Project dream

[Will Stuyvesant]

Suppose you have the time and the money to start a new project in
Python. What would you like to do?

I can think of:

- A civilization like game in Python, with multiplayer support via
twisted.

- An easy to use tool for drawing diagrams, typically various kinds of
arrows and circles and boxes, that produces nice .eps and .svg files.

- A roguelike in Python. Since there is still no portable curses, it
needs WConio or something like that, and also multiplayer would be
great.

- Something for weblogging and todo things, probably via CGI.

What would your favorite be?

 

[Andrew Dalke]

Will Stuyvesant:

Suppose you have the time and the money to start a new project in
Python. What would you like to do?

An integrated chemical/bioinformatics development and
exploration environment. Got a few million dollars to fund me?
(Actually, more like 10 million, but bootstrappable with only
a few million.)

- A civilization like game in Python, with multiplayer support via
twisted.

There's been a few civ clones -- I recall playing one in the
mid-90s for Python using CLIPS for the AI. Don't recall the
name now. There's also openciv (in Python, nearly complete,
no longer active) or freeciv (in C, active). What would the
advantage be to writing yet another clone?

- An easy to use tool for drawing diagrams, typically various kinds of
arrows and circles and boxes, that produces nice .eps and .svg files.

What about Sketch?
http://sketch.sourceforge.net/

- A roguelike in Python. Since there is still no portable curses, it
needs WConio or something like that, and also multiplayer would be
great.

Why does that need to be rewritten in Python? As I understand
it, the C version is very portable and may simply need just bindings
for Python.

- Something for weblogging and todo things, probably via CGI.

Aren't there a few dozen of those already? Falls into the category
of easy to write but without one clear way to do it. So there are
a lot of different implementations, all different, all focused on solving
the given author's needs.

What would your favorite be?

More important, what would *your* favorite be. It looks like
you want to do a project but don't know which one to focus on.
My answer then is to do any of these projects; they are all great
ones to learn how to do larger, more useful projects.

Andrew
(e-mail address removed)

 

[Ray Smith]

Suppose you have the time and the money to start a new project in
Python. What would you like to do?
[snip]

What would your favorite be?

I'd like to develop some business type of app, eg:

* like a personal accounting system,

* followed by commercial accounting,

* POS,

Python still "seems" to lack "mature" business applications.

Regards,

Ray Smith

 

[Rene Pijlman]

Will Stuyvesant:

Suppose you have the time and the money to start a new project in
Python. What would you like to do?

I was thinking of an open source personal information management
application, to store notes, knowledge, hyperlinks, snippets, documents
and such in one flexible integrated information structure. All the things
that now end up in unrelated documents in different folder hierarchies.

Something with the flexibility and visualisation of the Brain
(www.thebrain.com), the structuring capabilities of InfoHandler
(www.mdesoft.com/english.htm) and a Googleish search engine.

 

[Aahz]

I was thinking of an open source personal information management
application, to store notes, knowledge, hyperlinks, snippets, documents
and such in one flexible integrated information structure. All the things
that now end up in unrelated documents in different folder hierarchies.

Something with the flexibility and visualisation of the Brain
(www.thebrain.com), the structuring capabilities of InfoHandler
(www.mdesoft.com/english.htm) and a Googleish search engine.

You mean something like Chandler?
http://osafoundation.org/Chandler_Compelling_Vision.htm

 

[Irmen de Jong]

- A roguelike in Python. Since there is still no portable curses, it
needs WConio or something like that, and also multiplayer would be
great.

While not written *in* Python, it's scriptable with Python:
http://www.thangorodrim.net/pangband.html (PAngband).

waaay back someone even created an Amiga version of this
game based upon Amiga Angband and my AmigaPython port

--Irmen

 

[Brandon J. Van Every]

Suppose you have the time and the money to start a new project in
Python. What would you like to do?

I can think of:

- A civilization like game in Python, with multiplayer support via
twisted.

Will, are you aware that I'm beginning exactly this project? Well, not the
twisted bit, I don't even know what that is or why one would want it.
Please see my post "ProtoCiv: porting Freeciv to Python." I'd be interested
in your feedback even if my project goals don't match yours.

- A roguelike in Python. Since there is still no portable curses, it
needs WConio or something like that, and also multiplayer would be
great.

There's Umbra. http://kuoi.asui.uidaho.edu/~kamikaze/Umbra/ Graphics and
gameplay aren't great, but it is something to start with.

If you go to http://www.thangorodrim.net/ and search with the keyword
"Python," you will find some roguelikes that claim to have various amounts
of Python support in them. I haven't chased any of these down myself
though. At some point I gave up on the "do a RPG" project excuse, opting
instead for 4X TBS because it's more directly applicable to my Ocean Mars
project.

--
Cheers, www.indiegamedesign.com
Brandon Van Every Seattle, WA

20% of the world is real.
80% is gobbledygook we make up inside our own heads.

 

[Rene Pijlman]

Aahz:

You mean something like Chandler?
http://osafoundation.org/Chandler_Compelling_Vision.htm

Yes, that's what I meant. Although Chandler seems much more ambitious.
Very interesting project.

 

[Paul Rubin]

What would your favorite be?

A CVS-like system for email. You'd receive all your messages on a
server somewhere. You'd then be able to connect your laptop to the
internet, download ("check out") your mail, and read and reply to it
offline (not necessarily all of it). When you dial up again, the
replies get sent out and stored ("checked in") on the server, the
messages that you read get marked as read, the ones you didn't read
don't get marked, etc. The CVS-like aspect is that you can do the
same thing from your office computer, your friend's computer, etc., so
you have the same messages checked out on multiple clients at the same
time. The server automatically merges the "change sets" when you
check any in. Finally, the server shouldn't need any special protocol
to check messages in or out. It should be able to create a single
tarball or zipfile that you download, and accept a single tarball or
zipfile when you upload

I've been wanting for a while to write something like this. Everyone
I've mentioned it to wants to use it. I'm amazed it doesn't exist
already, at least in any well-known form.

 

[Mark Carter]

I was thinking of an open source personal information management

application, to store notes, knowledge, hyperlinks, snippets, documents
and such in one flexible integrated information structure. All the things
that now end up in unrelated documents in different folder hierarchies.

Something with the flexibility and visualisation of the Brain

Well, it doesn't have the flexibility of the brain, but you could take
a look at Leo (Literate Editor for Outlines):
http://sourceforge.net/projects/leo/

 

[Ville Vainio]

Suppose you have the time and the money to start a new project in
Python. What would you like to do?

Well, not completely *in* python, but about Python anyway:

- Rewrite Emacs (though this would probably need lots of C code for
that extra snappiness). Make it extensible in Python (natively,
i.e. w/o pymacs), use all the latest Gtk UI stuff, provide all the
intellisense:ish features of commercial IDEs (yes, I know about
Semantic).

- Port Python to Symbian OS > 6.0

- Implement fast native code compilation for Python (though this is
well beyond my capabilities). This might be interesting also in the
sense of creating a compiler in Python, even though it might not be
that fast...

 

[Ville Vainio]

A CVS-like system for email. You'd receive all your messages on a

[..]

This sounded quite fascinating. It also sounded quite doable w/
Subversion, w/ its "Properties" (metadata) and Python API.

check any in. Finally, the server shouldn't need any special protocol
to check messages in or out. It should be able to create a single
tarball or zipfile that you download, and accept a single tarball or
zipfile when you upload

No biggie, but it might still be useful for the system to be 100%
accessible from a normal Subversion client.

OT: Subversion (http://subversion.tigris.org) is finally going to hit
Beta and 1.0 Real Soon Now (tm). I read that it might be even during
this year, but that's probably not going to happen.

 

[Michael Hudson]

A CVS-like system for email. You'd receive all your messages on a
server somewhere. You'd then be able to connect your laptop to the
internet, download ("check out") your mail, and read and reply to it
offline (not necessarily all of it). When you dial up again, the
replies get sent out and stored ("checked in") on the server, the
messages that you read get marked as read, the ones you didn't read
don't get marked, etc. The CVS-like aspect is that you can do the
same thing from your office computer, your friend's computer, etc., so
you have the same messages checked out on multiple clients at the same
time.

Up to here, this sounds like the idealized IMAP experience. Using
Apple's mail.app offline isn't so different from this -- there are
some bogosities in the implementation (in Jaguar), but the intent is
clearly there.

The server automatically merges the "change sets" when you check any
in. Finally, the server shouldn't need any special protocol to
check messages in or out. It should be able to create a single
tarball or zipfile that you download, and accept a single tarball or
zipfile when you upload

Mail.app can import from most formats and will upload on sync... of
course, I like to have my mail on a machine where I have a shell too,
which makes all this a little bit trival (it also means that when I'm
online -- as now -- I use ssh + screen to acheive something a litle
bit like what you describe).

Cheers,
mwh

 

[Will Stuyvesant]

[Andrew Dalke"]

An integrated chemical/bioinformatics development and
exploration environment. Got a few million dollars to fund me?
(Actually, more like 10 million, but bootstrappable with only
a few million.)

I guess all that money is needed for getting high quality people into
it? Or other reasons? What techniques would you use?

There's been a few civ clones -- I recall playing one in the
mid-90s for Python using CLIPS for the AI. Don't recall the
name now. There's also openciv (in Python, nearly complete,
no longer active) or freeciv (in C, active). What would the
advantage be to writing yet another clone?

Getting one to 1.0, to *finish* one, to a stable version that is at
least as good as the original civ outofthebox. I don't think the C
based versions will finish, C is rather too low level to my taste.
Maybe Python can do it. I *know* it can, but for it to happen is
another thing.

What about Sketch?
http://sketch.sourceforge.net/

I was not aware of Sketch, thanks for the link. But it has to wait
until I have broadband again: needs GTK, GTK+, libart, etc. I am on
Windows and downloads are slow now.

Why does that need to be rewritten in Python? As I understand
it, the C version is very portable and may simply need just bindings
for Python.

It is, but I would like to program it. Design an AI, called Borg in
roguelikes. And I can not do C bindings (too long ago I did C).
Besides I don't *want* to C again

Aren't there a few dozen of those already? Falls into the category
of easy to write but without one clear way to do it. So there are
a lot of different implementations, all different, all focused on solving
the given author's needs.

Python can be used here to *find* the way to do it, because it is easy
to start all over from scratch. But so far I have not found a good
path.

More important, what would *your* favorite be. It looks like
you want to do a project but don't know which one to focus on.
My answer then is to do any of these projects; they are all great
ones to learn how to do larger, more useful projects.

Um, more useful? Can you give some more examples of what you think
useful?
I am looking for a fun project and inspiration and maybe even fun
people who join a project, you got that right

 

[Will Stuyvesant]

[Irmen de Jong]

While not written *in* Python, it's scriptable with Python:
http://www.thangorodrim.net/pangband.html (PAngband).

waaay back someone even created an Amiga version of this
game based upon Amiga Angband and my AmigaPython port

You did an AmigaPython port? Fun! How long does something like that
take?

How long do you think a new roguelike from scratch would take in
Python and where the most time is spent? I am afraid of a lot of work
on options like magic and +ToHit etc., while I am more interested in a
smaller set of options but with a really smart AI.

 

[Will Stuyvesant]

[Rene Pijlman]

Aahz:

Yes, that's what I meant. Although Chandler seems much more ambitious.
Very interesting project.

I wonder how healty it is: their wiki is empty, the Mitch Kapor blog
is stopped, there is not much recent news. But interesting certainly.

 

[Will Stuyvesant]

[Brandon J. Van Every]

Will, are you aware that I'm beginning exactly this project? Well, not the
twisted bit, I don't even know what that is or why one would want it.
Please see my post "ProtoCiv: porting Freeciv to Python." I'd be interested
in your feedback even if my project goals don't match yours.

I have that posting saved somewhered, rather too expensive to read
online (telephone). Will do.

...
There's Umbra. http://kuoi.asui.uidaho.edu/~kamikaze/Umbra/ Graphics and
gameplay aren't great, but it is something to start with.

Thanks for the link, going to look at it.

If you go to http://www.thangorodrim.net/ and search with the keyword
"Python," you will find some roguelikes that claim to have various amounts
of Python support in them. I haven't chased any of these down myself
though. At some point I gave up on the "do a RPG" project excuse, opting
instead for 4X TBS because it's more directly applicable to my Ocean Mars
project.

I don't know what TBS is.

Can you show something of the Ocean Mars project, link, screenshots?

 

[Paul Rubin]

How long do you think a new roguelike from scratch would take in
Python and where the most time is spent?

These things can be as simple or as complicated as you like. Remember
that the original Rogue ran on a PDP-11 so it can't have been that large.

I am afraid of a lot of work on options like magic and +ToHit etc.,
while I am more interested in a smaller set of options but with a
really smart AI.

Well, that's lots of work too.

 

[Will Stuyvesant]

[Paul Rubin]

A CVS-like system for email...
...It should be able to create a single
tarball or zipfile that you download, and accept a single tarball or
zipfile when you upload

I've been wanting for a while to write something like this. Everyone
I've mentioned it to wants to use it. I'm amazed it doesn't exist
already, at least in any well-known form.

Indeed an interesting project! And include newsgroups like
comp.lang.python too!

For another project I finally got the "upload zip", "unzip",
"base64encode" etc. thing right: users can send a .zip file via a HTML
INPUT file="type" and the CGI script unzips it, and can send zipped
parts to the next page, etc. Would like to warn you: it seems easy
but it took me some retries to get it right. If you are interested I
can put the code somewhere on the net.

 

[Ville Vainio]

I wonder how healty it is: their wiki is empty, the Mitch Kapor blog
is stopped, there is not much recent news. But interesting certainly.

Reading the blog at

http://blogs.osafoundation.org/mitch/

suggests that he is turning his energies to execution:

""" What remains is the sometimes unglamorous process building the
complete team and orchestrating the complex software development
process efficiently to realize the dreams we have created. It is to
that end that I am now trying to turn my energies to. """

So the project seems healthy. Even if the blog entry starts in the
style of writing that usually accompanies "we tried our best, we had
good people, but we're screwed" announcements.

The fact that they have new job postings on the Announcements page,
dating Dec 18, can't be a bad sign either.