PHP Throw Down - News from the DevNet Team Meet
Posted: Tue Jan 23, 2007 8:48 pm
PHP Throw Down is really happening!
Members
Present for the meeting (shown as phpBB usernames) in order of appearance:
Others that showed an interest in the thread previously:
a) dedicated to the project and will be contributing 8 hours or more
b) can program in PHP, well. Or are near expert in something else we will be using such as XML or MySQL.
I'm worried that too many cooks will spoil the broth and I've chaired game mod projects where that has happened. Thoughts?
Communication
We've all agreed that Skype is the way to go for communication but we are only using the chat features. You can set it to notify you only when certain strings appear in the messages. That could be handy. Also you can bookmark the main chat and close the window and all the previous chat history, that you weren't present for, will appear in the window as backlog so you can catch up even if you close the program (we tested it) although there may be a limited buffer. You can of course to PMs (backlog always present) as well or set up separate chats.
We have agreed to have another meeting. Because of timezone issues it is difficult to decide a time when this should happen. So we are leaving the chat window open and you can join anytime by downloading a copy of skype (v3 is recommended) and adding "irkengir" (me) or "kieran.huggins" (Kieran Huggins duh!) to your contact list. We'll add you to the chat. Unless we're both asleep or something. Incidentally, I'm on GMT and I'm going to be asleep soon after this post. It would be useful if people interested in the project could post their timezones here too.
Leadership
The role of leadership will largely be one of coordination and code design decisions. Decisions relating to features, for example: what to cut out of the spec if we are running short of time, will, I imagine, be done by the person who contributed the initial idea, not the leader.
There was a rumour that feyd would lead the project. feyd is this true, comments? I wouldn't mind leading but only if people back me. Kieran isn't interested in leading the project and we didn't really get round to asking Aaron and Dave. Who do you think should lead? Would you like to stand as a possible leader?
Team Name
Whilst I love "PHPDN Ninja Forum Space Crew Team" we've pretty much decided to go with "PHP Developer Network" its the full title of who we are and we won't be embarrassed about it in the years to come. If anybody has any strong views on this (because we didn't really) then make them known to us. Personally I think it is more important to fry the big fish that has "What application should we write?" etched into the side of it, first.
Said Fish
I don't particularly want to reveal all the ideas for applications in here because this is a public forum. @admins and mods: That private forum could be really useful right now because each idea could warrant it's own thread. Is that possible?
There were quite a lot of ideas, all of which were rejected. I think a number of these ideas still have life in them however, and I'd like to see further discussion until we are sure that they should be scrapped. Some of the ideas were rehashes of old one, some completely new, others new perhaps to PHP but implemented in other languages already. There seemed to be a lot of disagreement on exactly what makes a good idea and also how much we can achieve in a 24 hour period. Ideally we are looking for ideas that are
Technologies
Room for debate here. XSLT was pretty much decided though.
Stuff Not Discussed
...that probably should have been: -
Members
Present for the meeting (shown as phpBB usernames) in order of appearance:
- Kieran Huggins
- ole
- aaronhall
- DaveTheAve (he really wants to win!)
Others that showed an interest in the thread previously:
- The Ninja Space Goat
- [s]Maugrim_The_Reaper[/s] (not available)
- d11wtq
- seodevhead (for css)
- kaYak
a) dedicated to the project and will be contributing 8 hours or more
b) can program in PHP, well. Or are near expert in something else we will be using such as XML or MySQL.
I'm worried that too many cooks will spoil the broth and I've chaired game mod projects where that has happened. Thoughts?
Communication
We've all agreed that Skype is the way to go for communication but we are only using the chat features. You can set it to notify you only when certain strings appear in the messages. That could be handy. Also you can bookmark the main chat and close the window and all the previous chat history, that you weren't present for, will appear in the window as backlog so you can catch up even if you close the program (we tested it) although there may be a limited buffer. You can of course to PMs (backlog always present) as well or set up separate chats.
We have agreed to have another meeting. Because of timezone issues it is difficult to decide a time when this should happen. So we are leaving the chat window open and you can join anytime by downloading a copy of skype (v3 is recommended) and adding "irkengir" (me) or "kieran.huggins" (Kieran Huggins duh!) to your contact list. We'll add you to the chat. Unless we're both asleep or something. Incidentally, I'm on GMT and I'm going to be asleep soon after this post. It would be useful if people interested in the project could post their timezones here too.
Leadership
The role of leadership will largely be one of coordination and code design decisions. Decisions relating to features, for example: what to cut out of the spec if we are running short of time, will, I imagine, be done by the person who contributed the initial idea, not the leader.
There was a rumour that feyd would lead the project. feyd is this true, comments? I wouldn't mind leading but only if people back me. Kieran isn't interested in leading the project and we didn't really get round to asking Aaron and Dave. Who do you think should lead? Would you like to stand as a possible leader?
Team Name
Whilst I love "PHPDN Ninja Forum Space Crew Team" we've pretty much decided to go with "PHP Developer Network" its the full title of who we are and we won't be embarrassed about it in the years to come. If anybody has any strong views on this (because we didn't really) then make them known to us. Personally I think it is more important to fry the big fish that has "What application should we write?" etched into the side of it, first.
Said Fish
I don't particularly want to reveal all the ideas for applications in here because this is a public forum. @admins and mods: That private forum could be really useful right now because each idea could warrant it's own thread. Is that possible?
There were quite a lot of ideas, all of which were rejected. I think a number of these ideas still have life in them however, and I'd like to see further discussion until we are sure that they should be scrapped. Some of the ideas were rehashes of old one, some completely new, others new perhaps to PHP but implemented in other languages already. There seemed to be a lot of disagreement on exactly what makes a good idea and also how much we can achieve in a 24 hour period. Ideally we are looking for ideas that are
- Scalable - a simple implementation can be built and features added on to it in a modular fashion so that there is something to show for our efforts even if it goes tits up
- Original or with strong elements of originally - something like a spamless wiki isn't really going to cut it on the originality front.
- Useful - this seemed to be important to many during discussions but often in conflict with originality. Which do you think is more important?
- Possible to actually write in PHP/JS/XML/SQL - these are technical limitations we simply can't ignore. It should also not require complex installation (we'll lose points immediately for that)
- Be Web 2.0 - Generally this involves community generated content but there are more detailed definitions. I remember watching a really good podcast on this actually but I can't find it anymore

- Be Legal - One example: I suggested we create a PHP implementation of a board game but this can't be done unless the board game is of our own creation for licensing reasons.
Technologies
Room for debate here. XSLT was pretty much decided though.
- Zend Framework - This should speed development and provide us with many nice things straight out of the box. Would the learning curve of this be a problem or perhaps people are prepared to learn in the time before the compo? There are now many good tutorials on it. The initial difficulty of the MVC pattern is the only real challenge to people. Other things that could be useful in ZF: Db, Services, Config, ACL - none of which I've personally used.
- XSLT - to be used for all the view code. That would mean all model code will have to finish by parsing in a load of XML for the XSLT to transform. Kieran uses XSLT extensively in his work so he is experienced with it, I've used it for a few days only but I'm pretty confident with it. I have a couple of books available to me. Basically me and Kieran should be able to handle all the XSLT code. The advantages of using XSLT are that is creates a very real, separation of roles between model and view code and will totally prevent model programmers from writing view code. Also its very very fast to develop and should allow us to mock up interfaces and change our mind about interfaces several times throughout the 24 hour period. Also won't let you produce anything but valid XHTML output.
- JQuery - Seems to be accepted as the JavaScript library of choice. Once reason for this was that it plays well with other libraries. There was plenty of other praise about JQuery and, in fact, I've never actually heard a bad word about it.
Stuff Not Discussed
...that probably should have been: -
- Use of patterns and OOP versus alternatives
- SVN usage issues and conventions
- Unit Testing
- Coding Standards
- Code review / pair programming
- Code commenting
- Division of roles amongst individuals (probably need to decide what we're going to write first)