When do you think PHP 5 will become totally unusable for developers?

Ask about general coding issues or problems here.

Moderators: egami, macek, gesf

Post Reply
pivadu
New php-forum User
New php-forum User
Posts: 7
Joined: Wed Apr 07, 2021 6:44 am

Fri Apr 16, 2021 5:28 am

Most of the projects I've been given for a new job are PHP 5.6... I wasn't expecting this, and have had difficulty getting them to work locally. I know that some software still supports it without having to compile old libraries (MAMP Pro, maybe some things on Linux), but even then it seems to be waning. I'm afraid within the next year or two, that any new developer at a job trying to get a PHP 5 site working will just be... out of luck.

The problem is that a lot of businesses, many big businesses, are still running PHP 5... like the one I'm working for. They simply do not allocate resources to upgrading, no matter how often or strongly they're told that it's out of support and insecure. They may also be in a situation where upgrading may be too disruptive or costly, and will need to simply rewrite from new technology.

Anyway, what I'm asking is what your thoughts are on PHP 5 going out of support enough that it just won't be feasible for developers to set up repos for it locally. Do you have any historical experiences with PHP 4 becoming unusable? Also if there's any links I could show our payers, to maybe finally convince them that they need to upgrade, that would be wonderful. Right now it just appears that a behemoth vehicle we're all riding on is headed for a cliff and will drag us all down with them.
User avatar
Strider64
php-forum GURU
php-forum GURU
Posts: 421
Joined: Sat Mar 23, 2013 8:24 am
Location: Livonia, MI
Contact:

Fri Apr 16, 2021 9:20 am

pivadu wrote:
Fri Apr 16, 2021 5:28 am
Most of the projects I've been given for a new job are PHP 5.6... I wasn't expecting this, and have had difficulty getting them to work locally. I know that some software still supports it without having to compile old libraries (MAMP Pro, maybe some things on Linux), but even then it seems to be waning. I'm afraid within the next year or two, that any new developer at a job trying to get a PHP 5 site working will just be... out of luck.

The problem is that a lot of businesses, many big businesses, are still running PHP 5... like the one I'm working for. They simply do not allocate resources to upgrading, no matter how often or strongly they're told that it's out of support and insecure. They may also be in a situation where upgrading may be too disruptive or costly, and will need to simply rewrite from new technology.

Anyway, what I'm asking is what your thoughts are on PHP 5 going out of support enough that it just won't be feasible for developers to set up repos for it locally. Do you have any historical experiences with PHP 4 becoming unusable? Also if there's any links I could show our payers, to maybe finally convince them that they need to upgrade, that would be wonderful. Right now it just appears that a behemoth vehicle we're all riding on is headed for a cliff and will drag us all down with them.
I really can't answer you question, but I don't know why businesses would be reluctant to change as it will only cost them to upgrade the code and that something a programmer is doing already that is coding. It would only set them back a week or two updating the code. Most of the code would work on the updated versions of PHP anyways. The only big bottlenecks probably would be database tables and handshaking to the internet.
Life is a fig newton of your imagination! https://www.phototechguru.com/
pivadu
New php-forum User
New php-forum User
Posts: 7
Joined: Wed Apr 07, 2021 6:44 am

Fri Apr 16, 2021 11:16 am

> I don't know why businesses would be reluctant to change

How many businesses have you worked for? Almost every company I've worked for in my career has refused to update old software on at least 1 repo we've worked on (typically several), and many to the point of the software requiring finding old libraries to try to compile on the computer hopefully without compromising it. I've worked for dozens of tiny to huge companies (like tens of thousands in income per month to billions in a year).

> it will only cost them to upgrade the code

That's why they put upgrading on the back burner... right now the company I work for is looking like they're driving towards a cliff, and no developer has been able to convince them to change direction. I don't know where that cliff is with PHP 5 yet :/

> and that something a programmer is doing already that is coding

The coder is not upgrading the code while they're coding... they have to stick within the compatible syntax and library range. They may be able to get their local PHP 7.3 to work with a PHP 5.6 code, but that's completely dependent on the repo and its dependencies. See what I'm saying?

> It would only set them back a week or two updating the code. Most of the code would work on the updated versions of PHP anyways. The only big bottlenecks probably would be database tables and handshaking to the internet.

This is highly dependent on the repo, its dependencies, the compatibility of the software on the developer's system or other systems, and the availability of required software. If it really would just take a week to upgrade, then companies would just do it. But they don't feel that way, they instead put it off until they hit a wall. What I'm trying to do is figure out WHEN they will hit that wall, if I can.

Anyway, what I described about big businesses refusing to upgrade has not been the exception in my career, it's been the rule. Even when talking to my compatriots in real life and on forums throughout the years, this has been the norm. Financial and governmental businesses especially just do not want to upgrade, because they're afraid of the time and cost, realistic or not. The bigger problem is they hit a wall or drive off a cliff, and chaos happens... and it usually costs them more time and money, as well as aggravation or loss to their workers and buyers. This is evidence with huge companies being hacked due to old insecure software, or just not spending the time and money to properly secure their software with modern recommendations.

This is really a tangential moot point though... and doesn't address the question. I hope you realize that I can't convince the Big Company to upgrade, unless I can say "This is when your software will fail, and why". The closest 2nd is "This is when your developers will no longer be able to work on your software anymore". That's kind of what I'm trying to feel out with my question.

To clarify: PHP 5 will eventually just be dead... meaning, it will not be feasible to develop with anymore. Because I haven't seen a PHP 4 repo in almost a decade, I'd like to know, in people's experience here, how that phase-out went for them, or YOU. Did you recommend upgrading to your boss and then they did? Did you say "This won't work for us anymore" and they upgraded? Did you say "This is when it will fail" and they upgraded or changed?
User avatar
hyper
php-forum GURU
php-forum GURU
Posts: 1222
Joined: Mon Feb 22, 2016 5:52 pm

Sat Apr 17, 2021 4:50 am

It's difficult, the main features promoted with PHP 7 are that it is more secure, but the biggest difference in my mind is the speed increase, this has got to be a big plus by reducing the processing required and stop bounces while a user gets bored waiting for a page to load.

The way I see it is that you make your recommendations, either take the money and do what they are asking you to do (to the letter, as in get it in writing) or refuse to do the work and take the next job.

It's nice to see someone that actually cares about what they do, the truth of the matter is that the developers should have been upgrading over the last several years in a gentle transition - there's nothing that you can do about that. You will only bang your head against a wall trying to convince people that are put in jobs that they have no clue about and want the path of least resistance; least cost mainly; and as you point out - the most expensive in the long run, in terms of time needed as well as actual cash. It's the way companies work, from the largest to the smallest.

Keep up the good work, although it may be difficult to keep your head up when all around you are trying to keep theirs down, it is possible.
pivadu
New php-forum User
New php-forum User
Posts: 7
Joined: Wed Apr 07, 2021 6:44 am

Mon Apr 19, 2021 3:00 pm

> speed increase

This should be a nice incentive for a big company, but so far the Prod websites are either fast enough for them or not so slow that it's annoying the managers in regards to productivity or sales.

> you make your recommendations, either take the money and do what they are asking you to do (to the letter, as in get it in writing) or refuse to do the work and take the next job

I'm afraid this is the only thing that works with these companies, but only if enough developers leave or won't take the job. It's unfortunately very difficult for people to pass up jobs right now, and probably will be for a long time.

> It's nice to see someone that actually cares about what they do

It's not really about that... I'm not looking for a solution here to optimize anything or make myself feel better, I'm trying to figure out how to talk uppers into not driving off the cliff or hitting the wall where new developers can't even set up their environments on existing repos because of software that's too out of date, so therefore can't even be worked on to upgrade the software by them. See what I'm saying? It's a conundrum that creates a disaster for the company (and everyone that works there, including myself) when they hit that wall or something like it.

I haven't seen a PHP 4 repo in almost a decade, so I'm wondering if it just got so old that it was impossible to set up existing PHP 4 repos on new boxes, and the companies had to upgrade or rebuild from newer software. I'd like to know either when you think PHP 5 will hit that wall, or if it's not likely that it ever will. I guarantee you many companies will put off upgrading until it's too late, unless they see a definite start point for disaster.

> You will only bang your head against a wall trying to convince people that are put in jobs that they have no clue about and want the path of least resistance

I'm not looking to bang my head against the wall on anything, I'm just looking for better arguments than I have right now; such as "you know, PHP 4 just wasn't even installable on platforms at this point because it so old and incompatible, certain libraries and versions just disappeared - example 1, example 2, example 3 - and so likely PHP 5 will have the same problem at around the same time from deprecation"... that would be an argument if I could site examples.

There must be some reason PHP 4 just seems to have disappeared... so why hasn't PHP 5? It still seems to be all over the place, and I don't know when I can say it will be infeasible, or what I can tell a big company using it that they need to upgrade. They just don't see the cliff or wall, but I'm not sure I can say where it is either.
simonbrahan
php-forum Fan User
php-forum Fan User
Posts: 137
Joined: Mon Jun 08, 2020 2:00 am
Contact:

Tue Apr 20, 2021 1:51 am

There's no way to give an answer that won't make your managers' eyes glaze over. PHP 5 as an ecosystem has different usage characteristics to earlier versions, so there's no saying that it will become impossible to support over the same timescale that PHP 4 did. There will just be a day some point in the future when someone says "why are we so slow and insecure?" and you say "that's what all those memos were about".

I think the best approach here is to start refactoring your existing code as you work on it - well written PHP 5 code will run just fine in PHP 7 and 8, you just have to get to that point. Then when it really is impossible to continue with PHP 5, the switch over will be fairly straightforward. "We need to do all this work now" is a scary conversation to have when you don't have explicit data points to give; "We need to do this ongoing work to keep our customers safe" and a much easier sell.
pivadu
New php-forum User
New php-forum User
Posts: 7
Joined: Wed Apr 07, 2021 6:44 am

Fri Apr 23, 2021 2:38 pm

There is a way to give an answer that won't make your managers' eyes glaze over XD That answer is "PHP 5 will not be able to be set up on new computers on Aug 9 2021... if I can find that answer, or one like it.

Bosses like hard walls and stark lines of "do or will not do".

Because PHP 4 seems to have been completely erased from all websites... I want to know what got bosses to move from PHP 4 to 5. I of course may be wrong about PHP 4, but I haven't seen a job for it in over 5 years. Something convinced the bosses to upgrade... so saying there's no answer... well there was for PHP 4... so what's the one for PHP 5?

Anyway... the question for you is... what is your experience in a company deciding to upgrade from PHP 4?

Ours already upgraded from Symfony 2.8 to 3.4... I wasn't around when that happened... I don't know how the devs convinced the bosses to upgrade... but it worked. And if they didn't convince them, and they hit a hard wall of this will not work without upgrading... even if that meant burning money until it was upgraded, even if it was a preventable disaster where all hands went on deck to get the ship working again. I of course can't tell why they did it, nor do I need to, but the point is that if not upgrading will hit a wall, and it has before with PHP 4, I want to know how to determine what that wall will be, in any way.

I want to give them evidence of PHP 4 or something else just hitting a wall of unusability... then they'll go "OH, OK... here's when we need to upgrade" and they'll allocate time and resources. I know how this works in the business world, I've seen it many times before, but I can't tell with PHP 5. It's a rotting corpse infesting everywhere at the moment, but they don't see how it could cause the pipes to burst and I can't tell them when.
simonbrahan
php-forum Fan User
php-forum Fan User
Posts: 137
Joined: Mon Jun 08, 2020 2:00 am
Contact:

Sat Apr 24, 2021 9:32 am

https://museum.php.net/php1/

You can still get hold of any previous version of PHP. Previous versions peter out of usage as old applications are shut down or migrated, but there's no hard wall. I can't prove it but I'm fairly confident there's still a server out there running PHP 1.

You can only inform your managers about the trade off they're making - lower performance, less security, less willing developers to fill positions - and let them make the call. Some will let you do it, some won't.
pivadu
New php-forum User
New php-forum User
Posts: 7
Joined: Wed Apr 07, 2021 6:44 am

Tue Apr 27, 2021 8:53 am

> You can still get hold of any previous version of PHP. Previous versions peter out of usage as old applications are shut down or migrated, but there's no hard wall. I can't prove it but I'm fairly confident there's still a server out there running PHP 1.

OK thanks, that's helpful. I haven't seen any PHP versions less than 5 anywhere in almost 10 years, but I suppose they're still being worked on somewhere by somebody, and if it can still be installed and run on modern computers... then it's technically not dead for a company yet. That company may still use the excuse of "we don't have time" or "it's not the highest priority" to upgrade :/ which to me is just a scenario for the company running off the cliff and causing a disaster that costs them a lot of money and collateral damage with people involved.

> You can only inform your managers about the trade off they're making - lower performance, less security, less willing developers to fill positions - and let them make the call. Some will let you do it, some won't.

Thanks. I've informed everyone I know who will understand on the projects... and they've all agreed "yes, we should" but they don't make the decision, so apparently the giant companies relying on this archaic software that I could barely get working locally will just have to hit that wall and we'll see what happens.
pivadu
New php-forum User
New php-forum User
Posts: 7
Joined: Wed Apr 07, 2021 6:44 am

Tue Apr 27, 2021 8:57 am

> https://museum.php.net/php1/

What's hilarious about that link is that all of the PHP files before version 5 were uploaded in 2014, and the rest were 2020 onward.

I suppose in 2014, the PHP maintainers decided to zip them all up for storage... and then forgot about it for 6 years.

What's strange is that PHP 5 and 7 have the same file date... indicating they were also made at the same time, but I don't consider PHP 7 to be a museum piece yet.
simonbrahan
php-forum Fan User
php-forum Fan User
Posts: 137
Joined: Mon Jun 08, 2020 2:00 am
Contact:

Wed Apr 28, 2021 11:52 am

The 2014 / 2020 split seems to be a bug - there's no mention of the year 2020 in the PHP 5 folder itself. The last entry in the PHP 5 folder is the 9th January 2019 - one day before the final PHP 5.6 release.
pivadu
New php-forum User
New php-forum User
Posts: 7
Joined: Wed Apr 07, 2021 6:44 am

Mon May 03, 2021 7:44 am

I don't think the folder dates are bugs... it just appears they were created in those years.

Parent folder dates don't need to match sub-folder dates, and they don't even have to be after or the same as their sub-folder date years.

It all just depends on the person making the folders and moving things around. I think the dates may change depending on what the person did with the folder and the operating system they were using, but I don't know all the behavioral rules that would determine folder dates changing.
Post Reply