Developer forums (C::B DEVELOPMENT STRICTLY!) > Development

Time to release the CVS unstable? Or perhaps RC1?

<< < (2/4) > >>

rickg22:
Yes, just let us fix the "Phantom lines on closing project" bug first.

RShadow:

--- Quote from: rickg22 ---Now that we have a branch, I think it's time we should release a stable snapshot (binary, HEAD) so people can test it and *ahem* stop submitting outdated bugs.

Boy, we didn't count on this (the increasing popularity of Code::Blocks  :lol: ) when developing.

Or maybe... Yiannis, what bugs do you consider need to be fixed before releasing RC1?
--- End quote ---


As far as a binary release.. sure.. but I think we should also create a nightly "snapshot" of the source code for non CVS folks.  I'm in that boat.. the only cms I've ever used is subversion.. so when I check C::B out using cvs.. I really haven't a clue whats it doing.. I have some hacked together command line that I found.. what branch it's pulling.. I dunno..   A nightly snapshot of the source code would be very beneficial.

As far as new features.. I think all new features should be moved to the new branch.. leave the current one alone.. simply put bug/security fixes in the current 1.0 branch.  This is also why I would recommend that each release cycle the dev's get together and based on their input and users request come up with a "feature list"... this way you know when all the items have been implemented.. thats version x.x and its time to branch and start working on the next version..

darklordsatan:
Agree with RShadow, a nightly snapshot would be convenient; I could also checkout the cvs, but as you, I dont know what am I doing (yet).
I dont know for sure if this has to be done manually everytime, or if theres some automatic way to do it (probably not, since we'd need a daemon running all the time -where?- checking if its a determined time to create the .tgz or whatever). Notice Im talking about the fact of putting the snapshot online, not taking all the cvs files and packing them (this for sure can be done automatically, i think)

RShadow:

--- Quote from: darklordsatan ---Agree with RShadow, a nightly snapshot would be convenient; I could also checkout the cvs, but as you, I dont know what am I doing (yet).
I dont know for sure if this has to be done manually everytime, or if theres some automatic way to do it (probably not, since we'd need a daemon running all the time -where?- checking if its a determined time to create the .tgz or whatever). Notice Im talking about the fact of putting the snapshot online, not taking all the cvs files and packing them (this for sure can be done automatically, i think)
--- End quote ---


As far as creating the snapshot's automaticly it depends.. I don't know if you have access to cron on sourceforge or not.. If so its a simple matter of creating a cron script to zip up the contents of the CVS tree and put the file into whatever directory you need to put stuff to be downloaded.

If we don't have access to cron at sourceforge then its a matter of somebody with access to the SF account to create a nightly script to make a checkout of cvs and post the zipped file.

This is also why I would suggest using sourceforge as the distribution point, and using another service for actual development (preferably a serivice with a subversion server).

darklordsatan:

--- Quote from: RShadow ---
 I don't know if you have access to cron on sourceforge or not.. If so its a simple matter of creating a cron script to zip up the contents of the CVS tree and put the file into whatever directory you need to put stuff to be downloaded.

--- End quote ---


1. Do you have to ask sf staff for cron access?
2. If you had cron access, then how do you do it. You log on to your shell account, and leave the process running in the background forever?, doesnt sf gets mad for this?

Navigation

[0] Message Index

[#] Next page

[*] Previous page