Good moring, > interesting that you should point out the number of files as a > differentiation > criterion: why do you think C would be handling them better than Python? > and why do you think if speed is a priority, C should be chosen? Well, first of all Python code is interpreted and there are many reasons that python is slower than c(like calculations(*,/,+), python needs more time to be executed) . But in our project it isn't all about the language. We should select cleaver techniques to solve the problem and implement commands with an efficient way. What I enjoy doing is to select a lot of solutions for a problem and pick the best(based on complexity and memory space etc). In this way we will not have problem with the speed of our application and your note is quite right. But theoritically if we have the exact same implementations in c and python our application will be faster in c. System calls will behaved better than python code. On the other hand if we want to handle better our meta-data we should select C. It is very powerfull and with system calls we can optimize many things better than Python. I am quit familiar with c and system calls. In my opinion Python is ideal for our project. I thnik that students and scientists prefer Python to work with. In this way our application will have a better impact and it can be used from a variety of people. The goal is to build an open source appliaction that everyone can have an easy-access to use it and can interact with it easily. although data mining is useful in many domains, it has nowadays > become a buzzword, like machine learning; > what do you think "data mining" would produce > more than the already specified "produce statistics" functionality? > I understand than it could "mine" the information that files > ending in .mp3 are typically larger than file ending in .c > but smaller than files ending in .mp4, but that's not really novel info... I see your point. Having not a prior experience in data mining i can't answer to all these questions. I study this semester data mining and i find it interesting. I think that we can use some techniques in our application. As i have told you i will be more specific in about 2-3 months(i don't want to note something wrong) although a rudimentary web interface would indeed be easy to build, > one should actually consider the use case and its deployment > and answer questions like: where does it run? does it need > and external webserver? if it's on a remote machine, how are > the databases accessed? etc. etc. > or one could go towards the idea of implementing a server > that provides the raw data via an interface, and have > the web interface part be a separate program. > lots can be done in this space, but all should come after > the complete command-line utility. Well there are a lot of options. For example we can built a site that the user can have his own account and scan it's files from his computer. The meta-data will be saved in a data base. This data base can placed in his computer or in our machine. There are disadvantages and advantages of each solution. In my proposal i consentrate more in console application than web interfance. Do you think that i should go ahead ? What is more, i have made a draft proposal. I have completed the introduction section. Can I publish my proposal(as darft) or should I finished all the parts ? I think that i will have finished my proposal by the end of this week. But i want to publish it earlier and present some parts so as to get your feedback. Thank you again for your reply and your time, Panagiotis Petropoulakis <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> Απαλλαγμένο από ιούς. www.avast.com <https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail> <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2> On Wed, Mar 14, 2018 at 10:10 AM, Alexios Zavras <zvr+eellak [ at ] zvr [ dot ] gr> wrote: > Panagiotis Petropoulakis wrote [edited]: > > Well concerning technical issues: > > I believe that filesdb should be implemented with python(and sqllite) > than > > C. I think that the project can be easily done with python. Python is an > > easy language to learn(and write) and many students from universities > > prefer python. In this way students from all over the world can > understand > > our project and "play" with it. This is very important, because this > > project is very usefull and something different and new as you have said. > > Another idea is that we can build jupyter notebooks to help new comers. > On > > the other hand if we want to handle many files and speed is the number > one > > priority we should select c as language. > > interesting that you should point out the number of files as a > differentiation > criterion: why do you think C would be handling them better than Python? > and why do you think if speed is a priority, C should be chosen? > > > > Also i would like to note that we should build a console application than > > to build a web interface. A web interface can be build easily if we have > > completed the project. I think we should concentrate more in commands. > > Ofcourse if it goes well we could build a web interfance. Do you think > that > > i should insist on this and add it in my proposal ? > > although a rudimentary web interface would indeed be easy to build, > one should actually consider the use case and its deployment > and answer questions like: where does it run? does it need > and external webserver? if it's on a remote machine, how are > the databases accessed? etc. etc. > or one could go towards the idea of implementing a server > that provides the raw data via an interface, and have > the web interface part be a separate program. > lots can be done in this space, but all should come after > the complete command-line utility. > > > > Another idea is that we could use data mining to extract usefull > > information about files, system, creators etc. I have some ideas but i > > don't have engouh experience yet. In this semester i will attend a class > > for data mining and i will be able to produce more ideas in about 2 > months. > > although data mining is useful in many domains, it has nowadays > become a buzzword, like machine learning; > what do you think "data mining" would produce > more than the already specified "produce statistics" functionality? > I understand than it could "mine" the information that files > ending in .mp3 are typically larger than file ending in .c > but smaller than files ending in .mp4, but that's not really novel info... > > > -- > -- zvr -- > -- +---------------------------+ Alexios Zavras (-zvr-) > | H eytyxia den exei enoxes | zvr [ at ] zvr [ dot ] gr > +-----------------------zvr-+ >
---- Λαμβάνετε αυτό το μήνυμα απο την λίστα: Γενική λίστα αλληλογραφίας που απευθύνεται σε developers/contributors έργων ανοικτού λογισμικού - A general discussion list for developers/contributors of open-source projects, https://lists.ellak.gr/opensource-devs/listinfo.html Μπορείτε να απεγγραφείτε από τη λίστα στέλνοντας κενό μήνυμα ηλ. ταχυδρομείου στη διεύθυνση <opensource-devs+unsubscribe [ at ] ellak [ dot ] gr>.