@Michael Those are interesting ideas, I think you can register by yourself on the wiki, so you can add the RFC. OTOH, and again, I must say I really think the echo shortcut should be regarded as a separate issue, and now that there was some consensus we shouldn't deviate from the topic.
In the end I think it is going to be as bad to have short open tags turned off by default. Hosting services still meddle with the php.ini and some even let you make your own customizations to it (or at least to a subset of it). But again, can we first agree on the echo shortcut feature to be decoupled from short tags?. No agreements === no progress. BTW, what is your open source project? @Thomas I agree on dropping <% for good, I personally don't know any project that uses it and don't think there is currently any point to them anymore. Also, I do use '<? ' instead of '<?' (originall I thought it might throw an error, then it became habit), although I try not tu use <? as much as <?=. Having the space as a requirement (i.e. "<?for(..." would throw an error) would fix the XML situation, is this what you are saying?. @all Can we decide on decoupling <?= before going back to the general short tag matter? On Thu, May 19, 2011 at 10:35 AM, Thomas Hruska <thru...@cubiclesoft.com>wrote: > On 5/19/2011 12:23 AM, Arvids Godjuks wrote: > >> It's essentially the same what I said - move it out of stort_tags and >> make it "On" permanently. >> >> As I remember the decision to remove short_tags was made together with >> register_globals, magic_quotes and other legacy stuff. I can be that I >> remember wrongly, but really do people really use<% ? >> > > Would be **really** nice if '<?' with a whitespace as the next byte was > also detected (i.e. '<? ', '<?\n', etc). (Single-quotes have been added to > aid readability.) > > The '<? ' short tag is syntactic sugary convenience that is **widely** > used: Internal corporate servers, personal machines, and millions upon > millions of websites. The results and financial costs of cleaning up the > upgrade fallout of removing the '<? ' short tag are incalculable. > > Comparing short tags to magic_quotes/register_globals is apples to oranges. > The two are so vastly different and not in the same class. The latter is a > failed security measure. The former is a syntactic sugary convenience. > Every PHP userland developer I know understands the risks associated with > magic_quotes and register_globals but, at the same time, they use the '<? ' > short tag extensively wherever possible. > > Or, perhaps more simply put: If you remove the "syntactic sugary > convenience" of the '<? ' short tag, you'll have an army of developers > dropping by shortly after the release of PHP 6 and they will be incredibly > unhappy. But you just go ahead and remove the '<? ' short tag for PHP 6. > You'll be adding it back into PHP 6.0.1. > > The ONLY reason anyone types '<?php ' in the first place is because '<? ' > isn't guaranteed to work everywhere. And that rule really only applies to > open source software and certain web hosts, which is a very small segment of > the total PHP market share. It would probably be fine if you removed the > _option_ itself but merged '<? ' detection into the core. I don't know > anyone who uses anything but '<? ', so it won't likely be a huge loss for > anyone if '<% ' support is dropped (but I could be wrong about that). > > The important part of this discussion is making sure convenient > functionality doesn't just vanish for stupid reasons. I recognize there > will be breakage regardless because it is a new major version, but looking > ahead one extra byte isn't going to kill you. > > -- > Thomas Hruska > CubicleSoft President > > Barebones CMS is a high-performance, open source content management system > for web developers operating in a team environment. > > An open source CubicleSoft initiative. > Your choice of a MIT or LGPL license. > > http://barebonescms.com/ > > > > -- > PHP Internals - PHP Runtime Development Mailing List > To unsubscribe, visit: http://www.php.net/unsub.php > >
