This was a great talk and, imho, the best one for me at this conference. Not only was the content interesting and the slides and speaker engaging, but he actually showed REAL examples for everything from workers to how to get them working with queuing systems. Many other talks didn't do this - I'm a very visual learner so being able to see how everything worked together was great.
This talk really did it for me and unfortunately although I didn't get the chance to talk to Justin after the talk I hope this very positive review suffices. I've now got lots of interesting and cool stuff to have a play with in PHP.
Although the speaker had to put up with stifling heat in the room and no microphone, I thought he did a good job on (as his first conference talk) remaining relaxed and confident about the knowledge he was trying to put across. I thought he had spoken many times before. We went through the code slides rather fast and although I was able to keep up with what Chris was saying, I think others may have had a problem with keeping up to speed with the fast progression - I had never done anything with event sourcing before. I am looking forward to Chris's next talk.
Very useful talk on overall security requirements - I can see this being particularly useful for more beginner to intermediate developers. Could use code examples for a security failure and then how it would be better towards the last few points.
I find that there's usually one talk that stands out and really makes me thing and reconsider how I do something at a conference, and this talk did that for me.
Why he isn't more forthcoming with such useful knowledge outside of conferences is beyond me!
This was a great Keynote! It was funny, kept us all entertained and I learned a lot about the beginnings of the PHP community and how it's grown because of the amalgamation of photos from all the events throughout the years - and I haven't seen anything like that, even online, before. Also my face was shown without my permission so I suing. 5 stars!
The talk was a good length and showed different available technologies for machine learning. In fact, I'll refer to the slides (when Mariusz puts them up) for choosing the right technique for a given problem (like classification etc).
I've used machine learning before briefly, through the use of a Naive Bayes classifier. However, I felt that the talk was not at all suited to beginners to machine learning. Terms like "document", "model", "classification" etc would need to be explained and I felt some of the examples were arbitrary, although the salary one was good. If I could give some advice to the speaker for next time, it would be to come up with a use case and a problem right at the beginning, and show through the use of machine learning how it was solved, in-depth, throughout the talk. Like what all these averages etc are when writing the code.
I'm glad I went to this and I'm looking forward to seeing more machine learning talks at PHP conferences.
What a great talk! It was funny and thought provoking. The speaker didn't need to read from slides a lot and knew how to keep the audience engaged and interested.
Initially I was a little put off by Josh saying at the beginning this has nothing to do with PHP really, but he brought back a lot of the points to PHP and related technologies like MySQL.
Also more jokes about managers and ZF2 next time ;-)
A useful introduction to redis. The talk had a good problem and use case from the beginning that was solved by the end.
The only thing I can suggest is to make the text larger on one of the earlier slides (it was in a small, calligraphic style of font and was hard to read), and I think the large bit of code could've used comments or arrows to point out what was going on at a high level.
Please do a talk on caching strategies! :-)
The topic was good and there were some interesting slides on what a message broker is and does and what PubSub was. But after that intro and onto the main topic it went a little downhill.
I felt that the demo wasn't useful because it was just a large block of code with little explanation, sorry! The speaker could use being more enthusiastic about his topic, as he clearly knows a lot about it, by facing the audience more and not facing the side and screen with a laser pointer so much - the audience had disengaged and started chatting two thirds of the way throughout the talk as a result. Also the code was unreadable not due to lighting or background colour but because the font was small and grey instead of white.
I'm glad I came to this talk which is why I chose "worth hearing", however it was entirely theoretical. The slides were just bullet points with little-to-no code examples and was a little dry as a result. I think the AOP-side wasn't too well thought out as DI and AOP are not mutually exclusive.
I think some were lost in the Symfony parts because if you don't know the framework then everything else is a moot point - mentioning the kernel / event dispatcher etc. I think to resolve this the speaker merely needs to put an introductory slide for the big things he's mentioning even if people might already know it, which will help both newbies to the concept and allow more advanced devs to gloss over it as it's only a minute out.
I was hoping for a good anti-DI talk (I had my laptop ready to make notes for some quick retrofitting for my talk!!) but didn't really gain anything like the abstract promised.
I enjoyed this talk. I'm always open to hearing other good examples of the SOLID principles, but hearing examples on ducks and shapes - these are easy to find and repeat from online sources and the reason behind this is because it's just not easy to come up with a decent example we could all use in the real world for a few of these principles!
So those few principles that used ducks / shapes I felt was a bit of a cop out - hence the docking of one star, but the rest of the talk I can imagine would be useful for newbies to the principles and the speaker - this was his first conf talk - well, you'd never tell, he was confident and engaging!
Good shout out about Auryn (more an 'injector' than a 'container' - completely different so I'd recommend learning and making that distinction clear). It also doesn't shift the problem like other containers do.
So future tip - remove the arbritrary non-real-world examples which are present everwhere online and copied from Uncle Bob (easier said than done) and show us something concrete we may work with instead. Good talk!
Thanks for the talk Ciaran, I found the OO examples particularly useful. I already used Behat but actually doing it like TDD - from the tests first, changed my mindset a bit. There's always one talk at a conference that I go away learning something decent from and I believe this was the one for me at PHP UK :-)