T.R | Title | User | Personal Name | Date | Lines |
---|
908.1 | How to 4.0... | OASIS2::BERNARD | Have Gun - Will Travel | Mon Nov 16 1987 14:45 | 5 |
| How do you go about getting the upgrade? I have Lattice 3.10 and haven't
heard a word from them. (Yup... I sent in my registration...)
-John Bernard-
|
908.2 | BLINK GURU's here too | PUERTO::ALVAREZ | Miguel,from sunny Puerto Rico | Mon Nov 16 1987 15:37 | 11 |
| Re.0, BLINK GURU's,
I tried an example assembly program from AMIGA WORLD which appeared
sometime ago. I typed it and assembled correctly ( BTW, I use
a PD assembler). BLINK would GURU, but with ALINK I had no errors
and the program ran fine.
Conclusion, you are not alone, it's probably a BLINK problem.
Miguel A. Alvarez
|
908.3 | In search of the missing LINK | STAR::BANKS | In Search of Mediocrity | Mon Nov 16 1987 17:30 | 65 |
| I don't have the version of BLINK that comes with Lattice 4.0, but
I have come across the following three bugs in the public domain
version of BLINK:
1) If you have a module with too many global symbols in it (I don't
know how much is too many, but I do know that removing some globals
fixed it), BLINK will GURU.
2) Whenever I've asked for an XREF map from BLINK, I get a GURU.
3) Zero length modules with relocatable global symbol definitions
will cause BLINK to generate an incorrectly formatted load file,
although nothing GURUs. To be more specific about this one, an
assembly module like
XDEF Foo
Foo EQU $1
END
BLINKs just fine. However:
XDEF Foo
RORG 0
Foo:
END
Assembles fine, but causes BLINK to make a broken load file. ALINK
can handle this fine, and while some people might think it to be
questionable, it's not too pleasant for a linker to get upset at
something that the compiler/assembler is perfectly happy with.
I found this when debugging an EXEC library, and you can't believe
how much time it cost me.
The above three bugs cost me an entire day's waking time. It's
a large measure of how much I hate ALINK that I still use BLINK
after all this. (2) above was easy to figure out, but the other
two weren't. So, my "rules" for using BLINK without GURU visitations
are:
(1) Don't XDEF a lot of symbols
(2) Don't use the BLINK command line XREF option (not to be confused
with the assembler directive of the same name)
(3) Don't make zero length modules with relocatable symbol definitions.
That is, fix the broken example above by:
XDEF Foo
RORG 0
Foo: DC.B 0
END
Which, of course, makes the load file 4 bytes larger, but in
BLINK's defense, this exec library load file of mine that has
to be 4 bytes too large (for the above reason) is about 400
bytes smaller than what ALINK would produce. It's not that
ALINK isn't properly consolidating code hunks (since I loaded
them all into the same hunk name), it's that ALINK is incredibly
pessimal about how it produces Reloc32 fixup blocks.
Unlike the bugs with BLINK, the above mis-feature of ALINK can be
gotten around with a post processor that consolidates Reloc32 blocks
wherever possible, but I still hate ALINK enough that it ain't worth
the bother to me.
I don't know if any of these are what's been biting you, but I am
sure there are more lurking out there somewhere.
|
908.4 | $45 for 3.10 upgrade... | MVCAD3::BAEDER | | Tue Nov 17 1987 08:42 | 12 |
| re .1
I saw some info on the "NET" about this...I too have 3.10, and it
said the cost was $45 ($75 from 3.03, and previous), but you may
want to call them to check (ask for upgrade service).
In any case, I sent in the card, and the bucks this past weekend...will
post if any problems.....
General comment on the NET lately...Lots of FLUFF, and not much
meat, but a few gems like the tip on upgrade, etc. still cause me
to read it...oh well...
|
908.5 | | ANGORA::SMCAFEE | Steve McAfee | Tue Nov 17 1987 09:52 | 15 |
|
Yes, I paid $45 from 3.10 to 4.0 and it is $75 from 3.03 to 4.0. I
received a letter in the mail with a short form to get the upgrade. The
upgrade arrived about two weeks after I sent in the money/form.
This weekend I intend to chip away at this assembly program until
I figure out what the problem is. Maybe I'll connect to the BBS.
The new manual is really lacking on Assembler documentation. I
believe this is because it is now compatible with the documentation
in the Bantam AmigaDOS manual.
Thanks for the help. I'll let you know if I find out about any general
tips on using assembly and Blink.
- steve mcafee
|
908.6 | | STAR::BANKS | In Search of Mediocrity | Tue Nov 17 1987 12:15 | 17 |
| I may have misunderstood the problem in .0:
Are you getting the GURU when you run BLINK, or are you getting
it when you try to run the resultant load image. In either case,
do you have any better luck with ALINK?
If it isn't a GURU from BLINK, perhaps you could upload the code,
post it here, and we could all take shots at what the problem might
be.
Anything I've done on the Amiga since we got our "developer's kit"
two years ago has been in assembler. I never ran into anything
that made programming the Amiga in assembler more difficult than
C (other than anyone wanting to trot out the old arguments of high
level vs low level). I guess what I'm saying is that there aren't
any "secrets" to getting your assembler program to run that don't
also apply to C programs.
|
908.7 | BLINK gurus | ANGORA::SMCAFEE | Steve McAfee | Tue Nov 17 1987 13:01 | 10 |
|
re: -1
Yes you had the right idea. The problem is that BLINK gurus. I
do have other compiled assembly programs that BLINK does not guru
on however. I thought what I might do is try chipping pieces out
of the program (it's only 4 pages long) until BLINK accepts it.
(thank God for VD0: !!!).
- steve mcafee
|
908.8 | what a fine piece of software ... | STAR::BANKS | In Search of Mediocrity | Tue Nov 17 1987 13:12 | 2 |
| Well, good luck to you. Speaking as another veteran of the BLINK
lost weekend club.
|
908.9 | | BAGELS::BRANNON | Dave Brannon | Tue Nov 17 1987 15:20 | 7 |
| i had my first Blink Guru happen last night, the only change to
the program i was compiling was a couple printfs.
I was running the compiler, linker, etc. out of RAM:, i've never
had Blink Guru when running from floppy.
-dave
|
908.10 | v4 arrived! | MVCAD3::BAEDER | | Wed Nov 25 1987 08:30 | 7 |
| re: .4
came yesterday...about 10 days...not bad for UPS...anyway, complete
new manual, and 4 disks...now to get back home, and play with it!
scott.
|
908.11 | Are "enum's" supported ? | TEACH::ART | Art Baker, DC Training Center (EKO) | Mon Dec 07 1987 11:55 | 18 |
|
Regarding the Lattice-C compiler, does it support
enumerated types ? You know, stuff like:
enum color { red, blue, yellow };
.
.
.
enum color this_color;
.
.
.
this_color = red;
Manx (alas) does not...
Also, any further word on the BLINK problem ?
-Art
|
908.12 | T9����^Z�Z� | HPSTEK::SENNA | | Mon Dec 07 1987 12:08 | 4 |
| RE: .11
Sure'nough!
|
908.13 | I wasn't cursing it was noise! | HPSTEK::SENNA | | Mon Dec 07 1987 12:11 | 3 |
| re: .12
the title of that last note was curtousy of call waiting. Sorry!
|
908.14 | Export version | CHEFS::SKINNER | Andrew Skinner | Tue Feb 16 1988 11:54 | 12 |
| re .10
4 disks? Over here (UK) there seem to be ads for Lattice C V.4 (about
#130-180) and Lattice C Professional, or "Developers'" at about
#280, but its only V3.1. Apparently developers' C is still being
upgraded.
The blurb for V4.0 talks about 2 disks, whch suggests to me that
what you are all talking about is what's described as the
'Professional' package. Can anyone shed more light ? And what sort
of prices are we talking about ?
|
908.15 | 4.0 here | WJG::GUINEAU | The Mathematics of Reality ? | Tue Feb 16 1988 13:28 | 26 |
|
Well, I just picked up Lattice 4.0 at The Software Shop a few weeks back
for (I think) $150. I got the standard kit (not developers or professional)
The documentation mentions 3 disks, and an addendum mentions 4.
I got 4 disks: 1 is Workbench and Lattice Compiler, Assembler, Blink (linker)
OMD (object module dissassembler) OML (object module librarian) was not ready
for release or something so it was missing.
There's also FD2PRAGMA which converts Basic's .FD files into C pragma
files (routine interface specification's to help the compiler bypass
Amiga.lib)
Disk 2 was all the libraries in compressed form (.h for C, .i for ASM)
Disk 3 was all the libraries in un-compressed form and some examples.
Disk 4 was some extras and examples
And a good manual...
That's about all I can remember off the top of my head.
John
|
908.16 | more info...tmu included in new 4.0 package?? | MVCAD3::BAEDER | D. Scott DTN 237-2961 SHR1-3/E19 | Tue Feb 16 1988 16:27 | 13 |
| actually, the oml is included, but they are using a new indexed
library format...this is what the oml that is included can't handle
(not bad for off the top of your head)...the other major change
(as I recall) is the use of register addressing as the default,
and the default libs that are compiled this way (old libs also included
for compatibility)...some source (for startups, etc) and some examples
also included...
by the way, when I got 3.10, it included TMU (Text management
utilities) with it. Are these part of a new 4.0 package. there
not much...grep, ed (from the software tools book by kernigan, and
a few others to make script files, etc...)
|
908.17 | | WJG::GUINEAU | The Mathematics of Reality ? | Tue Feb 16 1988 16:32 | 13 |
|
> by the way, when I got 3.10, it included TMU (Text management
> utilities) with it. Are these part of a new 4.0 package. there
> not much...grep, ed (from the software tools book by kernigan, and
> a few others to make script files, etc...)
Hmmm. Didn;t see them? I'll have to look tonight.
BTW - I can type OMD and get it to run. If I type OML I get "unknown command"
I'll have to search the disks some more...
John (who apparently is missing some hair on the top of his head :-) )
|
908.18 | | STAR::BANKS | In Search of Mediocrity | Tue Feb 16 1988 17:05 | 5 |
| Slightly tangential question:
Does the Lattice 4.0 DOC set document their extensions to the standard
Amiga object file format, or are they keeping them secret? (Just
curious)
|
908.19 | Lattice 4.01 | ANGORA::SMCAFEE | Steve McAfee | Mon Apr 11 1988 13:46 | 9 |
|
There is a Lattice 4.01 update on the the Lattice BBS and I've
downloaded it an applied it to a set of disks. Looks like quite
a few bug fixes. Can I post this to save others the trouble of
phoning lattice? It's quite large. Maybe I could post the
list of bugs it fixes. For those who wish to download this
it's called something like AMFIX401.ARC on the lattice BBS.
- steve
|
908.20 | I'd like it | LDP::MCCARTHY | That's a mighty humble bumble! | Mon Apr 11 1988 13:55 | 6 |
| Post it! I was thinking of calling the Lattice BBS to download
it. Could you at least post the list of fixes to see if I need
the update?
Thanks,
Mike
|
908.21 | | PLDVAX::SMCAFEE | Steve McAfee | Mon Apr 11 1988 14:19 | 6 |
|
I just re-read my posting in .19 and I guess wasn't very clear.
I'm wondering whether it is legal for me to post the update.
I had to give my owner registration number to download this.
- steve
|
908.22 | Oh yeah... | LDP::MCCARTHY | That's a mighty humble bumble! | Mon Apr 11 1988 15:14 | 11 |
| Oh yeah, I hadn't thought of that. Good question. I assume that
the update is a patch, and couldn't be used by itself. Has
anyone seen the update appear on any of the commercial boards
(PLINK, Compuserve, etc.) If so, it's probably legal to post
it here. But then again, I'm not a lawyer...
Could you post the list of bug fixes. That should tell me if
it's worth my while to download from Lattice, or wait for the
next update.
Mike
|
908.23 | | FNYTC4::KENNEDY | Keith | Wed Apr 13 1988 03:46 | 7 |
|
Having just taken delivery of 4.0 I'd really appreciate it if
you could post the list of bugs that 4.1 fixes - it could save
me some trouble later.
KK
|
908.24 | pretty please? | WJG::GUINEAU | | Wed Apr 13 1988 09:18 | 5 |
|
Same here.
John (who *knows* the bug is in the compiler :-)
|
908.25 | | PLDVAX::SMCAFEE | Steve McAfee | Wed Apr 13 1988 10:32 | 5 |
|
Tomorrow night I'll post the list of bugs. I've got class tonight
and a few things to do afterwards...
- steve
|
908.26 | Don't Upload Patches | TLE::RMEYERS | Randy Meyers | Wed Apr 13 1988 13:10 | 34 |
| Re: .25
I am sure that Lattice would not want you to post the patches, am also
sure that it is illegal to do so.
First, Lattice C is not copy protected. One of the rationales in the
great copy protection debate is that if companies provided support,
pirates would become frustrated, and start buying copies of the products.
Lattice is providing support, but they are also demanding that you be
a registered owner of their compiler before you can obtain support.
When you dial into the Lattice BBS, you are permitted to download all files
except the product patches. You are only allowed to download product fixes
after you prove you are a registered owner. You prove you are a registered
owner by typing your registration number from your update card to a special
menu, and then waiting 24 hours for a Lattice employee to verify that number
against their records.
Since Lattice is restricting distribution of fixes to proven owners of
their products, I believe they would view it as a hostile act to post
the fixes to the world. (On the other hand, I am sure that they
wouldn't mind you giving out the fixes to anyone that could prove
to you that they bought the compiler.)
I recommend that everyone obtain the fixes. Lattice usually publishes
fixes as deltas from the latest fix. Since you will have to get the
first set of fixes if you want the future fixes, there is no reason
to wait.
Downloading the fixes isn't too bad. You will need to make a 5 minute
phone call to drop off your registration number with the system. It
will then only take about 15 minutes to download the fix arc file at
1200 baud when you call back on the next night. That is only about
a $5 phone bill.
|