RE: Unfulfilled promises... forever experimental extensions... all over again

php.internals

NAIK,ROSHAN (HP-Cupertino,ex1)

22 years ago
> -----Original Message----- > From: Rasmus Lerdorf [mailto:rasmus@php.net] > Sent: Wednesday, August 25, 2004 5:42 PM > To: Naik, Roshan > Cc: internals@lists.php.net > Subject: Re: [PHP-DEV] Unfulfilled promises... forever experimental > extensions... all over again > > > On Wed, 25 Aug 2004, Naik, Roshan wrote: > > Hopefully belts will be tightened up around the extensions and the > > responsibility taken up by their authors. > > Wow, that was quite a rant from someone who hasn't contributed a damn > thing. >
Really ? Are you sure ? http://www.zend.com/zend/week/week146.php Even if I were someone who never contributed anything to php and is just a user of it ... does my concern or that of thousands of others in the same boat count at all ? From that stmt one can only infer I guess they are all non-entities. This is not a community culture but a detrimental coterie culture.
> People volunteer their time and effort to this project. There are no > guarantees that any of this code won't make your cpu melt or > cause your > curtains to catch fire. Ranting at a bunch of volunteers is utterly > useless and counterproductive. If you have a genuine > interest in seeing > things improve, get involved. Pick your favorite extension that is > troubling you, talk to the authors and other users and figure out what > needs to be done to solve its remaining issues. If you do that, your > words might mean something. >
Precisely the attitude expected (indifference and excuses), as was pointed out concisely by another user. http://marc.theaimsgroup.com/?l=php-dev&m=103161229529458&w=2 The central point that these features never made it to production quality previously and that no reason to believe the new cool features advertised may meet the same fate is overlooked and left unaddressed. And as to what Andi wrote:
> Thanks for this quite useless and unproductive email. > I suggest to inform yourself slightly better on the tons of improvements > the PHP development team have made and where we are today vs. 4 years ago. > I suggest next time, if you want to be productive, you can contact us with > a concrete problem and proposed solution.
So that's not a concrete problem ? The proposed solution as I mentioned is to start tightening things up where too much things have been lax. And you get me wrong, I am not saying PHP5 is worthless and there are no useful features. There are clearly many nice improvements. Users simply like to see those things completed before shipping them. My intention is not to point fingers at people. It is to point finger at problems. A response that either refutes the existence of the problem or the affirms the existence of the problem and how (if at all) can be tackled is a more appropriate response. Rather than the usual "buzz-off" As the above quoted user posted... "Nothing prevents us from treating people with patience and courtesy. Except of course bad manners and bad attitude." In theory everything can be marked experimental... the zend engine (object model changed) and other the core parts too. But a certain quality level is indeed being enforced upon, even today, before its made available as a release quality product. Another solution is to leave such "not-yet-ready" things out in the other extension repositories, allow them to mature and stabilize there before shipping it with a standard distribution. That way there is an incentive for authors to complete their work and this encourages competing implementations. The healthy competition also enhances quality of the components in the product. The best one gets to be part of the standard distribution... a nice reward for the efforts. Rather than rewards first and work later (if he got a chance). -Roshan

George Schlossnagle

22 years ago
On Aug 25, 2004, at 10:22 PM, Naik, Roshan wrote:
> "Nothing prevents us from treating people with patience and courtesy. > Except of course bad manners and bad attitude."
Posting this quote in this thread is beyond ironic. George

Rasmus Lerdorf

22 years ago
On Wed, 25 Aug 2004, Naik, Roshan wrote:
>> Wow, that was quite a rant from someone who hasn't contributed a damn >> thing. > > Really ? Are you sure ? > http://www.zend.com/zend/week/week146.php
Ok, one self-serving fix.
> Even if I were someone who never contributed anything to php > and is just a user of it ... does my concern or that > of thousands of others in the same boat count at all ?
No, not really. Your complaint is basically one of process and whether experimental extensions should be widely distributed. We don't include extensions by default unless we feel the usefulness outweighs the hassle of the fact that it may not be completely finished and tested on every possible platform. Sometimes it takes a while to get it right as well for the more complicated extensions. That's just the way things go.
> The central point that these features never made it to > production quality previously and that no reason to believe the > new cool features advertised may meet the same fate > is overlooked and left unaddressed.
You are right, there is no guarantee that our code won't make your curtains catch fire. We don't actually write code for you. We write code for ourselves and happen to share it with you. If you think that is an excuse or a brush-off, then fine, you probably should give up on any and all open source projects because in the end this is how most of them operate. And you didn't propose a solution. Telling us to "tighten things up" is not a solution. "You suck. I propose that you stop sucking!" Well, thanks, that's helpful. Whether an extension is marked experimental is actually quite arbitrary and different developers have different criteria. It is nearly impossible to force a single set of criteria on people that specifies what "production-quality" is. I deploy plenty of extensions marked experimental on some of the busiest web servers in the world. Does that make these production-quality? In my particular case it does because they meet my particular criteria, but in another environment they may indeed blow up badly. So, how about proposing an actual solution instead of pissing off the very people who you will need help from to improve things? This solution may be a set of general criteria an extension has to meet to leave its experimental state, a set of test tools, or a group of developers that you convince with your obvious social skills to contribute to this effort. -Rasmus