Re[4]: [PEAR-QA] ping XML-RPC service

Hello Helgi, From Thursday, December 16, 2004, 12:41:04 AM, you wrote: >> Global " to ' replacing is a _very_ bad practice in CVS, because it doesn't >> provide any significant perfomance improvement (esp. in XML_RPC), but >> renders CVS annotations useless. HЮ> Just a quick side note, who actually cares about those annotation stuff HЮ> anyway ? HЮ> Okey it can be useful at a time, but not committing something to CVS HЮ> just because it could 'messup' the CVS annotation is kinda silly. I already said before, that this messed up patches in bugreports. And you think it's ok, right? =/ HЮ> Plus iirc he only applied one patch and made those CS changes and afaik HЮ> that patch was from the becoming developer on XML_RPC) Dan what's the HЮ> status on that btw. ? HЮ> In addition, I don't want to get into any argue why he did it, and what HЮ> impact that might have had on the other patches and etc. Annotations used to detect which commit caused possible problems and what patches were applied after that. If you use CVS intensively with several developers, using branch merges, locking and so on, you'll understand that more quickly. Of course when you use CVS solely for storage it doesn't matter, but when can you describe what do you need CVS for? If you fetch previous, not bugfixed version and try to diff against it to see what have changed since then - do you realise what will you get with all these CS changes? Take into account, that most patches were submitted against latest stable release. My proposal is to add small request for developers to do CS only while going from RC1 to release for big packages with lots of bugs to fix from last release. P.S. Helgi, if there is too much text for you to read carefully, please, look into open XML-RPC bugreports and try to apply some of the patches mentioned there against latest CVS. Almost every bugreport contains some. t --