From stuckey@mrcnext.cso.uiuc.edu Fri Dec 4 07:04:41 1992 Received: from calvin.sfasu.edu by umaxc.weeg.uiowa.edu (5.61.jnf/920629) on Fri, 4 Dec 92 07:04:39 -0600 id AA15014 with SMTP Received: from garcon.cso.uiuc.edu by calvin.sfasu.EDU with SMTP (5.59/25-eef) id AA07593; Fri, 4 Dec 92 06:53:58 CST Return-Path: Received: from mrcnext.cso.uiuc.edu by garcon.cso.uiuc.edu with SMTP id AA28490 (5.67a8+/IDA-1.5 for ); Fri, 4 Dec 1992 06:51:35 -0600 Received: by mrcnext.cso.uiuc.edu (NeXT-1.0 (From Sendmail 5.52)/NeXT-1.0) id AA18014; Fri, 4 Dec 92 06:51:16 CST Date: Fri, 4 Dec 92 06:51:16 CST From: stuckey@mrcnext.cso.uiuc.edu (Anthony J. Stuckey) Message-Id: <9212041251.AA18014@mrcnext.cso.uiuc.edu> To: hyperc-l@calvin.sfasu.edu Subject: Updated Bugs list... Status: R Just thought that I'd let everyone know the new things that have come to light. If anyone can disprove any of these bugs, I'd like to hear about it. If you've got behavior that you think is a bug, either email me or post to the list for general discussion. I'll probably re-post this list at various intervals. Should it be placed on calvin.sfasu.edu for ftp? So that people know what they're getting into? Anthony J. Stuckey stuckey@mrcnext.cso.uiuc.edu RR1 Box 153, Montrose, IL, 62445. Known Bugs in the standard distribution of HyperC. 1) ED. - Tabs are expanded to 4 spaces on changed lines of code. This is not only on the display (which is reasonable), but also in the save file. - characters which run into the 80th column are deleted. - files larger than available memory cannot be edited. - the Mark option to move or copy text is easy to confuse. You can get garbage characters entered into the file by using ESCape to abort a Mark operation for move or copy, then performing one for real. - It is possible to delete the contents of a file using the A)ppend option and moving between 3 or 4 different source files. :From: nwp90@ecs.soton.ac.uk (Nick Phillips) :You forgot an important one: :In insert mode, pressing a key (to insert a character) whilst :the screen is scrolling will cause the system to hang (and you :lose your work since the last save...) : :From: ST2975@siucvmb.siu.edu :In the editor, if you try to save a file to other than :the current working directory, you must type its full :pathname. It won't check to see if it is saving the :file to a regular file or subdirectory file and it won't :append the current filename to the new path. I've lost :a couple of subdirectories this way. :cp may or may not do this also. I don't exactly remember. 2) NAT.MAC - the JSR macro sub-optimally fixes the stack on return. Specifically, the lines "if ]1 \n isp ]2" should be changed to "if ]2 \n isp ]2". Subroutines with no parameters will now return faster. 3) PP #define != # define The standard says that whitespace should be allowed between #'s and directives, HyperC forces them to be abutted. :From: ST2975@siucvmb.siu.edu :PPF: :The preprocessor functions Extended, Double, and Single (Comp doesn't :matter here) will not convert a string like "-1.034" to floating point :format. The floating point constant is always positive. You must use :the fneg function to change it. :From: nwp90@ecs.soton.ac.uk (Nick Phillips) :SHELL: :cd doesn't report failure. :cp doesn't report failure (more worrying!) :if(ch=='\n') seems to allow ch to be 0x0a or 0x0d (annoying when trying to : write a filter to turn cr/lf into cr)... There's also a bug or series of bugs in the native code generation set. When you compile JUMP.C to CP-code, it works as expected. When you compile down to native code, the output file is mangled. Specifically, the first character of every line (and sometimes the first 2-3) are missing.