Gentoo Forums
Gentoo Forums
Gentoo Forums
Quick Search: in
Why has portage become so slow
View unanswered posts
View posts from last 24 hours

Goto page Previous  1, 2, 3, 4, 5, 6, 7  Next  
Reply to topic    Gentoo Forums Forum Index Portage & Programming
View previous topic :: View next topic  
Author Message
mv
Watchman
Watchman


Joined: 20 Apr 2005
Posts: 6396

PostPosted: Sat Oct 05, 2013 12:52 pm    Post subject: Reply with quote

TomWij wrote:
Please note that things like dependency string USE reduction are not cached in any way and are actually calculated. Embedded systems usually have smaller worlds; so, it scales down to them as well.

Given that most users will probably have much less USE entries than me, if even adding 200 entries makes almost no difference, caching instead of calculating will optimize nothing. It can even hurt performance on systems with low memory.
Back to top
View user's profile Send private message
TomWij
Retired Dev
Retired Dev


Joined: 04 Jul 2012
Posts: 1553

PostPosted: Sat Oct 05, 2013 12:56 pm    Post subject: Reply with quote

mv wrote:
TomWij wrote:
Please note that things like dependency string USE reduction are not cached in any way and are actually calculated. Embedded systems usually have smaller worlds; so, it scales down to them as well.

Given that most users will probably have much less USE entries than me, if even adding 200 entries makes almost no difference, caching instead of calculating will optimize nothing. It can even hurt performance on systems with low memory.


This is not about your local package.use USE flag entries, but rather about the USE flag state of the package; whether you use global USE or local package.use to change that, it doesn't matter.

Let me explain: Portage turns something like A? ( PKG1 ) B? ( PKG2 ) C? ( PKG3 ) D? ( PKG4 ) with USE="A B D" to PKG1 PKG2 PKG4 every time; this USE dependency reduction costs some time (10% according to first profile), as USE flag and dependencies don't change too often we could make a cache entry for the package stating that if the USE flags are still A B D to use PKG1 PKG2 PKG4 (which is a simple comparison as opposed to costly parsing and reducing).

On low memory systems @world will be much smaller, so this will not be problematic; it doesn't have to store much anyway.
Back to top
View user's profile Send private message
mv
Watchman
Watchman


Joined: 20 Apr 2005
Posts: 6396

PostPosted: Sun Oct 06, 2013 9:38 am    Post subject: Reply with quote

TomWij wrote:
Portage turns something like A? ( PKG1 ) B? ( PKG2 ) C? ( PKG3 ) D? ( PKG4 ) with USE="A B D" to PKG1 PKG2 PKG4 every time; this USE dependency reduction costs some time (10% according to first profile)

This is clear to me, but if the time to do one of the presumable most time-consuming steps in this task - namely the verification whether the USE-flag is active for the current package by testing the package name against a list of possibly wild-cards - does almost not increase when the corresponding list is dramatically extended, there is almost no potential to save time through caching. Especially since caching also has some overhead.
Quote:
On low memory systems @world will be much smaller

No, there is no reason to install less packages. Quite the opposite: On e.g. small laptops, one usually needs more packages like a larger collection of programs to connect to arbitrary networks (wireless, samba, printer drivers, modems, zeroconf, etc.) which is all not needed on a desktop but which does not mean that one does not want all of the other typical desktop tools (including - of course - libreoffice, tex, etc). E.g. I have a 512 MB laptop, and already now portage needs there 10 times longer because of swapping. Caching even more could render portage useless on such machines.
Back to top
View user's profile Send private message
TomWij
Retired Dev
Retired Dev


Joined: 04 Jul 2012
Posts: 1553

PostPosted: Sun Oct 06, 2013 1:24 pm    Post subject: Reply with quote

mv wrote:
there is almost no potential to save time through caching. ...

There is potential, because it saves out on parsing and reducing; whereas caching gives the result right away.

mv wrote:
No, there is no reason to install less packages. Quite the opposite: On e.g. small laptops, one usually needs more packages like a larger collection of programs to connect to arbitrary networks (wireless, samba, printer drivers, modems, zeroconf, etc.) which is all not needed on a desktop but which does not mean that one does not want all of the other typical desktop tools (including - of course - libreoffice, tex, etc). E.g. I have a 512 MB laptop, and already now portage needs there 10 times longer because of swapping. Caching even more could render portage useless on such machines.


Small laptops are not embedded; and due to their small size, they usually have more memory (from the money not spent on large laptop screen).

Furthermore, the caching doesn't even have to happen in memory (it is a bonus for high memory systems); the main point here is to replace "parsing and reducing the *DEPEND" by "reading the result from the cache", which can be a cache on either disk or memory.
Back to top
View user's profile Send private message
mackal
Tux's lil' helper
Tux's lil' helper


Joined: 04 Aug 2011
Posts: 87

PostPosted: Sun Oct 06, 2013 5:29 pm    Post subject: Reply with quote

Although I've noticed that portage does take a bit longer sometimes, it does not seem to be as ridiculous as some users seem to be describing it. Plus I like not having to run revdep-rebuild ever again, and counting the time it takes to run revdep-rebuild + emerge -uDN world, just having to run emerge is probably faster. (I still run revdep-rebuild, it just hasn't returned anything in AGES)

I also have a fairly nice computer, so maybe that's why I have no problems, but I also have a lot of packages installed and a few too many overlays, but I'm still getting sub 2 min emerge's, although today it returned 0 packages to update.
Back to top
View user's profile Send private message
mv
Watchman
Watchman


Joined: 20 Apr 2005
Posts: 6396

PostPosted: Sun Oct 06, 2013 6:19 pm    Post subject: Reply with quote

TomWij wrote:
mv wrote:
there is almost no potential to save time through caching. ...

There is potential, because it saves out on parsing and reducing; whereas caching gives the result right away.

You continue to ignore the fact that almost doubling what the algorithm has to do did practically not change the time.
You also continue to ignore the fact that caching has always overhead.
Quote:
Small laptops are not embedded

I spoke about low-memory systems.
Quote:
and due to their small size, they usually have more memory (from the money not spent on large laptop screen).

Great: Get a working portage by buying new hardware. If portage would no longer work reasonably with still not so ancient machines with 512MB, it would be a reason to dump portage completely.
Quote:
which can be a cache on either disk or memory.

Yeah, reading from disk will certainly save lot of time compared to applying an algorithm which currently costed almost no time.
Back to top
View user's profile Send private message
xaviermiller
Bodhisattva
Bodhisattva


Joined: 23 Jul 2004
Posts: 8621
Location: ~Brussels - Belgique

PostPosted: Sun Oct 06, 2013 7:00 pm    Post subject: Reply with quote

Hello,

I tested to call egencache on pro-audio and my local overlay. No outstanding enhancement.
_________________
Kind regards,
Xavier Miller
Back to top
View user's profile Send private message
TomWij
Retired Dev
Retired Dev


Joined: 04 Jul 2012
Posts: 1553

PostPosted: Sun Oct 06, 2013 7:54 pm    Post subject: Reply with quote

mv wrote:
TomWij wrote:
mv wrote:
there is almost no potential to save time through caching. ...

There is potential, because it saves out on parsing and reducing; whereas caching gives the result right away.

You continue to ignore the fact that almost doubling what the algorithm has to do did practically not change the time.


Sentences that contain "almost", "potential" and "practically" are no facts; since you are talking about doubling, you do not seem to be talking about caching.

mv wrote:
You also continue to ignore the fact that caching has always overhead.


Depends on the type of overhead you are talking about.

mv wrote:
Quote:
Small laptops are not embedded

I spoke about low-memory systems.


And I spoke about embedded systems.

mv wrote:
Quote:
and due to their small size, they usually have more memory (from the money not spent on large laptop screen).

Great: Get a working portage by buying new hardware. If portage would no longer work reasonably with still not so ancient machines with 512MB, it would be a reason to dump portage completely.


Don't make it black on white image; being a progressive as well as a configurable enhancement , this forms no problem.

mv wrote:
Quote:
which can be a cache on either disk or memory.

Yeah, reading from disk will certainly save lot of time compared to applying an algorithm which currently costed almost no time.


It has shown to take 10% of the time, which is quite a large share; I don't understand why you are talking about memory and disk, the storage medium is irrelevant to the simplification of the algorithm complexity.
Back to top
View user's profile Send private message
haarp
Guru
Guru


Joined: 31 Oct 2007
Posts: 531

PostPosted: Fri Oct 18, 2013 2:54 pm    Post subject: Reply with quote

Can the dependency stage be multithreaded? Fighting the current algorithm with all available cores might yield massive improvements.
Back to top
View user's profile Send private message
steveL
Watchman
Watchman


Joined: 13 Sep 2006
Posts: 5153
Location: The Peanut Gallery

PostPosted: Fri Oct 18, 2013 5:49 pm    Post subject: Reply with quote

TomWij wrote:
mv wrote:
You continue to ignore the fact that almost doubling what the algorithm has to do did practically not change the time.

Sentences that contain "almost", "potential" and "practically" are no facts; since you are talking about doubling, you do not seem to be talking about caching.

Doubling N is a fact.
Quote:
mv wrote:
You also continue to ignore the fact that caching has always overhead.

Depends on the type of overhead you are talking about.

Huh? You appear to be in la-la-land whereas mv is a real-world software engineer.
Quote:
Don't make it black on white image; being a progressive as well as a configurable enhancement , this forms no problem.

You do this on the mailing-list as well: you get into discussion and argument on a conceptual level, that has nothing whatsoever to do with implementing and making things work. I think you're trying to say "they can turn if off, so they don't need to use it, and the overhead won't affect them." In which case, just bloody well say so.
Quote:
mv wrote:
Yeah, reading from disk will certainly save lot of time compared to applying an algorithm which currently costs almost no time.
It has shown to take 10% of the time, which is quite a large share; I don't understand why you are talking about memory and disk, the storage medium is irrelevant to the simplification of the algorithm complexity.

Man you are off your tits. I haven't heard so much denial of reality for a long time. It's very relevant to the execution time, which is what this is about. Getting the job done. Which is your only use as a developer: to help the user get their job done, as effectively and efficiently as possible, where both are judged by the user, not you.

Get that clear in your head and don't ever forget it. That is what you are here for. The same applies to me or mv, and to everyone else who puts finger to keyboard and expects someone else to use the output.

And take a break, as you are burning out, as evinced by your rambling and incoherent tripe on both mailing lists and forums.
Utter nonsense.
Back to top
View user's profile Send private message
mv
Watchman
Watchman


Joined: 20 Apr 2005
Posts: 6396

PostPosted: Fri Oct 18, 2013 6:08 pm    Post subject: Reply with quote

steveL wrote:
mv is a real-world software engineer.

Nope. I'm a mathematician :wink:
Back to top
View user's profile Send private message
TomWij
Retired Dev
Retired Dev


Joined: 04 Jul 2012
Posts: 1553

PostPosted: Fri Oct 18, 2013 8:03 pm    Post subject: Reply with quote

steveL wrote:
TomWij wrote:
mv wrote:
You continue to ignore the fact that almost doubling what the algorithm has to do did practically not change the time.

Sentences that contain "almost", "potential" and "practically" are no facts; since you are talking about doubling, you do not seem to be talking about caching.

Doubling N is a fact.


No, "Doubling N" is a consequence; now the only missing part is the causation to turn it into a fact, so how does caching causes a double N?

steveL wrote:
Quote:
mv wrote:
You also continue to ignore the fact that caching has always overhead.

Depends on the type of overhead you are talking about.

Huh? You appear to be in la-la-land whereas mv is a real-world software engineer.


Depends on the la-la-land and real-world you are talking about. What are we even talking about? That's why I'm expecting elaboration on the overhead, it could mean anything; so it needs to be defined...

steveL wrote:
Quote:
Don't make it black on white image; being a progressive as well as a configurable enhancement , this forms no problem.

You do this on the mailing-list as well: you get into discussion and argument on a conceptual level, that has nothing whatsoever to do with implementing and making things work. I think you're trying to say "they can turn if off, so they don't need to use it, and the overhead won't affect them." In which case, just bloody well say so.


That's what it says, in one word, "configurable"; I'd rather not expand the length of my mails, unless asked for more details.

steveL wrote:
Quote:
mv wrote:
Yeah, reading from disk will certainly save lot of time compared to applying an algorithm which currently costs almost no time.
It has shown to take 10% of the time, which is quite a large share; I don't understand why you are talking about memory and disk, the storage medium is irrelevant to the simplification of the algorithm complexity.

Man you are off your tits. I haven't heard so much denial of reality for a long time. It's very relevant to the execution time, which is what this is about. Getting the job done. Which is your only use as a developer: to help the user get their job done, as effectively and efficiently as possible, where both are judged by the user, not you.

Get that clear in your head and don't ever forget it. That is what you are here for. The same applies to me or mv, and to everyone else who puts finger to keyboard and expects someone else to use the output.

And take a break, as you are burning out, as evinced by your rambling and incoherent tripe on both mailing lists and forums.
Utter nonsense.


Please do not use disrespectful personal ad hominem statements that don't reflect truth on this discussion forum as they are not constructive; if you want to counter argument me, then feel free to back up your contradiction with facts and references.
Back to top
View user's profile Send private message
Ant P.
Watchman
Watchman


Joined: 18 Apr 2009
Posts: 6875

PostPosted: Fri Oct 18, 2013 8:37 pm    Post subject: Reply with quote

TomWij wrote:
Please do not use disrespectful personal ad hominem statements that don't reflect truth on this discussion forum as they are not constructive; if you want to counter argument me, then feel free to back up your contradiction with facts and references.

With all due respect, 100% of flame wars outside of OTW lately have had you in common.

It's getting more than a little disappointing that I can guess the contents of a topic based on the fact it spans multiple pages and you were the last to post.
Back to top
View user's profile Send private message
TomWij
Retired Dev
Retired Dev


Joined: 04 Jul 2012
Posts: 1553

PostPosted: Fri Oct 18, 2013 8:56 pm    Post subject: Reply with quote

Ant P. wrote:
TomWij wrote:
Please do not use disrespectful personal ad hominem statements that don't reflect truth on this discussion forum as they are not constructive; if you want to counter argument me, then feel free to back up your contradiction with facts and references.

With all due respect, 100% of flame wars outside of OTW lately have had you in common.


Which statistics are that 100% based on? Please be serious and don't make up things. This thread is not the right place to do that, feel free to contact me instead...

Ant P. wrote:
It's getting more than a little disappointing that I can guess the contents of a topic based on the fact it spans multiple pages and you were the last to post.


Thank you for your feedback, I don't know what you mean by that but you are free to detail that by sending me an e-mail; could you now please get back on topic?
Back to top
View user's profile Send private message
TomWij
Retired Dev
Retired Dev


Joined: 04 Jul 2012
Posts: 1553

PostPosted: Sat Oct 19, 2013 12:00 am    Post subject: Reply with quote

haarp wrote:
Can the dependency stage be multithreaded? Fighting the current algorithm with all available cores might yield massive improvements.


The Global Interpreter Lock of Python makes this hard to implement, while you can do multithreading the lock slows things down tremendously;
another option is to implement this on PyPy instead, but I am not sure if that is going to be considered as a default interpreter soon.

On a side note, a patch has landed on gentoo-dev which claims a ~2% reduced run time improvement; see http://thread.gmane.org/gmane.linux.gentoo.devel/88547
Back to top
View user's profile Send private message
grey_dot
Tux's lil' helper
Tux's lil' helper


Joined: 15 Jul 2012
Posts: 142

PostPosted: Sat Oct 19, 2013 3:21 pm    Post subject: Reply with quote

TomWij wrote:

The Global Interpreter Lock of Python makes this hard to implement, while you can do multithreading the lock slows things down tremendously;
another option is to implement this on PyPy instead, but I am not sure if that is going to be considered as a default interpreter soon.


Why? Using PyPy as a default interpreter is not required. In fact it is only necessary to be sure that portage works with PyPy. I know several examples of software speed greatly improved after switching to a decent python implementation (e.g. stackless).

Also the very idea of implementing complex software in python seems just wrong because of the language downfalls.
Back to top
View user's profile Send private message
TomWij
Retired Dev
Retired Dev


Joined: 04 Jul 2012
Posts: 1553

PostPosted: Sat Oct 19, 2013 4:01 pm    Post subject: Reply with quote

grey_dot wrote:
TomWij wrote:

The Global Interpreter Lock of Python makes this hard to implement, while you can do multithreading the lock slows things down tremendously;
another option is to implement this on PyPy instead, but I am not sure if that is going to be considered as a default interpreter soon.


Why? Using PyPy as a default interpreter is not required. In fact it is only necessary to be sure that portage works with PyPy. I know several examples of software speed greatly improved after switching to a decent python implementation (e.g. stackless).

Also the very idea of implementing complex software in python seems just wrong because of the language downfalls.


That would be an eventual goal if we want every user to benefit as Python's GIL might become a show stopper for everyone towards the future. PyPy could be such a switch, but it'll be a while until we reach that.

Indeed and that might be the case, but we should look at short term goals to keep Portage usable; in the long term, we might or might not see the need for a change in language and implementing it from scratch.
Back to top
View user's profile Send private message
grey_dot
Tux's lil' helper
Tux's lil' helper


Joined: 15 Jul 2012
Posts: 142

PostPosted: Sat Oct 19, 2013 4:16 pm    Post subject: Reply with quote

TomWij wrote:

Indeed and that might be the case, but we should look at short term goals to keep Portage usable; in the long term, we might or might not see the need for a change in language and implementing it from scratch.


I can recall atleast two separate gentoo package management systems implemented in other languages - pkgcore and paludis. The former seems to be quite dead, the second one is moving that direction. Seems like nobody is really interested.
Back to top
View user's profile Send private message
TomWij
Retired Dev
Retired Dev


Joined: 04 Jul 2012
Posts: 1553

PostPosted: Sat Oct 19, 2013 4:28 pm    Post subject: Reply with quote

grey_dot wrote:
TomWij wrote:

Indeed and that might be the case, but we should look at short term goals to keep Portage usable; in the long term, we might or might not see the need for a change in language and implementing it from scratch.


I can recall atleast two separate gentoo package management systems implemented in other languages - pkgcore and paludis. The former seems to be quite dead, the second one is moving that direction. Seems like nobody is really interested.


pkgcore is mostly in Python, though there is indeed a small part in C (probably just the part that needs extra speed) and another part is Bash (for the ebuild); its founder has stopped working on it, in the middle of July two developers have picked it up but I'm not sure if they are going through with this. So yes, it is in its apparent dead state; but who knows we might see a change, EAPI 5 was almost finished in its current state so who knows we might see something soon. If Portage becomes unbearable and unfixable I might join them to help; will see, currently it is still fast enough to not be too bothered.

(https://code.google.com/p/pkgcore/source/browse/#git%2Fpkgcore%253Fstate%253Dopened)

Paludis indeed is in another language; but it makes itself somewhat quite different and stricter than Portage, so I think that might possibly scare of some people. There are also some developers that don't accept Paludis bugs and ask people to try again with Portage; because they want to be sure it is an actual bug, and not due to the package manager behaving different. However, there are sometimes bug reports here and there from Paludis so it is definitely is being used by a share of Gentoo users.

I think there would be more interest and acceptance for a new package manager that is implemented in a different language but lies close to what Portage was.
Back to top
View user's profile Send private message
mv
Watchman
Watchman


Joined: 20 Apr 2005
Posts: 6396

PostPosted: Sat Oct 19, 2013 4:48 pm    Post subject: Reply with quote

TomWij wrote:
I think there would be more interest and acceptance for a new package manager that is implemented in a different language but lies close to what Portage was.

I agree, but with portage relying more and more on python-specific things like ConfigParse (e.g. for sets.conf, layout.conf, overlays.conf), this is becoming increasingly more cumbersome. For C++ maybe some things can be stolen from eix (make.conf, *.conf, profile and maskfile(s) parsing is implemented there as well as proper BasicVersion handling) but it would still take a long path to a full-blown package manager, especially since USE-Flags and dependnencies are not even slightly supported yet, not to speak about the bash side of ebuild parsing which has also become more and more complex in the previous years. Moreover, eix's main philosophy is currently to get a reasonable output for any kind of input rather than a proper error reporting, so lots of the error handling would have to be worked over.
Back to top
View user's profile Send private message
steveL
Watchman
Watchman


Joined: 13 Sep 2006
Posts: 5153
Location: The Peanut Gallery

PostPosted: Sun Oct 20, 2013 7:05 am    Post subject: Reply with quote

TomWij wrote:
Please do not use disrespectful personal ad hominem statements that don't reflect truth

Those were not ad hominem: they were all about your statements and how you discuss, which more often than not seems to veer off into conceptual or cod-philosophical meta-argument: discussion about discussion and how nice the ideas appear. Not very useful, and ends up taking several emails or posts with very little content.

Personal statements would be about an individual and his hygiene for example, or about their character; I have not alluded to any such, and I never do, as such things are irrelevant, as well as being uncalled-for, and against what I believe in. You should learn to detach your ego from your output: it's something people get taught in Art schools before degree level, but is not something easy for anyone to do. In computing it normally takes a few big mess-ups before people get it, and it's not normally til their early 30s; depends when they started, and as ever on so many other factors.
Back to top
View user's profile Send private message
TomWij
Retired Dev
Retired Dev


Joined: 04 Jul 2012
Posts: 1553

PostPosted: Sun Oct 20, 2013 9:40 am    Post subject: Reply with quote

steveL wrote:
TomWij wrote:
Please do not use disrespectful personal ad hominem statements that don't reflect truth

Those were not ad hominem: they were all about your statements and how you discuss, which more often than not seems to veer off into conceptual or cod-philosophical meta-argument: discussion about discussion and how nice the ideas appear. Not very useful, and ends up taking several emails or posts with very little content.


How do we get to content, if references to outlined factual examples that reflect what you are trying to talk about are missing; you are now leading by example of how a thread veers off if you don't do that.

steveL wrote:
Personal statements would be about an individual and his hygiene for example, or about their character; I have not alluded to any such, and I never do, as such things are irrelevant, as well as being uncalled-for, and against what I believe in.


Then please stop responding to tone, because the tone is on the same line as a person's hygiene and character; as an example, you refer to my tone being "ego":

steveL wrote:
You should learn to detach your ego from your output: it's something people get taught in Art schools before degree level, but is not something easy for anyone to do. In computing it normally takes a few big mess-ups before people get it, and it's not normally til their early 30s; depends when they started, and as ever on so many other factors.


Which output being ego are you exactly talking about?

I can't learn anything from you without solid argumentation, facts and references.
Please assume good faith and stop attaching an ego connotation to my output...

So, please stop disrupting the service on the Gentoo websites and get back on the actual topic of this thread;
once elsewhere was enough, please stop doing it again on the forums.
Back to top
View user's profile Send private message
steveL
Watchman
Watchman


Joined: 13 Sep 2006
Posts: 5153
Location: The Peanut Gallery

PostPosted: Sun Oct 20, 2013 3:50 pm    Post subject: Reply with quote

You have completely misconstrued my point. I have not said anything in other media because I figured you'd get bored when you realised no-one was that interested in your latest deconstruction of the meaning of words.

Please do not try to make this into something which it is not: this is an informal discussion forum, not a University debating chamber. As such trying to tie it down with rules based on misinterpreted statements just makes it into something which it is not, though it is very easy to pick out and misconstrue if you are want to: after all these are mostly throwaway comments, and most of us do not want to have to go over everything we write to filter it in the same way we might a corporate press release.

AFAIC you are trying to make the forums into a place where a developer can talk nonsense, and no-one can point out that it is nonsense, as said developer cannot detach his words from himself, and takes criticism of his words, as criticism of himself personally.

That does not work, especially not on a forum like this, where it is natural for people to make mistakes. If we continue down this road, the forums will lose what makes it such a useful resource: the sense of community. Simply put: we all talk rubbish sometimes, and we're allowed to. Don't make such a big deal out of it, and don't automatically reach for calling criticism of your words, criticism of you. The same applies to your code.

You turn around and start implying threats, by first casting a criticism of your words as ad hominem, and then ignoring the mistake pointed out to you, to instead go further and translate that into a wider "interfering with the forums" when in fact it's just discussion, exactly the same type of discussion that has always happened on these forums. Personally I think that behaviour sucks, but that's just my opinion, as is everything else I come out with: that's usually taken as a given, or we'd have to qualify literally every post.

Anyhow, I assume you're intelligent enough to work out what I mean now, by "detach your ego from your output." If you still want to misconstrue an attack on your words as an attack on your person, there is nothing I can do about it, but wish you luck in your maturation as a human-being.

Regards,
steveL.
Back to top
View user's profile Send private message
TomWij
Retired Dev
Retired Dev


Joined: 04 Jul 2012
Posts: 1553

PostPosted: Sun Oct 20, 2013 4:17 pm    Post subject: Reply with quote

steveL wrote:
You have completely misconstrued my point. I have not said anything in other media because I figured you'd get bored when you realised no-one was that interested in your latest deconstruction of the meaning of words.


My interest is to have an on-topic discussion here, where what we discuss is clear thus having the meaning of words clarified in normal in the context of a discussion; I have not insinuated that you have said anything in the other media (please note that I am not addressing media where I was part of), though under your interpretation of that media you refer to a personal mail to me would have been appreciated as I do want to improve communication if it is perceived incorrectly due to something like a language barrier.

But I just ask want you to stop being off-topic here on the forums; in your own favor, as well as in the favor of the constructive discussion we were pursuing in this thread.

steveL wrote:
Please do not try to make this into something which it is not: this is an informal discussion forum, not a University debating chamber.


Then please get back on topic; this is not a place where we make statements about people participating in the discussion, but rather a place to discuss the regression in the performance of Portage.

steveL wrote:
As such trying to tie it down with rules based on misinterpreted statements just makes it into something which it is not, though it is very easy to pick out and misconstrue if you are want to: after all these are mostly throwaway comments, and most of us do not want to have to go over everything we write to filter it in the same way we might a corporate press release.


Asking for clarification is not the same as filtering for corporate press release; without facts and references, things here are merely thoughts and opinion without any form of weight to believe them.

When discussing performance regressions; we want to have this certainty, as we do not want to waste time on implementing a solution that has not sufficiently been judged.
But we rather want to spend time on something that we can all agree on; that way, we can collaboratively focus on a better implementation of that feature than to each work on our own...

The comments between you and me are throwaway, they do not belong in this Portage performance discussion and have been reported to be removed from this thread.

steveL wrote:
AFAIC you are trying to make the forums into a place where a developer can talk nonsense, and no-one can point out that it is nonsense, as said developer cannot detach his words from himself, and takes criticism of his words, as criticism of himself personally.


Could you elaborate what makes you believe so? I have perceived that you were talking about me personally in your first comment instead of talking about my words; I am happy to accept criticism of my words, but since you do not address the topic of this discussion near the end I have perceived that to be exclusively about me and that is no longer on the topic of my words or the topic of this thread...

steveL wrote:
That does not work, especially not on a forum like this, where it is natural for people to make mistakes.


Mistakes aren't bad, defending them is; not forgiving them is even worse.

steveL wrote:
If we continue down this road, the forums will lose what makes it such a useful resource: the sense of community.


A sense of community starts with professionalism and constructive discussion; if everyone's going to respond to the tone of other writers and not assume good faith, we cannot reach that.

steveL wrote:
Simply put: we all talk rubbish sometimes, and we're allowed to. Don't make such a big deal out of it, and don't automatically reach for calling criticism of your words, criticism of you. The same applies to your code.


We are not allowed to, as per the rules. I do not make a big deal out of it; the only deal that concerns me is that there's now a page of comments not on the topic of to the original discussion, and due to the nature of our comments we're not reaching consensus either so I do not see any reason for this to even continue.

steveL wrote:
You turn around and start implying threats, by first casting a criticism of your words as ad hominem


An ad hominem statement is not a threat, otherwise I would have labeled it a personal attack; which I do not consider it, it is rather a statement that is off-topic and distracts from the discussion being held in this thread.

steveL wrote:
and then ignoring the mistake pointed out to you, to instead go further and translate that into a wider "interfering with the forums" when in fact it's just discussion


Discussion that does not belong to this thread; you are free to start a new thread, although I expect these replies to be moved to such thread soon.

steveL wrote:
exactly the same type of discussion that has always happened on these forums.


But not in this forum thread.

steveL wrote:
Personally I think that behaviour sucks, but that's just my opinion, as is everything else I come out with: that's usually taken as a given, or we'd have to qualify literally every post.

Anyhow, I assume you're intelligent enough to work out what I mean now, by "detach your ego from your output." If you still want to misconstrue an attack on your words as an attack on your person, there is nothing I can do about it, but wish you luck in your maturation as a human-being.


By labeling my output as "ego", "misconstrued", "attack", "immature" and "unhuman" you are not assuming good faith; please assume it, and read my output again.

Please note that this does not imply that I assume that you assume bad faith; I know you don't assume bad faith as you are trying to have me improve my output,
I just want you to know I cannot change anything about my output if you don't assume good faith (as having my output labeled with such words does not buy my interest).
Back to top
View user's profile Send private message
TomWij
Retired Dev
Retired Dev


Joined: 04 Jul 2012
Posts: 1553

PostPosted: Sun Oct 20, 2013 8:23 pm    Post subject: Reply with quote

The above discussion has been resolved by directly communicating with steveL; it was based on misunderstanding each other, we have spend the last hour talking it through and we have agreed to agree and disagree.
Back to top
View user's profile Send private message
Display posts from previous:   
Reply to topic    Gentoo Forums Forum Index Portage & Programming All times are GMT
Goto page Previous  1, 2, 3, 4, 5, 6, 7  Next
Page 3 of 7

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum