Maybe I'm just being paranoid, but I've been seeing a lot of "I just updated to latest git and X broke" And that x is always something else... I'm not a seasoned coder on this magnitude, but perhaps we can develop a test suite (at least part of one) to test that a fresh install is functioning? A FULL test would probably be a huge undertaking and involve recqual for testing media, too... Also, each mod would need to have it's own test suite. Due to the rapid updating of FreeSWITCH this not be what you want to put your effort into, but... Thoughts? On Feb 4, 2011 4:04 AM, "Anirudha Shimpi" <anu@familytv.com> wrote:
Just downloaded (from git) and installed FS 1.0.7, however, every time I do originate user/1001 &park, FS crashes. Running Centos 5.5 64 bit.
I've been doing this a long time and it boils down to: 50% superstition/paranoia. 50% regressions from commits. Stop and think. Without people updating everyday and telling us when something is not right, we may never find out. Engaging in the community and this mailing list and irc is a community effort and its sometimes a BETA program too. Just like if this was a cool video game, you get to play it before everyone else. It's hard to follow every possible execution path and get into every nook and cranny of 3 million lines of code. Even with the best unit testing int the world, we can't do it without real people testing. If you stop updating or persuade others to stop, now we are in real trouble. Also, pay attention to the GIT commit messages: This particular problem (which is fixed now btw) was committed with a warning to please test. There was a bigger problem earlier this week that resulted in some minority invasive changes and its easy to catch the problems and in the end the software is now more stable. commit f60fdf653dd2d7f8d3eaa6a9086e1f68bd993c59 Author: Anthony Minessale <anthm@freeswitch.org> Date: Wed Feb 2 16:22:43 2011 -0600 fix possible bad pointer in global vars (please test) On Thu, Feb 3, 2011 at 11:46 PM, Avi Marcus <avi@avimarcus.net> wrote:
Maybe I'm just being paranoid, but I've been seeing a lot of "I just updated to latest git and X broke" And that x is always something else... I'm not a seasoned coder on this magnitude, but perhaps we can develop a test suite (at least part of one) to test that a fresh install is functioning? A FULL test would probably be a huge undertaking and involve recqual for testing media, too... Also, each mod would need to have it's own test suite. Due to the rapid updating of FreeSWITCH this not be what you want to put your effort into, but... Thoughts?
On Feb 4, 2011 4:04 AM, "Anirudha Shimpi" <anu@familytv.com> wrote:
Just downloaded (from git) and installed FS 1.0.7, however, every time I do originate user/1001 &park, FS crashes. Running Centos 5.5 64 bit.
_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
-- Anthony Minessale II FreeSWITCH http://www.freeswitch.org/ ClueCon http://www.cluecon.com/ Twitter: http://twitter.com/FreeSWITCH_wire AIM: anthm MSN:anthony_minessale@hotmail.com GTALK/JABBER/PAYPAL:anthony.minessale@gmail.com IRC: irc.freenode.net #freeswitch FreeSWITCH Developer Conference sip:888@conference.freeswitch.org googletalk:conf+888@conference.freeswitch.org pstn:+19193869900
FS team needs user as users need FS team... maybe think to a script (developped by all possible users) that test every new functionality... ----- Original Message ----- From: "Anthony Minessale" <anthony.minessale@gmail.com> To: "FreeSWITCH Users Help" <freeswitch-users@lists.freeswitch.org> Sent: Friday, February 04, 2011 9:58 AM Subject: Re: [Freeswitch-users] Auto test FS install - now necessary?
I've been doing this a long time and it boils down to:
50% superstition/paranoia. 50% regressions from commits.
Stop and think. Without people updating everyday and telling us when something is not right, we may never find out. Engaging in the community and this mailing list and irc is a community effort and its sometimes a BETA program too. Just like if this was a cool video game, you get to play it before everyone else. It's hard to follow every possible execution path and get into every nook and cranny of 3 million lines of code. Even with the best unit testing int the world, we can't do it without real people testing. If you stop updating or persuade others to stop, now we are in real trouble.
Also, pay attention to the GIT commit messages:
This particular problem (which is fixed now btw) was committed with a warning to please test. There was a bigger problem earlier this week that resulted in some minority invasive changes and its easy to catch the problems and in the end the software is now more stable.
commit f60fdf653dd2d7f8d3eaa6a9086e1f68bd993c59 Author: Anthony Minessale <anthm@freeswitch.org> Date: Wed Feb 2 16:22:43 2011 -0600
fix possible bad pointer in global vars (please test)
On Thu, Feb 3, 2011 at 11:46 PM, Avi Marcus <avi@avimarcus.net> wrote:
Maybe I'm just being paranoid, but I've been seeing a lot of "I just updated to latest git and X broke" And that x is always something else... I'm not a seasoned coder on this magnitude, but perhaps we can develop a test suite (at least part of one) to test that a fresh install is functioning? A FULL test would probably be a huge undertaking and involve recqual for testing media, too... Also, each mod would need to have it's own test suite. Due to the rapid updating of FreeSWITCH this not be what you want to put your effort into, but... Thoughts?
On Feb 4, 2011 4:04 AM, "Anirudha Shimpi" <anu@familytv.com> wrote:
Just downloaded (from git) and installed FS 1.0.7, however, every time I do originate user/1001 &park, FS crashes. Running Centos 5.5 64 bit.
_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
-- Anthony Minessale II
FreeSWITCH http://www.freeswitch.org/ ClueCon http://www.cluecon.com/ Twitter: http://twitter.com/FreeSWITCH_wire
AIM: anthm MSN:anthony_minessale@hotmail.com GTALK/JABBER/PAYPAL:anthony.minessale@gmail.com IRC: irc.freenode.net #freeswitch
FreeSWITCH Developer Conference sip:888@conference.freeswitch.org googletalk:conf+888@conference.freeswitch.org pstn:+19193869900
_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
http://en.wikipedia.org/wiki/Software_testing#Input_combinations_and_precond... "A very fundamental problem with software testing is that testing under all combinations of inputs and preconditions (initial state) is not feasible, even with a simple product. This means that the number of defects in a software product can be very large and defects that occur infrequently are difficult to find in testing. More significantly, non-functional dimensions of quality (how it is supposed to be versus what it is supposed to do)—usability, scalability, performance, compatibility, reliability—can be highly subjective; something that constitutes sufficient value to one person may be intolerable to another. Software testing is a very difficult thing. Testing every possible combination of actions and specifically covering the edge cases most likely to cause problems is pretty much impossible - to run through every possible action for something as large as FS would literally take longer than the age of the universe. As a result a script to test for all possible errors is pretty much impossible, especially for new functionality that'll be added before the script's updated to test them. There will always be some bugs, users testing and reporting is the best way to detect ones that the developers have missed. Generally FS is pretty good at avoiding them. Testing yourself is always a good idea before rolling out to a production system. That'll show any obvious ones up that can be fixed quickly. -Steve On 4 February 2011 15:13, Madovsky <infos@madovsky.org> wrote:
FS team needs user as users need FS team... maybe think to a script (developped by all possible users) that test every new functionality...
----- Original Message ----- From: "Anthony Minessale" <anthony.minessale@gmail.com> To: "FreeSWITCH Users Help" <freeswitch-users@lists.freeswitch.org> Sent: Friday, February 04, 2011 9:58 AM Subject: Re: [Freeswitch-users] Auto test FS install - now necessary?
I've been doing this a long time and it boils down to:
50% superstition/paranoia. 50% regressions from commits.
Stop and think. Without people updating everyday and telling us when something is not right, we may never find out. Engaging in the community and this mailing list and irc is a community effort and its sometimes a BETA program too. Just like if this was a cool video game, you get to play it before everyone else. It's hard to follow every possible execution path and get into every nook and cranny of 3 million lines of code. Even with the best unit testing int the world, we can't do it without real people testing. If you stop updating or persuade others to stop, now we are in real trouble.
Also, pay attention to the GIT commit messages:
This particular problem (which is fixed now btw) was committed with a warning to please test. There was a bigger problem earlier this week that resulted in some minority invasive changes and its easy to catch the problems and in the end the software is now more stable.
commit f60fdf653dd2d7f8d3eaa6a9086e1f68bd993c59 Author: Anthony Minessale <anthm@freeswitch.org> Date: Wed Feb 2 16:22:43 2011 -0600
fix possible bad pointer in global vars (please test)
On Thu, Feb 3, 2011 at 11:46 PM, Avi Marcus <avi@avimarcus.net> wrote:
Maybe I'm just being paranoid, but I've been seeing a lot of "I just updated to latest git and X broke" And that x is always something else... I'm not a seasoned coder on this magnitude, but perhaps we can develop a test suite (at least part of one) to test that a fresh install is functioning? A FULL test would probably be a huge undertaking and involve recqual for testing media, too... Also, each mod would need to have it's own test suite. Due to the rapid updating of FreeSWITCH this not be what you want to put your effort into, but... Thoughts?
On Feb 4, 2011 4:04 AM, "Anirudha Shimpi" <anu@familytv.com> wrote:
Just downloaded (from git) and installed FS 1.0.7, however, every time I do originate user/1001 &park, FS crashes. Running Centos 5.5 64 bit.
_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE: http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
-- Anthony Minessale II
FreeSWITCH http://www.freeswitch.org/ ClueCon http://www.cluecon.com/ Twitter: http://twitter.com/FreeSWITCH_wire
AIM: anthm MSN:anthony_minessale@hotmail.com <MSN%3Aanthony_minessale@hotmail.com> GTALK/JABBER/PAYPAL:anthony.minessale@gmail.com<PAYPAL%3Aanthony.minessale@gmail.com> IRC: irc.freenode.net #freeswitch
FreeSWITCH Developer Conference sip:888@conference.freeswitch.org <sip%3A888@conference.freeswitch.org> googletalk:conf+888@conference.freeswitch.org<googletalk%3Aconf%2B888@conference.freeswitch.org> pstn:+19193869900
_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
Testing yourself is always a good idea before rolling out to a production system. That'll show any obvious ones up that can be fixed quickly.
Not that this will always catch everything, but is there a "release notes" or some other simple document that outlines what parts of Freeswitch have recently been changed? We ended up building a sort of beta-test network, with a second server and some customers that are very benevolent about this sort of thing, and if I had a simple checklist of things that had been worked on, I can make sure that I test those areas more closely. I don't have any experience working with software development beyond my own little LUA and PHP scripts, so maybe this is already available and I just don't know where to look. Is there a tutorial somewhere on how to be a good user, from a developers perspective? -Tim
Best source of that information is the commit log http://fisheye.freeswitch.org/changelog/freeswitch.git There's also a changelog file that's periodically brought up-to-date. -Steve On 4 February 2011 15:58, Tim St. Pierre <fs-list@communicatefreely.net>wrote:
Testing yourself is always a good idea before rolling out to a production system. That'll show any obvious ones up that can be fixed quickly.
Not that this will always catch everything, but is there a "release notes" or some other simple document that outlines what parts of Freeswitch have recently been changed? We ended up building a sort of beta-test network, with a second server and some customers that are very benevolent about this sort of thing, and if I had a simple checklist of things that had been worked on, I can make sure that I test those areas more closely.
I don't have any experience working with software development beyond my own little LUA and PHP scripts, so maybe this is already available and I just don't know where to look.
Is there a tutorial somewhere on how to be a good user, from a developers perspective?
-Tim
_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
I understand well, but I think everything is possible, even if it seems huge work by the right method the huge work can be a reasonable work. It needs only to reflect deeply of how to do this.... ----- Original Message ----- From: Steven Ayre To: FreeSWITCH Users Help Sent: Friday, February 04, 2011 10:45 AM Subject: Re: [Freeswitch-users] Auto test FS install - now necessary? http://en.wikipedia.org/wiki/Software_testing#Input_combinations_and_precond... "A very fundamental problem with software testing is that testing under all combinations of inputs and preconditions (initial state) is not feasible, even with a simple product. This means that the number of defects in a software product can be very large and defects that occur infrequently are difficult to find in testing. More significantly, non-functional dimensions of quality (how it is supposed to be versus what it is supposed to do)—usability, scalability, performance, compatibility, reliability—can be highly subjective; something that constitutes sufficient value to one person may be intolerable to another. Software testing is a very difficult thing. Testing every possible combination of actions and specifically covering the edge cases most likely to cause problems is pretty much impossible - to run through every possible action for something as large as FS would literally take longer than the age of the universe. As a result a script to test for all possible errors is pretty much impossible, especially for new functionality that'll be added before the script's updated to test them. There will always be some bugs, users testing and reporting is the best way to detect ones that the developers have missed. Generally FS is pretty good at avoiding them. Testing yourself is always a good idea before rolling out to a production system. That'll show any obvious ones up that can be fixed quickly. -Steve On 4 February 2011 15:13, Madovsky <infos@madovsky.org> wrote: FS team needs user as users need FS team... maybe think to a script (developped by all possible users) that test every new functionality... ----- Original Message ----- From: "Anthony Minessale" <anthony.minessale@gmail.com> To: "FreeSWITCH Users Help" <freeswitch-users@lists.freeswitch.org> Sent: Friday, February 04, 2011 9:58 AM Subject: Re: [Freeswitch-users] Auto test FS install - now necessary? > I've been doing this a long time and it boils down to: > > 50% superstition/paranoia. > 50% regressions from commits. > > Stop and think. Without people updating everyday and telling us when > something is not right, we may never find out. > Engaging in the community and this mailing list and irc is a community > effort and its sometimes a BETA program too. > Just like if this was a cool video game, you get to play it before > everyone else. > It's hard to follow every possible execution path and get into every > nook and cranny of 3 million lines of code. > Even with the best unit testing int the world, we can't do it without > real people testing. > If you stop updating or persuade others to stop, now we are in real > trouble. > > Also, pay attention to the GIT commit messages: > > This particular problem (which is fixed now btw) was committed with a > warning to please test. > There was a bigger problem earlier this week that resulted in some > minority invasive changes and its easy to catch the problems and in > the end the software is now more stable. > > commit f60fdf653dd2d7f8d3eaa6a9086e1f68bd993c59 > Author: Anthony Minessale <anthm@freeswitch.org> > Date: Wed Feb 2 16:22:43 2011 -0600 > > fix possible bad pointer in global vars (please test) > > > > On Thu, Feb 3, 2011 at 11:46 PM, Avi Marcus <avi@avimarcus.net> wrote: >> Maybe I'm just being paranoid, but I've been seeing a lot of "I just >> updated >> to latest git and X broke" >> And that x is always something else... >> I'm not a seasoned coder on this magnitude, but perhaps we can develop a >> test suite (at least part of one) to test that a fresh install is >> functioning? >> A FULL test would probably be a huge undertaking and involve recqual for >> testing media, too... Also, each mod would need to have it's own test >> suite. >> Due to the rapid updating of FreeSWITCH this not be what you want to put >> your effort into, but... >> Thoughts? >> >> On Feb 4, 2011 4:04 AM, "Anirudha Shimpi" <anu@familytv.com> wrote: >>> Just downloaded (from git) and installed FS 1.0.7, however, every time I >>> do >>> originate user/1001 &park, FS crashes. Running Centos 5.5 64 bit. >>> >> >> _______________________________________________ >> FreeSWITCH-users mailing list >> FreeSWITCH-users@lists.freeswitch.org >> http://lists.freeswitch.org/mailman/listinfo/freeswitch-users >> UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users >> http://www.freeswitch.org >> >> > > > > -- > Anthony Minessale II > > FreeSWITCH http://www.freeswitch.org/ > ClueCon http://www.cluecon.com/ > Twitter: http://twitter.com/FreeSWITCH_wire > > AIM: anthm > MSN:anthony_minessale@hotmail.com > GTALK/JABBER/PAYPAL:anthony.minessale@gmail.com > IRC: irc.freenode.net #freeswitch > > FreeSWITCH Developer Conference > sip:888@conference.freeswitch.org > googletalk:conf+888@conference.freeswitch.org > pstn:+19193869900 > > _______________________________________________ > FreeSWITCH-users mailing list > FreeSWITCH-users@lists.freeswitch.org > http://lists.freeswitch.org/mailman/listinfo/freeswitch-users > UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users > http://www.freeswitch.org _______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org ------------------------------------------------------------------------------ _______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
On Fri, Feb 4, 2011 at 9:15 PM, Steven Ayre <steveayre@gmail.com> wrote:
Testing yourself is always a good idea before rolling out to a production system. That'll show any obvious ones up that can be fixed quickly.
On a related note, After thorough testing, if a system is put in production. Would it be 'best practice' not to update it, if no config changes or new functionality is ever desired?
-Steve
guru
On Fri, Feb 4, 2011 at 9:52 AM, guru singh <grsingh750@gmail.com> wrote:
On Fri, Feb 4, 2011 at 9:15 PM, Steven Ayre <steveayre@gmail.com> wrote:
Testing yourself is always a good idea before rolling out to a production system. That'll show any obvious ones up that can be fixed quickly.
On a related note, After thorough testing, if a system is put in production. Would it be 'best practice' not to update it, if no config changes or new functionality is ever desired?
This depends greatly on your scenario. Personally, I think that anyone who puts FS into production (i.e. in a place where it helps one make money) should *ALWAYS* have a backup/sandbox/test system for handling updates. If you are in a mission-critical scenario then it probably means that once you get it working you don't touch it unless you absolutely have to. That's where your test system comes in to play. Ideally you have the exact same hardware on your test system that you do in production. Then your test system can double as a warm standby. Just my $.02, but I think that commodity hardware being so cheap these days that there's no excuse for a for-profit venture using FS (or any other software, for that matter) not to have a test system for mission-critical servers. -MC
On Fri, Feb 4, 2011 at 11:52 PM, Michael Collins <msc@freeswitch.org> wrote:
This depends greatly on your scenario. Personally, I think that anyone who puts FS into production (i.e. in a place where it helps one make money) should *ALWAYS* have a backup/sandbox/test system for handling updates. If you are in a mission-critical scenario then it probably means that once you get it working you don't touch it unless you absolutely have to. That's where your test system comes in to play. Ideally you have the exact same hardware on your test system that you do in production. Then your test system can double as a warm standby.
Thanks for the advice, and also the awesome work you guys are doing. :) I, for one am more than happy with the 'latest is the best' philosophy.
Just my $.02, but I think that commodity hardware being so cheap these days that there's no excuse for a for-profit venture using FS (or any other software, for that matter) not to have a test system for mission-critical servers.
Hmmm, someday! -guru
And be stuck with a version that has known bugs? If they're minor bugs that don't effect you why not. Occasionally there may be one that's exploitable though and then you'd want to update. That's when the stable branch Anthm said is being planned could be useful. -Steve On 4 February 2011 17:52, guru singh <grsingh750@gmail.com> wrote:
On Fri, Feb 4, 2011 at 9:15 PM, Steven Ayre <steveayre@gmail.com> wrote:
Testing yourself is always a good idea before rolling out to a production system. That'll show any obvious ones up that can be fixed quickly.
On a related note, After thorough testing, if a system is put in production. Would it be 'best practice' not to update it, if no config changes or new functionality is ever desired?
-Steve
guru
_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
I do not see any piece of unit testing code neither. Reliability is more important than feature. Asterisk is a good example of such case. I would recommend developers take a pause to add unit testing cases and auto integration testings. Also enforce any future code must be accompanied by unit testing code. On Feb 3, 2011 9:47 PM, "Avi Marcus" <avi@avimarcus.net> wrote:
Maybe I'm just being paranoid, but I've been seeing a lot of "I just updated to latest git and X broke" And that x is always something else... I'm not a seasoned coder on this magnitude, but perhaps we can develop a test suite (at least part of one) to test that a fresh install is functioning? A FULL test would probably be a huge undertaking and involve recqual for testing media, too... Also, each mod would need to have it's own test suite. Due to the rapid updating of FreeSWITCH this not be what you want to put your effort into, but... Thoughts? On Feb 4, 2011 4:04 AM, "Anirudha Shimpi" <anu@familytv.com> wrote:
Just downloaded (from git) and installed FS 1.0.7, however, every time I do originate user/1001 &park, FS crashes. Running Centos 5.5 64 bit.
I guess we should try to be more like all the other open source projects where every single day of GIT HEAD is perfect. hmm.... The alternative is you can buy a soft-switch and live with it's faults and pay annually for an update. Or even wait for our official releases which are very infrequent and not well-supported once they are more than a few months old. Also don't forget we use FreeSWITCH in a real product so you don't even see the amount of testing and QA we inherit from that product as well as all the large carriers who use FreeSWITCH and do their best to try beta version for us. Thank you all..... We have plans to stable branch soon. I guess that means everyone will flock to that and it will be the end of my beta testers. We do not only get regressions from features we get them sometimes from fixes as well, other peoples patches, many factors. I guess what's really saddening since you bring up asterisk, is that I spent years using it with perpetual problems in the release or the daily snapshot and it sort of has a reputation for this kind of thing and all you hear about is how awesome it is from people. I don't appreciate being treated like a witch hunt over one regression especially when we can fix it in 1 minute. That double-standard is ridiculous. I should just make a FAQ about this and paste in the URL: We only have so much time to give to the world, we spend like 8-14 hours a day on average working on the code. Any suggestions or improvements (even the ones with the best intentions) are only viable if they come with volunteers. Bottom line is by downloading and building the code you are opting in to our community which involves occasional speed bumps. On Fri, Feb 4, 2011 at 10:34 AM, jesse <chat2jesse@gmail.com> wrote:
I do not see any piece of unit testing code neither. Reliability is more important than feature. Asterisk is a good example of such case.
I would recommend developers take a pause to add unit testing cases and auto integration testings. Also enforce any future code must be accompanied by unit testing code.
On Feb 3, 2011 9:47 PM, "Avi Marcus" <avi@avimarcus.net> wrote:
Maybe I'm just being paranoid, but I've been seeing a lot of "I just updated to latest git and X broke" And that x is always something else... I'm not a seasoned coder on this magnitude, but perhaps we can develop a test suite (at least part of one) to test that a fresh install is functioning? A FULL test would probably be a huge undertaking and involve recqual for testing media, too... Also, each mod would need to have it's own test suite. Due to the rapid updating of FreeSWITCH this not be what you want to put your effort into, but... Thoughts? On Feb 4, 2011 4:04 AM, "Anirudha Shimpi" <anu@familytv.com> wrote:
Just downloaded (from git) and installed FS 1.0.7, however, every time I do originate user/1001 &park, FS crashes. Running Centos 5.5 64 bit.
_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
-- Anthony Minessale II FreeSWITCH http://www.freeswitch.org/ ClueCon http://www.cluecon.com/ Twitter: http://twitter.com/FreeSWITCH_wire AIM: anthm MSN:anthony_minessale@hotmail.com GTALK/JABBER/PAYPAL:anthony.minessale@gmail.com IRC: irc.freenode.net #freeswitch FreeSWITCH Developer Conference sip:888@conference.freeswitch.org googletalk:conf+888@conference.freeswitch.org pstn:+19193869900
On 04/02/2011 10:01 AM, Anthony Minessale wrote:
Reliability is more important than feature. Asterisk is a good example of
such case. I snipped a very small part of this thread but the above line provided my comedy relief for the day. Thanks, I needed it.
-- Darren Wiebe Aleph Communications -------------------- Phone: 1-877-702-2900 Fax: 1-866-274-4506 Email: darren@aleph-com.net
I actually spilt my tea. :) On 4 February 2011 17:09, Darren Wiebe <darren@aleph-com.net> wrote:
On 04/02/2011 10:01 AM, Anthony Minessale wrote:
Reliability is more important than feature. Asterisk is a good example of
such case. I snipped a very small part of this thread but the above line provided my comedy relief for the day. Thanks, I needed it.
-- Darren Wiebe Aleph Communications -------------------- Phone: 1-877-702-2900 Fax: 1-866-274-4506 Email: darren@aleph-com.net
_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
Steven Ayre wrote:
I actually spilt my tea. :)
Honestly, I have no idea what your statement means. However, I am using both Asterisk and FreeSWITCH. The only reason I still keep my Asterisk is because it is configured with a BlackList and incoming calls with CIDs not listed in the BlackList will be rejected. When I find a way to do this, I definitely will decommission my Asterisk. ----- don't and stop are the ONLY two 4-letter words considered offensive to men, but not when used together. -- View this message in context: http://freeswitch-users.2379917.n2.nabble.com/Auto-test-FS-install-now-neces... Sent from the freeswitch-users mailing list archive at Nabble.com.
On 04/02/2011 10:34 AM, mazilo wrote:
Steven Ayre wrote:
I actually spilt my tea. :) Honestly, I have no idea what your statement means. However, I am using both Asterisk and FreeSWITCH. The only reason I still keep my Asterisk is because it is configured with a BlackList and incoming calls with CIDs not listed in the BlackList will be rejected. When I find a way to do this, I definitely will decommission my Asterisk.
----- don't and stop are the ONLY two 4-letter words considered offensive to men, but not when used together.
That should be fairly easy using either mod_xml_curl or lua. -- Darren Wiebe Aleph Communications -------------------- Phone: 1-877-702-2900 Fax: 1-866-274-4506 Email: darren@aleph-com.net
So it's a whitelist, and that is quite easy to do with FreeSWITCH, with some LUA, and a SQL DB. David Ponzone Direction Technique email: david.ponzone@ipeva.fr tel: 01 74 03 18 97 gsm: 06 66 98 76 34 Service Client IPeva tel: 0811 46 26 26 www.ipeva.fr - www.ipeva-studio.com Ce message et toutes les pièces jointes sont confidentiels et établis à l'intention exclusive de ses destinataires. Toute utilisation ou diffusion non autorisée est interdite. Tout message électronique est susceptible d'altération. IPeva décline toute responsabilité au titre de ce message s'il a été altéré, déformé ou falsifié. Si vous n'êtes pas destinataire de ce message, merci de le détruire immédiatement et d'avertir l'expéditeur. Le 04/02/2011 à 18:34, mazilo a écrit :
Steven Ayre wrote:
I actually spilt my tea. :)
Honestly, I have no idea what your statement means. However, I am using both Asterisk and FreeSWITCH. The only reason I still keep my Asterisk is because it is configured with a BlackList and incoming calls with CIDs not listed in the BlackList will be rejected. When I find a way to do this, I definitely will decommission my Asterisk.
----- don't and stop are the ONLY two 4-letter words considered offensive to men, but not when used together. -- View this message in context: http://freeswitch-users.2379917.n2.nabble.com/Auto-test-FS-install-now-neces... Sent from the freeswitch-users mailing list archive at Nabble.com.
_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
David Ponzone wrote:
So it's a whitelist, and that is quite easy to do with FreeSWITCH, with some LUA, and a SQL DB.
If anyone out here is willing to provide a howto, especially without an SQL DB, that will be cool. I did this on Asterisk using the built-in database command with the astdb file. ----- don't and stop are the ONLY two 4-letter words considered offensive to men, but not when used together. -- View this message in context: http://freeswitch-users.2379917.n2.nabble.com/Auto-test-FS-install-now-neces... Sent from the freeswitch-users mailing list archive at Nabble.com.
why not use a file like hosts.deny from a perl/php/other script called with system ? ----- Original Message ----- From: "mazilo" <Nabble@slickdeals.endjunk.com> To: <freeswitch-users@lists.freeswitch.org> Sent: Friday, February 04, 2011 2:27 PM Subject: Re: [Freeswitch-users] Auto test FS install - now necessary?
David Ponzone wrote:
So it's a whitelist, and that is quite easy to do with FreeSWITCH, with some LUA, and a SQL DB.
If anyone out here is willing to provide a howto, especially without an SQL DB, that will be cool. I did this on Asterisk using the built-in database command with the astdb file.
----- don't and stop are the ONLY two 4-letter words considered offensive to men, but not when used together. -- View this message in context: http://freeswitch-users.2379917.n2.nabble.com/Auto-test-FS-install-now-neces... Sent from the freeswitch-users mailing list archive at Nabble.com.
_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
Madovsky wrote:
why not use a file like hosts.deny from a perl/php/other script called with system ? IIRC, the host.deny only filters incoming calls from specific hosts. For instance, if you want to block crackers out there who try to exploit your FS system, perhaps this is not a bad idea, but not if you only want to allow some specific incoming CIDs to be able to reach your internal extensions.
----- don't and stop are the ONLY two 4-letter words considered offensive to men, but not when used together. -- View this message in context: http://freeswitch-users.2379917.n2.nabble.com/Auto-test-FS-install-now-neces... Sent from the freeswitch-users mailing list archive at Nabble.com.
I meant hosts.deny was an example, you can use your own file ----- Original Message ----- From: "mazilo" <Nabble@slickdeals.endjunk.com> To: <freeswitch-users@lists.freeswitch.org> Sent: Friday, February 04, 2011 3:23 PM Subject: Re: [Freeswitch-users] Auto test FS install - now necessary?
Madovsky wrote:
why not use a file like hosts.deny from a perl/php/other script called with system ? IIRC, the host.deny only filters incoming calls from specific hosts. For instance, if you want to block crackers out there who try to exploit your FS system, perhaps this is not a bad idea, but not if you only want to allow some specific incoming CIDs to be able to reach your internal extensions.
----- don't and stop are the ONLY two 4-letter words considered offensive to men, but not when used together. -- View this message in context: http://freeswitch-users.2379917.n2.nabble.com/Auto-test-FS-install-now-neces... Sent from the freeswitch-users mailing list archive at Nabble.com.
_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
If you misunderstood my point and ruined your keyboard, your own fault! My friend. On Feb 4, 2011 9:21 AM, "Steven Ayre" <steveayre@gmail.com> wrote:
I actually spilt my tea. :)
On 4 February 2011 17:09, Darren Wiebe <darren@aleph-com.net> wrote:
On 04/02/2011 10:01 AM, Anthony Minessale wrote:
Reliability is more important than feature. Asterisk is a good example of
such case. I snipped a very small part of this thread but the above line provided my comedy relief for the day. Thanks, I needed it.
-- Darren Wiebe Aleph Communications -------------------- Phone: 1-877-702-2900 Fax: 1-866-274-4506 Email: darren@aleph-com.net
_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
Missed the laptop thankfully ;) On 4 February 2011 22:04, jesse <chat2jesse@gmail.com> wrote:
If you misunderstood my point and ruined your keyboard, your own fault! My friend. On Feb 4, 2011 9:21 AM, "Steven Ayre" <steveayre@gmail.com> wrote:
I actually spilt my tea. :)
On 4 February 2011 17:09, Darren Wiebe <darren@aleph-com.net> wrote:
On 04/02/2011 10:01 AM, Anthony Minessale wrote:
Reliability is more important than feature. Asterisk is a good example of
such case. I snipped a very small part of this thread but the above line provided my comedy relief for the day. Thanks, I needed it.
-- Darren Wiebe Aleph Communications -------------------- Phone: 1-877-702-2900 Fax: 1-866-274-4506 Email: darren@aleph-com.net
_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE: http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
I think you misunderstand me. I meant asterisk is a good example of bad code quality system. On Feb 4, 2011 9:12 AM, "Darren Wiebe" <darren@aleph-com.net> wrote:
On 04/02/2011 10:01 AM, Anthony Minessale wrote:
Reliability is more important than feature. Asterisk is a good example of
such case. I snipped a very small part of this thread but the above line provided my comedy relief for the day. Thanks, I needed it.
-- Darren Wiebe Aleph Communications -------------------- Phone: 1-877-702-2900 Fax: 1-866-274-4506 Email: darren@aleph-com.net
_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
Just to say I think you all do a great job with FreeSWITCH. :) -Steve On 4 February 2011 17:01, Anthony Minessale <anthony.minessale@gmail.com>wrote:
I guess we should try to be more like all the other open source projects where every single day of GIT HEAD is perfect. hmm....
The alternative is you can buy a soft-switch and live with it's faults and pay annually for an update. Or even wait for our official releases which are very infrequent and not well-supported once they are more than a few months old.
Also don't forget we use FreeSWITCH in a real product so you don't even see the amount of testing and QA we inherit from that product as well as all the large carriers who use FreeSWITCH and do their best to try beta version for us. Thank you all.....
We have plans to stable branch soon. I guess that means everyone will flock to that and it will be the end of my beta testers. We do not only get regressions from features we get them sometimes from fixes as well, other peoples patches, many factors.
I guess what's really saddening since you bring up asterisk, is that I spent years using it with perpetual problems in the release or the daily snapshot and it sort of has a reputation for this kind of thing and all you hear about is how awesome it is from people. I don't appreciate being treated like a witch hunt over one regression especially when we can fix it in 1 minute. That double-standard is ridiculous.
I should just make a FAQ about this and paste in the URL:
We only have so much time to give to the world, we spend like 8-14 hours a day on average working on the code. Any suggestions or improvements (even the ones with the best intentions) are only viable if they come with volunteers.
Bottom line is by downloading and building the code you are opting in to our community which involves occasional speed bumps.
On Fri, Feb 4, 2011 at 10:34 AM, jesse <chat2jesse@gmail.com> wrote:
I do not see any piece of unit testing code neither. Reliability is more important than feature. Asterisk is a good example of such case.
I would recommend developers take a pause to add unit testing cases and auto integration testings. Also enforce any future code must be accompanied by unit testing code.
On Feb 3, 2011 9:47 PM, "Avi Marcus" <avi@avimarcus.net> wrote:
Maybe I'm just being paranoid, but I've been seeing a lot of "I just updated to latest git and X broke" And that x is always something else... I'm not a seasoned coder on this magnitude, but perhaps we can develop a test suite (at least part of one) to test that a fresh install is functioning? A FULL test would probably be a huge undertaking and involve recqual for testing media, too... Also, each mod would need to have it's own test suite. Due to the rapid updating of FreeSWITCH this not be what you want to put your effort into, but... Thoughts? On Feb 4, 2011 4:04 AM, "Anirudha Shimpi" <anu@familytv.com> wrote:
Just downloaded (from git) and installed FS 1.0.7, however, every time I do originate user/1001 &park, FS crashes. Running Centos 5.5 64 bit.
_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
-- Anthony Minessale II
FreeSWITCH http://www.freeswitch.org/ ClueCon http://www.cluecon.com/ Twitter: http://twitter.com/FreeSWITCH_wire
AIM: anthm MSN:anthony_minessale@hotmail.com <MSN%3Aanthony_minessale@hotmail.com> GTALK/JABBER/PAYPAL:anthony.minessale@gmail.com<PAYPAL%3Aanthony.minessale@gmail.com> IRC: irc.freenode.net #freeswitch
FreeSWITCH Developer Conference sip:888@conference.freeswitch.org <sip%3A888@conference.freeswitch.org> googletalk:conf+888@conference.freeswitch.org<googletalk%3Aconf%2B888@conference.freeswitch.org> pstn:+19193869900
_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
I hope that we don't ever return to the old days of spending a year or more regression testing PBX firmware between each release. That was okay, if you did not mind having a product that worked really well but did not really do much and was always about a year behind the cutting edge. I applaud all of the FS Devs, I believe that everyone is doing their best to avoid introducing new issues as work is done to improve the code base and add new features and functionality. We are happy to participate in the testing and feedback process. We hope to be able to make meaningful contributions down the road when we are more familiar with things. I have never seen a major issue that was not corrected with lightning speed. It seems that even whenever someone reports an issue that is important to them, they are likewise addressed very rapidly. In my opinion, this is just the way that software development goes. Do your best to modify (improve the project) code without causing other issues. Unfortunately, sometimes this happens. Especially when there so many people working to make contributions to grow and improve the product for all. Frankly, I think the project is progressing very well. On 02/04/2011 10:01 AM, Anthony Minessale wrote:
I guess we should try to be more like all the other open source projects where every single day of GIT HEAD is perfect. hmm....
The alternative is you can buy a soft-switch and live with it's faults and pay annually for an update. Or even wait for our official releases which are very infrequent and not well-supported once they are more than a few months old.
Also don't forget we use FreeSWITCH in a real product so you don't even see the amount of testing and QA we inherit from that product as well as all the large carriers who use FreeSWITCH and do their best to try beta version for us. Thank you all.....
We have plans to stable branch soon. I guess that means everyone will flock to that and it will be the end of my beta testers. We do not only get regressions from features we get them sometimes from fixes as well, other peoples patches, many factors.
I guess what's really saddening since you bring up asterisk, is that I spent years using it with perpetual problems in the release or the daily snapshot and it sort of has a reputation for this kind of thing and all you hear about is how awesome it is from people. I don't appreciate being treated like a witch hunt over one regression especially when we can fix it in 1 minute. That double-standard is ridiculous.
I should just make a FAQ about this and paste in the URL:
We only have so much time to give to the world, we spend like 8-14 hours a day on average working on the code. Any suggestions or improvements (even the ones with the best intentions) are only viable if they come with volunteers.
Bottom line is by downloading and building the code you are opting in to our community which involves occasional speed bumps.
In my personal experience, I also switched from asterisk to freeswitch. From engineering perspective, a good process will befits the project in the long run, especially as it becomes more complicated and involves more people. On Feb 4, 2011 9:34 AM, "Phone" <phone.bytes@gmail.com> wrote:
I hope that we don't ever return to the old days of spending a year or more regression testing PBX firmware between each release. That was okay, if you did not mind having a product that worked really well but did not really do much and was always about a year behind the cutting edge.
I applaud all of the FS Devs, I believe that everyone is doing their best to avoid introducing new issues as work is done to improve the code base and add new features and functionality.
We are happy to participate in the testing and feedback process. We hope to be able to make meaningful contributions down the road when we are more familiar with things.
I have never seen a major issue that was not corrected with lightning speed. It seems that even whenever someone reports an issue that is important to them, they are likewise addressed very rapidly.
In my opinion, this is just the way that software development goes. Do your best to modify (improve the project) code without causing other issues. Unfortunately, sometimes this happens. Especially when there so many people working to make contributions to grow and improve the product for all. Frankly, I think the project is progressing very well.
On 02/04/2011 10:01 AM, Anthony Minessale wrote:
I guess we should try to be more like all the other open source projects where every single day of GIT HEAD is perfect. hmm....
The alternative is you can buy a soft-switch and live with it's faults and pay annually for an update. Or even wait for our official releases which are very infrequent and not well-supported once they are more than a few months old.
Also don't forget we use FreeSWITCH in a real product so you don't even see the amount of testing and QA we inherit from that product as well as all the large carriers who use FreeSWITCH and do their best to try beta version for us. Thank you all.....
We have plans to stable branch soon. I guess that means everyone will flock to that and it will be the end of my beta testers. We do not only get regressions from features we get them sometimes from fixes as well, other peoples patches, many factors.
I guess what's really saddening since you bring up asterisk, is that I spent years using it with perpetual problems in the release or the daily snapshot and it sort of has a reputation for this kind of thing and all you hear about is how awesome it is from people. I don't appreciate being treated like a witch hunt over one regression especially when we can fix it in 1 minute. That double-standard is ridiculous.
I should just make a FAQ about this and paste in the URL:
We only have so much time to give to the world, we spend like 8-14 hours a day on average working on the code. Any suggestions or improvements (even the ones with the best intentions) are only viable if they come with volunteers.
Bottom line is by downloading and building the code you are opting in to our community which involves occasional speed bumps.
_______________________________________________ FreeSWITCH-users mailing list FreeSWITCH-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
participants (12)
-
Anthony Minessale
-
Avi Marcus
-
Darren Wiebe
-
David Ponzone
-
guru singh
-
jesse
-
Madovsky
-
mazilo
-
Michael Collins
-
Phone
-
Steven Ayre
-
Tim St. Pierre