, Lawrence, try addressing the
point instead of pulling these intellectually dishonest pony tricks, Lawrence.