Darren New <dne### [at] sanrrcom> wrote:
> outint'endian = little;

  Exactly which language is this? And have you actually corroborated by
examining the resulting machine code that it indeed does as you say?

-- 
                                                          - Warp

Post a reply to this message

On Wed, 12 Mar 2008 09:15:42 +0000, Invisible wrote:

>>>> But if I wanted it to work all the time, I'd write it in C and be
>>>> done with it - then the shell doesn't come into play. :-)
>>> Yah, as I said, when it gets messy, I use Tcl. :-)
>> 
>> Really the best option - remove the shell. :-)
> 
> Cool. I just hope your filenames don't contain any of the characters
> that Tcl considers to be "special" either. :-P

Well, I use C, and it deals with literal strings, but I suspect TCL is 
the same in that regard.

Jim

Post a reply to this message

Nicolas Alvarez wrote:

> And also one reason why command-line tools are better for many things. 
> (image editing probably not one of them)

Funny, there's a set of commandline tools for just that called 
ImageMagick. :D

Post a reply to this message

Mike Raiford wrote:
> Nicolas Alvarez wrote:
> 
>> And also one reason why command-line tools are better for many things. 
>> (image editing probably not one of them)
> 
> Funny, there's a set of commandline tools for just that called 
> ImageMagick. :D

Image "editing"? Or simple image transformation operations?

-- 
http://blog.orphi.me.uk/
http://www.zazzle.com/MathematicalOrchid*

Post a reply to this message

>> Cool. I just hope your filenames don't contain any of the characters
>> that Tcl considers to be "special" either. :-P
> 
> Well, I use C, and it deals with literal strings, but I suspect TCL is 
> the same in that regard.

Tcl encodes all data types as strings. I've been bitten by this 
before... You write a Tcl script, it works fine, and then it encounters 
some input that happens to contain a curly-bracket or a dollar sign and 
Weird Crap happens...

[I'm sure there's a way to avoid this. It's just tricky, and it makes me 
nervous.]

-- 
http://blog.orphi.me.uk/
http://www.zazzle.com/MathematicalOrchid*

Post a reply to this message

Mike Raiford escribió:
> Nicolas Alvarez wrote:
> 
>> And also one reason why command-line tools are better for many things. 
>> (image editing probably not one of them)
> 
> Funny, there's a set of commandline tools for just that called 
> ImageMagick. :D

Well yeah. But you can't do as much as you can in photoshop.

Can you make a good-looking POV-Ray scene *without ever looking at the 
result*? :)

Post a reply to this message

On Wed, 12 Mar 2008 19:02:00 +0000, Orchid XP v7 wrote:

>>> Cool. I just hope your filenames don't contain any of the characters
>>> that Tcl considers to be "special" either. :-P
>> 
>> Well, I use C, and it deals with literal strings, but I suspect TCL is
>> the same in that regard.
> 
> Tcl encodes all data types as strings. I've been bitten by this
> before... You write a Tcl script, it works fine, and then it encounters
> some input that happens to contain a curly-bracket or a dollar sign and
> Weird Crap happens...

That's one of the potential pitfalls with most scripting languages.

Jim

Post a reply to this message

Eero Ahonen wrote:
> Well ok, after that you wrote about chopping the file, which practically 
> is what split does, but it should be able to work without creating 
> millions of (temp)files in the middle of the process.

The problem is the size of the file. If I split it after the fact, I now 
have *two* files, the original and the new one. So now instead of half 
an hour of deleting directory entries, I have three hours of copying 
data, followed by five minutes of deleting one directory entry.

Split doesn't split a file. It writes new files, each of which contain 
pieces of the previous file.

-- 
   Darren New / San Diego, CA, USA (PST)
     "That's pretty. Where's that?"
          "It's the Age of Channelwood."
     "We should go there on vacation some time."

Post a reply to this message

Invisible wrote:
> Cool. I just hope your filenames don't contain any of the characters 
> that Tcl considers to be "special" either. :-P

I have, on occasion, got caught by a file whose name starts with ~. But 
none of the other characters are special in Tcl file names. In 
particular, spaces and brackets and such just don't come into play.

-- 
   Darren New / San Diego, CA, USA (PST)
     "That's pretty. Where's that?"
          "It's the Age of Channelwood."
     "We should go there on vacation some time."

Post a reply to this message

Orchid XP v7 wrote:
> Tcl encodes all data types as strings. I've been bitten by this 
> before... You write a Tcl script, it works fine, and then it encounters 
> some input that happens to contain a curly-bracket or a dollar sign and 
> Weird Crap happens...

Only if you don't treat your lists as lists and your strings as strings. 
Otherwise, the rules are extremely straightforward.

> [I'm sure there's a way to avoid this. It's just tricky, and it makes me 
> nervous.]

I never get bitten by this. The only time this happens is when you pass 
a string to something that indexes into a list, without first splitting 
the string into lists appropriately.

-- 
   Darren New / San Diego, CA, USA (PST)
     "That's pretty. Where's that?"
          "It's the Age of Channelwood."
     "We should go there on vacation some time."

Post a reply to this message

Warp wrote:
> Darren New <dne### [at] sanrrcom> wrote:
>> outint'endian = little;
> 
>   Exactly which language is this? 

Ada.  I am pretty sure it supports this.

> And have you actually corroborated by
> examining the resulting machine code that it indeed does as you say?

No. Never needed to. It's not the sort of thing you leave out of a 
language you're programming flight control software in.

-- 
   Darren New / San Diego, CA, USA (PST)
     "That's pretty. Where's that?"
          "It's the Age of Channelwood."
     "We should go there on vacation some time."

Post a reply to this message

Darren New wrote:
> Warp wrote:
>> Darren New <dne### [at] sanrrcom> wrote:
>>> outint'endian = little;
>>
>>   Exactly which language is this? 
> 
> Ada.  I am pretty sure it supports this.

It turns out the syntax is

    for My_Type’Bit_Order use High_Order_First;

(The ' indicates a compile-time value, in much the same sense that 
sizeof() is a compile-time function in C. Ada lets you read dozens of 
these and set others, like the address, volitility, atomicity, etc.)

It also turns out it's not *quite* so straightforward especially for . 
But clearly it's better to let the compiler do it in cases where it can 
than to code different routines with #ifdef in there.  Autoconf doesn't 
make C portable - it rewrites your code to be non-portable to that 
specific machine. :-)

-- 
   Darren New / San Diego, CA, USA (PST)
     "That's pretty. Where's that?"
          "It's the Age of Channelwood."
     "We should go there on vacation some time."

Post a reply to this message

>> Tcl encodes all data types as strings. I've been bitten by this 
>> before... You write a Tcl script, it works fine, and then it 
>> encounters some input that happens to contain a curly-bracket or a 
>> dollar sign and Weird Crap happens...
> 
> Only if you don't treat your lists as lists and your strings as strings. 
> Otherwise, the rules are extremely straightforward.
> 
>> [I'm sure there's a way to avoid this. It's just tricky, and it makes 
>> me nervous.]
> 
> I never get bitten by this. The only time this happens is when you pass 
> a string to something that indexes into a list, without first splitting 
> the string into lists appropriately.

I'm pretty sure I remember several occasions where if a certain 
expression happened to return only 1 item, it tried to interpret the 
item as a list. IOW, when the expression finds multiple matches, 
everything works fine, but when it finds exactly 1 match it malfunctions 
horribly. Took me ages to figure out why...

-- 
http://blog.orphi.me.uk/
http://www.zazzle.com/MathematicalOrchid*

Post a reply to this message

>> Tcl encodes all data types as strings. I've been bitten by this
>> before... You write a Tcl script, it works fine, and then it encounters
>> some input that happens to contain a curly-bracket or a dollar sign and
>> Weird Crap happens...
> 
> That's one of the potential pitfalls with most scripting languages.

Now you know why I dislike scripting languages in general. ;-)

-- 
http://blog.orphi.me.uk/
http://www.zazzle.com/MathematicalOrchid*

Post a reply to this message

Darren New <dne### [at] sanrrcom> wrote:
> Autoconf doesn't make C portable

  It only makes the C program to compile and work in a multitude of
different architectures. Yeah, it doesn't make it portable, only portable.

-- 
                                                          - Warp

Post a reply to this message

"Darren New" <dne### [at] sanrrcom> wrote in message 
news:47d6ca98@news.povray.org...

> Again, I tend to work with tens of thousands of files at once.

      Tens of THOUSANDS?? I'm not that shocked at what a programmer can do, 
but fair play for handling that.

       ~Steve~







>   Darren New / San Diego, CA, USA (PST)
>     "That's pretty. Where's that?"
>          "It's the Age of Channelwood."
>     "We should go there on vacation some time."

Post a reply to this message

St. wrote:
>       Tens of THOUSANDS?? I'm not that shocked at what a programmer can do, 
> but fair play for handling that.

Yeah. A couple weeks of raw audio, in 20-second chunks. Sadly, we throw 
most of it away. :-)

-- 
   Darren New / San Diego, CA, USA (PST)
     "That's pretty. Where's that?"
          "It's the Age of Channelwood."
     "We should go there on vacation some time."

Post a reply to this message

Invisible wrote:
> I'm pretty sure I remember several occasions where if a certain 
> expression happened to return only 1 item, it tried to interpret the 
> item as a list. IOW, when the expression finds multiple matches, 
> everything works fine, but when it finds exactly 1 match it malfunctions 
> horribly. Took me ages to figure out why...

You would need to either return a list, or return one item. If you want 
to return either, you need to return a list of one item. Otherwise, yes, 
if you sometimes return a list, and sometimes return one item that looks 
like a malformed list, then you're going to confuse yourself.

if {$count > 1} {return $a}
if {$count == 1} {return [list $singleton]}

# Wrong:
# if {$count == 1} {return $singleton}


-- 
   Darren New / San Diego, CA, USA (PST)
     "That's pretty. Where's that?"
          "It's the Age of Channelwood."
     "We should go there on vacation some time."

Post a reply to this message

Darren New escribió:
> Eero Ahonen wrote:
>> Well ok, after that you wrote about chopping the file, which 
>> practically is what split does, but it should be able to work without 
>> creating millions of (temp)files in the middle of the process.
> 
> The problem is the size of the file. If I split it after the fact, I now 
> have *two* files, the original and the new one. So now instead of half 
> an hour of deleting directory entries, I have three hours of copying 
> data, followed by five minutes of deleting one directory entry.
> 
> Split doesn't split a file. It writes new files, each of which contain 
> pieces of the previous file.

With dd you could copy any section of the file into a new one.

Post a reply to this message

Darren New wrote:
> Invisible wrote:
>> I'm pretty sure I remember several occasions where if a certain 
>> expression happened to return only 1 item, it tried to interpret the 
>> item as a list. IOW, when the expression finds multiple matches, 
>> everything works fine, but when it finds exactly 1 match it 
>> malfunctions horribly. Took me ages to figure out why...
> 
> You would need to either return a list, or return one item. If you want 
> to return either, you need to return a list of one item. Otherwise, yes, 
> if you sometimes return a list, and sometimes return one item that looks 
> like a malformed list, then you're going to confuse yourself.

If was more that I was processing a list, filtering out some of its 
content, and if it turned out that the result was exactly 1 element the 
script behaved in a different way than it did when the filtering yielded 
several elements.

But anyway, why worry? Just stop using Tcl. ;-)

-- 
http://blog.orphi.me.uk/
http://www.zazzle.com/MathematicalOrchid*

Post a reply to this message

Orchid XP v7 escribió:
> But anyway, why worry? Just stop using Tcl. ;-)

For some reason that reminded me of...

<FarSide> i had a guide that told u to put the 1.1.3 firmware file on 
the touch
<Brethil> It will be way faster even if you decrypt it on your computer 
and put the decrypted.dmg file on your iPod (should work, but I'm not 
100% sure
<FarSide> then run the upgrade
<PovAddict> FarSide: I did that to save myself from a hour-long download
<PovAddict> but the iPod still did decryption, extraction, patching, etc
<FarSide> i chose the ron popeal method...set it and forget it
<FarSide> i just went to bed
<FarSide> lol
<PovAddict> okay, time to try this again
<PovAddict> ah wait... to copy the firmware I need ssh
<PovAddict> to install ssh I need BSD subsystem
<PovAddict> BSD subsystem messes with Cydia (which I haven't installed yet)
<FarSide> yup
<PovAddict> or, I could install Cydia and get saurik's ssh
<Brethil> you can also use iPHUC
<Brethil> to copy over usb
<FarSide> i just suggest suicide
<FarSide> a lot easier
<FarSide> lol

Post a reply to this message

Nicolas Alvarez wrote:
> With dd you could copy any section of the file into a new one.

But again, it would take a long time to copy that.  The point is to not 
slog a half a terabyte of data from one disk to another in order to do 
the processing.

It works out OK. It just takes surprisingly long. :-)  It's automated at 
this point, unless I specifically take it off automation, so it's not 
like I'm waiting for it. It just kind of surprised me that deleting a 
big file could take over a minute. Must be the ext3 journaling or 
something going on. Ext was never really optimized for big files.

-- 
   Darren New / San Diego, CA, USA (PST)
     "That's pretty. Where's that?"
          "It's the Age of Channelwood."
     "We should go there on vacation some time."

Post a reply to this message

Darren New escribió:
> Nicolas Alvarez wrote:
>> With dd you could copy any section of the file into a new one.
> 
> But again, it would take a long time to copy that.  The point is to not 
> slog a half a terabyte of data from one disk to another in order to do 
> the processing.
> 
> It works out OK. It just takes surprisingly long. :-)  It's automated at 
> this point, unless I specifically take it off automation, so it's not 
> like I'm waiting for it. It just kind of surprised me that deleting a 
> big file could take over a minute. Must be the ext3 journaling or 
> something going on. Ext was never really optimized for big files.

What do you do with the small section once you cut it? Can't you use 
pipes to avoid so many temporary files?

Post a reply to this message

Warp wrote:
> Darren New <dne### [at] sanrrcom> wrote:
>> Autoconf doesn't make C portable
> 
>   It only makes the C program to compile and work in a multitude of
> different architectures. Yeah, it doesn't make it portable, only portable.
> 

That's the idea behind metaprogramming, after all.  You write something 
that gets turned into code for whatever your actual situation is :)

-- 
...Ben Chambers
www.pacificwebguy.com

Post a reply to this message

Nicolas Alvarez wrote:
> What do you do with the small section once you cut it? Can't you use 
> pipes to avoid so many temporary files?

We don't know which ones we're keeping until some time later. Normally 
about 10 hours later, if it's automated, or as much as a week if I let 
it pile up while I'm doing something else.

As for what we're doing with it, I'll have to defer answering that until 
our company goes "live" in a couple weeks.

http://www.skyclix.com/

-- 
   Darren New / San Diego, CA, USA (PST)
     "That's pretty. Where's that?"
          "It's the Age of Channelwood."
     "We should go there on vacation some time."

Post a reply to this message

Chambers wrote:
> That's the idea behind metaprogramming, after all.  You write something 
> that gets turned into code for whatever your actual situation is :)

And if the metaprogramming was built into the compiler, I'd agree that C 
is nicely portable.  :-)

C isn't portable. C + autoconf + the dozen tools that autoconf depends 
on + knowing all the arcane invocations you need to include to use 
autoconf is what's portable.

I mean, heck, 6502 assembler is portable, if you write a program to 
interpret it on a bunch of other machines, yes?

-- 
   Darren New / San Diego, CA, USA (PST)
     "That's pretty. Where's that?"
          "It's the Age of Channelwood."
     "We should go there on vacation some time."

Post a reply to this message

"Darren New" <dne### [at] sanrrcom> a écrit dans le message de news: 
47da9a40@news.povray.org...
> As for what we're doing with it, I'll have to defer answering that until 
> our company goes "live" in a couple weeks.
>
> http://www.skyclix.com/

According to Google it's about delivering targeted ads based on 
automagically recognised audio (rather than text) samples.

G.


-- 
*****************************
http://www.oyonale.com
*****************************
- Graphic experiments
- POV-Ray, Cinema 4D and Poser computer images
- Posters

Post a reply to this message

On Thu, 13 Mar 2008 09:28:04 +0000, Invisible wrote:

>>> Tcl encodes all data types as strings. I've been bitten by this
>>> before... You write a Tcl script, it works fine, and then it
>>> encounters some input that happens to contain a curly-bracket or a
>>> dollar sign and Weird Crap happens...
>> 
>> That's one of the potential pitfalls with most scripting languages.
> 
> Now you know why I dislike scripting languages in general. ;-)

They work for what they're designed for - I use awk quite a bit, perl 
occasionally, and shell scripts on rare occasions.

But if I wanted to write something that was usuable by a general computer 
user, I'd look to develop in a compiled language, as a general rule.

Jim

Post a reply to this message

>> Now you know why I dislike scripting languages in general. ;-)
> 
> They work for what they're designed for - I use awk quite a bit, perl 
> occasionally, and shell scripts on rare occasions.
> 
> But if I wanted to write something that was usuable by a general computer 
> user, I'd look to develop in a compiled language, as a general rule.

The kind of tasks I typically want to do involve heavy number crunching. 
Scripting languages aren't terribly suitable to this...

-- 
http://blog.orphi.me.uk/
http://www.zazzle.com/MathematicalOrchid*

Post a reply to this message

Gilles Tran wrote:
> According to Google it's about delivering targeted ads based on 
> automagically recognised audio (rather than text) samples.

Damn. So much for "stealth mode."  Even our partners are ratting us out. :-)

-- 
   Darren New / San Diego, CA, USA (PST)
     "That's pretty. Where's that?"
          "It's the Age of Channelwood."
     "We should go there on vacation some time."

Post a reply to this message

"Darren New" <dne### [at] sanrrcom> a écrit dans le message de news: 
47dadf76$1@news.povray.org...
> Gilles Tran wrote:
>> According to Google it's about delivering targeted ads based on 
>> automagically recognised audio (rather than text) samples.
>
> Damn. So much for "stealth mode."  Even our partners are ratting us out. 
> :-)

BTW I'm wondering what kind of audio input would be likely to cause a flow a 
Viagra ads via SMS. Will it recognise panting and moans? I can't wait for 
spammers to get hold of this technology ;)

G.


-- 
**********************
http://www.oyonale.com
**********************
- Graphic experiments
- POV-Ray, Cinema 4D and Poser computer art
- Posters

Post a reply to this message

Gilles Tran escribió:
> "Darren New" <dne### [at] sanrrcom> a �crit dans le message de news: 
> 47dadf76$1@news.povray.org...
>> Gilles Tran wrote:
>>> According to Google it's about delivering targeted ads based on 
>>> automagically recognised audio (rather than text) samples.
>> Damn. So much for "stealth mode."  Even our partners are ratting us out. 
>> :-)
> 
> BTW I'm wondering what kind of audio input would be likely to cause a flow a 
> Viagra ads via SMS. Will it recognise panting and moans?

Of course not. If there are moans, Viagra wasn't needed.

It will just recognize male voice saying "I swear this is the first time 
this happens to me". (or female voice saying "*that* is all?", etc; you 
get the idea).

Post a reply to this message

"Nicolas Alvarez" <nic### [at] gmailisthebestcom> a écrit dans le 
message de news: 47daea78$1@news.povray.org...
> Gilles Tran escribió:
>> "Darren New" <dne### [at] sanrrcom> a ?crit dans le message de news: 
>> 47dadf76$1@news.povray.org...
>>> Gilles Tran wrote:
>>>> According to Google it's about delivering targeted ads based on 
>>>> automagically recognised audio (rather than text) samples.
>>> Damn. So much for "stealth mode."  Even our partners are ratting us out. 
>>> :-)
>>
>> BTW I'm wondering what kind of audio input would be likely to cause a 
>> flow a Viagra ads via SMS. Will it recognise panting and moans?
>
> Of course not. If there are moans, Viagra wasn't needed.
>
> It will just recognize male voice saying "I swear this is the first time 
> this happens to me". (or female voice saying "*that* is all?", etc; you 
> get the idea).

Yes, but if the system is really smart it will detect the duration and 
intensity of the moans. For instance low intensity moans during less than 30 
s will get all your gizmos from cell-phones to PDAs to start flashing both 
penis enlargement and Viagra ads.

I just hope that the tech is smarter than Google Adsense, that still insists 
in displaying ads about "Free mature models of 1000s of Beautiful Mature 
Women" because my website contains 3D computer models.

G.


-- 
**********************
http://www.oyonale.com
**********************
- Graphic experiments
- POV-Ray, Cinema 4D and Poser computer art
- Posters

Post a reply to this message

"Darren New" <dne### [at] sanrrcom> wrote in message 
news:47d99d76$1@news.povray.org...
> St. wrote:
>>       Tens of THOUSANDS?? I'm not that shocked at what a programmer can 
>> do, but fair play for handling that.
>
> Yeah. A couple weeks of raw audio, in 20-second chunks. Sadly, we throw 
> most of it away. :-)

     Really? Well, that's still cool. What you've pointed out is still 
beyond my grasp, but I still admire what you guys and girls can do. Me, I'm 
just a computer user.  :o/   (Null points there then!)   ;)


          ~Steve~



>
> -- 
>   Darren New / San Diego, CA, USA (PST)
>     "That's pretty. Where's that?"
>          "It's the Age of Channelwood."
>     "We should go there on vacation some time."

Post a reply to this message

Darren New wrote:
> Chambers wrote:
>> That's the idea behind metaprogramming, after all.  You write 
>> something that gets turned into code for whatever your actual 
>> situation is :)
> 
> And if the metaprogramming was built into the compiler, I'd agree that C 
> is nicely portable.  :-)
> 
> C isn't portable. C + autoconf + the dozen tools that autoconf depends 
> on + knowing all the arcane invocations you need to include to use 
> autoconf is what's portable.
> 
> I mean, heck, 6502 assembler is portable, if you write a program to 
> interpret it on a bunch of other machines, yes?
> 

Absolutely, and I mean that quite literally.

It just goes to show that sometimes, our ideas of what words should 
mean, and the reality of the situation, are very different things.

After all, what is the *point* of portability?  Being able to use the 
same program on different machines with minimal work.  Sometimes the 
easiest way to do that will surprise you.

-- 
...Ben Chambers
www.pacificwebguy.com

Post a reply to this message

Chambers wrote:
> Absolutely, and I mean that quite literally.

I think portability isn't a binary quality.  You seem to be saying that 
any turing-complete language is equally portable because you could write 
an interpreter for it in some other language.  So why do people complain 
about C# being microsoft-only? :-)

-- 
   Darren New / San Diego, CA, USA (PST)
     "That's pretty. Where's that?"
          "It's the Age of Channelwood."
     "We should go there on vacation some time."

Post a reply to this message

Darren New wrote:
> Chambers wrote:
>> Absolutely, and I mean that quite literally.
> 
> I think portability isn't a binary quality.  You seem to be saying that 
> any turing-complete language is equally portable because you could write 
> an interpreter for it in some other language.  So why do people complain 
> about C# being microsoft-only? :-)
> 

Because they're lazy-a** b****es who would rather complain about ideals 
than get real work done?

After all, what if someone made a CPU that could execute Java bytecode 
directly?  (In fact, I think someone *has* done this)  At this point, 
the Java bytecode is now machine language.  Does that make it any less 
portable?

He**, is the Java bytecode any less portable for requiring an 
interpreter to begin with?

-- 
...Ben Chambers
www.pacificwebguy.com

Post a reply to this message

Chambers <ben### [at] pacificwebguycom> wrote:
> Darren New wrote:
> > I think portability isn't a binary quality.  You seem to be saying that
> > any turing-complete language is equally portable because you could write
> > an interpreter for it in some other language.  So why do people complain
> > about C# being microsoft-only? :-)
>
> Because they're lazy-a** b****es who would rather complain about ideals
> than get real work done?

No, I'd say it's because C# the language is about as useful as Java the
language:  not at all!  Both don't work outside their humongous associated
libraries/frameworks.

So, commiting to implemente said "languages" -- for them to be useful to
developers -- is commiting to implement not just a parser anb compiler but also
all the cruft people expect.  And guess what?!  C# is an ECMA standard, but not
..NET!  Thus, implementing a C# compiler by itself is not that useful.  Unless
you wish to follow a "standard" dictated by a single company with its
well-known forced upgrade cycle and be always left way behind.

Or simply ship a free compiler and help spread C# apps that require a licensed
framework to run.

Post a reply to this message

Chambers wrote:
> After all, what if someone made a CPU that could execute Java bytecode 
> directly?  (In fact, I think someone *has* done this) 

Actually, the Dolphin computer had programmable microcode and was used 
to run Smalltalk early on, interpreting Smalltalk bytecodes directly. So 
yah, like that.

-- 
   Darren New / San Diego, CA, USA (PST)
     "That's pretty. Where's that?"
          "It's the Age of Channelwood."
     "We should go there on vacation some time."

Post a reply to this message

Chambers escribió:
> After all, what if someone made a CPU that could execute Java bytecode 
> directly?  (In fact, I think someone *has* done this)

Yes, I think there are native CPUs for java bytecode; mainly for 
embedded devices where it's better to have a slow-ish CPU running Java 
directly than a fast CPU running a software interpreter (and sucking 
more power off the battery).

Post a reply to this message

nemesis wrote:
> No, I'd say it's because C# the language is about as useful as Java the
> language:  not at all!  Both don't work outside their humongous associated
> libraries/frameworks.

Personally, I'm looking into Erlang right now, thinking about how it 
might work for the code I have at work. Looks pretty cool, but it's one 
of those things that it seems most of the stuff makes sense at the 
fundamental assignment/arithmetic/functioncall level, and then there are 
libraries for things like "Build a release of several applications that 
let you upgrade a running system along with the distributed database 
schema without shutting it down." So like the "kernel" and "stdlib" 
documents are together 800 pages (something like 60 pages of one-line 
descriptions for routines in Kernel), and all the stuff in between is 
just huge.

Contrast with C#, where each library is pretty well distinct, for the 
most part. You can read about the regexp library without even knowing 
about the existence of the dynamic-code-loading library.

I think for anything particularly useful, you're going to have big 
libraries anyway. It's hard to build an "enterprise" class application 
without a whole bunch of stuff, so why not design a library for 
enterprise-class applications and actually use it in your system? C#, I 
think, started out with the idea, rather than building it up without 
extensive language support like Java did.

Got the book on order, tho, so that should help.

-- 
   Darren New / San Diego, CA, USA (PST)
     "That's pretty. Where's that?"
          "It's the Age of Channelwood."
     "We should go there on vacation some time."

Post a reply to this message

Darren New <dne### [at] sanrrcom> wrote:
> Personally, I'm looking into Erlang right now... the "kernel" and "stdlib"
> documents are together 800 pages (something like 60 pages of one-line
> descriptions for routines in Kernel), and all the stuff in between is
> just huge.
>
> Contrast with C#, where each library is pretty well distinct, for the
> most part. You can read about the regexp library without even knowing
> about the existence of the dynamic-code-loading library.

Why is it any different with the Erlang library?  I've just taken a look at
stdlib in http://erlang.org/doc/pdf/ and, though a big document, the various
modules it describes as part of the "standard library" are clearly all distinct
and separate from each other, with an occasional reference to related
functionality in other modules.

Perhaps what bothered you is that the unmaintained docs are kept in PDF files
without links rather than a cross-referenced tree-based Windows Help file?

> I think for anything particularly useful, you're going to have big
> libraries anyway.

I didn't say otherwise.  I said .NET is closed and non-standardized and C#
without it is like bread and no butter -- well, much worse actually...

> It's hard to build an "enterprise" class application
> without a whole bunch of stuff, so why not design a library for
> enterprise-class applications and actually use it in your system?

Actually, the thing that p*** me off about Java libs or .NET or even Python etc
is that they're all just a bunch of redundant -- but perhaps a bit more
convenient -- wrappers to existing OS functionality.  In the case of Java or
Python at least, this leads to cross-platform ease.  In the case of .NET, it's
just Windows2 on top of Windows1, really, for no reason other than captivating
an evergoing mass of programming enterprise drones that prefer to type
FileAccess.ChangeDirectory(dir) rather than cd(dir) or something...

Post a reply to this message

nemesis wrote:
> Darren New <dne### [at] sanrrcom> wrote:
>> Personally, I'm looking into Erlang right now... the "kernel" and "stdlib"
>> documents are together 800 pages (something like 60 pages of one-line
>> descriptions for routines in Kernel), and all the stuff in between is
>> just huge.
>>
>> Contrast with C#, where each library is pretty well distinct, for the
>> most part. You can read about the regexp library without even knowing
>> about the existence of the dynamic-code-loading library.
> 
> Why is it any different with the Erlang library? 

Look at anything other than the stdlib module.  I'm still wading thru 
kernel, but stuff like understanding the supervisor model, or the 
release application, or the mnesia stuff, requires a whole bunch of 
background.  (Not that it's hard to find the background. It's just 
there's a whole lot of it.)

C# probably has a lot of similar stuff, like when you're doing code 
generation and such. Understanding "assemblies" for example takes a lot 
of background, but doing something with assemblies that needs that 
understanding that stuff is unusual rather than step one.

In other words, I can't imagine why you'd want to use Erlang for a 
system other than a system that takes advantage of all the difficult 
stuff, while I can easily imagine using C# to do nothing more than a bit 
of simple XML parsing or some file management sort of thing.

It's entirely possible I'm reading Erlang docs in the wrong order. :-) 
I'm still trying to figure out where the parser is described. Various 
routines seem to imply you can add your own parsing to the language 
("behaviors" and "records" come to mind), as well as manipulating 
parse-trees, and I'm curious to see what that's all about.

> Perhaps what bothered you is that the unmaintained docs are kept in PDF files
> without links rather than a cross-referenced tree-based Windows Help file?

No, they have fine cross-references. It's just hard to know where to 
start. In retrospect, starting with kernel and stdlib is probably best, 
but I figured that out by reading the part in the release management app 
that said you had to include those in every release. :-)

>> I think for anything particularly useful, you're going to have big
>> libraries anyway.
> 
> I didn't say otherwise.  I said .NET is closed and non-standardized and C#
> without it is like bread and no butter -- well, much worse actually...

Yeah. And Erlang without the standard libraries (including gen_server, 
sasl, etc) isn't especially useful either.

> Actually, the thing that p*** me off about Java libs or .NET or even Python etc
> is that they're all just a bunch of redundant -- but perhaps a bit more
> convenient -- wrappers to existing OS functionality.

I would disagree. Certainly a large part of what you might use in some 
types of programs is like that. But if you're writing an ASP.NET 
program, most of the stuff you're doing has zero to do with the OS. 
Indeed, for that stuff, I understand that the executables are directly 
portable between MS's .NET and Mono.

> In the case of Java or Python at least, this leads to cross-platform ease.  

Yeah, right. Let me know when you have cross-JVM ease, let alone 
cross-platform ease, with Java.

-- 
   Darren New / San Diego, CA, USA (PST)
     "That's pretty. Where's that?"
          "It's the Age of Channelwood."
     "We should go there on vacation some time."

Post a reply to this message

Darren New wrote:
> nemesis wrote:
>> Darren New <dne### [at] sanrrcom> wrote:
>>> I think for anything particularly useful, you're going to have big
>>> libraries anyway.
>>
>> I didn't say otherwise.  I said .NET is closed and non-standardized 
>> and C#
>> without it is like bread and no butter -- well, much worse actually...
> 
> Yeah. And Erlang without the standard libraries (including gen_server, 
> sasl, etc) isn't especially useful either.

Yes, except Erlang's libraries aren't closed technologies dictated by a 
single monopolistic company trying to pass it as free and open by 
standardizing just the language and very basic library.  Anyone can 
benefit, anyone can contribute and anyone can use without fear of 
dependencies of closed technologies that change depending on 
market/financial mood.

>> In the case of Java or Python at least, this leads to cross-platform 
>> ease.  
> 
> Yeah, right. Let me know when you have cross-JVM ease, let alone 
> cross-platform ease, with Java.

I run most Java or Python programs unmodified both at the WinBox at work 
and at my Linux home box.  The guy coding them doesn't have to worry 
about hairy ifdefs or memory management to get it working on different 
platforms.  The only requirement is that he uses the standard 
cross-platform API and doesn't rely on specific platform services.

Post a reply to this message

nemesis escribió:
> I run most Java or Python programs unmodified both at the WinBox at work 
> and at my Linux home box.  The guy coding them doesn't have to worry 
> about hairy ifdefs or memory management to get it working on different 
> platforms.  The only requirement is that he uses the standard 
> cross-platform API and doesn't rely on specific platform services.

I saw a post in worsethanfailure of a Java method to copy files. It 
created a .bat file with a 'copy' command, then ran it, and finally 
deleted the .bat.

Post a reply to this message

nemesis wrote:
> I run most Java or Python programs unmodified both at the WinBox at work 
> and at my Linux home box. 

You're probably not doing too much hairy. I have to have three different 
JVMs installed at work to use three different libraries from three 
different partners.

I was always amused by the taglines that said things like
    100% Java! (Requires Java 1.5.2)

-- 
   Darren New / San Diego, CA, USA (PST)
     "That's pretty. Where's that?"
          "It's the Age of Channelwood."
     "We should go there on vacation some time."

Post a reply to this message

Nicolas Alvarez wrote:
> nemesis escribió:
>> I run most Java or Python programs unmodified both at the WinBox at 
>> work and at my Linux home box.  The guy coding them doesn't have to 
>> worry about hairy ifdefs or memory management to get it working on 
>> different platforms.  The only requirement is that he uses the 
>> standard cross-platform API and doesn't rely on specific platform 
>> services.
> 
> I saw a post in worsethanfailure of a Java method to copy files. It 
> created a .bat file with a 'copy' command, then ran it, and finally 
> deleted the .bat.

yes, even Java with all its stupid "enterprise" limitations is 
completely fool-proof!

Post a reply to this message

On Fri, 14 Mar 2008 18:38:46 +0000, Orchid XP v7 wrote:

>>> Now you know why I dislike scripting languages in general. ;-)
>> 
>> They work for what they're designed for - I use awk quite a bit, perl
>> occasionally, and shell scripts on rare occasions.
>> 
>> But if I wanted to write something that was usuable by a general
>> computer user, I'd look to develop in a compiled language, as a general
>> rule.
> 
> The kind of tasks I typically want to do involve heavy number crunching.
> Scripting languages aren't terribly suitable to this...

Scripting languages aren't designed for performance, as a general rule, 
so yes, that wouldn't be appropriate for a scripting language if you have 
complex calculations or lots of data to process.

Jim

Post a reply to this message

Jim Henderson wrote:

>> The kind of tasks I typically want to do involve heavy number crunching.
>> Scripting languages aren't terribly suitable to this...
> 
> Scripting languages aren't designed for performance, as a general rule, 
> so yes, that wouldn't be appropriate for a scripting language if you have 
> complex calculations or lots of data to process.

This is the problem with POV-Ray's SDL. People *have* built physics 
engines with it, but it's really not fantastically suitable for that task...

-- 
http://blog.orphi.me.uk/
http://www.zazzle.com/MathematicalOrchid*

Post a reply to this message

On Wed, 26 Mar 2008 11:12:04 +0000, Orchid XP v7 wrote:

> Jim Henderson wrote:
> 
>>> The kind of tasks I typically want to do involve heavy number
>>> crunching. Scripting languages aren't terribly suitable to this...
>> 
>> Scripting languages aren't designed for performance, as a general rule,
>> so yes, that wouldn't be appropriate for a scripting language if you
>> have complex calculations or lots of data to process.
> 
> This is the problem with POV-Ray's SDL. People *have* built physics
> engines with it, but it's really not fantastically suitable for that
> task...

Well, I don't know that I'd call it a problem per se, more of a 
limitation to be aware of.

Jim

Post a reply to this message