Date: Sun, 1 Nov 1998 16:54:02 +0200 (IST) From: Eli Zaretskii X-Sender: eliz AT is To: Ralph Proctor cc: djgpp AT delorie DOT com Subject: Re: "port" In-Reply-To: <3.0.1.16.19981101083112.1c0ff2f6@shadow.net> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Reply-To: djgpp AT delorie DOT com On Sun, 1 Nov 1998, Ralph Proctor wrote: > But when somebody says "I uploaded a port of ___________to djgpp" then > one would expect more. You cannot realistically expect more than this: that the uploaded package will build in exactly the same environment as the one used by the porter. Anything else requires either access to many different platforms, or a lot of time devoted to testing, or a lot of experience in porting to DJGPP (usually, all of the above ;-). The above might seem like an extreme, but I came to this conclusion based on several dozens of ports I've done in recent years. It is just too damn hard to do a clean job that will work for everybody (even though, to my personal amazement, I seem to succeed in that lately). Anybody who tried knows that it takes several iterations to get things right; you might try this yourself some day to fully appreciate the amount of gotchas involved. > But sometime in the future, IMHO, since this is a very good math > accessory to EMACS, I do think the porting problem should be > solved for DJGPP. If you'd agree to do the mundane job of packaging and uploading this, I promise to help you get the Makefile to work. It worked for me. > A "Here's how I got _______________to work with DJGPP" put in the > mail archives would do just as well. This is a vast job, if done right. (Simply listing all the hacks one needed to do just to make a package compile isn't my idea of such an entry.)