| T.R | Title | User | Personal
Name | Date | Lines |
|---|
| 2770.1 | They need Help! | AIMTEC::BUTLER_T | | Fri May 28 1993 13:31 | 11 |
| I'll open my mouth - I used to programmers with funny foreheads
and scraped knuckles.
From and API point of view, it is definitely better in V3.0. Their
use of the .IF as an example confirms their coding standards!
I think they are just pounding their chests to distract from the real
problem.
Tim
|
| 2770.2 | compilation errors | SUBURB::CLEPHANEB | | Fri May 28 1993 13:40 | 9 |
|
I've seen ALL-IN-1 complain about .IF statements when compiling the CM
TXL. Boilerplates/Scripts that compiled OK under V2.4 now produce errors
under V3.0
I'll try and find an example,
Bruce
|
| 2770.3 | The rumour came from DEC... | FLEX7::ALLINGHAM_PD | Permenantly Peaking! | Fri May 28 1993 13:41 | 18 |
| I know what you mean... but this came from *within* DEC - attributed to
someone who really knows his stuff...
Personally, I think that because the software is seriously high profile
they are trying to employ the 'if it isn't broken then don't fix it'
methodology - though as I said it's not exactly quality stuff as it is.
It is very much in our (Digital's) interest to move this stuff off the
2.4 m/c, trash it, and have everything working in 3.0 land. I have
tried to contact the proported originator of the rumour but with no
success - does anyone else have any input on this?
Ta much
Peter.
P.S. It's great to be back in ALL-IN-1 notes after a break of nearly 4
years - and nice to see some of the old names still in harness!
|
| 2770.4 | The Plot Thickens. | FLEX7::ALLINGHAM_PD | Permenantly Peaking! | Fri May 28 1993 13:42 | 4 |
| Bruce, .3 was written before I saw your .2 ... so it may be true... all
further information gratefully received.
Peter.
|
| 2770.5 | I'm trying to be open minded about this! | AIMTEC::BUTLER_T | | Fri May 28 1993 14:11 | 18 |
| The .IF is an example of things fixed that now cause something
that was kludge in the first place to work.
Most cases I have seen have been where they got it to work and do not
understand why. The fix comes along and ...
While I do appreciate this situation, I do not feel we should support
something like this as portable. I have worked with some non-knucklers
who have been in this situation and have redone their code and felt
much better about it and ALL-IN-1.
In addition, a good example might be the TM subsystem and the functions
and special dsab(s) there are alot of changes that will cause
portability problems. I'll let our DPC expand on that.
Tim
|
| 2770.6 | a DPC cue | AIMTEC::ZANIEWSKI_D | Why would CSC specialists need training? | Fri May 28 1993 14:37 | 6 |
| Tim's correct about the TM subsystem. If the customer has done
almost any customization in pre-v3.0, it won't work at all in
v3.0. Several programming functions and at least 1 dsab have been
retired.
Dave Zaniewski
|
| 2770.7 | Killing the .IF lie . . | IOSG::HULIN | Ian Hulin, IOSG: REO, DTN 830-6141 | Fri May 28 1993 18:26 | 14 |
| O.K, here's the truth about .IF/TXL compilation in V3.0.
In V2.4, the TXL compiler used to ignore any syntax errors and plough on.
The compiled code then stood a random 50/50 chance of working as desired as
it produced an invariant result of either 1 or 0.
V3.0 now tells you about errors encountered so that API programmers (however
prominent their brow or trailing their knuckles) can correct them and have a
fighting chance of getting their applications working to spec.
If this note sounds upset, it's because I am.
Ian (V3.0 .IF maintainer)
|
| 2770.8 | Maybe me | SIOG::T_REDMOND | Thoughts of an Idle Mind | Fri May 28 1993 19:04 | 6 |
| I may be the one who told BT (I have told lots of customers this) that
ALL-IN-1 V3.0 does not accept badly written code and try to do its best
with the rubbish any longer. I stand over these words and applaud the
efforts of IOSG to make ALL-IN-1 a predictable programming environment.
Tony
|