[Lazarus] Application Information

[Mark Morgan Lloyd]

Reinier Olislagers wrote:
> On 26/12/2013 09:26, Mark Morgan Lloyd wrote:
>> Reinier Olislagers wrote:
>>> On 25/12/2013 20:16, Mark Morgan Lloyd wrote:
>> I don't pretend to understand all the cross-platform issues and I was
>> trying to avoid work :-)
> Very understandable ;)

There'll be something there in a few minutes. After that I'm trying to 
get my teeth into transcribing http://retro-b5500.blogspot.co.uk/ from 
the original Klingon into Pascal.

> In the end, any contribution is better than none - people knowledgeable
> about other platforms can always contribute about theirs.
> At least the trunk resource units I linked to are cross-platform, so
> that shouldn't be a problem.
> 
> And as usual, I'd be happy to to demolish^H^H^H^H go over the page you
> created and update what I can...

I don't think there's much urgency once it shows up in the Wiki search 
results, except for deleting (or at least marking as deprecated) 
anything that you know is obsolete.

>> OK, I'll do a very basic page "Show Application Title, Version, and
>> Company" i.e. similar to the existing one that somebody's filled with
>> code- I don't want to mess around with that until we know where it came
>> from.
>>
>> Somebody was asking about getting version info into the final (i.e.
>> executable) filename. I've put date info into filenames on Linux using
>> Compilation -> Execute After -> Command (with mv, `date` and so on) but
>> if access to the IDE-generated version info is now settling down
>> presumably it wouldn't be too difficult to read it out as the final
>> stage of the build process- even if this takes a separate 'brand' program.
> Could be. Haven't looked at file version info etc lately.

I'll append a couple of hacks that I find useful. They're better in the 
Wiki than discussed in email etc. since they rely on accurate backticks 
and quoting.

>> I suppose the next question is going to be how to sign an executable so
>> that collaborating programs can check each others' authenticity. I
>> believe that ELF doesn't have provision in the header for anything like
>> that, but presumably it could be stuck into the resource info e.g. in
>> the FileDescription field.
> Or a separate resource actually. I presume that's how it done on
> Windows, OSX too.
> Don't know if there are any initiatives about that in the Linux world...

Not as far as I know. But I'm very slowly trying to put together 
something collaborative, and it would be a useful feature if programs 
working at different levels (i.e. at different places relative to a tree 
of database servers) could validate each other to ensure that nobody had 
inserted code to spoof a logon.

If something (an IDE plugin?) inserted two hashes (one plain, the other 
salted) then it would be possible both to validate that code hadn't been 
corrupted in transit and that it hadn't been built by somebody unauthorised.

-- 
Mark Morgan Lloyd
markMLl .AT. telemetry.co .DOT. uk

[Opinions above are the author's, not those of his employers or colleagues]