[Mageia-dev] Please test: initscripts+systemd in updates_testing

[Thomas Backlund]

Colin Guthrie skrev 31.10.2011 19:54:
> [Sorry for multiple replies, missed this bit in the last one :s]
>
> 'Twas brillig, and Thomas Backlund at 31/10/11 17:06 did gyre and gimble:
>> But I guess that's the development trend nowdays: "I cant be bothered to
>> fix things properly so I just call it "depreceated"... and go ahead
>> and break things just as I like"
>
>
> I think that's a little unfair in the representation of this problem (I
> won't comment if this *is* an overall trend).
>

Yeah, I guess I'm blowing some steam based on some misunderstandings 
regarding this problem :)

On the overall trend I think it still stands, atleast for many companies 
that I have had to deal with when they deliver
apps/"cloud *"/... that does not work even closly to how it's supposed 
to... :/

<ot>
One _big_ company built a new ticket ordering system, and when they
send the tickets by mail they will be sent as pdf. so far so good.
But then the guy responsible for the coding pretty much thougth
"hey its a file that is in pdf format, so lets set the mimetype to 
file/pdf". and even if I have pointed out many times that there is
no such thing, it's still not fixed.

An other companys booking system showes the whole booking info
that belongs in the mail body in the mail header. and when I
pointed out the problem, the response was pretty much "it's too
hard to fix", and since mailservers can cope with it, I guess it
will never be fixed...

I could go on, but this is just the two last incidents I have
had to work around and it seems it just happends more and more...
</ot>

> I think there has been a significant push recently to break from
> "traditions" that are holding things back. If something doesn't work *by
> design* then coming up with an alternative way of working and
> deprecating the old way seems like a perfect way of driving progress IMO.
>

Yeah, in some cases deprecating is unavoidable, that's true, in other 
cases it seems to have been done too easily.

> The "I can't be bothered to fix things properly" could easily be applied
> to the current status quo, which has a carefully ordered sequence of
> otherwise parallelisable tasks and a careful placement of certain key
> files, just because no one can "be bothered" to address the underlying
> design issues.
>

touché :)

> I guess it's likely just a matter of perspective/opinion as to which
> philosophy "can't be bothered".

Yeah,
but it's always good to get perspective/opinions from several points in 
order to get a good product in the end :)

--
Thomas