BGAFC Vendégkönyv

Σ: 1 post

TCH (statz)

#1, Főfasz (10443)

4073 | #4f07 | ^ | Idézet | Mon, 10 Feb 2020 19:23:41 +01

188.143.*.*

*.pool.digikabel.hu

@Carlos:
About the multiple choice selection, you're right. It isn't really an issue. But "all but no" I have only to click a single time. No I have to click yes, ?, partial. But again, it isn't an issue.

Yes, that is true, it is simpler that way. Luckily when i had constructed the new code, i was thinking of 100% backward compatibility, so the old parameters of ft#=x are still working and also, the option #4 for argument 3 and 5 also have a wrapper to check the corresponding fields. I now simply put back the items into the list and if you check them, then it will check the other three after filtering. This was not planned at all, it is simply a byproduct of a backward compatibility wrapper. Backward compatibility paid off again.

@Carlos:
Regarding SUSE, if I make this selection:
oses4ppc.php?ft%5B1%5D%5B1%5D=on&ft%5B3%5D%5B1%5D=on&ft%5B3%5D%5B2%5D=on&ft%5B3%5D%5B3%5D=on&ft%5B7%5D%5B4%5D=on

Then only Tumbleweed appears... Ahh ... got it.

If I make this selection:
oses4ppc.php?ft%5B1%5D%5B0%5D=on&ft%5B3%5D%5B1%5D=on&ft%5B3%5D%5B2%5D=on&ft%5B3%5D%5B3%5D=on&ft%5B7%5D%5B4%5D=on

Then Leap 15.1 apears, too. But if I select highest supported version, Leap 15.1 should appear in any case. Because it's the latest and or highest supported Leap version :-)

Okay, i see now what is the problem. The highest/latest is done by SQL GROUP BY and it was ordered by version number which brought 42.3 as top, but since that is not supported and you filtered unsupported entries, the secondary filter dropped it. If you do not filter unsupported ones, it appears. This not a bug as no code error is present here, this rather is a conceptional flaw in the approach...if we can call the lack of anticipation that the lower version occasionally can be the higher one as a conceptional flaw...as in OpenSUSE Leap 42.3 < OpenSUSE Leap 15.1... However the flaw also applies if the lower one is the lower one, but it is supported, when the higher one is not. Example: Here are all the Ubuntus what are still supported and supports BE systems. Now, if we tell the list to list only the highest version, then we got nothing, because there are higher versioned Ubuntus, but they dropped BE systems.

Now, about fixing these flaws...the first one is simple, it cannot be fixed from code, as it defies arithmetics at it's very base; yet, since the real ordering is not done by ordering '42.3' and '15.1', but a hidden field, which contains '42003000' and '15001000', i can change the first one to '42003' and thus fix it from data. It will look stupid, but it works. The second one can be fixed from code, but i'll have to think about it as i have to reconstruct the entire grouping part.

Update: Fixed it. The grouping was tricky, but thanks to Taryn, i now learned the needed trick, to group and order a filtered list and get only top rows.

@Carlos:
Ah and I just saw SLES is at version 15, also SLED.

Thanks. I added them and also SLES version 11 too, because i have just found out, that it is available for POWER. And SLES 11 and 12 are still supported.

@Carlos:
and another question: I was pretty sure I saw Windows NT on your list, too. Did you remove it or am I mistaken?

You were mistaken, there were never were any kind of windows on this list...this is still the Bil Geec Anti-Fan Club, you know. :)