Hi Jaques, I am not sure if this is helpful for your comments below .. but i know of multiple software packages (git for windows which has msysgit for example) that provide native linux tools on windows including grep and ssh (for local and remote work).
Furthermore you can still download powerful text editors on windows. Vim for example, is extremely fast with massive text files which we opened in the past and regex through that thing is a breeze. So I guess I'm trying to emphasize that the OS choice is not a big issue with proper tools. Taher Alkhateeb On Sep 9, 2014 8:08 AM, "Jacques Le Roux" <[email protected]> wrote: > > Le 08/09/2014 23:31, Scott Gray a écrit : > >> On 9/09/2014, at 2:59 am, Jacopo Cappellato <[email protected]> >> wrote: >> >> On Sep 8, 2014, at 3:51 PM, Jacques Le Roux < >>> [email protected]> wrote: >>> >>> Then, if nobody mind, I'd like to re-add the error.log concept >>>> >>> I prefer you don't but if I am the only one I would not object. >>> >>> Jacopo >>> >> I spend a lot of time debugging production instances and I never use it, >> so I'd prefer if it wasn't there as well. Using grep to list errors is a >> simple alternative that doesn't require additional files. Individual >> errors are often worthless without the surrounding INFO and DEBUG >> information IMO. >> > > I had also my share of reading myself blind at production logs and I must > admit I then rarely used the error.log, but indeed greped. Let me list the > cases I found error.Log useful, first I must say my platform of choice is > Windows. > When in development and test phase, using either File Explorer on Windows > (development) or WinSCP on servers (test). It's then the fastest way to get > an idea on what is going wrong. The same applies with OFBiz demos on the > OFBiz-VM at the ASF. Once the file (error.log) is open (I use Scite as > local text editor) either from File Explorer or WinSCP, it's a breeze to > refresh it, because normally this file is lighter than ofbiz.log (else you > have a big problem). > > Also we removed console.log. In some cases I found it very convenient. On > the other hand I can agree that having a lot of logs written can slow down > the system. And I guess that's the reason why Jacopo and you are against > multiplying them. I believe, contrary to console.log, simply re-adding > error should not have an important impact in most circumstances, hopefully > you have not that much error in your instance. > > The idea is I don't want to have to manually re-add it each time I want > it. For console.log these cases are less frequent, so doing it by hand then > will not be a problem. I will even maybe add a patch for console.log in the > repo, ready to use... > > What others think? > > Jacques > > >> Regards >> Scott >> >> >> >>
