Lot 49: Greg Beaver's blog - Comments

Zoe Keating: Transporting a cello: buy a seat? check it? Definitive answers on how to fly with a cello

nospam@example.com (Zoe Keating) — Fri, 05 Feb 2010 00:16:43 -0700

Really great article, thank you! I have to fly a lot with my cello and fragile electronic gear (including a laptop for stage). I dare not check my gear. (Ironically, even though my cello is worth more than I care to think about, when it gets temporarily lost by an airline, its far easier for me to find a cello than it is to get the exact same electronics) I find that getting around airports, and on and off planes, with both cello and gear is too difficult for me. So, knock on wood, the last 4 years I've been flying regularly around the world with my cello case in a BAM flight cover. I feel pretty confident about it, but I realize that there is always a risk.

bank: Setting up your own PEAR channel with Chiara_PEAR_Server - the official way

nospam@example.com (bank) — Wed, 30 Dec 2009 23:17:04 -0700

I have a problem when installing Pear, It has dialog box"php.exe -Unable to locate component This application has failed to start because php_mbstring.dll was not found. re-installing the application may fix this problem."

Nate Abele: How to put the FAIL in open source

nospam@example.com (Nate Abele) — Wed, 25 Nov 2009 02:09:35 -0700

Manuel, At this point, pretty much everyone is ready move away from the original name, since people have interpreted it to mean much more than we had originally intended. See here: http://groups.google.com/group/php-standards/browse_thread/thread/166de4c66846d4a8 and here: http://groups.google.com/group/php-standards/browse_thread/thread/9c02083319dc12b3 Our group is only interested in coming together on a few common ideas that each of our projects can implement to ease the burden of interoperating with each other. If any other projects would like to join and benefit as well, so much the better. None of us is interested in the backlash that would ensue if anyone tried to suggest any *actual* standards for PHP (i.e. JCP-style standards). Taken as a whole (and I mean everybody, including the crazy-bug-report-submitters), the PHP community is far too anarchistic for promoters of such ideas to escape with their lives, let alone have any chance of success.

Manuel Lemos: How to put the FAIL in open source

nospam@example.com (Manuel Lemos) — Sat, 21 Nov 2009 12:41:25 -0700

Roman, don't be silly, nobody alone can start a JCP like effort if there is not enough people that join in so it gains traction. What you are suggestion is such a non-sense like "getting married with myself and form a family with nobody else". I am not calling the framework people group "php standards group". It was them that gave that name. While you say that it is just a php project interoperability, it seems that when they call themselves "php standards group" they have their greater ambitions to set their definitions as the "de facto" php standards. I did not come here to jump on anybody else initiative. I just came here to express that I agree with the concerns of Greg, and that the way to go for a standardization initiative have success is to follow the good example of JCP. It is not a new idea that I am having. I already written about it more than 3 years ago: http://www.phpclasses.org/blog/post/52-Recommended-PHP-frameworks.html I think it would be useful for PHP to have a JCP process but as I explained above, it cannot be done with just myself, nor just gathering an elite that does not allow the process to be transparent and democratic.

romanb: How to put the FAIL in open source

nospam@example.com (romanb) — Sat, 21 Nov 2009 02:59:57 -0700

Manuel, I am one of those strange guys who do not need to drop one language in order to use another. As such I always use several languages in parallel. I did not say "I don't care about PHP", I said that "I dont care that much". That means I dont care enough in order to spend a lot of time in forming and pushing some standards group and specifications. I only use PHP for pet projects. You should really read more carefully (or slowly). As such I'm surely not making fun of anybody. If I understand you right, you would like to see something similar to the JCP in PHP but apparently you dont want to do or start it yourself. OK, so you just want to jump on board when someone else is doing it. At the same time you are upset about a group that you still call "php standards group" even though I told you its a "php project interoperability group". Its not even trying to be like the JCP and it certainly dont forces standards on anybody. Just some people from different projects who agree on some conventions between their projects. I say, stop looking at this group you're so upset about and instead start a JCP-like initiative yourself with all the things you would like to see. If people like it they will surely jump on board. I will let this discussion end here in order not to abuse Greg's blog as a discussion forum / mailing list

Manuel Lemos: How to put the FAIL in open source

nospam@example.com (Manuel Lemos) — Thu, 19 Nov 2009 23:37:11 -0700

Roman, if you do not care about PHP, what are you doing here? Making fun of all of us that care about PHP? As for JCP, as you may read in the overview page, everybody can nominate themselves to serve on the Expert Groups, propose, review, and comment JSR, as long as you are a member. You can apply to be member as an individual. The point is that everybody has an equal chance to participate. Whether it works exactly like this in practice in JCP, I do not care because I do not want to participate in JCP, as I do not work with Java. What matters is that the model of equal opportunity and transparence is the way to go for PHP. When the criteria to participate in a standards group is not clear, discussions take place in private lists, people are accepted and rejected without transparence that prevent us to verify whether everybody had equal opportunity, all this leads to resent and complaints from renown members of the PHP community Greg, that wrote the post that starts this discussion. This is not a good thing. It is just a sign that things are ill fated from the beginning. The solution is to start over, make everything open and transparent, give everybody a chance to nominate themselves to participate without the usual silly excuses that elitists make to keep people out of their way. This is just my opinion. Anyway, I am not interested in participating in this so called PHP standards group, but others want to but have been excluded.

romanb: How to put the FAIL in open source

nospam@example.com (romanb) — Thu, 19 Nov 2009 15:56:10 -0700

Manuel, I'm not sure I understand what you're trying to tell me. Have you ever participated in the JCP? If so, let me know how it turned out, whether your voice was heard. I'm a Java developer (among other things) but have never participated in the JCP but from what I've heard and read from others, its not as democratic as you think. You dont just get into an expert group easily, and for good reason. Also, despite of what I wrote, you still did not seem to understand what this "php standards group" is currently doing and you apparently dont even know that the name "php standards group" is a legacy. Its about php project interoperability and, again, all that is discussed so far is class loader interoperability between some projects. Its extremely irritating how you can apparently get so upset about something that trivial. Cartel? Please... come down. If you want to see a process similar to the JCP in PHP then go start it and dont wait for others to do it. I dont care that much as php is not really my living.

Manuel Lemos: How to put the FAIL in open source

nospam@example.com (Manuel Lemos) — Thu, 19 Nov 2009 11:13:23 -0700

Roman, it does not seem you know very well how JCP works. If you read the participation overview you can clearly read that: 1. propose new or revised Java API specification projects by filing a JSR, 2. nominate themselves to serve on the Expert Groups that create or revise Java specifications (when the JSR is initiated, or during JSR Review), 3. review and comment on all specifications developed using the JCP before Public review (Community Review). http://www.jcp.org/en/participation/overview It seems to me that this self-called PHP standards group is afraid to be overruled by a majority of interested PHP developers that they want to prevent them from joining. In my dictionary that is called a cartel and so cannot be compared with JCP. http://en.wikipedia.org/wiki/Cartel I am afraid that this cartel will end alone because they excluded themselves from being fair and democratic. I don't think that is a good thing for anybody.

Herman Radtke: How to put the FAIL in open source

nospam@example.com (Herman Radtke) — Thu, 19 Nov 2009 10:21:56 -0700

I thought the whole reason this standards group was even formed was to make it easier for the community to use frameworks. Whether someone wants to use two or more frameworks in conjunction or switch from one framework to another, having the same standard obviously makes this easier. So if this is being done for the community, then why is the community being left out?

Greg Beaver: How to put the FAIL in open source

nospam@example.com (Greg Beaver) — Thu, 19 Nov 2009 07:35:07 -0700

James, I also support what is being done (which should be very clear from the first paragraph of the blog post), but you are naive to think that the standards won't be enforced. I write code for pear.php.net and pear2.php.net. Whatever is decided will be enforced there. If you write code for ZF, Cake, Doctrine, or any other project, it will be enforced on your code. The flaw in your reasoning is that I *have* been in contact with the representatives from PEAR, and the real issue is that PEAR's representatives are not representing the actual wishes of the PEAR community, simply because they don't know what they are. The debate is an external thing, and most developers simply don't know what it is. Want proof? Travis Swicegood, who started the standards process as a representative of PEAR on the PEAR group supposedly with a standard originating in PEAR, was actually voted out of the PEAR group after PEAR's developers got a look at what was actually being proposed and how far it was from the existing PEAR standard (and in fact the existing official PEAR2 standard). See http://pear.php.net/election/info.php?election=12&results=1 I fully support the effort to create a PHP standard, and was very active on the php.standards list, but a standard must have community input even if a small group votes on it, especially in a project with a community as strong as PHP at large.

Greg Beaver: How to put the FAIL in open source

nospam@example.com (Greg Beaver) — Thu, 19 Nov 2009 07:26:41 -0700

Hi Roman, You yourself gave 2 examples of the kind of process that actually works, even if you disagree on the degree of effectiveness: Everyone can comment, a few people can vote. The php standards group has decided pre-emptively that only people who can vote can comment or suggest anything, and have expressed annoyance at people who try to comment who aren't voters. If you re-read my post, you'll see I am simply asking for all input to be heard in an official way, not that everyone have a chance to vote. This is a VERY important distinction. All successful open source projects have a few "elite" who make the final decisions, but at the same time, all successful open source projects allow anyone, no matter how experienced (or smart, for that matter) to comment. Good leaders are willing to tolerate a little noise to find the gems that no one else would have thought of. Fearful leaders censor and quench any possible dissent before there is a problem to be solved. As an example of the problems inherent in this model, I have pointed out several potential problems in the standard in an off-list exchange which have never been even responded to. On the other hand, everything I posted publicly has been addressed, and at the very least debated. Without openness in the input process (voting matter much less), you lose actual value, not just some vague principles (otherwise it wouldn't really matter at all).

romanb: How to put the FAIL in open source

nospam@example.com (romanb) — Thu, 19 Nov 2009 03:09:46 -0700

Manuel, it *could* move in that direction but it is in a much earlier state. Really all that has been discussed so far is a fundamental agreement on how to handle class loading in a uniform manner. Without such fundamental technical things theres not much of a point in starting to create specifications. As such I can't understand all the uproar around this at all. What is being discussed is a very simple agreement between many projects on how to structure class files for uniform class loading. Thats far from being a coding standard. Noone cares about tabs vs spaces, lower vs uppercase, opening curly brace on the same line or next line or what not. I dont care about the open vs closed debate personally. It's just obvious that given the "anti-standard" or "I do it my way" or "Cowboy" mentality, as Greg put it, that is so prevalent in the php world it is very unlikely that anything productive would come out of a discussion where everyone can have their say (and vote). Elitism or not, democracy (as in "everyone gets a vote and is heard no matter who they are") doesnt work well in software development. You need to have a) a common/unified vision and b) some people who have the last word on things. With too many people you have too many different visions and even if you get a concensus out of it, it wont be any good any more because the vision is lost. Its just a huge compromise, a patch work. And the development of PHP itself is not different. Being subscribed to the internals list for at least some years now, however democratic and "open" it is supposed to be, at the end there is "elitism" and a certain set of people have the final word. And if you ask me there is not enough elitism in the development of php which results in there being no clear vision of where the language is supposed to be heading. It is more of a patchwork, too. It's much more important that a project has a clear vision and goals than that everyone likes it and has a vote. Consistency matters. The JCP doesnt work that way either. Anyone can propose a standard but only the formed "expert group" (elitism?) really decides where things are going in the end. And no, I'm not (any more) on the (closed) list.

Manuel Lemos: How to put the FAIL in open source

nospam@example.com (Manuel Lemos) — Wed, 18 Nov 2009 16:32:48 -0700

I deslike elitism. It hurts people feelings needlessly. It only makes people boycott the elites every time they can behind the scenes, start competing projects, in sum avoiding cooperation. PHP does not need more elitist efforts. I agree with Lukas, elitist efforts will die of natural causes soon or later. I thought this php.standards effort was for creating high level specifications, for instance for a message queueing API, just like JCP (Java Community Process). From the specification, different teams could implement compatible implementations, just like you have different but (more or less) compatible implementations of Java application servers from different vendors: Sun, JBoss, IBM, Oracle, etc... The actual code standards that are used by each implementation do not matter as long as the API is compatible with the standard specification. In the end the users win because they have different offers to implement compatible API. They can pick the best implementation but still have the freedom to switch to another at later point if they want. That is what I wrote here: http://www.phpclasses.org/blog/post/96-PHP-standards-discussion-group-opens-to-the-world.html http://www.phpclasses.org/blog/post/52-Recommended-PHP-frameworks.html Unfortunately I was wrong. It seems that this php.standards effort is not going that direction. If anybody is willing to start a JCP-like initiative for PHP, count me in. Just let me know if you like to move ahead. mlemos at acm.org

James Rodenkirch: How to put the FAIL in open source

nospam@example.com (James Rodenkirch) — Wed, 18 Nov 2009 15:50:22 -0700

Greg, I think you are over reacting a bit. I am not on the list, nor do I have a vote, but I fully support the effort that this group is trying to accomplish. This is not an attempt to enforce standards on the community, but an effort to help increase interoperability between the projects involved. It is very common for projects/teams to accept different coding standards based on what best supports the need of the project - look at PEAR and Zend Framework - their coding standards are similar, but not exactly the same. You are free to adopt any coding standard you choose for your projects, these developers are choosing to draft something that will help the community integrate multiple tools in a easier/more standard manner. If you are an active member/user of any of the involved projects, you can certainly contact the lead developers to make requests or offer suggestions.

Greg Beaver: How to put the FAIL in open source

nospam@example.com (Greg Beaver) — Tue, 17 Nov 2009 22:29:26 -0700

Who am I? I am a nobody. That's the whole point of my post. Nobodys have plenty to offer to an open source project, ideas or code (both are integral parts of designing PHP source, as I'm sure you would agree). I find it laughable that you think I am imposing anything on anyone when it is you who is imposing moderation and a closed list on the rest of the world. All I am doing is asking other like-minded people to help jolt some sense into the standards group. You may note that I spent months attempting to do this both on standards@lists.php.net and also off-list. As I said early on, if this is not a PHP-project wide standards initiative, then you are misleading at best and dishonest at worst to call it "PHP Standards." This was raised within the first week when it was apparent that you sought to limit input rather than respond to it. Perhaps I am arrogant, perhaps not. I have absolutely no problem with that label in this case because it is irrelevant to the issue. The problem here is not that you or I are arrogant, but that you are alienating a large segment of the people who will be expected to follow a standard. Let's not kid ourselves: the point of a standard is for people to follow it. If that isn't your goal, then you aren't drafting a standard. Let me put this another way: if the people who so kindly got together from their own volition to tell us how we should write our code had the humility to imagine that their original proposal could be flawed, that would help. Then, having the vision to imagine that others might have better solutions for the problem and open up the process to avoid losing that great idea outside the mainstream, well then we'd be getting somewhere.