I'd be more than happy to run some benchmarks if you decide to move forward with the Service Routing update later on.
Also, the new validation module is truly a masterpiece. I'm looking forward to seeing it integrated into the framework as its own module!
On a side note, if all goes according to plan, my first video will be launching tomorrow evening (EST) on the official Trongate YouTube channel.
2 months ago
#31
2 months ago
#32
Thank you - that’s amazing.
Just to let you know, the weekend is generally a duff time to post videos on YouTube, so the response will probably be a bit slow. But let me say publicly exactly what I’ve already said to you in private.
We’re playing the long game. It’s going to take time to gain traction - perhaps even months. However, I trust you completely and I think you’re going to do great.
Here’s something else.
“The Count” - the mysterious self-made millionaire who has been funding my life for the past year - had a look at your YouTube interview. This one: https://www.youtube.com/watch?v=RWvhzQ40iOw
On Tuesday I spoke with him on the phone and you came into the conversation. Having watched the video, he said, “That guy is exactly the kind of person the framework needs.”
I’m sure I speak for everyone when I say we’re all right behind you and wishing you the very best. There’s no need to seek my approval before or after posting something. You’re your own man.
So, I wish you the very best of luck. I’ll post a video on my own channel to let people know the vibe (just not at the weekend - it’s too slow).
Your video will be one small step for a man, but one giant leap for the framework.
To anyone reading this and feeling left out - don’t. The Trongate channel is going to be run like Free Code Camp, so I want as many people involved as possible. If you’re thinking about getting involved with the Trongate YouTube channel, I want to hear from you. I’m actively encouraging as many people as I can.
In fact, as I’m typing this, I’ve just had an idea...
Just to let you know, the weekend is generally a duff time to post videos on YouTube, so the response will probably be a bit slow. But let me say publicly exactly what I’ve already said to you in private.
We’re playing the long game. It’s going to take time to gain traction - perhaps even months. However, I trust you completely and I think you’re going to do great.
Here’s something else.
“The Count” - the mysterious self-made millionaire who has been funding my life for the past year - had a look at your YouTube interview. This one: https://www.youtube.com/watch?v=RWvhzQ40iOw
On Tuesday I spoke with him on the phone and you came into the conversation. Having watched the video, he said, “That guy is exactly the kind of person the framework needs.”
I’m sure I speak for everyone when I say we’re all right behind you and wishing you the very best. There’s no need to seek my approval before or after posting something. You’re your own man.
So, I wish you the very best of luck. I’ll post a video on my own channel to let people know the vibe (just not at the weekend - it’s too slow).
Your video will be one small step for a man, but one giant leap for the framework.
To anyone reading this and feeling left out - don’t. The Trongate channel is going to be run like Free Code Camp, so I want as many people involved as possible. If you’re thinking about getting involved with the Trongate YouTube channel, I want to hear from you. I’m actively encouraging as many people as I can.
In fact, as I’m typing this, I’ve just had an idea...
2 months ago
#33
Hello DC,
I wanted to let you know that the Trongate YouTube Channel has the old horse logo.
Dan
I wanted to let you know that the Trongate YouTube Channel has the old horse logo.
Dan
2 months ago
#34
Okay, this is huge!
I've gone through the entire tg_helpers thing and carried out the switcheroo that is described here:
https://trongate.io/essays/how-to-make-the-fastest-php-framework-even-faster
I brought in some help with an AI tool for this (otherwise it would have taken until next Christmas). The AI tool ran an assortment of tests that (it insists) have all been passed. However, I haven't had time to perform any deep testing. Frankly, I'm too scared to even try.
Please do check out the new 'validacious' repo on GitHub:
https://github.com/trongate/validacious
If all has gone according to plan, the 'tg_helpers' directory should now be about 80% smaller and the framework should be approximately 5% faster.
To our glorious testers - Balazs and Dafa: This is your thing. I await any feedback you have with great anticipation.
Hail!
I've gone through the entire tg_helpers thing and carried out the switcheroo that is described here:
https://trongate.io/essays/how-to-make-the-fastest-php-framework-even-faster
I brought in some help with an AI tool for this (otherwise it would have taken until next Christmas). The AI tool ran an assortment of tests that (it insists) have all been passed. However, I haven't had time to perform any deep testing. Frankly, I'm too scared to even try.
Please do check out the new 'validacious' repo on GitHub:
https://github.com/trongate/validacious
If all has gone according to plan, the 'tg_helpers' directory should now be about 80% smaller and the framework should be approximately 5% faster.
To our glorious testers - Balazs and Dafa: This is your thing. I await any feedback you have with great anticipation.
Hail!
2 months ago
#35
I have run the benchmarks comparing the newly updated “Validacious” app to Trongate v2.
The result: Validacious is ~ 3.4% faster. It’s a little less than what the AI predicted but I think it’s still a major win. Congratulations!
The result: Validacious is ~ 3.4% faster. It’s a little less than what the AI predicted but I think it’s still a major win. Congratulations!
2 months ago
#36
That's great. I think Gemini had predicted between 2 to 5 percent, and I was just being a blind optimist by constantly quoting 5%. As a matter of fact, here's a direct quote from last week's article:
Source: https://trongate.io/essays/how-to-make-the-fastest-php-framework-even-faster
In fairness, the AI engines got it right. Their predictions were good.
Even though we never made it all the way to 5%, I think it's a resounding success story because:
1). We now have some daylight between Trongate and Phalcon.
2). The engine directory has now been massively reduced in size, giving us an opportunity to have "(quick) paste and go" AI assistance.
I think this is a goer.
I'll be working with it today. If all goes well then I may send this out pretty quickly.
Thank you so much for that! Once again, you've done us all a huge service.
DC
Source: https://trongate.io/essays/how-to-make-the-fastest-php-framework-even-faster
In fairness, the AI engines got it right. Their predictions were good.
Even though we never made it all the way to 5%, I think it's a resounding success story because:
1). We now have some daylight between Trongate and Phalcon.
2). The engine directory has now been massively reduced in size, giving us an opportunity to have "(quick) paste and go" AI assistance.
I think this is a goer.
I'll be working with it today. If all goes well then I may send this out pretty quickly.
Thank you so much for that! Once again, you've done us all a huge service.
DC
2 months ago
#37
Well done guys, every percent counts.
Apologies for being a bit quiet, am flat out with work at the moment with very little time to do anything else.
Cheers
Apologies for being a bit quiet, am flat out with work at the moment with very little time to do anything else.
Cheers
2 months ago
#38
Of course, DC. I'm glad I could provide some factual data to confirm the AI's predictions. A 3.4% increase sits almost exactly in the middle of that 2–5% range.
Previously, Trongate held a ~2% advantage over Phalcon; with this update, that lead is set to grow to over 5%. This is a huge achievement that further establishes Trongate as the most performant PHP framework in the world! 🏆
Previously, Trongate held a ~2% advantage over Phalcon; with this update, that lead is set to grow to over 5%. This is a huge achievement that further establishes Trongate as the most performant PHP framework in the world! 🏆
2 months ago
#39
I tried to do some testing with the updated Validacious module today. I couldn't get too far. When trying to log in at /tg-admin, I get this Fatal Error:
To me, it looks like the cause of the error is that the method signature has changed from:
to this:
When calling form_open() from trongate_administrators/views/login.php, it is using the old method signature, hence the error.
I'm not creating a pull request because this might be just something the AI messed up (by adding a new parameter) and the rabbit hole goes deeper.
To me, it looks like the cause of the error is that the method signature has changed from:
to this:
When calling form_open() from trongate_administrators/views/login.php, it is using the old method signature, hence the error.
I'm not creating a pull request because this might be just something the AI messed up (by adding a new parameter) and the rabbit hole goes deeper.
2 months ago
#40
Well, thanks for that. I am glad you caught it when you did.
The AI has changed the function headers, and they will need to be reverted so that the arguments and doc blocks match precisely what we originally had.
Unfortunately, I cannot find an AI system that I trust to handle this correctly. I think the only reliable solution is to go through every single function in tg_helpers/ and restore the original function headers. Then, of course, everything will need to be tested.
So, instead of:
We should have:
Unless someone can think of a better way to do this quickly, I believe it's going to be a very slow process. There are, after all, hundreds of functions.
I am afraid the launch of this new thing will have to be delayed. It could take as long as a month. The reason is that I want to ensure both the documentation and the reference guide are completely accurate. It is a substantial amount of work.
On the positive side, the theory works. We can definitely use this approach to make the framework faster and easier to understand.
Many thanks again for spotting that.
DC
The AI has changed the function headers, and they will need to be reverted so that the arguments and doc blocks match precisely what we originally had.
Unfortunately, I cannot find an AI system that I trust to handle this correctly. I think the only reliable solution is to go through every single function in tg_helpers/ and restore the original function headers. Then, of course, everything will need to be tested.
So, instead of:
We should have:
Unless someone can think of a better way to do this quickly, I believe it's going to be a very slow process. There are, after all, hundreds of functions.
I am afraid the launch of this new thing will have to be delayed. It could take as long as a month. The reason is that I want to ensure both the documentation and the reference guide are completely accurate. It is a substantial amount of work.
On the positive side, the theory works. We can definitely use this approach to make the framework faster and easier to understand.
Many thanks again for spotting that.
DC