[opensuse-project] Upstream Tracker Project - GSoC 2012 - Week 1 Report
Hello! This is my first week work report for the Upstream Tracker Project. I have chosen to start my work by implementing the web interface first. This is because, by having a functional web interface, we could start crowd sourcing the data to proceed further. Also, the base work for the python scripts are already in place. Here is the list of things I have managed to do this week. 1. Created a rails application to manipulate records. Each record consists of data required to obtain the latest version of an application such as package name, method, URL etc. 2. Added support for XML and JSON Formats. 3. Added a page to view records which are erroneous after being processed. 4. Added color coding on main page to indicate record status. Red for Error, Green for OK and None for not processed. 5. Added Basic validation for user input. 6. Wrote pycurl scripts to Add, Delete, Update and Show records. 7. Set up a home server to host the rails application for feedback and testing. This can be found at nbprashanth.no-ip.org. List of things planned for the upcoming week: 1. Evaluate logging requirements and implement a solution. 2. Update the records model in the rails app to include version information after processing is complete. 3. Fill in gaps and test the rails app. the code is hosted at : 1. https://github.com/nbprashanth/Upstream-Tracker-Rails 2. https://github.com/nbprashanth/Upstream-Tracker-Python The project description can be found at http://www.google-melange.com/gsoc/proposal/review/google/gsoc2012/nbprashan... -- Regards, N.B.Prashanth -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org To contact the owner, email: opensuse-project+owner@opensuse.org
Hi, On 28.05.2012 19:36, N B Prashanth wrote:
This is because, by having a functional web interface, we could start crowd sourcing the data to proceed further. Also, the base work for the python scripts are already in place.
how about an importer for Debian watchfiles since they are already available for thousands of packages? They can be conveniently downloaded as a BDB at http://qa.debian.org/data/mole/db/dscfiles-watch.db (I have some python code for that for a project of my own). -- Guido Berhoerster -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org To contact the owner, email: opensuse-project+owner@opensuse.org
how about an importer for Debian watchfiles since they are already available for thousands of packages? They can be conveniently downloaded as a BDB at http://qa.debian.org/data/mole/db/dscfiles-watch.db (I have some python code for that for a project of my own).
You mean to say that debian has way to indicate that a library in debian is suspect as per user experience. Then this seems to be the best way go ahead. -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org To contact the owner, email: opensuse-project+owner@opensuse.org
On 29.05.2012 12:56, Sujit Karatparambil wrote:
how about an importer for Debian watchfiles since they are already available for thousands of packages? They can be conveniently downloaded as a BDB at http://qa.debian.org/data/mole/db/dscfiles-watch.db (I have some python code for that for a project of my own).
You mean to say that debian has way to indicate that a library in debian is suspect as per user experience. Then this seems to be the best way go ahead.
Sorry, I don't quite understand what you're saying above? The above database has metadata in form of URLs with regexes [*] for monitoring around 6500 upstream projects, so using these rather than collecting them on our own would probably be quite a leap forward. You could probably use distromatch data in order to match Debian with openSUSE packages. [*] the format is documented here: http://wiki.debian.org/debian/watch/ -- Guido Berhoerster -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org To contact the owner, email: opensuse-project+owner@opensuse.org
On 29.05.2012 13:07, Guido Berhoerster wrote:
On 29.05.2012 12:56, Sujit Karatparambil wrote:
how about an importer for Debian watchfiles since they are already available for thousands of packages? They can be conveniently downloaded as a BDB at http://qa.debian.org/data/mole/db/dscfiles-watch.db (I have some python code for that for a project of my own).
You mean to say that debian has way to indicate that a library in debian is suspect as per user experience. Then this seems to be the best way go ahead.
Sorry, I don't quite understand what you're saying above?
The above database has metadata in form of URLs with regexes [*] for monitoring around 6500 upstream projects, so using these rather than collecting them on our own would probably be quite a leap forward. You could probably use distromatch data in order to match Debian with openSUSE packages.
[*] the format is documented here: http://wiki.debian.org/debian/watch/
FYI, here is a basic script I've been using to extract the data from the BDB which produces a list in the format "Debian Name" "Base URL" "Version Regex". -- Guido Berhoerster
Le mardi 29 mai 2012, à 11:04 +0200, Guido Berhoerster a écrit :
Hi,
On 28.05.2012 19:36, N B Prashanth wrote:
This is because, by having a functional web interface, we could start crowd sourcing the data to proceed further. Also, the base work for the python scripts are already in place.
how about an importer for Debian watchfiles since they are already available for thousands of packages? They can be conveniently downloaded as a BDB at http://qa.debian.org/data/mole/db/dscfiles-watch.db (I have some python code for that for a project of my own).
That's already part of the plan :-) Vincent -- Les gens heureux ne sont pas pressés. -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org To contact the owner, email: opensuse-project+owner@opensuse.org
Sorry, I don't quite understand what you're saying above? I meant something like sonar, I am a Java/J2EE Developer I see sonar being utilized quite a lot in companies, seems to be in thing. It does a basic analysis of the java code using rules, for example something like whether there is any replication/scope for reuse in the code.
What I meant is that, is there any similarity between the two? -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org To contact the owner, email: opensuse-project+owner@opensuse.org
On 29.05.2012 14:23, Sujit Karatparambil wrote:
Sorry, I don't quite understand what you're saying above? I meant something like sonar, I am a Java/J2EE Developer I see sonar being utilized quite a lot in companies, seems to be in thing. It does a basic analysis of the java code using rules, for example something like whether there is any replication/scope for reuse in the code.
What I meant is that, is there any similarity between the two?
No, I think you misunderstood, this project is about keeping track of new upstream releases of software packaged in openSUSE. -- Guido Berhoerster -- To unsubscribe, e-mail: opensuse-project+unsubscribe@opensuse.org To contact the owner, email: opensuse-project+owner@opensuse.org
participants (4)
-
Guido Berhoerster -
N B Prashanth -
Sujit Karatparambil -
Vincent Untz