kiddo

[Log in to get rid of this advertisement]

Note to moderators: please move this topic accordingly if I am in the wrong place.

Hi folks, I'm starting up a project which, I think, might interest some of you. The thing is that we are only beginning, so we need to get some initial developpers to give opinions, "prepare the terrain", or should I say, help build the foundation for it. (There is no code yet)

It is planned to be done in Python. Please have a look at the website here:
http://specto.sourceforge.net

I, myself, have no programming skills whatsoever, I'm only a designer, the project manager, and the weird kid that had that idea since I don't want to clutter that topic, I will not paste a complete description of the software (which is called Specto), but instead give a small exerpt:
I have setup the sourceforge project, the project page, the blog and wiki, and I'm now entering the recruiting phase. Feedback is much appreciated! ^^ besides, if you got other suggestions as to "how to admin it all", say 'em! I'm totally new to this, and willing to learn.

Even if you don't want to get involved, you may share your ideas and questions here. They're important.

mrGenixus

please, please, please, not another project in puthon. Please use a compile language so that the project doesn't lag my entire system. C is good, so is C++. Pyton is the VB of linux.

Nad0xFF

Let´s make something on C

kiddo

I guess that's something to be debated, and that will depend on the devs joining in I guess from what I understand of programming, of course it could be done in C; would be helluva faster, but it may also be a lot harder to code than python. One example I have in mind is pitivi.org; the guy is rewriting it from scratch in python, because he had enough of multiple-level C bugs spawning everywhere (but there are parts that are kept in C for performance reasons).

Additionally, here's a chat I just had with someone about this, feel free to comment back So, if you got something to precise I'm gladly taking comments! Thanks for the feedback already ^ ^

Nad0xFF

Good programmer doesn't make buffer overflows

Hko

True,...theoretically.

ahwkong

Hi, Kiddo,

It is interesting that you would like to manage a project while you does not have the programming skill.

I do mean no offence. Kiddo, you could be an excellent manager if you were my boss. Because, afterall, you can order me to do whatever you want to do, and at the end a great product may roll out. However, my point of view is that the nature of OSS requires people to take part voluntarily. And whenever there is a technical issue surfaced, the issue has to be resolved in accordance to technical merit and OSS principle (not any executive order).

So, kiddo, you would be put in a very disadvantaged position because there is a risk that you would be "push around" in these technical arguments.

For example,

-- "Option A is good", "No, Option A sucks. Option B is good".
-- "We should do A", " No, because of XYZ we cannot"

Being a manager and yet unable to settle this issue when needed greatly handicap your degree of freedom.

I think you are on the right track to suggest python because it is fast to build prototype in python and hence can prove your concept. Besides, it is relatively easy to learn, and probably your programming skill (at least "read only" skill) can grow with the project's progress. With C I would think you will be distracted by many issues (e.g. QT or GTK? C or C++? cross-platform issues...)

On the other hand, when you have a working prototype, you can worry about the speed then. You may decide to convert it to C for example.

I think now there is already a first test of you as a lead: "C/C++" or "python". Can you make a choice and stick to it? Can you explain your choice? Can you stay focus on issues?

kiddo

I'm really more interested in python, first because it's intended to be a lot more simple (read: human readable) and flexible (with other languages... if there's something you can't do with it, well, you can add some part in C). I also prefer a programming language that's "younger" (C and C++ are.. 30+ years old?). This being a "not so complicated" desktop app, why roll out something that would take a lot more work from devs? I think someone coming from C to python can get along quite well, the contrary would be less likely. Now, on a more technical aspect, C is more ... manual, I'd say. And I believe automation by computers is not a drawback, but a feature; of course you can tweak your engine a little bit faster if you do it by hand, but at the expense of how much tweaking time to get that extra performance? (I'm currently thinking of building a car from scratch, or linux from scratch if you prefer. Yeah you can do something real fast. But I think I'd like to see something appear BEFORE longhorn does ). Having memory allocation automatically managed (not by humans), you don't have to worry about weird errors that behave differently everywhere you go and can focus on the actual software, not bugs dating back to the '70s (sorry C fans, I'm not trying to insult/flame anything/anyone here).

And finally.. I might learn some bits of python just by reading up what's written (and I would be very happy to learn something from this project!), while I don't stand a chance with C/C++/C# I might get less C fans to develop for specto, but I think python really is more "open" to anyone.

Thank you for pointing this to me, ahwkong!

keefaz

I want just add a note : What will render a human readable script is not the programming
language but the programmer style of coding.

You can make readable a C script even a Perl script

bilboed

Hi all,

To put it into context, I'm the main developer of PiTiVi. To put a stop to all this python/C[|++|#] flaming take into account the following facts (from experience after doing several open/proprietary python and C programs):

• Python is way faster to code than in any other language, forget C# too ==> Faster development time
• Python programs are easier to modify ==> Add features faster, correct bugs faster
• Python's core is C-coded (and so are several bindings) ==> Runs slightly slower than a program made in C
• Python code is easy to understand ==> Easier adoption by "newbies", easy to understand by non-python coders
• Python is fun
• Kiddo is not a developer to start with, but seems to have a sense of organization that a lot of us developers lack (or at least lacked at the beginning) ==> He'll spend a few days/weeks before being able to hack in python, and a few months before being able to compile a C program without error

I've got serious doubts as to whether the people flaming python have actually used it at all, whether for a small hacking or for a more serious project. When you want fast high-level application development, and you are weak in every languages (because you're a beginner), languages like python are the way to go.

I do understand the credo of the C.. advocates, but give it a chance with python first, at least the concept will be proven, and if it's really slow (doubtly), critical parts of it can be converted to compiled languages. Oh, I almost forgot, making plugins is very easy if you use python, you can also easily make python<->C bindings for it.

Take care, good luck

Nad0xFF

Yes, python is very fun. I've coded on it a couple of times :-D. But I think, that C is more suitable, because with python we can make only scrpit, not a program. Also very funny are functional languages. I'm fond of it

bilboed

Nad0xFF: You can't use python to make programs ????????????

Go and try flumotion (www.flumotion.net) , a streaming server capable of handling thousands of simultaneous connections, done in python

And your "we can only do scripts in python" argument just proves that you don't have a clue what python really is. Functions, classes, modules (equivalent of libraries), network framework (twisted), multimedia framework (gst-python), etc ...... What do you want more to do a programm ???

What you surely mean is that is not compiled, which is way different.... it's even better, don't have to wait an eternity for the compilation when you're working on a big project. You can even reload parts of your program for it's file without having to stop the program from running !!

Bilbo

Nad0xFF

Hmm. I think that interpreted programs are very slow. And also, programming language needs to be a little hard . I don't think that python is very bad, but I don't believe to Interpreters

Nad0xFF

From program I want to be very fast and be able to have direct memory access and so on. And not a 5-minute script. This is for personal use.

ahwkong

This thread is heading in the wrong direction...

It is not about the language!

It is about if the project "specto" interests you or not. If yes, will you devote your time and take part in development.

To "C-advocate": if you win the "C vs python" debate here, will you join Kiddo to develop then? Not leaving the project unfinished at stage of 0.01-alpha?

(OK. Usually I enjoy this kind of debate -- fun to read, but maybe we should not steal the thread from kiddo. He wants some help for his project. )