We have moved our forum to GitHub Discussions. For questions about Phalcon v3/v4/v5 you can visit here and for Phalcon v6 here.

Vote for Phalcon's integration into Google App Engine

Google App Engine, the world's most robust, stable and scalable deployment environment has recently introduced support for PHP ( https://goo.gl/GtH6n ). However, due its very specific nature, no custom extension can be installed there. While GAE is already extremely fast, integrating Phalcon into the mix would bring everything to a whole new level. This is why I've made a plea for Phalcon's integration into GAE's PHP runtime. To support this notion, please just star the issue on their issue tracker, here: https://code.google.com/p/googleappengine/issues/detail?id=9336

To maximize effect, please only star it - do not comment with "+1" and try to comment only if you have something constructive to add - people tend to un-star to avoid spam if too many comments appear, so true support is shown by starring it only.

Here is my full initial post about making the issue: https://goo.gl/VdPCV

Added my star!

+1 and starred. :)

It seems that there will be issues with Volt in GAE thought.

Thank you Claudio.

I don't think that there will be problems with volt if the developer has the templates already compiled and then uploads them.

We do need to check though mbstring and implications of it not being present.


I'm adding support for writing compiled templates to APC, so that will make Volt faster and it will not have this problem of upload compiled templates.

@Phalcon good thinking but please note "We'll most likely turn off the caching part of APC at some point because we want people to use memcache and APC caching does not scale across instances."

+1 and starred :) But i think that they will have more problems to add phalcon to GAE, than cpanel. I know that they could have the same framework on every instances, but it is Big Company, slow decisions. They need to be sure that phalcon is stable and don't change very often i think, so at this stage of development and roadmap is not possible. I saw that in your request any FlameBoyTroll reply :(

Please don't consider what one of the last posters in the google thread suggested. Hacking specific configuration settings into the framework which act differently based on some local variables is not nice. If a dev is too stupid to adapt his bootstrap to the hosting environment in use ... well, not anyones fault but his. Just put the specs in the docs.

edited Mar '14

@Phillip Actually, Phalcon uses syslog functionality provided by PHP instead of the one provided by the OS (which in Unix is the same).

I would like to understand why mbstring/mcrypt aren't allowed, what alternatives do we have to supply that functionality?

Looking at the phpinfo() output, iconv does not seen to appear there.

@Philip I did not mean that suggestion as an obligatory hack, but as an option that makes it easy to do a one-step setup of Phalcon on GAE. Of course it wouldn't make sense to strictly enforce it - but a proper warning should be given when using Phalcon's logger in a manner that's appropriate for regular development setups. That said, I understand your point, if the docs made explicit mention of this, there would be no problems.

@Phalcon That's something that should be brought to their attention - those are important extensions for any PHP app and effort should be made to include them. Perhaps you could bring it up in the issue?

There are also issues regarding iconv and mcrypt

https://code.google.com/p/googleappengine/issues/detail?id=9340 https://code.google.com/p/googleappengine/issues/detail?id=9342

Their status has been accepted so I presume that they will be included in the near future.

Nice. Those are two major hurdles out of the way. Still, it couldn't hurt starring them, just to expedite their realization.

@Swader I agree that warnings ("This logger requires an environment which supports syslog") are acceptable, but Phalcon should be very cautious about not becoming a huge, bloated, hard to maintain and in-transparent framework. Once specific support for GAE is hacked into the core, others will ask for similar hacks (PaaS X, PaaS Y, PaaS Z) for their respective environments. I'm certainly no technology evangelist but please keep in mind that the major USP of Phalcon for PHP is 'performance', not 'works without you using your brain' ;-)

That becomes especially true when I see the question from @Phalcon about "how do we supply mbstring or iconv if it's not available at GAE". Who cares? If GAE does not even supply the essentials of PHP extensions, it's certainly not the job of Phalcon to spend time working around the (short-sighted) limitations of GAE which is making its first steps in the PHP world just now and certainly far from being the first choice of anyone for hosting a critical PHP application in the cloud.

I agree that support from major hosts is critical for the success of Phalcon, but I believe that it would be a bad, bad deal to achieve that support by soiling the frameworks core.

edited Mar '14

I'm thinking that maybe (at this moment) we should not invest resources and time on adapting Phalcon to GAE, despite its experimental state, I'm very surprised about the current implementation of PHP, I don't know if before releasing PHP for GAE, they tried to test at least common software like Wordpress or Drupal running there.

I can understand that Phalcon must be starred, because it's young, it is a framework, etc, but, Why do we should star issues about missing extensions like mbstring/iconv/mcrypt/openssl?

This is like: hey, star this issue about enable the "+" operator, so you can do sums in GAE,

This missing functionality is essential, not just for Phalcon, any professional application in PHP need these extensions.

I hope that in some months, the implementation of PHP become more mature and so we can know exactly what to adapt or improve to run Phalcon properly there. (Assuming that Phalcon will be accepted)

The mcrypt, iconv and mbstring proposals have been accepted so they will be introduced in the GAE PHP land.

I think we need to get Phalcon in there for more publicity and also to allow people to have a good hosting platform for their applications. In the meantime I nudged cPanel to see what goes on with that.

If Phalcon gets approved by GAE it will be quite a bit before we can actually say that yes version x.x will work with it and here are the instructions. There is a long road ahead but every star helps :)

Nikolaos is quite right, the sooner we start pushing en-masse, the better. You definitely don't want a rival framework / extension to get there first just by bringing in a bigger audience and having more stars. Phalcon's integration into GAE, in whichever form, can only help matters. The incoming stream of both C and PHP developers such a move would bring would be absolutely priceless and would propel the already awesome Phalcon project to entirely new levels. Heck, with this jump, Phalcon would, at least for me, get the status of being the "missing PHP6 version", that's how much I've gotten to appreciate it.

Regarding the test of Drupal and Wordpress, yes, there are demos running Wordpress in the announcement video. Wordpress is just a bunch of procedural hacks, though, so it doesn't exactly depend on quality extensions. Allegedly someone tried (and succeeded) to install Drupal too, which I assume has fallbacks. However, there definitely shouldn't be any compromises - if mcrypts, iconv and mbstring aren't there, Phalcon shouldn't be either. This is why we, as a community, should push for faster realization of those proposals too, by starring them - even though they are already accepted.


I agree with you guys, the votes need to continue, we can fix technical issues later.

I mentioned in the issue my concerns about those missing extensions, and the answer from Google was:

star https://code.google.com/p/googleappengine/issues/detail?id=9342 & https://code.google.com/p/googleappengine/issues/detail?id=9340

I see both mbstring/mcrypt are already accepted, great :)

We've made it to the top 100 feature requests, but we're a long way from the priority list. If we just get 100 more votes, we'll have made top 30. Nikolaos, any news about cPanel? I believe integration with cPanel would also help us get votes for GAE due to rise of awareness.

Nothing as of yet. I did post something today in the relevant thread (https://features.cpanel.net) asking for updates. Lets see how that goes.


There are a lot of comments but only 7 likes for adding PhalconPHP to cPanel:


Maybe we should add some more :)

@pavlov That is really weird because we had over 100 (or more) wants and were the second or third most requested feature. I don't know why it went to 7....

Very strange..... cpanel is... no comments....

Whoa, last time i saw it there was around 140 upvotes on CPanel... what happened? Are they cheating?

I opened a ticket with cPanel. I will keep you guys updated.

4 stars to 200, this is the 26th most requested feature on GAE now. Let's make it happen people!

We're one vote shy of being the 17th most requested GAE feature.

And now we're 17th :) just voted.

+1 and starred :) Go Phalcon!

This is progressing nicely, and getting some new labels now. It's among the top 15 requested features, and with full docker and VM support coming to GAE soon, it just might happen sooner rather than later. If you know people who still haven't voted but love Phalcon, have them drop a star!


Any updates on this?