Hi Committers, I want to propose a change in the current tracker movement on sourceforge when we assign for peer review.
Currently sf-robot is closing the tracker after 15 days on pending status.
I would propose that we configure the trackers to NOT close automatically by sf-robot.
Now, when a committer solve a bug or integrate a contribution/patch then committer will put the tracker in Pending status and assign it to his peer-reviewer.
When the peer-review accept it, he assign the tracker again to original committer and close the tracker.
What do you think?
It will be easier to control which trackers are pending by peer-review.
Regards,
Carlos Ruiz
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
>I would propose that we configure the trackers to NOT close automatically by sf-robot.
>
>Now, when a committer solve a bug or integrate a contribution/patch then committer will put the tracker in Pending >status and assign it to his peer-reviewer.
>
>When the peer-review accept it, he assign the tracker again to original committer and close the tracker.
It sounds good to me.
[+1] vote
Trifon
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi Committers, I think this proposed procedure is not working, and this is affecting our statistics since july.
Most of committers are putting the solved issues in Pending, but mostly peer review is not happening, and if it's happening they are not putting the tracker in closed status.
I think is better to back again to the previous state where the sf-robot close the tracker after 15 days.
What do you think?
Regards,
Carlos Ruiz
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
>Hi Committers, I think this proposed procedure is not working, and this is affecting our statistics since july.
>
>Most of committers are putting the solved issues in Pending, but mostly peer review is not happening, and if it's >happening they are not putting the tracker in closed status.
>
>I think is better to back again to the previous state where the sf-robot close the tracker after 15 days.
>
>What do you think?
[+1] Less rules -> more contributors -> more contributed code.
Kind regards,
Trifon
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
> [+1] Less rules ->
> more contributors ->
> more contributed code.
:-) Hope you're not misunderstanding the issue.
It looks like you're worried by quantity - meanwhile I keep worried about quality.
So, we need to find balance - and that's what we try to do permanently.
I would like to see many more contributed code - but for that we also need to have many more peer reviewers, or at least to have some commitment from the few reviewers we have.
Regards,
Carlos Ruiz
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I'm against automatic closure of any issue, bug or patches.
Wathever they have inside, they are real requests and should be reviewed.
I believe, that commiters could do closing them self, if they are sure issue is resolved.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
>I'm against automatic closure of any issue, bug or patches.
>Wathever they have inside, they are real requests and should be reviewed.
>I believe, that commiters could do closing them self, if they are sure issue is resolved.
Yes this is good wish, but question is reviewed by who?
Original question of Carlos was:
>but mostly peer review is not happening, and if it's happening they are not putting the tracker in closed status.
Kind regards,
Trifon
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
To clarify the situation, Red reports a bug, Blu fixes the bug and sets the status to Pending so that Red can agree it is fixed. But Red has now what he needs and forgets to close the report, or doesn't care enough to go do it.
Carlos is proposing that Bugs placed in Pending (which means the fix is there, just needs review from original person) will be closed automatically after a period.
+1 for this, it's good for stats. It can be re-opened or reported again if another issue surfaces.
Thanks,
Joel S
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
HI Matjaž Godec,
Closed threads are often opened again for review when need arise. This has been happening well so far. Nevertheless all our rules are open to review, so this new revert to an old rule is subject to change if it is not working.
But its good u give your opinion for everyone to take note and take action as required.
Trifon, i dont want to argue about rules. I would say:
MORE QUALITY > MORE BUSINESS > MORE MONEY > MORE CONTRIBUTION > MORE HOLIDAY
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Trifon,
I thought your mathematics is superb. I am saying with the symbols '>' - more than, not '=' equals to. :-).
Ok, ok, actually i think you caught me there! I like this. Removes boredom from my life of just eating and typing. Ok, how do i extricate myself from another Trifon's very smart Bulgarian chess move?
It is like this brother:
Carlos Ruiz is (like you) a very important contributor. And contributors are priceless. Thus whatever they do or say are very important and priceless. Whatever such ppl say or do, we pay real deep attention before we reply (just like chess grandmasters) and when we reply wrongly, we get smashed and will be very sorry (just like grandmaster). They keyword Carlos used was 'quality' as versus 'less rules' which can compromise that. Of course to me that measurement is subjective because in the bazaar culture, no rules can ultimately brings highest quality! Here are the chess moves:
NO RULES > ABSOLUTE CHAOS > ABSOLUTE LOSS OF QUALITY > LOTS OF BUGS > LOTS OF WORK FOR HOBBYISTS > LOTS OF CODERS CROWD AROUND > LOTS OF EYEBALLS MAKES ANY BUG SHALLOW -- last one from Linus Torvalds (my great great grandmaster)
But Carlos ask for balance so i put it in a balancing chess equation:
If there are some balanced rules without compromising the bazaar spirit where the balance also means in between Community and Commerce, then the more quality should attract more business/money and in turn that allows more contributors to work on it in many ways:
1) Company that gets free quality software begins to sponsor contributors like you are getting alot nowadays, from what i read :-)
2) Newbies who sees that this makes money so are drawn to become like you too
I can go on and on, but i know what u driving at my knight not my bishop. "Why red1 puts the knight and bishop in one line. Previously he always sees them as two battlefields. Do not talk about money when contributing. Well. I think i must have been influenced by listening to a lot of your comments in SF. I now am thinking:
a) Get more money
b) Pay the poor contributors
c) Get more quality features
d) Sell more services
e) Go back to (a)
red1 to f3
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Actually an equation of No Rules <> Attracts Hoobyists should bring about an 'S' wave curve in the graph of quality. Which means at first it should drop for a while and looks bad software and then suddenly with so many hobbyists it curves up sharply. I assumed Linux went throught that in early phase. But i cannot find any study of those very early phase from that aspect. If anyone knows if this is true, please share.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Well, I find strange how you keep thinking in communities as no-rules, and how you keep thinking in no-rules as more-quality, and how you keep putting linux and wikipedia as examples.
No rules in wikipedia?? Are you joking?
Just try to open your biography page to see what happens in a matter of minutes.
Just see the history of how Redhuan fought to open Adempiere wikipedia page.
Hmmm, do you see what I see? The quality of wikipedia is maintained by an ARMY of volunteers that keeps making peer review to ensure the wikipedia policies.
And you know even such approach is criticized by one of his creators (Larry Sanger) who forked and created http://en.citizendium.org to harden the rules.
No rules in linux?? Are you joking?
Hmmm, this is repetitive, I remember Heng Sin recently shared lots of links with us showing the strict ruling that linux have.
And again, what do we have in linux? An army of developers following the rules trying to get their patches in the main tree, and a small contingent of paid peer-reviewers that enforce the rules.
Well, I just imagine how would be the quality of linux or wikipedia without those strict rules.
I think you confuse openness with no-rules.
And here in Adempiere we have a BIG disadvantage (tending to decline).
Just compare the skills you need to make a peer review in wikipedia against the skill you need to do that in linux or adempiere.
And I said tending to decline - as you see this small contingent growing every day - hoping to become an army of peer reviewers :-)
Just one example: I'm glad to see a very active Tony enforcing the numbering rule for migration scripts, and keeping the trunk compilable and runnable, along with some important contributions.
NO-RULES remembered me the principles of anarchism - well, if you were closer to anarchism you must noted that anarchism meant to be some sort of self-ruling, in other words - there were rules, but they are enforced by the individuals on their own, not by a committee or government.
And here is where Mark words sound strong on the self-ruling that we need to be adempiere committers.
"At the end of the Day: I'll stay responsible for my changes i put into trunk - reviewed or not. If i break it, i'll fix it - fast!"
So, finally we arrive to the OLD known conclusion: to have autonomy you need a high degree of responsibility
And that's enforced mostly by the community - if you act irresponsible your autonomy is restricted, and this is mostly a natural process. Like when you live in a community and you leave your garbages all over the site.
Regards,
Carlos Ruiz
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
/me moves bishop and knight out of Carlos Castle's X and Y controlling flanks.
(I didnt know you play chess. But i see Colombian chess is very mid-game. Too much dependence on castles:-> )
Your points are mostly valid. It is just that there is always two schools of thought. Read any good book in economics or chess. Or even Java programming or mafia story. We can either put rules first or put rules last.
Wikipedia was born from the story of Nupedia, an effort between Jimmy Wales and Larry Sanger (see http://en.wikipedia.org/wiki/Nupedia ). It started as an important argument and parting of ways to the present legend. What did Larry said when Jimmy one day come running to Larry very excited that he found WikiMedia as a tool to allow anyone even anon to edit and then other can review. Larry thinks this is nuts and dangerous as encyclopedic content is more classic than java codings. There is only facts but dry facts. And Larry knows that if u allow 1000 professors to argue, you will have afhganistan all over.
So Jimmy and Larry parted ways. That was before. Now AFTER. Rules began to form, but they evolve with peer review. Just like what you doing right now. You do not make a decision and say, "Obey or die". You ask forums many times, and some times ppl do not answer. Happens to me all the time. Boring chess game when no one wants to play. You evolved your landed tracker by manual and now back to landing by auto-pilot. And it doesnt seem dictatorial at all. No one flames u, or criticise incessantly. Because u ask first, to review a rule. Now that is what i meant by anarchy. Yeah, its still a government, but violence has different context in cyberspace.
Linus is said to be promiscously open. Doesnt means he sleeps around with the committers, but that he very quickly lets peers in and try to see beyond the rules. Which is exactly what u advocate and have been doing very well AFAIS. What we are doing actually is playing the same chess board, but from different angles, and of course different colors.
/me offers gambit
red1
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
You didn't offer a gambit, you offered a sacrifice :D
That sounded really different, it's different to say "Less rules" or "No-rules" that saying "More logic and natural rules open to be reviewed when necessary"
Yes, like the Bobby Fischer criticism to chess, he preferred a random chess, with more flexible rules: http://www.chessvariants.com/diffsetup.dir/fischer.html
never tried to play that - but I suppose is more consistent with real life, where the rules are changing and need to be reviewed from time to time.
Now I remember that's one of the the most common question from people
- what are the rules to contribute?
- where are the rules to commit?
- etc
And we lack from proper documentation of the ruling - very possibly with more well documented, logic and natural rules the contribution process will be easier.
Regards,
Carlos Ruiz
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi all,
This is my take from my own experience.
I have been a professional software developer for more than 10 years, and I have been hacking Adempiere code the last couple of months.
From experience, it takes time for a developer, however good, to come to grasp with the intricasies of a large software project.
We as a firm have solved that by dividing our project into areas only which certain people can make changes to, and other general areas that a new person could make changes to.
As relates to Adempiere, and yes I am writing as a new contributor, we need controls on the submission of code. I have been in the game long enough to know that I don't know Adempiere well enough to go changing / fixing bugs in the Adempiere Core coz while I might fix a bug, I might break things unknown to me.
Considering the importance of this project, we could maybe have two types of members. The experienced ones in the project and newbies. A newbies code has to be reviewed and when they fix a bug, it cannot be automaticaly closed until a senior member approves.
As for the experienced ones, they can be automatically closed after a certain time.
Regards
Ngigi Waithaka
Alliance Technologies
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
But to support Trifon's statement i am advocate that less rules is the ideal, but how less? For example i like to just add developers into our list without question and kick them also without answers, but i dont think u like that anarchic practice. You like them to show u their work and it must be contributed seriously and not as marketing ploy.
We need rules but we must show that rules are made to protect contributors and avoid disruption to their work. I think this is what Carlos is looking for. We have rules but we approve very easily. See how the last few days we admitted many victims, i mean new developers.
And here is proof that we have no rules against contrarian opinion. They can object frankly without fear of been shut up. This is the bazaar air that we have. Can u breathe well in here?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi all,
I want to tell you a little bit about my life before metas and ADempiere
- for those who don't want to read it, just jump to the end, theres my vote ;-):
In December 2007 it was 9 years ago, since i started working as Software-Developer (i began with Programming in Cobol ;-) - greetings to red1) in a big insurance company. From the first day on i was able to commit changes and Bugfixes into production. And i did it. It was very exiting in the beginning, because when you failed as committer the production was corrupted and the eyes of the whole company would be looking at you. If this happened it was the highest prioriy to fix it again. As you can see, i survived the employment and the big insurance company has even become much bigger since then. ;-)
If i was unsure about Fixes it was possible to ask others to review my work. At the Beginning i asked a lot of people. At the end people asked me alot. But i never ended asking others to review my work when i decided it would be better. But important thing was always the same: If reviewed or not, the committer was always responsible for what he puts in production.
I am sure that all committers listed under http://www.adempiere.com/wiki/index.php/Committers are trusted to have a very good quality awareness and are willing to ask someone else for a peer review of a Bugfix if they decide that it would be better. This review should then be done in a short timeframe (<15 days).
Now that I'm a new committer for ADempiere, i feel a bit like having a time travel 9 years back. I have a mentor - thanxs alot Carlos - and i'll ask him to review (and the others don't be afraid, i'll ask you too ;-) ).
At the end of the Day: I'll stay responsible for my changes i put into trunk - reviewed or not. If i break it, i'll fix it - fast!
Therefore my vote:
[+1] for closing the Bug-Requests in status "pending" after 15 Days.
Greetings
Mark
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi Committers, I want to propose a change in the current tracker movement on sourceforge when we assign for peer review.
Currently sf-robot is closing the tracker after 15 days on pending status.
I would propose that we configure the trackers to NOT close automatically by sf-robot.
Now, when a committer solve a bug or integrate a contribution/patch then committer will put the tracker in Pending status and assign it to his peer-reviewer.
When the peer-review accept it, he assign the tracker again to original committer and close the tracker.
What do you think?
It will be easier to control which trackers are pending by peer-review.
Regards,
Carlos Ruiz
Good idea Carlos,
[+1] from me.
Regards,
Karsten
+1
Regards,
Low
Hi Carlos,
>I would propose that we configure the trackers to NOT close automatically by sf-robot.
>
>Now, when a committer solve a bug or integrate a contribution/patch then committer will put the tracker in Pending >status and assign it to his peer-reviewer.
>
>When the peer-review accept it, he assign the tracker again to original committer and close the tracker.
It sounds good to me.
[+1] vote
Trifon
[+1] from me
Regards,
Armen
That's OK for me..
+1
Fernando
Hi Committers, I think this proposed procedure is not working, and this is affecting our statistics since july.
Most of committers are putting the solved issues in Pending, but mostly peer review is not happening, and if it's happening they are not putting the tracker in closed status.
I think is better to back again to the previous state where the sf-robot close the tracker after 15 days.
What do you think?
Regards,
Carlos Ruiz
Hi Carlso,
>Hi Committers, I think this proposed procedure is not working, and this is affecting our statistics since july.
>
>Most of committers are putting the solved issues in Pending, but mostly peer review is not happening, and if it's >happening they are not putting the tracker in closed status.
>
>I think is better to back again to the previous state where the sf-robot close the tracker after 15 days.
>
>What do you think?
[+1] Less rules -> more contributors -> more contributed code.
Kind regards,
Trifon
> [+1] Less rules ->
> more contributors ->
> more contributed code.
:-) Hope you're not misunderstanding the issue.
It looks like you're worried by quantity - meanwhile I keep worried about quality.
So, we need to find balance - and that's what we try to do permanently.
I would like to see many more contributed code - but for that we also need to have many more peer reviewers, or at least to have some commitment from the few reviewers we have.
Regards,
Carlos Ruiz
Hi,
-1
I'm against automatic closure of any issue, bug or patches.
Wathever they have inside, they are real requests and should be reviewed.
I believe, that commiters could do closing them self, if they are sure issue is resolved.
hi Matias,
>I'm against automatic closure of any issue, bug or patches.
>Wathever they have inside, they are real requests and should be reviewed.
>I believe, that commiters could do closing them self, if they are sure issue is resolved.
Yes this is good wish, but question is reviewed by who?
Original question of Carlos was:
>but mostly peer review is not happening, and if it's happening they are not putting the tracker in closed status.
Kind regards,
Trifon
Hi agenda_gm,
To clarify the situation, Red reports a bug, Blu fixes the bug and sets the status to Pending so that Red can agree it is fixed. But Red has now what he needs and forgets to close the report, or doesn't care enough to go do it.
Carlos is proposing that Bugs placed in Pending (which means the fix is there, just needs review from original person) will be closed automatically after a period.
+1 for this, it's good for stats. It can be re-opened or reported again if another issue surfaces.
Thanks,
Joel S
+1. We are always open to such review.
Hi Carlos,
Yes, you're right.
[+1] for me.
Regards,
Alejandro
HI Matjaž Godec,
Closed threads are often opened again for review when need arise. This has been happening well so far. Nevertheless all our rules are open to review, so this new revert to an old rule is subject to change if it is not working.
But its good u give your opinion for everyone to take note and take action as required.
Trifon, i dont want to argue about rules. I would say:
MORE QUALITY > MORE BUSINESS > MORE MONEY > MORE CONTRIBUTION > MORE HOLIDAY
hi Red1,
>Trifon, i dont want to argue about rules. I would say:
>
>MORE QUALITY > MORE BUSINESS > MORE MONEY > MORE CONTRIBUTION > MORE HOLIDAY
Red1 it is the first time you write that contributions come from money.
Are you sure?
Kind regards,
Trifon
Trifon,
I thought your mathematics is superb. I am saying with the symbols '>' - more than, not '=' equals to. :-).
Ok, ok, actually i think you caught me there! I like this. Removes boredom from my life of just eating and typing. Ok, how do i extricate myself from another Trifon's very smart Bulgarian chess move?
It is like this brother:
Carlos Ruiz is (like you) a very important contributor. And contributors are priceless. Thus whatever they do or say are very important and priceless. Whatever such ppl say or do, we pay real deep attention before we reply (just like chess grandmasters) and when we reply wrongly, we get smashed and will be very sorry (just like grandmaster). They keyword Carlos used was 'quality' as versus 'less rules' which can compromise that. Of course to me that measurement is subjective because in the bazaar culture, no rules can ultimately brings highest quality! Here are the chess moves:
NO RULES > ABSOLUTE CHAOS > ABSOLUTE LOSS OF QUALITY > LOTS OF BUGS > LOTS OF WORK FOR HOBBYISTS > LOTS OF CODERS CROWD AROUND > LOTS OF EYEBALLS MAKES ANY BUG SHALLOW -- last one from Linus Torvalds (my great great grandmaster)
But Carlos ask for balance so i put it in a balancing chess equation:
If there are some balanced rules without compromising the bazaar spirit where the balance also means in between Community and Commerce, then the more quality should attract more business/money and in turn that allows more contributors to work on it in many ways:
1) Company that gets free quality software begins to sponsor contributors like you are getting alot nowadays, from what i read :-)
2) Newbies who sees that this makes money so are drawn to become like you too
I can go on and on, but i know what u driving at my knight not my bishop. "Why red1 puts the knight and bishop in one line. Previously he always sees them as two battlefields. Do not talk about money when contributing. Well. I think i must have been influenced by listening to a lot of your comments in SF. I now am thinking:
a) Get more money
b) Pay the poor contributors
c) Get more quality features
d) Sell more services
e) Go back to (a)
red1 to f3
Actually an equation of No Rules <> Attracts Hoobyists should bring about an 'S' wave curve in the graph of quality. Which means at first it should drop for a while and looks bad software and then suddenly with so many hobbyists it curves up sharply. I assumed Linux went throught that in early phase. But i cannot find any study of those very early phase from that aspect. If anyone knows if this is true, please share.
CarlosRuiz move - castle
Well, I find strange how you keep thinking in communities as no-rules, and how you keep thinking in no-rules as more-quality, and how you keep putting linux and wikipedia as examples.
No rules in wikipedia?? Are you joking?
Just try to open your biography page to see what happens in a matter of minutes.
Just see the history of how Redhuan fought to open Adempiere wikipedia page.
No rules? In fact wikipedia is a compendium of rules - and an army of editors forcing "community" to accomplish the rules.
http://en.wikipedia.org/wiki/Category:Wikipedia_official_policy
Hmmm, do you see what I see? The quality of wikipedia is maintained by an ARMY of volunteers that keeps making peer review to ensure the wikipedia policies.
And you know even such approach is criticized by one of his creators (Larry Sanger) who forked and created http://en.citizendium.org to harden the rules.
No rules in linux?? Are you joking?
Hmmm, this is repetitive, I remember Heng Sin recently shared lots of links with us showing the strict ruling that linux have.
And again, what do we have in linux? An army of developers following the rules trying to get their patches in the main tree, and a small contingent of paid peer-reviewers that enforce the rules.
Well, I just imagine how would be the quality of linux or wikipedia without those strict rules.
I think you confuse openness with no-rules.
And here in Adempiere we have a BIG disadvantage (tending to decline).
Just compare the skills you need to make a peer review in wikipedia against the skill you need to do that in linux or adempiere.
And I said tending to decline - as you see this small contingent growing every day - hoping to become an army of peer reviewers :-)
Just one example: I'm glad to see a very active Tony enforcing the numbering rule for migration scripts, and keeping the trunk compilable and runnable, along with some important contributions.
NO-RULES remembered me the principles of anarchism - well, if you were closer to anarchism you must noted that anarchism meant to be some sort of self-ruling, in other words - there were rules, but they are enforced by the individuals on their own, not by a committee or government.
And here is where Mark words sound strong on the self-ruling that we need to be adempiere committers.
"At the end of the Day: I'll stay responsible for my changes i put into trunk - reviewed or not. If i break it, i'll fix it - fast!"
So, finally we arrive to the OLD known conclusion: to have autonomy you need a high degree of responsibility
And that's enforced mostly by the community - if you act irresponsible your autonomy is restricted, and this is mostly a natural process. Like when you live in a community and you leave your garbages all over the site.
Regards,
Carlos Ruiz
/me moves bishop and knight out of Carlos Castle's X and Y controlling flanks.
(I didnt know you play chess. But i see Colombian chess is very mid-game. Too much dependence on castles:-> )
Your points are mostly valid. It is just that there is always two schools of thought. Read any good book in economics or chess. Or even Java programming or mafia story. We can either put rules first or put rules last.
Wikipedia was born from the story of Nupedia, an effort between Jimmy Wales and Larry Sanger (see http://en.wikipedia.org/wiki/Nupedia ). It started as an important argument and parting of ways to the present legend. What did Larry said when Jimmy one day come running to Larry very excited that he found WikiMedia as a tool to allow anyone even anon to edit and then other can review. Larry thinks this is nuts and dangerous as encyclopedic content is more classic than java codings. There is only facts but dry facts. And Larry knows that if u allow 1000 professors to argue, you will have afhganistan all over.
So Jimmy and Larry parted ways. That was before. Now AFTER. Rules began to form, but they evolve with peer review. Just like what you doing right now. You do not make a decision and say, "Obey or die". You ask forums many times, and some times ppl do not answer. Happens to me all the time. Boring chess game when no one wants to play. You evolved your landed tracker by manual and now back to landing by auto-pilot. And it doesnt seem dictatorial at all. No one flames u, or criticise incessantly. Because u ask first, to review a rule. Now that is what i meant by anarchy. Yeah, its still a government, but violence has different context in cyberspace.
Linus is said to be promiscously open. Doesnt means he sleeps around with the committers, but that he very quickly lets peers in and try to see beyond the rules. Which is exactly what u advocate and have been doing very well AFAIS. What we are doing actually is playing the same chess board, but from different angles, and of course different colors.
/me offers gambit
red1
You didn't offer a gambit, you offered a sacrifice :D
That sounded really different, it's different to say "Less rules" or "No-rules" that saying "More logic and natural rules open to be reviewed when necessary"
Yes, like the Bobby Fischer criticism to chess, he preferred a random chess, with more flexible rules:
http://www.chessvariants.com/diffsetup.dir/fischer.html
never tried to play that - but I suppose is more consistent with real life, where the rules are changing and need to be reviewed from time to time.
Now I remember that's one of the the most common question from people
- what are the rules to contribute?
- where are the rules to commit?
- etc
And we lack from proper documentation of the ruling - very possibly with more well documented, logic and natural rules the contribution process will be easier.
Regards,
Carlos Ruiz
Hi all,
This is my take from my own experience.
I have been a professional software developer for more than 10 years, and I have been hacking Adempiere code the last couple of months.
From experience, it takes time for a developer, however good, to come to grasp with the intricasies of a large software project.
We as a firm have solved that by dividing our project into areas only which certain people can make changes to, and other general areas that a new person could make changes to.
As relates to Adempiere, and yes I am writing as a new contributor, we need controls on the submission of code. I have been in the game long enough to know that I don't know Adempiere well enough to go changing / fixing bugs in the Adempiere Core coz while I might fix a bug, I might break things unknown to me.
Considering the importance of this project, we could maybe have two types of members. The experienced ones in the project and newbies. A newbies code has to be reviewed and when they fix a bug, it cannot be automaticaly closed until a senior member approves.
As for the experienced ones, they can be automatically closed after a certain time.
Regards
Ngigi Waithaka
Alliance Technologies
But to support Trifon's statement i am advocate that less rules is the ideal, but how less? For example i like to just add developers into our list without question and kick them also without answers, but i dont think u like that anarchic practice. You like them to show u their work and it must be contributed seriously and not as marketing ploy.
We need rules but we must show that rules are made to protect contributors and avoid disruption to their work. I think this is what Carlos is looking for. We have rules but we approve very easily. See how the last few days we admitted many victims, i mean new developers.
And here is proof that we have no rules against contrarian opinion. They can object frankly without fear of been shut up. This is the bazaar air that we have. Can u breathe well in here?
Hi all,
I want to tell you a little bit about my life before metas and ADempiere
- for those who don't want to read it, just jump to the end, theres my vote ;-):
In December 2007 it was 9 years ago, since i started working as Software-Developer (i began with Programming in Cobol ;-) - greetings to red1) in a big insurance company. From the first day on i was able to commit changes and Bugfixes into production. And i did it. It was very exiting in the beginning, because when you failed as committer the production was corrupted and the eyes of the whole company would be looking at you. If this happened it was the highest prioriy to fix it again. As you can see, i survived the employment and the big insurance company has even become much bigger since then. ;-)
If i was unsure about Fixes it was possible to ask others to review my work. At the Beginning i asked a lot of people. At the end people asked me alot. But i never ended asking others to review my work when i decided it would be better. But important thing was always the same: If reviewed or not, the committer was always responsible for what he puts in production.
I am sure that all committers listed under http://www.adempiere.com/wiki/index.php/Committers are trusted to have a very good quality awareness and are willing to ask someone else for a peer review of a Bugfix if they decide that it would be better. This review should then be done in a short timeframe (<15 days).
Now that I'm a new committer for ADempiere, i feel a bit like having a time travel 9 years back. I have a mentor - thanxs alot Carlos - and i'll ask him to review (and the others don't be afraid, i'll ask you too ;-) ).
At the end of the Day: I'll stay responsible for my changes i put into trunk - reviewed or not. If i break it, i'll fix it - fast!
Therefore my vote:
[+1] for closing the Bug-Requests in status "pending" after 15 Days.
Greetings
Mark
> In December 2007 it was 9 years ago
Wonder if this is a y2k bug?! *lol* Just joking... at least it shows i read every post!