Python--"Just a Language for Smart People"
zero
Python--"Just a language for smart people"
This is in response to the current discussion we are having regarding
"The Ultimate Python IDE"
A couple of weeks ago, someone posted "Prayer of a Python Newbie". A
post aimed at inspiring us to take our language to a higher level and
expanding it to a broader user base. I quickly jumped into the tread,
as did many others. After reading everyone's responses it seems no one
heard the prayer. Someone yesterday wrote a joke relating the recent
thread to an IQ quiz. The joke implied the discussions relating to IDEs
and Text Editors were for the simple minded, that real trial blazers
wouldn't waste time with such meaningless discussions. Many still
recommended Emacs (truly the most wonderful "TEXT EDITOR" ever
invented). Again, most of us completely missed the point of the
original post.
Prayer of a Python Newbie was a cry to the Python Gods (you know who you
are), to do something revolutionary.
Let's take a quick look at the current popular development tools.
__C/C++__
Used for everything--but isn't "everyone" trying to come up with
something better?
(Still the only choice for most application development)
__Visual Basic__
Microsoft's successful attempt to capture and win over a large and high
level developer base. (Implements a powerful IDE. If you can read its
manuals, it will pay your rent)
__Perl__
The "First tool to market" (remember “first” to market) to solve many
Unix shell level needs overcoming C/C++'s shortcomings of handling text.
(First tool to handle text well in the Unix environment)
__Java__
The pet project of a lot of smart people at SUN with a lot of free time
and nothing better to do. Stole the good stuff from C++ and left the
trash behind. "C++
Done Right" --Robert Kennedy, Santa Monica College 1997 (Right Place at
the Right Time)
__Python__
The world's best language for doing everything except telling the future
which it may soon be able to do. (Too difficult for new programmers who
need immediate industrial strength solutions. Possibly the most
powerful language in the hands of experts)
__Power Builder__
PowerSoft's better version of Visual Basic optimized for client server
development. (Strong user support, will be followed by PowerJ to keep up
with the growth of the internet. PowerJ is a full Java implementation.
The CASE tool functionality's of Power Builder has been ported to a Java
RAD tool. The tool is too new to have strong user support but should by
the end of the year as current Power Builder users convert.)
There are obviously more great tools and if I forgot your favorite tool
or crutch I apologize.
In response to some of my comments made following “Prayer of a Python
Newbie”, Andrew Kuchling wrote:
"No, what Python needs is more reasons for people to use
it,
and more applications written in it. For example, install the
Python
GTk module and write a GNOME application. Read the FSF's task
list,
or see what applications are requested in the Linux newsgroups,
and
write one of them. Or write some other interesting and useful
program,
like Gadfly's SQL database in Python, or the Mailman mailing
list
manager.
If people have a *reason* to use a language, they will
use it
regardless of whether there are fancy IDEs or not. IDEs will
appear
if people want them, but no one says "Hey, this language has an
IDE;
let's use it!""
I'm sure Andrew could shred on me and probably the rest of us in any
language, yet I disagree with his opinions. I believe that languages
grow and die based on the
ease of which they provide solutions. I believe that an advanced IDE
like Visual Basic's would give people “more reasons” to use Python. And
I believe more
applications would be written using Python if the standard install came
with a powerful IDE.
The new comer to Python is faced with the hassle of a somewhat difficult
install, considering all of the modules and extra interpreters and class
libraries that need to be separately downloaded and installed. Too many
external applications need to be separately installed in order to use
Python effectively.
I don't believe people will choose Python to develop just because there
are great programs written with it. There are a lot of great programs
written with COBOL, but that doesn't make me want to start developing in
COBOL. Sure, if something has already been written in COBOL then I may
need to extend it, however, I turn down COBOL programming jobs. I don't
want anything to do with COBOL. I believe shops chose languages based
solely on whether they provide easy solutions to their problems. Today,
this almost always means the inclusion of what Andrew Kuchling called a
“fancy IDE”. As a consultant, what matters to me most is being able to
solve my clients problems quickly and efficiently, not using Python or
Perl, or C++. I need a tool that solves problems quickly. A tool I
don't have to fight with just run a GUI “hello world”. Yes, I have run
the “Tkinter hello world”, but not without installing extra non-Python
utilities. For GUI widgets I want a tool that offers me built in drag
and drop functionality. I don't want to download anything extra. I
want everything in one build. For writing hooks into databases I want
built-in wizards complete with ODBC & JDBC (in JPython) drivers.
In the end it doesn't matter what language one uses, what matters is
solving problems quickly and easily. If something needs to be changed I
don't want to be intimidated. I want to be able to say “no problem,
I'll have that done ASAP. I care about meeting my deadlines.
__CASE Tools__
When I'm at a client's site, I want a tool that makes me look
competent. A tool that quickly solves the clients problems and gives me
plenty of time to test and
change implementations. I want an environment with context sensitive
help like the one found in Power Builder which allows me to cut and past
code from the help screens. I want to be able to say to a client "no
problem, I'll rewrite that for you before I go home today", and deliver
on my promises. I want to look good, more than I want to use Python or
any other language. No one in industry writes code with just a text
editor anymore. It's time for the Python community to wake up and
realize this. This is the nineties, CASE tools are as common to
developers as Nike shoes are to basketball players.
__Prototyping__ __
Python would be a great prototyping language if it weren't so hard to
get everything installed with all of the necessary external applications
and libraries. The standard install doesn't have enough power to solve
real problems. I believe an advanced IDE with a diverse built in
library would solve this problem.
Andrew Kuchling also said:
"Ha! Write Java code that does the equivalent of:
if hasattr(self, 'cmd_display'):
method = getattr(self, 'cmd_display)
apply(method, ("arg1", "arg2"))
With the reflection API, this is certainly possible (JPython
does it)
but it's nowhere near as compact and simple to read. You pay a
runtime penalty for this flexibility, but that's like any other
engineering tradeoff. Java isn't really either a scripting
language
or a rapid prototyping language."
First of all, as far as being easy to read I have no idea of what this
code does without comments. I'm sure it does something really amazing
for the amount of code
written, but if the above code were worth using it could easily be
imported as a class or package in any language. Tight code isn't as
important as code which is easy to write and debug. I'm sure Python is
the most efficient code, for almost anything line for line. However, if
I'm going against Sybase and I'm trying to create a view to help my
client coordinate "just in time inventory", I know Python wouldn't be my
first or second choice, even for prototyping, something Python is
proclaimed to excel in. I say this not because of the language, but
because of Python's lack of a powerful IDE and standard library set.
Actually the standard library is awesome, but there isn't a quick way of
implementing it. It requires too much look-up time. What a shame,
Python's library goes unused by so many.
I think Python is far too great a language not to be used more regularly
in business. I know its used in laboratories by the world's top
scientists, but unless
something is done to make Python easier to use, it will never be used by
anyone but computer-science geeks and world-class scientists.
I think if the Python Deities looked at Smalltalk and Oberon/Blackbox as
a starting point for a new IDE, they could develop something
revolutionary! Something that would shock the development community.
Perl was the first powerful full bodied scripting tool to hit the
market/public. Being the first has enabled Perl to become entrenched.
It's doubtful Python or any other language will ever catch up to it;
unless someone does something extraordinary. The same goes for TCL, who
wants to work with TCL when there is so much Perl to leverage. Java
wouldn't have made headlines if it hadn't introduced applets. Java
introduced something extraordinary, Python needs to do the same. The
Python user community needs to start a revolution. Just like Perl, just
like Java, otherwise we will remain a "small" band of brothers,
wondering why are language wasn't on the list in the next language
survey.
The computer was once Time Magazine's "Man of the Year". What will be
on the cover of January 1’s issue of Wired magazine in the year 2000?
Wouldn't it be awesome to see something like:
[ "JPython--an old language with a new following, squeezes the cost out
the E-Commerce". The Net Generations Language of Choice! Page 17 ]
Too many of us live in a small world. We fail to think outside the
box. Look to the future. What could Python become if we let ourselves
act on our dreams?
I believe the hour is now. I believe the call has gone out. Stop and
listen the the “Prayer of a Python Newbie”--give us an IDE that sends
Visual Basic, back to the
stone age. No more using Tkinter and borrowing libraries from our
scripting neighbors--we need to write our own code, our own libraries,
our own IDEs. We have to do something extraordinary. It's time to stop
answering the same tired questions about what our language can and can
not do, and start changing the way people use computers. It's time for
a Python IDE with drag and drop controls like Oberon’s. It's time for
all of the class libraries to be built into one standard development
environment like Smalltalk’s It's time for the Python community to
claim world recognition by changing the status quo like Netscape did
when it spawned a new net-society. It's time to impact the world with
the greatest language ever.
Until we become revolutionaries, Python will remain "Just a Language for
Smart People"--what a shame not to share computing's greatest secret
with the rest of the world.
I WANT MY PYTHON IDE
Consultant
*zERO*
Markus Fleck
zero <ze...@ecom.net> wrote:
> I think if the Python Deities looked at Smalltalk and Oberon/Blackbox as
> a starting point for a new IDE, they could develop something
> revolutionary!
IMHO, looking at Smalltalk and Oberon is amost the perfect
*counter-example*: they may have great IDEs, but (almost)
nobody uses them.
> [Perl]
> It's doubtful Python or any other language will ever catch up to it;
> unless someone does something extraordinary.
Right. But that most likely *isn't* an IDE, but rather a
cool, well-integrated library (such as the new XML library).
> give us an IDE that sends Visual Basic, back to the stone age. No more
> using Tkinter and borrowing libraries from our scripting neighbors--we
> need to write our own code, our own libraries, our own IDEs.
No. Two things:
1) There are literally dozens of IDEs for both UNIX and Windows that
support Python very well (including "on-line" help and sometimes
even class browsing)
2) IMHO, it's an example for a "cultural clash" between Windows and
UNIX users: Windows users want to have *one* IDE that integrates
everything at one place. UNIX people prefer to leave things modular
and be able to choose their favorite IDEs (Emacs, xwpe, nedit, etc.).
Unlike the "monolithic" Windows approach, the UNIX approach makes
it possible to integrate Python support into your *favorite* IDE,
not forcing you to learn an IDE just to use a certain programming
language.
So, while PTUI should definitely be enhanced to become even better (e.g.
by integrating the graphical Python debugger), an IDE isn't everything. It
*may* be more important for Windows people, but then it might be possible
to just integrate Python nicely into MS's "visual studio" and viola, done.
> It's time to stop answering the same tired questions about what our
> language can and can not do, and start changing the way people use
> computers. It's time for a Python IDE with drag and drop controls like
> Oberon's.
Well, sorry: this sounds too much just like marketing hype to me.
Also, the idea of IDEs isn't new, and there is *no* need to write an
IDE from scratch. Instead, existing popular IDEs should be configured
or extended to support Python better.
> Until we become revolutionaries, Python will remain "Just a Language for
> Smart People"--what a shame not to share computing's greatest secret
> with the rest of the world.
Hmm. Wouldn't the point of the vi/emacs joke be that
"smart people" don't bother to become language missionaries?
Yours,
Markus.
--
"Don't let the fact that it is impossible stop you from doing cool stuff."
Michael Toy, mt...@netscape.com, http://www.mozilla.org/apology.html
Historically Tcl has always stored all intermediate results as strings.
(With 8.0 they're rethinking that. Of course, Perl rethought that from
the start.) -- Larry Wall in <1997100717...@wall.org>
Mark Hammond
Markus Fleck wrote in message <6muqgi$15...@news.rhrz.uni-bonn.de>...
>zero <ze...@ecom.net> wrote:
>> It's time to stop answering the same tired questions about what
our
>> language can and can not do, and start changing the way people
use
>> computers. It's time for a Python IDE with drag and drop
controls like
>> Oberon's.
>
>Well, sorry: this sounds too much just like marketing hype to me.
>Also, the idea of IDEs isn't new, and there is *no* need to write
an
>IDE from scratch. Instead, existing popular IDEs should be
configured
>or extended to support Python better.
While I agree with Markus, I think the "marketting hype" is what is
necessary here. As Markus so rightly pointed out, Python _already_
has a number of IDE options for many platforms, but that is not the
point.
The coolest IDE on the planet wont make any difference until we
start beeping on peoples radar. A little bit of "marketting hype"
could be just the snake-oil we need to get the interest and funding
to get one of these :-)
Mark.
David Arnold
-->"zero" == zero <ze...@ecom.net> writes:
zero> <...> but unless something is done to make Python easier to
zero> use, it will never be used by anyone but computer-science
zero> geeks and world-class scientists.
if Python was a commerical software product, then all those license
fees would pay for developers to duplicate the functionality provided
by commercial IDEs. but python is free software, and you're expected
to do a little work in return for the benefits of that.
practically, i suggest the users wanting an IDE invest the time to
- install XEmacs (with OO-Browser), and the Python Info pages. you
don't get drag'n'drop GUI building, however ...
- write a drag'n'drop GUI builder for Tk (i'd suggest that if you're
prepared to organise this, you'll get help from a whole lot of
people).
perhaps you'd like to create a FAQ/HOWTO for those wanting to assemble
an IDE for Python, with instructions on how to install the required
stuff and configure it appropriately? you could even package it all
up and write an installer ...
-- David Arnold ,=================================================
=================' +617 33654310 (voice)
CRC for Distributed Systems Technology +617 33654311 (fax)
University of Queensland dav...@pobox.com (email)
Australia <http://www.pobox.com/~davida> (web)
free newton? the GNUton project http://www.pobox.com/~davida/gnuton/
Guido van Rossum
Funny you should address the Python Gods. I have no idea who we are :-).
Anyway. I was just saying to Jim Hugunin this afternoon, "What Python
needs is a good IDE." He responded, "What JPython needs is a good
IDE." (We also discussed how good Visual Basic's IDE is.) If you
read between the lines of Fredrik Lundh's web site, you'll read, "What
Python needs is a good IDE." Fredrik has been saying that for
several years, and is putting his money where his mouth is.
So I don't see where you get the impression that we don't care about
IDEs.
--Guido van Rossum (home page: http://www.python.org/~guido/)
Michael Scharf
zero wrote:
>
> I WANT MY PYTHON IDE
Use ours :-)
Have I ever mentioned, that the IDE SNiFF+ uses python as
internal scripting language? You can use SNiFF+ as development
tool for Python (debugger integrated) and you can program
and extend SNiFF+ with Python. In the next release 3.0 many
dialogs are written in python, ConfigManagemant Adaptors,....
See also
http://starship.skyport.net/crew/scharf/sn4py/
Michael
--
''''\ Michael Scharf
` c-@@ TakeFive Software
` > http://www.TakeFive.com
\_ V mailto:Michael...@TakeFive.co.at
Stefan Harms
Oh what is hope a powerful thing!
I have hope again, because some-one heard and understood the Prayer of a
Python Newbie. Apparently he does not claim to be one of the Gods, but at
least he is a Consultant:
Consultant *zERO* !
He wrote:
>Too many of us live in a small world. We fail to think outside the
>box. Look to the future. What could Python become if we let ourselves
>act on our dreams?
YES!!
>Until we become revolutionaries, Python will remain "Just a Language for
>Smart People"--what a shame not to share computing's greatest secret
>with the rest of the world.
YES!!
Talking of secrets, I wondered how you understand that a intuitive IDE
like Appware/Microbrew managed to stay hidden?
Is it purely the way that the company is managing it, does not advertise it?
Or is it because there is a "weak foundation" to an otherwise very
powerful idiom?
If you want to see what I mean, please look at some screen shots of
AppWare/MicroBrew:
http://www.mbnet.mb.ca/~harms/python/pythonprayer.html
Somehow, I suspect, Python could be the engine of a powerful tool - if I
could only SEE it!
--
Stefan Harms @ Squirrel's End
Tim Peters
> > I think if the Python Deities looked at Smalltalk and
> > Oberon/Blackbox as a starting point for a new IDE, they
> > could develop something revolutionary!
> IMHO, looking at Smalltalk and Oberon is almost the perfect
> *counter-example*: they may have great IDEs, but (almost)
> nobody uses them.
Ditto Interlisp long before, in the sense of WPE (wonderful programming
environment). "Revolutionary" anything scares people off! Better than,
e.g., improving on Visual Basic would be to make an exact VB clone -- a
trick Microsoft has pulled in the other direction more than once <0.6 wink>.
> Hmm. Wouldn't the point of the vi/emacs joke be that
> "smart people" don't bother to become language missionaries?
I mapped it like this:
IQ 172: Writes linear-time parser for van Wijngaarden's 2-level Algol68
grammar in head, which unfortunately runs in quadratic time.
IQ 103: Writes general continuation to continuation-passing-style
transformer in Scheme, which unfortunately fails when applied to itself.
IQ 51: Writes tic-tac-toe program in Perl because Tcl syntax was "too
rigid", which unfortunately gets in an infinite loop because it forgot to
ask for input.
IQ 0: Writes intelligent browser in Python after 5-minute introduction from
Fredrik, retires the next week off the proceeds, and spends the rest of his
life happily if moronically posting to comp.lang.python under the name of
"tim".
truth-is-stranger-than-fiction-ly y'rs - tim
RSteelandt
Dans l'article <3592E00E...@ecom.net>, zero <ze...@ecom.net> écrit :
>Python--"Just a language for smart people"
>
>This is in response to the current discussion we are having regarding
>"The Ultimate Python IDE"
>
Here are some interesting statements, alltough i'm do not completly agree with
them.
Here's my point of view:
I'm running a software development company, like a lot you here i suppose.
So what is the primary goal of it? It is to developpe professional software as
quick (wich means as cheap <G>) as possible.
To achive this, when we have a new project,the first thing to do is to choose
the development tool(s) wich we're gonna use to develope the project.
What i want to say here is, that we choose the development tool in function of
the specs of the project and NOT the other way around!!! When i read all the
mesages, style 'Python vs ....', 'Perl vs....' i have the impression that some
people here want to use a specific language for the sake of absolutly using it
!??
IMHO that can't be the goal !!
The major part of our projects are administratif software with heavy multi-user
database acces. I think everybody here agree with me that alltough we could do
it in python, it is certainly not the BEST tool to developpe this type of
software in it.
(We use powerbuilder for it, and i dont want to start a discusion here about
powerbuilder).
So how did we come to Python? I discovered it by accident together with a
installation of LINUX. Had a look at it and my first impression was, ok another
scripting language, forget it. Until a saw in a bookshop the book written by
Mark Lutz wich i bought. And there, i discovered that it was much more then
this. So started to learn it and using it for some inhouse tools.
A certain moment we did have an order to write a OLTP remote system wich
connects to a central system. And that wasn't obvious to do in powerbuilder.
After a discussion with our developpers the first conclusion was to write it in
C, until i said when we can write it in C why not write it in python? So it was
python. I don't want to explain here all the starting problems we had to do it
in python, especially to find all the librarys we needed. What i want to point
out here is the difficulty to find something when you need librarys (ex. serial
communication, ODBC....). I suppose nobody here wants to pay my internet phone
bill to find and collect it all <G>.
So i come to the point:
IMHO it's a more primary goal to set up a good info sheet for starters with
python then to discuse abouth IDE for python. Those IDE's exists, take ptui for
example, alltough it laks some features it is quit usable.
That info sheet should help starters to get up a development system wich can do
the most common development without browsing on the net for hours.
This could have the form of:
1. Where to find python (that everybody knows <G>
2. Databases :
- Different possibilitys (ODBC, gadfy,....) and WHERE to get them.
3. Other DLL calls
- Where to get them
4. Serial IO
5. GUI's, types,advantages of each and where to get them.
6. IDE's (<G>)
7....
That sort of info sheet will certainly help to get started with a development
in python and will avoid hours of searching on the net (in the starship p.e.,
do you know how many members are in there who wrote very good librarys, but you
have to find them !!!!)
If somebody has an idea how to setup this info sheet i'm willing to collaborate
(collect the info) for it, put it in a cetain form and ask the guys from
python.org to publish it somewhere.
PS : sorry for the spelling mistakes, i'm french so consider it as a accent <G>
(are there other french people here???)
Rony STEELANDT
e-mail : RStee...@aol.com
Greetings from france :)
Fredrik Lundh
> Someone yesterday wrote a joke relating the recent thread to
> an IQ quiz. The joke implied the discussions relating to IDEs
> and Text Editors were for the simple minded, that real trial blazers
> wouldn't waste time with such meaningless discussions.
(didn't that poster just refer to the "my text editor is better
than your text editor" part of that thread?)
Personally, I'm 110% convinced that a well-designed development
environment can add a whole new dimension to Python -- if that's
being simple minded, I'm proud to be simple minded.
And as Guido mentioned, we're trying to do something about it.
I'll follow up on his post later; in the meantime, take a look at:
http://www.pythonware.com/secretlabs/opal.htm
Btw, just sent copies of your post to every member of the Opal
development team. I think we all agree with everything you
wrote (you don't work here, do you?).
Cheers /F
fre...@pythonware.com
http://www.pythonware.com
"Requiring a separate class of programmers who translate user
requirements into algorithms is not only expensive but ultimately
frustrates our ability to make effective use of the technology. It
is this requirement for the specification of behaviors and effective
algorithms that is at the heart of the societal change. Just as there
weren't going to be enough phone operators, there aren't enough
programmers to add all the little bits of intelligent behavior we are
going to expect of the infrastructure."
-- Bob Frankston, in "Beyond Calculation: The Next
Fifty Years of Computing"
Jean-Claude Wippler
[Zero? Is that a boy's name or a girl's name? Sigh. Oh, well...]
FYI: the title of this post is what made me follow up, the language
comparisons, I can do without. There are too many toes to step on...
> [the recent post:] Prayer of a Python Newbie was a cry to the Python
> Gods (you know who you are), to do something revolutionary.
[... lots of languages, lots of toes ... no Tcl/Tk toes? ok, fine... ]
> I believe more applications would be written using Python if the
> standard install came with a powerful IDE.
Sounds logical to me.
> [...] I don't believe people will choose Python to develop just
> because there are great programs written with it.
Me neither.
> [...] As a consultant, what matters to me most is being able to
> solve my clients problems quickly and efficiently, not using Python or
> Perl, or C++. I need a tool that solves problems quickly.
And which is solid, portable in some cases, readable in the long run,
hookable to anything that happens to be of use (or already *in* use),
not evolving too fast (i.e. mature), and, and, and ...
> [...] Python is far too great a language not to be used more regularly
> in business. I know its used in laboratories by the world's top
> scientists, but unless something is done to make Python easier to use,
> it will never be used by anyone but computer-science geeks and
> world-class scientists.
There *have* been others with Truly Awesome Solutions to Truly Important
Problems - living on a Truly Beautiful Island of their own creation.
[... a plea for an IDE ... as a good way to reach out ...]
In the end, the acceptance of something new is determined by people who
don't know anything about it. In the case of Python, that excludes all
the regulars on this group. Wanna see Python reach the masses? Easy.
Just figure out who they are, and what meets their needs. Then bridge
the gap. THEIR way, because a handful of Python people are never going
to change the world. If you think otherwise, go back to your island.
Assuming of course, that "reaching the masses" is an important goal.
I'm very curious what the precise reasoning *is* for such an assumption.
So much from someone who has written no more than a few dozen lines of
Python code - a certified complete idiot on the Python scale... :)
Regards,
Jean-Claude
Piet van Oostrum
>>>>> Markus Fleck <fl...@informatik.uni-bonn.de> (MF) writes:
MF> zero <ze...@ecom.net> wrote:
>> I think if the Python Deities looked at Smalltalk and Oberon/Blackbox as
>> a starting point for a new IDE, they could develop something
>> revolutionary!
MF> IMHO, looking at Smalltalk and Oberon is amost the perfect
MF> *counter-example*: they may have great IDEs, but (almost)
MF> nobody uses them.
That is certainly not true for Smalltalk. It is used especially in Europe.
--
Piet van Oostrum <pi...@cs.uu.nl>
URL: http://www.cs.uu.nl/~piet [PGP]
Private email: Piet.van...@gironet.nl
John Grayson
> > [...] Python is far too great a language not to be used more regularly
> > in business. I know its used in laboratories by the world's top
> > scientists, but unless something is done to make Python easier to use,
> > it will never be used by anyone but computer-science geeks and
> > world-class scientists.
>
For the last 3-1/2 years, the Florida Dept. of Motor Vehicles in Dade,
Broward and West Palm Beach counties has been running an application
linking Voice Response and an X/Motif appointment booking system written
entirely in Python.
Other business-critical applications have been developed using Tkinter
(can't say where...).
Python was an effective development tool for these applications.
However, its not the tool that solves a problem - its the individual
using the tool.
-----
Original Message: http://www.findmail.com/list/python-list/?start=39270
Start a FREE e-mail list at http://www.FindMail.com/
Kevin Dahlhausen
"
>> Someone yesterday wrote a joke relating the recent thread to
>> an IQ quiz. The joke implied the discussions relating to IDEs
>> and Text Editors were for the simple minded, that real trial blazers
>> wouldn't waste time with such meaningless discussions.
>
>(didn't that poster just refer to the "my text editor is better
>than your text editor" part of that thread?)
Yes, my intention was to poke fun at the my-X-is-better-than-your-Y
for (X,Y) in (editors, languages, hand-tools, browsers, ...). Didn't
mean to offend anyone talking about IDEs.
I just realized I used the 'for' Python construct in this post. Oh
no.........
Richard Smol
In a previous article, ze...@ecom.net (zero) says:
>Many still
>recommended Emacs (truly the most wonderful "TEXT EDITOR" ever
>invented). Again, most of us completely missed the point of the
>original post.
Will you finally snap out of it. Calling Emacs a mere text-editor is like
calling Unix "a sort of MS-DOS". Sure, it can be used as a text-editor
only, but it's programmability make it reach far beyound that. If you
don't believe that, just look at the numerous elisp-modules and libraries
invented for it.
Writing a Python IDE would be a snap for Emacs (just look at what has been
done for Java).
Greetz,
RS
Vladimir Marangozov
[Fredrik]
> Personally, I'm 110% convinced that a well-designed development
> environment can add a whole new dimension to Python
Absolutely!
>
> And as Guido mentioned, we're trying to do something about it.
> I'll follow up on his post later; in the meantime, take a look at:
>
> http://www.pythonware.com/secretlabs/opal.htm
>
As soon as I knew about this project, I registered to the mailing list
because I'm particularily keen to see (and use!) what has been developed
in those cold Finnish labs. And the more we discuss this topic, the
more I have troubles to digest the time remaining to IPC7.
So, Fredrik, hurry up, please!! ;-)
Since I had similar ideas some time ago, but haven't found the time to
implement my thoughts since then (and even if I had the time, my
Tkintering is likely to be worse than F/'s), if there's a way to
contribute (and push the project) with something in a spare hour or two,
anonymously, with void claims in return, feel free to drop me a line ;-).
I'm unclear (understandably) on the features, but as I mentioned in
a previous thread in c.l.py, I'm willing to see some means to have
structured (global) vision of the code (the software system), and
facilities to manage this structure. If it isn't there yet, you can
expect some noise from me to include such facilities ;-)
[unrelated to the subject -- some thoughts on how I imagined my tool].
Among other cool stuff, proper to any modern IDE,
3 simultaneous views, kept always, tightly in sync:
(a) browsable component hierarchy - global view of all levels
(b) the current level of the hierarchy (box-and-line diagram, a la UML)
(c) program text
Thus, what I'd like to be able to do is:
- If I prefer to type (or annotate) my program, I use the textual view
(as I type, (a) & (b) are deduced, constructed and updated)
- If I prefer to design components quickly, I use (b)'s drawing facilities
(as I draw my box-and-line diagram, (a) is updated, (c) is generated)
- If I want to inspect the software system (conceptually), I use (a)
(as I walk, (b) & (c) show me the details)
This way I can choose my preferred way to design my app, I'll switch
views constantly and I'll use one view or another, depending on how fast
I can reach the design goal. The graphical views are essential, because
the human mind is visually oriented. We understand/explain better when
we draw pictures of the concept.
To get a rough idea on what the the component's hierarchy looks like,
you may think of the actual Python hierarchical namespaces. Each
scope can be considered as being a component.
[and now, forget about all this]
I'm almost sure, Pythonware has something even better! Good luck!
(counting days opally ;-)
--
Vladimir MARANGOZOV | Vladimir....@inrialpes.fr
http://sirac.inrialpes.fr/~marangoz | tel:(+33-4)76615277 fax:76615252
Andrew Kuchling
zero writes:
>language, yet I disagree with his opinions. I believe that languages
>grow and die based on the
>ease of which they provide solutions.
Certainly they do. But an IDE doesn't solve anything of
itself just by existing; it just makes it more convenient to use the
language. Would you use a COBOL compiler if it came with a really
cracking good IDE? No? To put it another way, IDEs decrease the
slope you have to climb to learn the language, but you still need a
reason to bother climbing the slope in the first place, no matter how
gentle it is.
Accordingly, I don't believe that having an IDE will magically
attract users to the language. They need a compelling reason to adopt
it in the first place; that reason may be simplicity or power or
existing interfaces to other components or to modify an existing
program that's very useful (how many people have learned Scheme just
to write GIMP plug-ins?), or whatever.
--
A.M. Kuchling http://starship.skyport.net/crew/amk/
It is not my intention to denounce modern education. If it is bad, it may be
said that all education is bad which is not self-education, and quite a lot of
self-education is going on today -- some of it in our schools, under the very
noses of the teachers!
-- Robertson Davies, _A Voice from the Attic_
zero
Andrew Kuchling wrote:
> zero writes:
> >language, yet I disagree with his opinions. I believe that languages
> >grow and die based on the
> >ease of which they provide solutions.
> Certainly they do. But an IDE doesn't solve anything of
> itself just by existing; it just makes it more convenient to use the
> language.
I guess I'm of the opinion that Python is much too good a tool, not to be made
more convenient for both development and end use. As it is, it's much too
difficult for me to configure one computer in order to use Python effectively let
alone configuring and maintaining a multitude of end user's machines spanned
across North America. The way it stands now, I'll never be successful
implementing Python software the way I want to (in the business community) given
the amount of overhead involved. In fact, I wouldn't want the task of supporting
it, given its current fractured state. I guess Andrew what I'm saying is, I wish
Python didn't YELL FREE WARE every time I showed it to someone. We both know what
can be done with it. The hard work contributed to its development deserves a
better front end. Something to reduce the "slope" not of learning so much as
using, maintaining and implementation. Perhaps my argument is falling on deaf
ears, perhaps Python was only meant to work on single user machines, one unique
build from another. Maybe Python will never be more than a second scripting
choice to perl. I hope not.
> Would you use a COBOL compiler if it came with a really cracking good IDE?
No?If it made working with COBOL easy I might not shy away from it the way I do.
The language itself has too many fundamental shortcomings in by today's standards,
I don't think any IDE could overcome them enough to make me want to use it.
However, I see Python from the opposite end. I see Python as this really awesome
framework that I want to use to reprogram everything just for the sake of
programming. Yet, there's that dame hassle of all the loose-leaf elements to deal
with when working with Python. Think how much time you put into your current
build. Like most of us, you have Python on two or three machines, each one
slightly optimized for something different. You know how much time it took to get
each one to its current state. What a tremendous hassle, programming shouldn't be
this way, at least not with a language designed to take the hassle out of
programming--don't you think?
> To put it another way, IDEs decrease the
> slope you have to climb to learn the language, but you still need a
> reason to bother climbing the slope in the first place, no matter how
> gentle it is.
Andrew, Python by nature reduces the "slope" by giving us an alternative to C/C++
and other heinous tools. I'm attracted to Python because it's easier to use than
any other language. My point is: why can't we build an IDE to complement what
Guido started?
> Accordingly, I don't believe that having an IDE will magically
> attract users to the language. They need a compelling reason to adopt
> it in the first place; that reason may be simplicity or power or
> existing interfaces to other components or to modify an existing
> program that's very useful (how many people have learned Scheme just
> to write GIMP plug-ins?), or whatever.
Once someone uses Python why would they use anything else? It's simply the best
tool for most RAD kinds of work. The way I see it, the overhead involved with
installing and applying its diverse and powerful library discourages folks from
staying with it. I think anyone after having tried it would have all the
compelling reasons needed. I think where you and I have been misunderstanding
each other has been over someone's desire to use the language. You seem to think
I'm arguing that Python needs an IDE to inspire people to use it. I believe
anyone in their right mind would choose Python wherever it made since, if it
weren't for all the hassles one needs to go through just to do anything useful.
Python feels too much like free ware. Why can't free ware install and lend itself
to "helping" people solve problems like regular over priced software. What I'm
saying is Python needs an IDE to keep people from turning away to other "less
superior languages" which provide solutions with "less overhead". I'd rather
work with "Visual Basic" for most projects because it's more of a "high speed low
drag" tool (due to its IDE & hooks to wintel) even though I hate it compared to
Python. Can someone help me with the point I'm trying to make.
Thanks
*zERO*
Dave Mitchell
On Fri, 26 Jun 1998, zero wrote:
> ... As it is, it's much too
> difficult for me to configure one computer in order to use Python effectively let
> alone configuring and maintaining a multitude of end user's machines spanned
> across North America. The way it stands now, I'll never be successful
> implementing Python software the way I want to (in the business community) given
> the amount of overhead involved. In fact, I wouldn't want the task of supporting
> it, given its current fractured state. I guess Andrew what I'm saying is, I wish
> Python didn't YELL FREE WARE every time I showed it to someone. We both know what
> can be done with it.
Ahh yes, the old flamefest heats up. So, can you explain what it
is that you've discovered is hard about "configuring and maintaining"
python on various machines? I've installed it tons of times on
various platforms and can't think of once that it was actually
more than just running configure and make. And if you're
distributing a commercial python app, wouldn't you give 'em
a frozen app or something so as to not require a python installation
at all?
>.... Perhaps my argument is falling on deaf
> ears, perhaps Python was only meant to work on single user machines, one unique
> build from another. Maybe Python will never be more than a second scripting
> choice to perl. I hope not.
uhh. single user machines? personally I thought it works on unix as
well as nt and dos...? your arguments are definitely not falling
or deaf ears, rather likely many people are reluctant to wade into
an obviously flameful subject as "here's the REAL reason python's not
popular." I think you got some decent responses early on from people
who agreed that an IDE would sure be a nice thing, if someone wrote it!
> .... Think how much time you put into your current
> build. Like most of us, you have Python on two or three machines, each one
> slightly optimized for something different. You know how much time it took to get
> each one to its current state.
umm.. you really have two or three differntly optimized versions of
python on different machines? aside from people who just can't
leave well enough alone (like M-A ;) why would you want to fiddle
with the interpreter itself? I sure can't think of even one time
i've played with anything except for C and python library modules..
I agree I get a little ugly when I have to recompile the C ones
for a newer version of python to get rid of the version mismatch
messages.. but you're not exactly talking about a daily occurrence
here.
...
> I'm arguing that Python needs an IDE to inspire people to use it. I believe
> anyone in their right mind would choose Python wherever it made since, if it
> weren't for all the hassles one needs to go through just to do anything useful.
> Python feels too much like free ware. Why can't free ware install and lend itself
> to "helping" people solve problems like regular over priced software. What I'm
> saying is Python needs an IDE to keep people from turning away to other "less
> superior languages" which provide solutions with "less overhead".
umm again. I always thought perl was one of the main competitors
for Python applications. Does perl have this drag and drop IDE you're
wishing for? Is that why it's so popularly used? Personally, I thought
it was because a) it's been around a lot longer and b) its already
installed in some version or other on prety much every unix machine
right out of the box. And B is starting to be true with Python
for example RedHat which uses it for some darn nice sysadmin tools..
Not sure your personal level of python experience, but all while
i've been reading this list it always seems like the people who
really havent spent much time with it are the most vocal about how
to do things like adding braces, removing space delimiting, needing
a drag-n-drop IDE, etc..
I agree, a nice IDE is a cool thing to wish for, and it must
be a cool thing to have if people like /F are actually doing it,
but how does that have anything to do with a language's popularity?
yours,
dave mitchell
da...@magnet.com
Mark Jackson
"Mark Hammond" <MHam...@skippinet.com.au> writes:
> The coolest IDE on the planet wont make any difference until we
> start beeping on peoples radar. A little bit of "marketting hype"
> could be just the snake-oil we need to get the interest and funding
> to get one of these :-)
"Python - why settle for snake oil when you can have the *whole* snake?"
--
Mark Jackson - http://www.alumni.caltech.edu/~mjackson
But that's the beauty of the game. At this very moment, your
absurd vicarious defeat is being perfectly counterbalanced by
some opposing fan's absurd vicarious triumph.
- Robert Mankoff
Damien Morton
I agree. IDE's dont make languages succeed in market and mindspace, but they
certainly do make things easier.
I started programming FORTRAN using punched cards, and I have to say that
Python, as a language, is certainly a a less painfull experience.
Nevertheless, banging Python in on a 80x25 DOS window, line by line, isnt
that much of an improvement in the experience.
Id have to say that the thing that makes Python attractive and easy to learn
is the immediacy with which you can get results: Any time you are curious
about a language feature you simply try it out and see what the result is.
In C and C++, the process is much more involved. Ideally an IDE would retain
that immediacy
Andrew Kuchling wrote in message
<13715.60256....@newcnri.cnri.reston.va.us>...
>zero writes:
>>language, yet I disagree with his opinions. I believe that languages
>>grow and die based on the
>>ease of which they provide solutions.
>
> Certainly they do. But an IDE doesn't solve anything of
>itself just by existing; it just makes it more convenient to use the
>language. Would you use a COBOL compiler if it came with a really
>cracking good IDE? No? To put it another way, IDEs decrease the
>slope you have to climb to learn the language, but you still need a
>reason to bother climbing the slope in the first place, no matter how
>gentle it is.
>
> Accordingly, I don't believe that having an IDE will magically
>attract users to the language. They need a compelling reason to adopt
>it in the first place; that reason may be simplicity or power or
>existing interfaces to other components or to modify an existing
>program that's very useful (how many people have learned Scheme just
>to write GIMP plug-ins?), or whatever.
>
Doug Stanfield
A.M. Kuchling wrote:
Accordingly, I don't believe that having an IDE will
magically
attract users to the language. They need a compelling reason to
adopt
it in the first place; that reason may be simplicity or power or
existing interfaces to other components or to modify an existing
program that's very useful (how many people have learned Scheme
just
to write GIMP plug-ins?), or whatever.
You're right, just having an IDE isn't compelling enough. But combined
with the other wonderful things about Python it will help make the
language accessible to another layer of users, a term I use
deliberately. I can't consider myself a programmer because that isn't
the primary thing I do at work. Its not even a secondary thing, but
every once in a while I need to hack together something that exceeds the
capabilities of wallowing around with MS Office apps. I'd like to use
Python for those things, Why? Because I recognize the elegant
simplicity and functionality. Because it has so many pieces that do so
many things. Because I (even I) can get things to work on multiple
platforms.
My problem is that I have to go through a learning curve every time I
pick it up and try to do something. Its usually important enough (and
Python learning curve shallow enough in the first place) that its worth
it. Having an IDE to act as the "crutch" I need to find all the docs
and appropriate functionality and organize my work would help me use
Python more for those things with less priority. That would be a big
win for me.
I think there are a lot of folks in similar straits. Most of the
constant contributors to this list don't need the same IDE crutch,
although it might make them even more prolific. But they're the tip of
the iceberg, and having a 'good' IDE will just remove another barrier to
acceptance for many more.
Looking forward to Opal with bated breath -
-Doug-
Doug Stanfield Oceanic Cable
Data Networking Manager 200 Akamainui St.
do...@oceanic.com Mililani, HI 96789
Tim Peters
[Damien Morton]
> ..
> I started programming FORTRAN using punched cards,
Same here. Hand the deck over the counter to the computer operator, and
come back 8 hours later to pick up a listing showing the syntax errors <0.1
wink>.
> and I have to say that Python, as a language, is certainly a less
> painful experience.
In some ways <wink>.
> Nevertheless, banging Python in on a 80x25 DOS window, line by line, isnt
> that much of an improvement in the experience.
I'm not sure that's Python's fault. For example, they have things called
"text editors" now that let you, in effect, create a whole deck of punched
cards *in software*! No cardboard, no sorting machines, and you don't even
have to punch sequence numbers on these virtual cards -- the text editor
remembers what order they came in all by itself. You store the virtual deck
in something called "a file" that lives on your disk drive, and Python can
be told to connect to this file and treat it just as if it were a deck of
input cards. It's like magic. And if you move up to Win95, you can even
get an 80x50 DOS box to run it in!
twice-the-fun-for-50x-the-memory-ly y'rs - tim
Jimmie Houchin
On Sat, 27 Jun 1998 04:05:11 GMT, Doug Stanfield <DO...@oceanic.com>
wrote:
>A.M. Kuchling wrote:
>> Accordingly, I don't believe that having an IDE will magically
>> attract users to the language. They need a compelling reason to
>> adopt it in the first place; that reason may be simplicity or power or
>> existing interfaces to other components or to modify an existing
>> program that's very useful (how many people have learned Scheme
>> just to write GIMP plug-ins?), or whatever.
>
>You're right, just having an IDE isn't compelling enough. But combined
>with the other wonderful things about Python it will help make the
>language accessible to another layer of users, a term I use
>deliberately.
><snip>
>
IDE may not be compelling enough. However to those of us who see the
solution to a problem. An IDE can enable us to provide that solution.
>My problem is that I have to go through a learning curve every time I
>pick it up and try to do something. Its usually important enough (and
>Python learning curve shallow enough in the first place) that its worth
>it. Having an IDE to act as the "crutch" I need to find all the docs
>and appropriate functionality and organize my work would help me use
>Python more for those things with less priority. That would be a big
>win for me.
I agree.
An IDE can act in many helpful ways.
1. As the crutch you describe above.
Or as I would prefer to call it an enabler.
It enables people to become immediately productive.
It enables much in the same way Python is an enabler over many
other (name your favorite) language(s).
Immediate productivity is very seductive.
Example:
My first experience in attempting to learn C++ (some people's
favorite language) was with a powerful IDE called Optima++ which
later became Power++. Optima/Power++ is a RAD C++, similar to
C++Builder which came later by Borland.
Optima++ enabled me to quickly develop some ideas I had.
It provided structure and support and basically allowed me
to use C++ to "glue" my app together. I provided the C++
logic to the event/functionality behind the GUI.
This is similar to VB, Delphi, etc. I just happened to choose one
where my language choice did not tie me down to Windows.
2. Productivity.
For those who do not "require" the structure and support of a
high quality IDE, there may be a tremendous productivity boost.
Yes, I understand this is debatable to some, but I don't find it
so. Productivity is an individual characteristic. People have
different productivity levels with different things. This is a
personal preference.
So goes the phrase, "Different strokes for different folks."
IDEs like computers, libraries (libs), modules, classes,
batch files, macros, etc. can boost productivity by automating
repetitive processes. An IDE is not the only way to automate
these processes as is adequately shown above.
However it does enable some to become more productive.
3. Leveraging.
Much like productivity. Leveraging enables those of us who
currently only aspire to become one of the "great ones" use the
wisdom, knowledge and understanding of those who have achieved
"greatness". Python as a language has accomplished much in this
area. Python with an IDE will enable those of us below to
accomplish many things "using other peoples knowledge".
Example: Opal
Opal will allow it users to utilize the knowledge of Fredrik, and
others at Pythonware.
How wonderful. :)
4. Structure and Support.
I've discussed this some above, but I believe a special treatment
is due.
In several threads in this newsgroup I have seen questions posted
as to, "How do I...?" Many of the replies have been, "How do you
want to...?", "If you come from 'x' (choose language) background,
you would do it as such, ..., if 'y' then ..., etc. For those who
are using this excellent language to learn programming, some
structure is beneficial. To those who are already proficient
programmers this capacity to program from multiple paradigms is
infinitely beneficial, however to one who is not it can sometimes
frustrate.
An IDE can provide a step by step programming process that can be
useful for enabling beginners and those early in the process to
get up to speed more quickly.
5. Templated Processes.
Templated Processes is a part of all of the above. It enables the
above to be realized. It can provide the structure and support
leveraging the wisdom, knowledge and understanding of the "great
ones" to enable those who are proficient increased productivity
and those who are learning, a crutch to lean on.
>I think there are a lot of folks in similar straits. Most of the
>constant contributors to this list don't need the same IDE crutch,
>although it might make them even more prolific. But they're the tip of
>the iceberg, and having a 'good' IDE will just remove another barrier to
>acceptance for many more.
Amen brother.
>Looking forward to Opal with bated breath -
Yes.
Currently Opal is targetted at Windows and Unix users. Mac users need
to speak up so that Fredrik knows there are those who care about Opal
for the Macintosh. I would even add Topaz to that request.
Mac users speak up.
I believe that a Python IDE should be a cross/multi-platform as Python
itself is.
And to balance out this post, a message from Guido:
I believe you understand and care. And many others too.
Well in order to stay brief I'll stop here. :) <wink>
Jimmie Houchin
Dirk Heise
zero wrote:
> I guess Andrew what I'm saying is, I wish
> Python didn't YELL FREE WARE every time I showed it to someone.
What exactly do you mean? Show it to whom?
- the programmers? Well, i am one, it was no problem to get started.
- the decision-makers? I have no clue how anyone can convince one of them
of anything. You?
- the customers of your python-driven app? They wouldn't really care.
Maybe you mean that not ALL kinds of programmers can be convinced easily.
There are the VB-like people who are really superficial and will not dwell
into any purely
text-oriented tool, usually, they don't even know grep.
Then, there are the tool-addicted like me. No mountain too high, no
problems.
(sees no problem, thinks for a while)
Alternative suggestion:
Why don't we collect the enthusiastic stories of newbies on the web site,
like
"Since i use Python, my life has changed completely. It was so easy to do,
i just did bla bla bla and it worked instantly and blablabla and now i know
VB suxxx."
Maybe we can fake the stories, i mean, all people selling p*nis enlargement
tools that turn out to be magnification lenses do it like that.
This would also make up some interesting how-to stories.
I'm the last to say IDEs are uncool, but Python can be tweaked with all
tools
that are around, so this is not exactly the stepping stone concerning the
subject of this message - rather, you sound to me like someone wanting
an "installation and maintenance assistance tool" and/or "HowTo Guides".
So, manuals. "Authentic" stories of people who managed to get happy
with Python. Before/After pictures.
If the python.org webmaster finds this idea interesting, i'd submit one,
if you want, i will it with the heaviest german accent in der world write.
You must know, that the german grammar very different word sequences
requires,
which always very nice looks even if the words are correct.
Oh, and we could buy advertisement space on p*rno sites.
Dirk
Tim Peters
[nick belshaw]
> ...
> Looking at this thread with everyone on the same side but
> still disagreeing ??????
I've noticed that too -- pretty weird.
> - seems to me that some don't want an IDE so they can keep
> the wonderful secret of Python to themselves <1.5wink>.
Am I being winked at? I can't take what I dish out, you know -- be kind.
I confess to not displaying proper enthusiasm for "IDEs are the answer!"
threads. That's just me, though. In my experience, IDEs are, more than
anything else, another layer of very complicated stuff that screws up when I
can least afford it -- whether it's InstallShield generating plausible but
incorrect boilerplate, or MSVC getting its baffling array of secret caches
out of synch and consequently fooling the debugger into lying about what's
actually happening. Honestly believe I've spent far more time worming
around bugs in IDEs than they've ever saved me.
BUT, and this is a biggie, since the beginning I've mostly used Python for
algorithm research. A class browser would be a minor aid in that, and a
supernaturally intelligent visualization package would be a huge aid, but
that's about it. There's nothing else cookie-cutter about it. So I wish
the Python IDE efforts all the luck in the world, but from a distance. In
return, I never ask that anyone show the slightest enthusiasm for anything
*I* do with Python -- not even myself, most days <wink>.
absence-of-enthusiasm-is-not-presence-of-hostility-ly y'rs - tim
Harri Pasanen
Dave Mitchell wrote:
>
> I agree, a nice IDE is a cool thing to wish for, and it must
> be a cool thing to have if people like /F are actually doing it,
> but how does that have anything to do with a language's popularity?
>
To my opinion this is really a marketing issue.
For newbies have some amount of programming experience,
especially those newbies coming from Windows environments,
an IDE is familiar concept.
If in a company there is one programmer interested in Python,
but the rest are at best disinterested, sometime apprehensive,
it is a marketing issue in the company. If Python has
a good programming environment, it is easier to sell the idea
of using it.
If a company is making the IDE and is expecting to make some
money out of it, they will devote some resources into marketing
it, which is also marketing for Python.
So even if the IDE is just wax and polish on a used car, it
is worth it. But a good IDE can be much more than that, an
actual tune up kit that will double your hp and make you the
coolest kid in the block ;-)
Just my 2 centimes,
Harri
Cameron Laird
In article <35975378...@news.bctel.ca>, <inv...@address.com> wrote:
.
.
.
>That is, it can serve as an application's macro language -- the role
>that Visual Basic for Applications is serving (and dominating) in the
>Win95 market; and a role that is a hodgepodge mess in the Unix world;
>and a role that I suspect is not being adequately serviced in the Mac
>world (unless Applescript is powerful enough).
Applescript isn't. Applescript has limited goals,
and, as usual, Apple has maintained sufficiently
tight control that Applescript hasn't grown past
them.
.
.
.
>It can serve as the Internet's server language -- the role played by
>CGI and Perl right now, and perhaps even the role performed by
>Javascript and Java. And for the mass Joe Blow, who wants to post web
I argue there's no profit in fighting JavaScript.
It's massively flawed, and will only be realist-
ically usable a few years from now, when ECMAscript
compatibility can safely be assumed, but it *will*
be the winner in that niche.
Java: Python complements Java. I'm not sure what
Java's real role is, but whatever it is, Python
complements it. Just keep repeating that.
>page pictures of his dog, it'd be nice for him to be able to use the
>same scripting language he's using in his word processor.
>
>And, finally, it can serve as as a full-fledged application
>programming language -- the role played by C++ and C in the main; and
>a handful of other languages in the minority. With hardware becoming
>senselessly fast, it might not even require a Python-to-machinecode
>compiler; it could remain interpreted.
Oh, heck; Python performs perfectly well for kajillions
of applications right now, on 80286-class machines.
Python performance should be a non-issue.
.
.
.
You left out GUI development, where Python has an inter-
esting story to tell, but one which the masses are little
inclined to hear.
--
Cameron Laird http://starbase.neosoft.com/~claird/home.html
cla...@NeoSoft.com +1 713 996 8546 FAX
Ivan Van Laningham
Hi All--
Tim Peters wrote:
>
[snip]
> Am I being winked at? I can't take what I dish out, you know -- be kind.
>
oh, <wink><wink><wink><wink><wink><wink><wink><wink><wink><wink><wink>
;-)
> I confess to not displaying proper enthusiasm for "IDEs are the answer!"
> threads. That's just me, though. In my experience, IDEs are, more than
> anything else, another layer of very complicated stuff that screws up when I
> can least afford it -- whether it's InstallShield generating plausible but
> incorrect boilerplate, or MSVC getting its baffling array of secret caches
> out of synch and consequently fooling the debugger into lying about what's
> actually happening. Honestly believe I've spent far more time worming
> around bugs in IDEs than they've ever saved me.
>
We're forced to use Borland C++ here at CallWare, in the interests of
maximal sausage extrusion. ... It was decided to take a large 16-bit
application and take it to 32-bit for Windows 95/NT. We expected
complications and gotchas and a Gorgo-sized Murphy incarnation, but we
did not expect that it would be 6 weeks (6 weeks!) before we could even
get the very first dialog onscreen. We spent four of those weeks
unsuccessfully trying to use debuggers, and the final two weeks
twiddling little tiny settings in the Borland's IDE that even Borland
didn't seem to know would send you down in flames if they weren't set
right. Naturally, Microsoft didn't have a clue. ...
[more snips]
> absence-of-enthusiasm-is-not-presence-of-hostility-ly y'rs - tim
<it-is-if-we're-talking-to-Microsoft-about-Borland>-ly y'rs,
Ivan ;-)
----------------------------------------------
Ivan Van Laningham
CallWare Technologies, Inc.
iva...@callware.com
http://www.pauahtun.org
----------------------------------------------
Damien Morton
Tim Peters wrote in message <000101bda18c$923ce680$8b9e2299@tim>...
>[Damien Morton]
>> ..
>> I started programming FORTRAN using punched cards,
>
<snip>
>>
>> Nevertheless, banging Python in on a 80x25 DOS window, line by line, isnt
>> that much of an improvement in the experience.
>
>I'm not sure that's Python's fault. For example, they have things called
>"text editors" now that let you, in effect, create a whole deck of punched
>cards *in software*! No cardboard, no sorting machines, and you don't even
>have to punch sequence numbers on these virtual cards -- the text editor
>remembers what order they came in all by itself. You store the virtual
deck
>in something called "a file" that lives on your disk drive, and Python can
>be told to connect to this file and treat it just as if it were a deck of
>input cards. It's like magic. And if you move up to Win95, you can even
>get an 80x50 DOS box to run it in!
I say "Lets move on from this punched card metaphor". You can try to hide it
behind fancy peices of software, behind editors and filesystems, behind
windows and mice and interpreters and super fast computers - but its STILL
there. You are still banging your code in - one line at a time, in sequence,
and seeing what happens when you run it.
It may not be Pythons fault, but... We are at the dawn of a new millenium,
and frankly 50s technology is kinda retro (by definition).
Anyone for Python parts and components. Python Beans? Visual Python?
Anyone know the story of the snake that swallowed an elephant?
Dave Kirby
>"Python - why settle for snake oil when you can have the *whole* snake?"
Great quote - it has just been added to my sig file.
Dave K
--------------------------------------------------
All great ideas start as heresy and end as dogma.
dkirby@ <-figure this out, spambots!-> Dave.Kirby@
bigfoot. My opinions are my own, psygnosis.
com but I'm willing to share. co.uk
Tim Peters
[Damien Morton]
> I say "Lets move on from this punched card metaphor". ...
> You are still banging your code in - one line at a time,
> in sequence, and seeing what happens when you run it.
Well, I tried banging it in two lines at a time, out of sequence, and that
didn't work nearly as well <wink>.
haven't-tried-a-big-ellipse-yet-though-ly y'rs - tim
Bill Anderson
Ivan Van Laningham wrote:
>
> Hi All--
>
> Tim Peters wrote:
> >
> [snip]
>
> > Am I being winked at? I can't take what I dish out, you know -- be kind.
> >
>
> oh, <wink><wink><wink><wink><wink><wink><wink><wink><wink><wink><wink>
> ;-)
>
You know, they can fix that twitch of yours with some good drugs ...
mostly. ;)
> > I confess to not displaying proper enthusiasm for "IDEs are the answer!"
> > threads. That's just me, though. In my experience, IDEs are, more than
> > anything else, another layer of very complicated stuff that screws up when I
> > can least afford it -- whether it's InstallShield generating plausible but
> > incorrect boilerplate, or MSVC getting its baffling array of secret caches
> > out of synch and consequently fooling the debugger into lying about what's
> > actually happening. Honestly believe I've spent far more time worming
> > around bugs in IDEs than they've ever saved me.
> >
>
I think having a good IDE would be an advantage to Python as a whole,
but for me, all I really need is somehting to golor highlight my syntax,
to keep my fingers honest. <wink> And Xemacs/emacs do that well enough
now, though I would like to see them recognize more keywords, and better
string parsing, but, as I said, it works well enough for me. I prefer to
learn how to create classes, etc. by hand, so I know just what the heck
I am doing!
pardon-my-usenet-piggybacking-ly y'rs
--
Bill Anderson Cyberhighway Internet Services, SYCON
"Pluralitas non est ponenda sine neccesitate."
Plurality should not be assumed without neccesity.
Tim Peters
[Bill Anderson]
When I was about 12, I used to hang out with my neighborhood buddy Bill
Anderson! We'd strip off our clothes, climb up a tree, and smoke cigarettes
he stole from his dad. Any relation? If you *are* him, you'll probably
deny it <wink>.
> I think having a good IDE would be an advantage to Python as a whole,
Same here! Especially if Fredrik is working on it == especially if it's a
really good IDE == especially if it's an IDE better than any I've ever used
<0.9 wink>.
> but for me, all I really need is somehting to golor highlight my syntax,
> to keep my fingers honest. <wink> And Xemacs/emacs do that well enough
> now, though I would like to see them recognize more keywords, and better
> string parsing,
Your *are* using python-mode.el, yes? Barry Warsaw can fix that. All he
needs is motivation. Like cash, or examples of problem cases.
> but, as I said, it works well enough for me. I prefer to learn how
> to create classes, etc. by hand, so I know just what the heck I am doing!
That's the problem. In almost every IDE I've used, the system screws up
soon after you leave the tutorial and try to do something the teensiest bit
different. Then you not only have to learn how to do it the low-level way,
but also master the bizarrely mechanical low-level ways in which the IDE
does it, *and* learn the secret and/or barely documented IDE hooks that
allow you to intervene at precisely the right moments to stop it from
screwing up. Then the next release of the IDE does it all in different
ways, and you're back at square zero <0.5 wink>.
Hmm. Now that I think of it, exactly the same arguments apply to every
compiler I've used. Therefore, from now on I'm sticking to assembly
language <wink>.
back-where-i-started-ly y'rs - tim
Barry A. Warsaw
>>>>> "TP" == Tim Peters <tim...@email.msn.com> writes:
TP> Your *are* using python-mode.el, yes? Barry Warsaw can fix
TP> that. All he needs is motivation. Like cash, or examples of
TP> problem cases.
You can keep the problem cases, I'll take the ca..., oops I mean the
other way 'round.
Seriously though, be sure you are using the latest version of
python-mode.el: <http://www.python.org/emacs/python-mode/>
-Barry
Dave Pawson
After debugging the given script, it works well. <0.something winks>,
then guessing what you have to do with custom.el,
and all its baggage.
It could be clearer....
[this under win95]
DaveP
Bill Anderson
Tim Peters wrote:
>
> [Bill Anderson]
>
> When I was about 12, I used to hang out with my neighborhood buddy Bill
> Anderson! We'd strip off our clothes, climb up a tree, and smoke cigarettes
> he stole from his dad. Any relation? If you *are* him, you'll probably
> deny it <wink>.
>
> > I think having a good IDE would be an advantage to Python as a whole,
>
> Same here! Especially if Fredrik is working on it == especially if it's a
> really good IDE == especially if it's an IDE better than any I've ever used
> <0.9 wink>.
>
> > but for me, all I really need is somehting to golor highlight my syntax,
> > to keep my fingers honest. <wink> And Xemacs/emacs do that well enough
> > now, though I would like to see them recognize more keywords, and better
> > string parsing,
>
> needs is motivation. Like cash, or examples of problem cases.
Got the new version ..mmmm works much better ....