| 1 |
On Tuesday 06 September 2005 17:33, Philip Webb wrote: |
| 2 |
> Sorry, but that's neither adequate nor polite as a response |
| 3 |
> to the genuine problem which I raised as the original poster. |
| 4 |
> The answer to your question should be clear from the rest of my message |
| 5 |
> -- the present warning is misleading, as everyone has agreed -- |
| 6 |
> & your other response needs some explanation on your part. |
| 7 |
> Someone else has suggested that Portage can't handle a N/Y of this kind, |
| 8 |
> but in that case please offer some confirmation at least. |
| 9 |
|
| 10 |
Okay. Let's take it one step through at a time. |
| 11 |
|
| 12 |
> >> 050905 Jason Stubbs wrote: |
| 13 |
> >>> it's possible that unmerging slotted packages of the one key |
| 14 |
> >>> may break your system. |
| 15 |
|
| 16 |
I explain the reasoning behind the current behaviour. |
| 17 |
|
| 18 |
> >> 050905 Jason Stubbs wrote: |
| 19 |
> >>> How's about not warning if there's more than one installed cat/pkg |
| 20 |
> >>> (rather than cat/pkg-ver) satisfying the profile atom that is being |
| 21 |
> >>> triggered? |
| 22 |
|
| 23 |
I suggest how to alter that behaviour to cover the former reasoning and |
| 24 |
address the current issue with it. |
| 25 |
|
| 26 |
> >> -- patch snipped -- |
| 27 |
|
| 28 |
I supply a working and tested patch 45 minutes later. |
| 29 |
|
| 30 |
> >> I'ld say that the behaviour should be left alone |
| 31 |
|
| 32 |
You imply that the patch isn't suitable... |
| 33 |
|
| 34 |
> >> pending a larger rewrite of Portage's handling of this kind of thing. |
| 35 |
|
| 36 |
... for no other reason that it should wait for some unknown "larger |
| 37 |
rewrite" of "this kind of thing". |
| 38 |
|
| 39 |
> >> Meanwhile, simply amend the warning to read in red letters eg |
| 40 |
|
| 41 |
Then you suggest that it would be better to frighten the users even more... |
| 42 |
|
| 43 |
> >> "WARNING : removing this package may break your system !! |
| 44 |
|
| 45 |
... by first summarizing the warning you suggest should follow after the |
| 46 |
user hits yes ... |
| 47 |
|
| 48 |
> >> Have you checked that you have a proper alternative installed ?? |
| 49 |
|
| 50 |
... then suggesting that the user should check what my patch would have |
| 51 |
checked for them anyway ... |
| 52 |
|
| 53 |
> >> If you are not certain what you are doing, please STOP NOW !! |
| 54 |
|
| 55 |
... and then suggesting to the user that they are possibly inept. |
| 56 |
|
| 57 |
> >> Do you want to proceed (NO/yes) ? " |
| 58 |
|
| 59 |
"NO"? s/possibly/very likely/ in the last sentence then. |
| 60 |
|
| 61 |
> Do you respond so abruptly to your neighbours or colleagues at work ? |
| 62 |
|
| 63 |
Should I have used more words that "Why?" to ask you to explain? That was |
| 64 |
half of my email - and the important half at that - but you never responded |
| 65 |
to it. I believe in expediency and your lack of response to the "why?" |
| 66 |
proves (even if it's only in my mind) that using more words would have been |
| 67 |
a waste of time. |
| 68 |
|
| 69 |
Should I have used more words than "No." to explain why your solution is not |
| 70 |
suitable? I don't think so. It didn't seem thought out at all and I don't |
| 71 |
think it should be my job to help you learn how to think. |
| 72 |
|
| 73 |
If it was a neighbour or colleage that came to me with something like the |
| 74 |
above, I'd just laugh and likely never interact with them again. Perhaps I |
| 75 |
should have done that in this case too. |
| 76 |
|
| 77 |
-- |
| 78 |
Jason Stubbs |
Archives