[Info-vax] What's up with DECUS HTNotes ?

[hb]

On 5/16/21 2:51 AM, Stephen Hoffman wrote:
> On 2021-05-15 23:16:04 +0000, Robert A. Brooks said:
> 
>> On 5/15/2021 5:56 PM, Simon Clubley wrote:
>>> On 2021-05-15, <kemain.nospam at gmail.com> <kemain.nospam at gmail.com>
>>> wrote:
>>>>
>>>> Not related to original q, but I tried installing htnotes_021b3.zip on
>>>> latest VSI OpenVMS V8.4-2L3 on rx2600.
>>>>
>>>> $  cc/version
>>>> VSI C V7.4-001 on OpenVMS IA64 V8.4-2L3
>>>>
>>>> Followed install notes, but I get an error on the last build
>>>> statement in
>>>> the build section
>>>>
>>>> $ link/exe=[.client] [.client]ht_client, [.source]ht_client.opt/opt
>>>> sys$library:notes$share/share
>>>> %ILINK-W-COMPWARN, compilation warnings
>>>> shareable image: SYS$COMMON:[SYSLIB]NOTES$SHARE.EXE;1
>>
>>> Also, I'm surprised to see a DEC product shipped with user visible
>>> compiler warnings in the product...
>>
>> It was probably freeware at that point; not a saleable product.
>>
>> A shareable image should never be built with compiler warnings, even
>> if it's a free product . . .
> 
> Notes was pulled into OpenVMS development after its cancellation as a
> product, due to the use of Notes by OpenVMS.
> 
> This once I was able to locate the source code, and to also acquire the
> Notes patches that had become available.
> 
> One of the other developers took it on at the time and had sorted out
> various issues, though there were other... complications.
> 
> This eventually led to the following kits becoming available:
> 
> https://www.digiater.nl/openvms/freeware/v70/notes/
> https://www.digiater.nl/openvms/freeware/v70/notes_ft/
> 
> Based solely on the version numbers and the Itanium, that looks like the
> "beta" kit.
> 
> And given that's a warning, not an error, either ignore it or patch the
> shareable to remove the there-be-build-errors flag, if it bothers you.
> 
> Whether VSI acquired the source code to Notes from then-HP? And if VSI
> did acquire the source code, whether VSI plans to release an updated and
> VSI-branded kit?
> 
> 

The linker issued a warning, not an error.

Without a linker map of NOTES$SHARE you do not know what kind of
link-time warning that was. It could be that one (or more) module(s)
have compile-time warnings. It could be that the linker could not
resolve one (or more) symbol(s). In case of compile-time warnings,
without the compiler listings, you do not know what kind of warnings
that were.

In both cases, when linking the main image, the linker issues a COMPWARN.

You probably know, that you can ignore the link-time warning and (try
to) run the image. It will be activated regardless of the warning(s).
Whether it works is a different question.

Unrelated, you will get an EOMWARN signaled from lib$find_image_symbol
if you try to find a symbol in the activated shareable image.

You can patch the (main and) shareable image with PATCH/ABSOLUTE to set
the completion code to SUCCESS. ANALYZE/IMAGE is your friend. You can
overwrite the link-time warning of an IA64 image with "SET IMAGE/SUCCESS".

PS: The version of the C compiler doesn't matter, here.