While this could be useful, I think there are better things which development time could be spent on. A re-validate email every * days would be good. :)

Excuse me, you have a big IPS logo in your signature and I lept.

An unfortunate crutch for me all too often it seems -_-

I do not see this as being a costly system since the code IS AVAILABLE in phplist already and I am assuming the license would permit its reuse.

What is "worth while" for open source is not always the same for close source paid products. Open source scripts and those available on places like hotscripts do not always (most likely don't) take in to account multiple systems, configurations, and so on. So yes, they could take something as a base and it'll be "easy". Then they have to track bugs in software they aren't familiar with, develop interoperability with multiple configurations, and so on. While "grabbing code" might work for a single person or company only using it for themselves, it does not work for a company like IPS that mass-distributes software. Thus, ground up or not, it would result in a costly, time oriented, procedure.

I don't see your objection to a simple POP based system. You seem to imply that if 100% of the world doesn't use POP, there's no point in coding for it. But POP is a HUGE standard with incredible penetration. If it was really to be done right perhaps it would also read mbox format, etc. I really don't see this as more than a day - at most - of code which would really help. Not unlike the idea of supporting https for admin logins.

Then make a mod. From my understanding of the suggestion, it could be built as a module/component with only a single, minor edit to the ACP (if at all) regarding email settings. My objection is simple, it isn't viable to all parties. Because of that, with my experience in programming software, it will require IPS to find a way to make it viable to all parties. Given the situation, it is not possible. They'd have to start teaching people how to administer servers and use POP accounts, and I find that well beyond the scope of IPS's purpose and support.

Asking for revalidation is not intrusive in the least. All it does it make sure that you have the most up to date information at your disposal. It could be anything from a simple once a month reminder to update their email accounts and go as far as requiring a re-validation every year (dates for everything configurable). I find this as a good solution and easily integrated. It also serves all parties and doesn't require IPS change its business model to suit training people on server administration.

I do not think it reasonable to send out 50,000 reminders every six months. And if you think the solution to a big hole in the program is to tell me to go and write my own mods...why do you think I paid for this software in the first place?

I do not think it reasonable to send out 50,000 reminders every six months.

It wouldn't send out 50k emails at once every 6 months and you could set the time frame to whenever. The re-validations could be staggered based on join date or last re-validation time (which is something I implied) in order to keep mass mailings from happening. It also wouldn't be an automatic generation (or could, but yet another option) but rather be triggered by a visit from the member after the anniversary of whatever date is decided above. When the member visits and logs in (automatic or not) they could be redirected to a page explaining the occurrence and telling them to enter their new email or send the validation to the old account. Once completed it can be forgotten until the next anniversary.

And if you think the solution to a big hole in the program is to tell me to go and write my own mods...why do you think I paid for this software in the first place?

Windows doesn't support nix built files by default. Should I not have bought that? The purpose of module/component and add-on type systems is to allow a way for people like you with unpopular ideas to accomplish their goals in an easy organized manner. No software company (or any company for that manner) can possibly add every single feature suggested by every individual. It's suicide.

With 50,000 members, it sure would, if I did them every six months. Whether they're stggered or not, it's a lot of mail.

"Windows doesn't support nix built files by default" ... so? Is POP suddenly a UNIX only system?

By your logic, IPS would be a barebones structure like phpBB.

Ooook. So what you are saying is that you'd rather have 10-100 times the amount of returned mail from 10s of thousands of topic tracking emails sent to sift through rather than a simple system that makes sure that users have their emails up to date?

"Windows doesn't support nix built files by default" ... so? Is POP suddenly a UNIX only system?

No, I was making an example to your comment here:

And if you think the solution to a big hole in the program is to tell me to go and write my own mods...why do you think I paid for this software in the first place?

It is your opinion that this is a "big hole". Not mine. I said nix files are supported under windows. I guess I shouldn't use windows then. Or, I can get/build a separate program (in the web software world this would equate to a module or component) to handle the missing abilities. I never said that POP is a *nix only protocol.

By your logic, IPS would be a barebones structure like phpBB.

No. It wouldn't be. I don't even know where you even get that idea. There is a big difference between "not adding every suggestion mentioned by customer x" and "not adding any feature at all".

Oh, you're one of those "if I want it it's urgent, anything else is pointless and nobody cares about it" people. Well, no point in arguing, then. You won't listen and it's a waste of my time.

Oh, you're one of those "if I want it it's urgent, anything else is pointless and nobody cares about it" people. Well, no point in arguing, then. You won't listen and it's a waste of my time.

Actually Digi is totally the opposite of how you just described him, he just takes a balanced view, looking at the whole picture.

Actually Digi is totally the opposite of how you just described him, he just takes a balanced view, looking at the whole picture.

Thank you for that. That is what I try to accomplish in any feature related post. :)

What if you use Yahoo or MSN mail? Last I checked they don't give you POP login information.

What if you don't run GD2 libraries? (Lots of IPB features are only available to those who have it)

Not all features apply to all users setups. :P

Right, but there are alternatives to GD available for the capcha. You just can't use the "advanced capcha image".

To do this feature at all Piping would be the best way to do this, otherwise IPB would need to "poll" an email account to know an email was recieved. With Piping the email goes directly to the IPB module.

Also you don't need to construct the email message at all. All you would need to do is include a string in the emails going out to serve as a marker. Something like:

IPB:##:##########

(IPB, Member ID, Timestamp when it was sent).

Then all you would need to do is have IPB search the entire message for this string using regular expressions:

if( preg_match( "/IPB\:([0-9]+)\:([0-9]{10})/" ) ) { }

It's true that in order for a human to read an email the way it needs to be seen you need to decode the email... But, for the most part, these "parts" are still in readable format to find the string, so you wouldn't need to decode the email. Some email "parts" are base64 encoded, but that really only happens with attached files. Even if you needed to decode emails, there is a free PEAR library that you can use to do just that, and it would not be the first time a PEAR library was used in IPB.

That being said, it's extremely easy to do. The only drawback is that not all "bounce backs" contain the original email that was sent.... But most of them do, so it would be worth doing.

Another thing to note is that most hosting providers will give you an email address and forward, so not having one is not an issue. I don't know if piping is available in any other environments other than cPanel though...

With all of that said I don't believe it's something IPS needs to do. It could easily be done as a small mod.

I use the PEAR lib you are speaking of. It is outdated and no longer developed. There are many things wrong with it. It could be used, but yeah. That still doesn't fix the fact that piping in not something you can do under windows. At least, not email piping like *nix. I've search for months on end to find a solution to a windows email piping scheme and have come up with nothing.

I agree with your other assessments. :D

While this could be useful, I think there are better things which development time could be spent on. A re-validate email every * days would be good. :)

This is 100% the PRIMARY point I was getting at. I didn't say this additional is simply not possible - I said when weighed against the 3000 other suggestions in this forum, it's not NEARLY as critical or as useful as some of the other ideas brought up.

To do this feature at all Piping would be the best way to do this, otherwise IPB would need to "poll" an email account to know an email was recieved. With Piping the email goes directly to the IPB module.

Also you don't need to construct the email message at all. All you would need to do is include a string in the emails going out to serve as a marker. Something like:

IPB:##:##########

(IPB, Member ID, Timestamp when it was sent).

Then all you would need to do is have IPB search the entire message for this string using regular expressions:

if( preg_match( "/IPB\:([0-9]+)\:([0-9]{10})/" ) ) { }

It's true that in order for a human to read an email the way it needs to be seen you need to decode the email... But, for the most part, these "parts" are still in readable format to find the string, so you wouldn't need to decode the email. Some email "parts" are base64 encoded, but that really only happens with attached files. Even if you needed to decode emails, there is a free PEAR library that you can use to do just that, and it would not be the first time a PEAR library was used in IPB.

That being said, it's extremely easy to do. The only drawback is that not all "bounce backs" contain the original email that was sent.... But most of them do, so it would be worth doing.

Another thing to note is that most hosting providers will give you an email address and forward, so not having one is not an issue. I don't know if piping is available in any other environments other than cPanel though...

With all of that said I don't believe it's something IPS needs to do. It could easily be done as a small mod.

Again, it's possible - but just something that would be difficult to support and maintain from our standpoint. User X knowing that POP is a sound and nothing more, wondering why "3" has suddenly been added to their favorite Rice Crispy character, is going to come along trying to set something like this up. Something like this could keep our tech support department extremely busy trying to help customers use a feature we provided - when in reality as has been stated before, email bounce management is outside the scope of a forum system largely.

{snip}

With all of that said I don't believe it's something IPS needs to do. It could easily be done as a small mod.

:)

Yeah, after I replied, I realized I quoted your whole block, and then my reply seemed to be against what you were saying. :P

I was mostly just commenting on the thoughts you put forth. :)

What about the revalidation plan I posed Brandon?

It's a possibility, though I think I'd go more for just a "Is this email still correct? If not, please update it" type deal. That way those people whose email never changes don't have to go through the hassle (and the burden is pulled off the server) while a reminder is still in place for someone who forgot to update their address. Best of both worlds.

And spread across all IPB boards this will result in millions of reminders sent out each month, adding to the general bounceback-and-junk-mail level of the Net, when a simple plug-in or module would have solved it all the right way from the start. Well, I guess my opinion doesn't matter at all and none of this is important because you don't think it is. Why did I pay for this software again?

*sigh*

Why is it that everytime I/we disagree with a suggestion "we don't listen, your opinion doesn't matter", etc? I simply can't add it just because YOU think it's a good idea and simple. If it's simple - code it and submit it as a mod to IPS Beyond.

That said, I'm done replying to suggestions in this forum. :) I will silently take in everything said, but I can't continue to be coming off as not caring everytime I try to explain WHY we won't add a suggestion. So from now on, from me at least, it will be a silent dismissal with no discussion as to why - because it seems that people interpret their suggestion not being considered as personal for some reason. :huh:

And spread across all IPB boards this will result in millions of reminders sent out each month, adding to the general bounceback-and-junk-mail level of the Net, when a simple plug-in or module would have solved it all the right way from the start. Well, I guess my opinion doesn't matter at all and none of this is important because you don't think it is. Why did I pay for this software again?

Since when did your opinion not count? IPS took your opinion into consideration. They can't agree with everyones opinion can they?

Because all my arguments are completely ignored, that's why I think my opinions don't matter. I get rubbish like "code it yourself" and

As far as I see it, outside of that it is, and should be, up to you as the administrator to get people to update their emails regularly. I would be ok with a new feature that asked every X days for the user to re-validate their email address. As that is the only thing left that the above 2 features do not cover. However, I feel that those who are actively subscribing to topics are likely to keep their emails up to date anyway. If they don't, they'll notice that they are not receiving topic notifications in the first place and fix it themselves.

Nowhere in any of this do I see where the development of a costly and time oriented system to pipe/deliver bounce-back to the forum is necessary.

...

Obviously it is apparent that you haven't developed software for mass distribution. To not consider what is necessary for EVERYONE to use it is silly. That's like saying IPS should only support *nix-based servers. Sure, they could do it, but don't you think that would be a silly move on their part?

In other words, if it's easy enough for IPB to code, they don't need to code it, and if it's hard to code, there's no justification. Again, NONE OF MY ARGUMENTS WERE RESPONDED TO DIRECTLY.

I'm glad to see that Brandon has been driven away from suggestion topics :(

I am a PHP developer myself, and have been for a long time - I fully appreciate that building a solution to solve this problem for 100% of your users will be a very hard task. A lot of people don't appreciate how long some things can take to develop, the time actually writing the code often is a small percentage of the effort required to implement something; R&D and testing usually take up far more time. So I can appreciate where you're coming from with your answers.

Having said that... That's no reason to ignore the problem entirely. IMO, a lot of administrators with a substantial number of users will be in a position to re-shape their email system to fit your requirements. Those that can't... well isn't that a perfect opportunity for you guys to make money out of a hosted service? :)

Here is a link to information about VERP, which I'm sure your developers already know about, but perhaps some users migth find interesting

http://en.wikipedia.org/wiki/Variable_envelope_return_path

In my opinion, it is certainly the responsibility for the forum software to manage it's users correctly, and I believe email address validation (not just whilst signing up, but continuing during the lifetime of the membership) to be well within the scope of this management. In systems with 10s of thousands of users, doing this manually is not an option.

I have recently taken control of a forum that's been operating for 5 years, and has a membership count in the 10's of thousands. In a recent mail out to all members about new site news, a very large percentage of members had email addresses causing bounces - there's no way I can manually process the 4000+ bounces. A system such as VERP would significantly help solve the problem.

Thank you for your thoughtful post - I should point out that I have 35,000 members on one of my forums, and would probably be automatically labelled as a spammer if I tried sending out confirmations to everyone every six months!

IPB has been a great forum for the most part, but there are three things that have always bugged me (aside from the occassional code bloat). Those are not having SSL logins as an option; no automatic handling of email bounces, especially considering that there are mass-mailing features; and of course breaking skins so easily when upgrading, which has been addressed more and more with each version.