Re[6]: [PEAR-QA] Documentation..

Hello Justin, From Tuesday, December 7, 2004, 3:02:42 AM, you wrote: >> JP> Stale? Closed? Are you trying to be a troll? If something hasn't been >> JP> updated then point it out and it will be dealt with. QA is doing a >> JP> good job trying to get maintainers for abandoned packages. *People are >> JP> busy*. You have to understand that this is all done *for free*. JP> Let me start here by saying that you need a serious adjustment. You JP> have no faith in others. If you have a problem, *do not* just say JP> "it's stale so don't count on them". As I have said before, try asking JP> why it is a certain way and having a positive discussion about how to JP> better things instead of trolling. And, yes, even if you give a good JP> idea it may never (or very slowly) get put into motion unless *you* JP> are the driving force and spend some of your own time fixing things. Well, I will try to ask it in a positive way, you're right. Even if I've got a new idea, I need to argue with unbeatable arguments like "we already have another person to handle this", "we can't allow anyone...", "I don't want to do it", "that's complicated - I won't do this and __ask__others__ not to do this". In some quite common cases I've got no response at all. That's why I am asking for tracker (1) I am the driving force, but I have my limits too. I've already spent some of my own time fixing things and you can track this in XML-RPC (2), thread about go-pear (3), perhaps regexp with "bugs" linking in pearweb (very demonstrative). Also I've got some other ideas, a lot of them if you ask Helgi =) but I want to close these mentioned here first. >> Stale means that there's no indication of PEAR's activity or it's >> status. Stale also mean, that users can see the result of they work. >> Their efforts are most of the time wasted and they can't get recognition. >> Stale means that 7 voices is enough to rule another 200. Stale means, >> that everybody seems to prefer to skip the problem than to solve it, >> because everyb. has no time. Stale also means that there's no trust to >> users. And hence no trust to developers. JP> "No indication of activity"? Excuse me? There are new releases of JP> packages all of the time. Commits to CVS happen all of the time. This JP> is all easily found on the website and you can even get e-mails (the JP> PEAR-general list gets release announcements, the CVS list gets all JP> CVS changes). There is activity, indeed, but according to RFC's and TODO's amount (esp. in QA area) this activity is more verbal than practical. I'm not subscribed to PEAR general and CVS list, because these are like a very rough information flow hard to follow. Information is not a bunch of letters with symbols - it is useful data you potentially interested into. I speaking from my own experience with XML-RPC and File::Search_and_Replace, from reading stuff on a site and from some experiments I've made before going to conclusions. Negative feeling for one "member" of community is not that should hamper overall progress and ruin benefits for end-users, which some of the people here are not taking into account. I agree - there is activity, but I'd like to see this activity to be more productive, more organized, more open for new ideas in the sence of "making it yourself". In every other community I'd be welcomed if developers don't have time to implement something and they will not judge if it good or bad as long as it works properly. Only here there are people constantly judging others. Good example - HTTP::Cache proposal. It was deleted, but as a user I've got a strong need in such a package and I want to place my vote on it, but I can't. Others said to look into HTTP::Download, HTTP::Header, but give me an example - how to use them like HTTP::Cache with one function to cache some content, one function to disable that cache, to generate Etag's for my images and Last-modified for my SQL queries interface? This proposal was deleted without even asking users - do _they_ need that package? Closed developers caste (no more than 10 people) decided, that package should not exists, because there are already solution known only to them. Do not take it personal - take a look from outside. JP> "users can see the result of they work". What users and what work? If JP> users contribute bug reports, they *are* handled (albeit not JP> super-quick, but again, people don't have time to be on-call for a JP> free project). If users contribute patches it happens all the faster. JP> The power is in your hands. My power was already described above. =) I just want to close bugs in XML-RPC. I don't think that everybody has that power like to struggle HTTP::Cache developer for his package being present in PEAR. It may be interesting for me, but others will prefer PHPClasses or SF or some of the PEAR alternatives, which just don't have official php.net status and even if it is a good package - there will be not too much people to find it. JP> "7 voices is enough to rule another 200". This is completely bogus. JP> All of these discussions are carried out on the pear dev list and all JP> concerns are taken seriously. You weren't around for it. Oh well, JP> that's your loss. Just because something was passed by 7 votes doesn't JP> mean that people did *not* want it. If people were against it, they JP> would have voted against it. There are only 7 votes because only 7 JP> people cared enough to vote. That's it. Stop assuming that there is a JP> ruling elite. *THERE IS NOT*. Could you explain this to other hundred users who constantly flow from and to PEAR website. To me democracy is when at least 50% voted in and I believe PEAR votings are intended to be democratic (that means objective) for wide public also: http://pear.php.net/pepr/pepr-proposal-show.php?id=60 JP> Who is skipping problems? I can't put a blame on the people, who do write PEAR like a hobby, so I'd like to hear opinion first - Does PEAR need tracker and simple votings on QA issues (everything concerning PEAR community, esp. ideas, which can improve PEAR service for both developers and end-users)? Should this votings be open for end-users too? JP> It is true that you have to be an experienced PEAR developer to vote. JP> This is done to make sure that voters are well informed and are a JP> serious part of the community. I completely disagree. You pointed a vital point in this discussion. Please, place some critics on your own arguments and try to analyse your point of view by answering these simple questions. Let's make some terminology agreements: _developers_ here - PEAR developers _users_ - developers, who use PEAR code, but are not PEAR developers 1. For whom PEAR is storing code? 2. What does PEAR for you as for user? 3. How can you help PEAR as user? 4. How can you assure your work will be noticed? 5. What can you do if your work is still unnoticed? 6. Can you help other users as user? 7. What does PEAR for you as for developer? 8. How can you help PEAR as developer? 9. How can you impede others work? Instead of learning users you prefer to limit them and thus satisfy your ambitions. You have an ability to easily justify others work giving your person more social value, perhaps even some status. _you_ means every member of community in this context, answering your quote. 10. Why do you need users to be serious part of community to vote? 11. For whom PEAR is storing code? 12. Do every developer is serious part of community? 13. What is "to be informed" and why users are not informed? 14. Why users need to be informed to express their ideas? 15. What kind of developers as people PEAR draws? 16. Do PEAR need to limit people with personal ambitions? 17. Can PEAR satisfy this ambitions in other way then voting and ruling other's ideas and what needs to be done for that? 18. What of your answers are untrue? =) >> Closed means, that only the person who contributed package can vote and >> decide that other members can or cannot do. Users can't vote on new >> features, can't propose ideas, can't share code or work on web-site >> improvements. They can't work on getting PEAR better at all and those >> people who can are too busy to do it themselves and the worse is that >> these people do not allow others to do it. Why? JP> Nobody votes on new features...users submit a feature request and the JP> package dev decides whether or not to add it. The developers are JP> usually pretty welcoming of new features if they have the time to deal JP> with it. JP> Users who have no proposal karma can submit a proposal to the list. If JP> they wish to create a proposal all they have to do is ask for a JP> pearweb account. Most people are given accounts who are serious about JP> proposing something. You may have to prove that you're serious by JP> actually writing most of the proposal before you get the account, but JP> this is simple accountability. What is proposal? I would like a new feature on File::SearchandReplace, for example. I left RFE which is left unnoticed. While asking for PEAR account - it clearly stated, what I don't need an account and I must leave RFE in appropriate bugtracker. What should I do next? Maybe not the best example, but I would like to place a voting on this RFE, so other users can also easily participate if they feel it is a valuable feature. JP> How can "Users" not share code? You're encouraged to create bug JP> reports, feature requests, and proposals which link to your own code JP> and/or patches. *ANYONE* can submit a bug report or feature request in JP> the bug system with *THEIR OWN* patches. I can't make that any JP> clearer. If you have a package you want to be part of PEAR, you JP> propose it and the community can comment on it. I left pretty much XML-RPC patches, which are still in bugreports. More to add that latest XML-RPC CVS CS commits made them almost unusable. JP> If what you meant is "everyone can't submit their own code without it JP> going through the proposal system" then yes, this is true. The JP> proposal system and coding standards are here for a reason. If you JP> don't like it, you can post whatever you want to hotscripts or JP> phpclasses. I'm not trying to push you away here, but it sounds like JP> you want a free-for-all, which PEAR is not. No, commiting new packages through proposal system is Ok, but you don't allow people, who really does something to submit their patches on a packages. Also you don't allow people to decide if some idea is valuable, but it must be users - not only developers who decide that. I propose to make PEAR more open. Not CVS-free-for-all as you afraid of, but participation-free-for-all style. >> Yes, I am trolling. I'd like you to think about why am I doing this >> and what signal I want to get on the output. I've got a solid POV and >> I want you to make me change my mind. >> But also I act like a user, who doesn't ask questions. He walks >> through the site, reads what he want, downloads package and leaves. >> Then he returns to the site to report something, leaves a note, which >> stays for months, he hasn't got an appropriate response and he will >> never return back. He is not a member of community. So far I'm here >> to report something. JP> You admit that you're trolling. Trolling has no purpose. Trolling is JP> complaining or bashing in a negative way without proposing what to do JP> to fix things. If you want to fix things, try being constructive. JP> Point out a shortcoming and then propose a new way. Be convincing JP> instead of rude. If you're rude no one is going to listen. Why not to include this into PEAR community guidelines? If you pretend to be a community - it is good to explain others what kind of community you are. You have all my letters with all possible negative factors users can think about while acquainting theirselves with PEAR (at least I've tried to express and gather them together). If you neutralize this negative with some positive FAQ or "Golden Rules" or RFC or whatever documents users can find easily, you will make a great step towards true community. Or I need to write the docs for you as Dan suggested at the beginning of this thread? =) JP> So Joe User reports something and hears nothing back. If nothing has JP> happened, the user can ask why nothing has happened. It may be that JP> the developer of the package hasn't had the time to deal with it. It JP> could be that the package has been abandoned. What you do here is send JP> a polite note to the package maintainers and the pear-dev (or users) JP> list asking if anyone knows what's up. In the past you could get JP> nothing back. *Now* we have QA which will handle such things by either JP> fixing things themselves or appointing a new lead for the package to JP> do it. We know we had this problem in the past. This is what QA is JP> for. They are still working out the details, it's a very new group. My proposal is to make QA person not by status/votings, but for the work she/he is made to close questions concerning QA issues. Also make tracker to monitor this activity, to make it possible to "estimate" this amount of work. QA will be more alive this way and will drive people's attitude into other, more constructive channel. QA issues - everything concerning PEAR community, esp. ideas, which can improve PEAR service for both developers and end-users Feel free to offer your definition. >> JP> This e-mail wasn't meant as a personal attack. Lukas suggested that >> JP> you create a bug report with the example. It's the best place to put >> JP> things like this. He was informing you, not criticising you. >> >> His solution resembles a temporary stub for this "special case". It >> seems like an intention to skip nuisance. I just wanted to inform, that >> RFE is not a proper place for such Examples. I like critique and in >> fact every my letter is it, because I can't do anything except >> critisize (if you have already noted, I'm not a developer). Letters, >> what can't help or doesn't contain any good ideas are spam. If any of >> my letters is nonsense - point it. Even if after all my intentions are >> not clear just ask me to leave and I will not return here. Perhaps >> I'll post a patch or two, but this will be all I can do for yours. JP> It may not be the perfect place for examples, but there is nowhere JP> else to put it now. You can e-mail the example to the maintainer. This JP> is the best way. If there is no response, make an RFE so that it JP> doesn't get lost. In the future we will have user comments and you can JP> put it there. When in the future? Where is the problem to do it now? Why can't you allow me to know this problem? Is it free time? Is it hard to find 5 minutes? Why it is not 5 minutes? Why can't you allow to do this somebody else? Why smb. else can't do it? What can you do to help smb. else do it for you? t --