Re: Re: preg_replace/str_replace Re: [PEAR-DEV] Re: IT[X] fork, round 2

On Wed, 12 Feb 2003 18:45:34 -0500 rocha@i-node.com.br wrote: > And going back to the problem, I replace preg to str_replace, and yes, > it works ok, in my case. The only way it could be a problem is if i > use regex inside arguments, which i dont, in this case. Even only one case is too much. > So, before living this egotrip. Since we have, if I understand, a huge > patch to be applied on one side, potentially huge BC problems on the > other and two guys standing on feet in front of each other, like 'dont > go any further', could we open a NEW (oh boy!) template subdir and let > people go work? That's the 1st proposition that I have done, propose an alternative package/template, no BC problem. > (Alexey i am pretending you rewrite the code from > scratch, it looks like that if the patch is the size of the package). > And yes, i prefer that to cause a possible major BC broken version > change. Christian and a few others people ask for a set of small patches, far easier to check and valid, that's the basic step, and, indeed, always better then a "hello guy, drop everything I'm comin' !" ;-) So, here is the status: The cache is not needed inside both IT and IT[x]. They are template modules, we are all use others caching methods with these classes. Droping the preg is not possible, that will break BC; even if that's not the case for 95% of the users base, I do not agree with any break. As a suitable alternative (said many times the past 2 days) is to use a flag to choose str_replace or preg, preg is used by default. 'By default' means without any additionnals args or calls as it is actually. That fit the current requests and does not break anything. This is my p.o.v, as yesterday, today, and certainly tomorrow too :) Besides this problem, apologize for my last mail, but irrespect, injuries and all these things just make me feel really bad, the authors, imho, cannot expect anything in return, regardless to what they propose. - pierre.