bill@hao.UUCP (Bill Roberts) (11/03/86)
I'm starting to use TML Pascal on a Mac+ (with DataFrame 20 hard disk) i and tried using the SetPaths DA included with the TML 2.0 release. First off, what is 'SetPaths' supposed to do? And secondly, what ever it is supposed to be doing, I don't think it is! I use SetPaths to setup search paths for files referenced in the source code of the file that I am trying to compile. Everything is fine until I try to compile. As soon as the compiler starts to work on the target file the system hangs! So I guess the question is, is 'SetPaths, ver1.1' incompatable with the Mac+ or is it intrinsically buggy? And what *exactly* is SetPaths supposed to be doing? Thanks in advance. Bill Roberts NCAR/HAO Boulder, CO !hao!bill
bezanson@gumby.WISC.EDU (Brian Bezanson) (11/04/86)
In article <286@hao.UUCP>, bill@hao.UUCP (Bill Roberts) writes: > I'm starting to use TML Pascal on a Mac+ (with DataFrame 20 hard disk) i > and tried using the SetPaths DA included with the TML 2.0 release. > First off, > what is 'SetPaths' supposed to do? Set Paths is intended to help programs - specifically TML Pascal, find files on an HFS system. When an open (or similiar) call is done by the compiler to find one of the Units you are using and it is not in the same folder as the Pascal application, then Set Paths steps in and looks in the 5 alternate search areas you have declared with the DA. There were several bugs in the previous versions, but they have been corrected in version 1.3 which is on GEnie. > And secondly, what ever it is supposed to be doing, I don't think it is! > I use SetPaths to setup search paths for files > referenced in the source code of the file that I am trying to compile. > Everything is fine until I try to compile. As soon as the compiler starts to > work on the target file the system hangs! The system shouldn't hang, even without set paths, the compiler should put up an SFGet box so that you can try and hand-select the appropriate files. The TML 2.0 update information is not great at best, I was one of the first to receive the upgrade - I got the new manual with a separate supplement (a friend received his 2.0 upgrade the other day with the supplement bound into the new manual. Anyway, TML 2.0 is now up to 2.01 with several bug fixes. Did you read/see that you need to compile the Units in the Pascal System folder before you go about compiling your programs. This was not overly clear from a quick scan of the new manual. One you have compiled all the units, quit the compiler and rerun it, this will pre-load the MacIntf Unit for superfast compiling. Also, when you get the 1.3 version of SetPaths there is also a doc file that explains its function and use better. Hope this helps any TML 2.0 users who are having similiar problems. -- Brian Bezanson {seismo,ihnp4,allegra,topaz,harvard}!uwvax!gumby!bezanson Manta Software Corp. bezanson@gumby.wisc.edu
chas@ssc-bee.UUCP (Chas Boyd) (11/08/86)
> > I'm starting to use TML Pascal on a Mac+ (with DataFrame 20 hard disk) i >and tried using the SetPaths DA included with the TML 2.0 release. First off, >what is 'SetPaths' supposed to do? And secondly, what ever it is supposed to be >doing, I don't think it is! I use SetPaths to setup search paths for files >referenced in the source code of the file that I am trying to compile. >Everything is fine until I try to compile. As soon as the compiler starts to >work on the target file the system hangs! > So I guess the question is, is 'SetPaths, ver1.1' incompatable with the >Mac+ or is it intrinsically buggy? And what *exactly* is SetPaths supposed to >be doing? Thanks in advance. > > > Bill Roberts > NCAR/HAO > Boulder, CO > !hao!bill I, too, have had this problem on my Mac+. Depending on the paths specified, the compiler or the linker bombs the system. I have also been unable to use TML 2.01 at all from a ramdisk. I think this problem may also be related to SetPaths/HFS. I have yet to try the version 1.3 of SetPaths that was posted to net.micro.mac, but that may fix both problems. -- Chas. - - - - - - - - UUCP^ (uw-beaver|fluke)!ssc-vax!ssc-bee!chas (Charles Boyd) ARPA^ ssc-vax!ssc-bee!chas@uw-beaver WORK^ (206) 773-3878 HOME^ (206) 271-7216 "It wasn't me. I've never been there. I was under orders."
bezanson@gumby.WISC.EDU (Brian Bezanson) (11/09/86)
In article <645@ssc-bee.UUCP>, chas@ssc-bee.UUCP (Chas Boyd) writes: > > So I guess the question is, is 'SetPaths, ver1.1' incompatable with the > >Mac+ or is it intrinsically buggy? And what *exactly* is SetPaths supposed to > >be doing? Thanks in advance. > I, too, have had this problem on my Mac+. Depending on the paths specified, > the compiler or the linker bombs the system. I have also been unable > to use TML 2.01 at all from a ramdisk. I think this problem may also > be related to SetPaths/HFS. I have yet to try the version 1.3 of > SetPaths that was posted to net.micro.mac, but that may fix both problems. Use only SetPaths 1.3. Paul Snively - the author of setpaths and #1 Beta tester of TML 2.0 says do not use 1.2 or 1.2, because they are buggy and 1.3 isn't. On using SetPaths, Snively never intended it to work with RAM disks. He did have it tested on RAM disks for 1.3 but only for Data Storage. The Pascal System still needs to be run off a floppy. He also said that TML will need "at least one more non-trivial revision" before it can work properly with MacApp - this is a big bummer. The good news is that I heard on the APDA graipvine that updates to MPW 1.0 will be sent out free to all Beta users/owners (maybe Larry R. from Apple can confirm this). MPW 1.0 is also said to be near completion and ready for release. Anyway, hope the info on SetPaths 1.3 helps. Snively also mentioned that TML put too much emphasis on SetPaths and not looking for files itself. I found this out big time when my new TML 2.0 wouldn't work on a 800K MFS MacServe volume. 1.2 worked fine, but 2.0 was so dependant of SetPaths that it couldn't even look for file in its own volume/folder. And SetPaths 1.2 wasn't any help either because it only wants HFS folders as paths, not volume names. 1.3 seems to have sold the old problems (I haven't tried putting Pascal back on volumes yet - just use it fine from HFS). -- Brian Bezanson {seismo,ihnp4,allegra,topaz,harvard}!uwvax!gumby!bezanson Manta Software Corp. bezanson@gumby.wisc.edu