flat assembler
Message board for the users of flat assembler.
Index
> Main > ? : Goto page 1, 2, 3 Next |
Author |
|
Tomasz Grysztar 27 Nov 2010, 22:03
Hmm, should be "reserved word used as symbol", I think.
Nice catch. |
|||
27 Nov 2010, 22:03 |
|
ouadji 27 Nov 2010, 22:19
no ... "invalid operand" i would understand that both compile, i would understand that both don't compile ... but one and not the other ... it seems a bit strange and a bit illogical nice catch .. it's not a "catch" ... it's completely by chance, I must all try for wink i think both should be allowed ... or not ... but not "one ok and not the other" what do you think about it Tomasz ? |
|||
27 Nov 2010, 22:19 |
|
Tomasz Grysztar 27 Nov 2010, 22:52
I meant that it should be "reserved word used as symbol" message for the "?:" line, just like with any other reserved word.
|
|||
27 Nov 2010, 22:52 |
|
ouadji 27 Nov 2010, 23:02
ok, I agree. "?:" <--- This should not be allowed (but with last fasm 1.69.27, this is allowed) |
|||
27 Nov 2010, 23:02 |
|
Tomasz Grysztar 27 Nov 2010, 23:08
I made a silent update with a fix.
|
|||
27 Nov 2010, 23:08 |
|
ouadji 27 Nov 2010, 23:26
all ok now. thank you Tomasz. |
|||
27 Nov 2010, 23:26 |
|
killasmurf86 28 Nov 2010, 09:54
Tomasz Grysztar wrote: I made a silent update with a fix. Please don't do that next time. Need to update FreeBSD port now.... before someone noticed, that it's broken now |
|||
28 Nov 2010, 09:54 |
|
Tomasz Grysztar 28 Nov 2010, 10:25
I did implement the time marking system to make it possible to detect minor updates to packages - I often do ones that do not change the name of file, though usually it is the case for Windows package only (as it contains macro packages and IDE sources that often get updated before the new core release).
As for the change in the core causing the "silent update" - I only do that (very rarely) if I have some additional correction on the very same day when the last core version was uploaded. If you allow yourself a one day delay to make the package "settle down", then it may be safe to rely on Linux/Unix packages not changing, because they don't contain any IDE right now. |
|||
28 Nov 2010, 10:25 |
|
revolution 28 Nov 2010, 10:32
Maybe a fourth digit?
1.69.27.02 |
|||
28 Nov 2010, 10:32 |
|
Tomasz Grysztar 28 Nov 2010, 10:47
Date serves the same purpose. That's what I put it there for.
|
|||
28 Nov 2010, 10:47 |
|
revolution 28 Nov 2010, 10:49
But the date is only available on the website. We don't see the date in the program when it is running.
It's not like the numbers are limited, we won't run out of them, so we don't need to conserve them or anything. |
|||
28 Nov 2010, 10:49 |
|
Tomasz Grysztar 28 Nov 2010, 11:17
revolution wrote: But the date is only available on the website. We don't see the date in the program when it is running. Note that most of these updates do not touch the core at all, so the version of core is not relevant there. Thus the date is a real marker for the version of the package (also fasmw has its own versioning, but its not shown on the page, as I think the date is enough). |
|||
28 Nov 2010, 11:17 |
|
revolution 28 Nov 2010, 11:35
Well when someone reports a problem there can be a situation when the difference in version/time is important. And the date of one's download is not necessarily easy to ascertain. Time zone differences and zip extraction software can mask the date of the internal files. And even if I do know the date of the download I would still have to convert it back to your local time zone before I can match to the website date. Time is a trickier thing to deal with than a simpler, unambiguous, version number.
|
|||
28 Nov 2010, 11:35 |
|
ouadji 28 Nov 2010, 11:53
I'm agree. "Silents updates" are not a good thing. it is difficult to see them. This requires multiple downloads often unnecessary. It's not a good way, really. |
|||
28 Nov 2010, 11:53 |
|
Tomasz Grysztar 28 Nov 2010, 11:56
revolution wrote: Well when someone reports a problem there can be a situation when the difference in version/time is important. revolution wrote: Time is a trickier thing to deal with than a simpler, unambiguous, version number. As for the versions inside the package - well, right, there are some components that lack its own version number now, like macros or AsmEdit core. I may consider maintaining version numbers for them as well. |
|||
28 Nov 2010, 11:56 |
|
Tomasz Grysztar 28 Nov 2010, 12:02
ouadji: by "silent update" I mean the situation when I do some minor correction in the core without changing the version of the core - this term is from the very old times when this was more frequently happening than now, since fasm's development was in a much more active stage then. I agree this may be misleading, so I should refrain from doing that.
However there are many other components that get constantly improved - that's why you get lots of other kinds of updates that do not change the version of core, by which package is titled. |
|||
28 Nov 2010, 12:02 |
|
revolution 28 Nov 2010, 12:06
Using time as a version number is ambiguous. My time zone changes frequently. Other people are in different time zones from GMT. FF bowser downloads are time stamped at the local time of download not the server file GMT time. If someone were to ask me if I have the latest fasm version then I would find that a very difficult question to answer without going through various measures to try to determine what version I actually currently have.
|
|||
28 Nov 2010, 12:06 |
|
Tomasz Grysztar 28 Nov 2010, 12:15
revolution wrote: Using time as a version number is ambiguous. revolution wrote: If someone were to ask me if I have the latest fasm version then I would find that a very difficult question to answer without going through various measures to try to determine what version I actually currently have. That's why the "silent update" (as in: silently updating the core) is actually a bad thing - nowadays I only decide to do it when it is something so minor that should not make any difference in real applications. I may as well stop doing that. |
|||
28 Nov 2010, 12:15 |
|
ouadji 28 Nov 2010, 12:17
minor update or major update, a update is a update. about the "core", or not ... not matter ! Silents updates are not easy. It's a problem for us, users. thank you Tomasz |
|||
28 Nov 2010, 12:17 |
|
Goto page 1, 2, 3 Next < Last Thread | Next Thread > |
Forum Rules:
|
Copyright © 1999-2024, Tomasz Grysztar. Also on GitHub, YouTube.
Website powered by rwasa.