| View previous topic :: View next topic |
| Author |
Message |
Naib Advocate


Joined: 21 May 2004 Posts: 3891 Location: UK - Birmingham
|
Posted: Mon Mar 31, 2008 6:48 am Post subject: |
|
|
| dleverton wrote: | | UberPinguin wrote: | | Let's see.. invalid atoms in the world file | Not invalid, just not recognised by portage. And again, you'll notice that all it did was print a warning message. Not that this is relevant to pkgcore, it would have died with a different traceback if it had looked in world. | riiiight and you don't think that a packagemanager that is suppose to "replace" portage is BROKEN when it does things that
| Quote: |
| UberPinguin wrote: | | portage trying to resolve things from incompatible repositories (/var/paludis/repositories/*) | Er, WTF? Are you saying I should make a seperate copy of all my repositories for portage? |
Well since you pointed out that portage/pkgcore doesn't accept this (ie paludis has done something completly different) yer otherwise yr tests are skewed and irrelvant
chroot for each 3 from a stage3 tarball or you are skewing!
| Quote: |
| UberPinguin wrote: | | and eclass overrides that invalidate the portage cache. | Again, nothing to do with Paludis. | BUT portage and pkgcore don't do this from a sane layout ero something is afoot with what paludis is doing
yr tests are skewed and irrelvant
chroot for each 3 from a stage3 tarball or you are skewing!
| Quote: |
| UberPinguin wrote: | | To my knowledge, -foo/bar is invalid; did you perhaps mean =foo/bar ? | No I didn't. | so a nice paludis uniqness again, mmm how is that good for gentoo? How easy is it for someone that had been running paludis for say... a year, with the Portage-compat turned off, how easy is it to go back to portage or are you creating a virtual lockin?
yr tests are skewed and irrelvant
chroot for each 3 from a stage3 tarball or you are skewing!
| Quote: |
| UberPinguin wrote: | | Because that would create a proper testing environment in which one could be certain that the three different package managers being benchmarked are not interfering with each other. | Tell that to patrick. | I think he get the idea of an isolate/fair test alot more then you could possibly imagine concidering yr attepts at one
trying to run portage/pkgcore (which are virtually drop-in replacements to each other in file syntax and command option) with paludis mangled config files funnies thing I have heard today, but I am sure you will exceed this by day's end
"haha pkgcore/portage is slow.... oh all those exceptions/errors popping out from pkgcore/portage on this paludis-ified system.... don't worry abt them, they couldn't possibly mess with how they work)
yr tests are skewed and irrelvant
chroot for each 3 from a stage3 tarball or you are skewing! _________________
| Quote: | | Voting holds no real power, he who counts the votes has the true power. |
Weaver Projects
whats the difference between 9/11 and a cow?
u stop milking a cow after 10 years |
|
| Back to top |
|
 |
dleverton Guru

Joined: 28 Aug 2006 Posts: 517
|
Posted: Mon Mar 31, 2008 8:07 am Post subject: |
|
|
| energyman76b wrote: | | since portage is still THE package manager for gentoo, atoms not recognised by portage are invalid. Just because Ciaranm twists PMS so that Paludis looks good does not change that simple fact. | Both sentences wrong.
| energyman76b wrote: | | nope, just slowing down portage, making it look bad. You really try to artificially make Paludis look better than it is. | If portage slows down significantly in this situation, that's portage's fault, and a valid point to consider when comparing them.
| energyman76b wrote: | | dleverton wrote: | | UberPinguin wrote: | | To my knowledge, -foo/bar is invalid; did you perhaps mean =foo/bar ? | No I didn't.
| so you are using invalid stuff to prove that poopludis is better? | -foo/bar is entirely valid.
| dleverton wrote: | | his tests looks better than yours. Starting with clean stage3 and stuff. Instead of a mangled system full of poopludis friendly, invalid garbage. | There is nothing invalid on my system. |
|
| Back to top |
|
 |
dleverton Guru

Joined: 28 Aug 2006 Posts: 517
|
Posted: Mon Mar 31, 2008 8:15 am Post subject: |
|
|
| UberPinguin wrote: | | Except that these missing digests result in portage going through additional logic and printing more information to console, which slows it down and skews the benchmarks. I really don't care if you don't maintain digests for your local overlay, but your environment is inherently poor for comparing portage against paludis. | As with the eclasses "issues", if portage behaves poorly in certain situations, that's portage's fault, and a (rather minor in this case) reason not to use it.
| Uberpinguin wrote: | | No, I'm saying that your repositories are incompatible with portage, and therefore make a poor environment with which to benchmark it. This would go to the previous argument about why one would want to use a clean environment for each package manager being tested. | I have portage configured to look in /var/paludis/repositories for the shocking reason that that's where my repositories are. Are you saying that portage is incompatible with any repository stored in a path with "paludis" in its name? If so, it's even worse than I thought.
| Uberpinguin wrote: | | So, why else are you overriding eclasses? Even if it's not for paludis, it has been noted previously that the presence of these overrides puts portage at an automatic disadvantage on your system. | I'm not overriding anything, those eclasses are overridden by the overlays I use. Since those overlays are also designed to be used with portage, it's fair to compare how portage treats them when making a comparison.
| UberPinguin wrote: | | What, exactly, is -foo/bar meant to accomplish in package.mask, then? I don't see how it would be useful to negate a package in package.mask. | Did you miss the part about overlays? Do you also think it's acceptable for pkgcore to crash and burn, refusing to perform any operation except possibly sync (I haven't tested that one), on a perfectly valid construct just because "UberPinguin doesn't see how it would be useful"?
| UberPinguin wrote: | | Or, you could stop pointing fingers and come up with some real benchmarks of your own. | That's really not the point. You know what they say about lies, damned lies and those other things.
| UberPinguin wrote: | | Like these: | That's nice for you. Plenty of other people have different results. |
|
| Back to top |
|
 |
dleverton Guru

Joined: 28 Aug 2006 Posts: 517
|
Posted: Mon Mar 31, 2008 8:22 am Post subject: |
|
|
| Naib wrote: | riiiight and you don't think that a packagemanager that is suppose to "replace" portage is BROKEN when it does things that  | Do you know what "replace" means?
| Naib wrote: | | dleverton wrote: | | Er, WTF? Are you saying I should make a seperate copy of all my repositories for portage? | Well since you pointed out that portage/pkgcore doesn't accept this (ie paludis has done something completly different) | Er, completely wrong. portage accepts it just fine, and pkgcore's problems have nothing to do with the location of the repositories.
| Naib wrote: | | yer otherwise yr tests are skewed and irrelvant chroot for each 3 from a stage3 tarball or you are skewing! | Again, you're missing the point completely
| Naib wrote: | BUT portage and pkgcore don't do this from a sane layout ero something is afoot with what paludis is doing
yr tests are skewed and irrelvant | portage and pkgcore don't do what? eclasses in overlays are there because of the overlay maintainer, not because of anything the package manager does.
| Naib wrote: | | dleverton wrote: | | UberPinguin wrote: | | To my knowledge, -foo/bar is invalid; did you perhaps mean =foo/bar ? | No I didn't. | so a nice paludis uniqness again | Wrong. It is supported by portage, and has been for quite a long time. It's paludis that only supported that recently, but before then, it was ignored with a warning rather than a crash.
| Naib wrote: | | [snip random inane ranting] |  |
|
| Back to top |
|
 |
dleverton Guru

Joined: 28 Aug 2006 Posts: 517
|
Posted: Mon Mar 31, 2008 8:28 am Post subject: |
|
|
| energyman76b wrote: | | poopludis | Hey, look, I can do that too! "poopage" and "poopgcore"! See how clever and funny I am?!  |
|
| Back to top |
|
 |
Naib Advocate


Joined: 21 May 2004 Posts: 3891 Location: UK - Birmingham
|
Posted: Mon Mar 31, 2008 8:31 am Post subject: |
|
|
yr tests are skewed and irrelvant
chroot for each 3 from a stage3 tarball or you are skewing! _________________
| Quote: | | Voting holds no real power, he who counts the votes has the true power. |
Weaver Projects
whats the difference between 9/11 and a cow?
u stop milking a cow after 10 years |
|
| Back to top |
|
 |
bonsaikitten Apprentice

Joined: 01 Jan 2003 Posts: 213 Location: Aachen, Germany
|
Posted: Mon Mar 31, 2008 8:51 am Post subject: tEh! |
|
|
So, uhm, hi
Thanks for motivating me to reenable my forums account
| dleverton wrote: |
| UberPinguin wrote: | | Or, you could stop pointing fingers and come up with some real benchmarks of your own. | That's really not the point. You know what they say about lies, damned lies and those other things.
| UberPinguin wrote: | | Like these: | That's nice for you. Plenty of other people have different results. |
So you're saying that because we all get consistent results they must be wrong? Sorry, can't follow.
Also, please stop insulting me, it makes you look bad and only ends up with forum moderators having more work to do. I don't mind you disagreeing with reality, but if we were in a pub drinking beer you'd have my beer on your shirt and the table on your nose by now. |
|
| Back to top |
|
 |
steveL Veteran

Joined: 13 Sep 2006 Posts: 1441 Location: The Peanut Gallery
|
Posted: Mon Mar 31, 2008 8:54 am Post subject: |
|
|
| Moloch wrote: | | I don't feel like doing a side by side comparison because my system is incompatible with portage. |
That's the bit that worries me; so effectively once you've used paludis for a while, that's it: you can't go back to the official package-manager.
How on earth is this helping? The whole project seems deliberately divisive for the sake of it.
And sorry, but --long-opts-just-because-you-can-are-a-PITA-for-usrs although paludis devs have a long history of pouring scorn on users (til ciaranm got kicked out and suddenly became interested in "protecting user's systems" usually by denigrating the QA his co-lead was supposed to be in charge of, though he never did anything. Go figure. Thankfully QA has a new team who are far more proactive, and far less vindictive; in a word, more professional.)
@ ShiRhoHaiSun: As for the so-called "personal attacks" it's standard operating procedure for paludis bods; from where I'm sitting it's clear you at least were not being malicious. |
|
| Back to top |
|
 |
Naib Advocate


Joined: 21 May 2004 Posts: 3891 Location: UK - Birmingham
|
Posted: Mon Mar 31, 2008 9:14 am Post subject: |
|
|
So let me get this straight
IF you go and use Paludis for your package manager you cannot (easily) go back to using portage? _________________
| Quote: | | Voting holds no real power, he who counts the votes has the true power. |
Weaver Projects
whats the difference between 9/11 and a cow?
u stop milking a cow after 10 years |
|
| Back to top |
|
 |
dleverton Guru

Joined: 28 Aug 2006 Posts: 517
|
Posted: Mon Mar 31, 2008 9:21 am Post subject: |
|
|
| Naib wrote: | yr tests are skewed and irrelvant
chroot for each 3 from a stage3 tarball or you are skewing! |
Nice to see you don't have any answers to my points, and continue to repeat what I've already told you misses the point entirely.
| Naib wrote: | | So you're saying that because we all get consistent results they must be wrong? Sorry, can't follow. | Not really. Just saying that your results aren't the same as other people's. Computers are funny like that.
| steveL wrote: | | That's the bit that worries me; so effectively once you've used paludis for a while, that's it: you can't go back to the official package-manager. | Not as far as I'm aware. "Incompatible" was probably an exaggeration.
EDIT: more specifically, perhaps he would have needed to update his portage configuration files, or remove any installed packages that aren't compatible with portage (BTW, paludis has an option to mask those).
| steveL wrote: | | And sorry, but --long-opts-just-because-you-can-are-a-PITA-for-usrs | Nope. Use bash/zsh completion or $PALUDIS_OPTIONS. And it's not "because we can", it's because there are only so many short options available, especially if we want to keep them understandable.
| steveL wrote: | | although paludis devs have a long history of pouring scorn on users | Nope.
| steveL wrote: | | usually by denigrating the QA his co-lead was supposed to be in charge of, though he never did anything | Perhaps you're missing the point of the QA team.
| steveL wrote: | | where I'm sitting it's clear you at least were not being malicious | He called me a ricer and accused me of using Sabayon, both without the slightest shred of evidence whatsoever. |
|
| Back to top |
|
 |
Naib Advocate


Joined: 21 May 2004 Posts: 3891 Location: UK - Birmingham
|
Posted: Mon Mar 31, 2008 9:29 am Post subject: |
|
|
| dleverton wrote: | | Naib wrote: | yr tests are skewed and irrelvant
chroot for each 3 from a stage3 tarball or you are skewing! |
Nice to see you don't have any answers to my points, and continue to repeat what I've already told you misses the point entirely.
|
I havn't answered yr points because you refuse to answer my intial points
IF you have issues with the results provides then do it yourself. Skewing results due to config mangling that is non-portage compatible does not count
WHEN you figure that out and post the results then we can discuss. UNtil then you are just wasting space in RAM/HD
3chroots each for paludis,pkgcore,portage to provide the benchmark. Unless you provide that you are just skewing results with paludis mangling and making forum noise to try to distract from the truth _________________
| Quote: | | Voting holds no real power, he who counts the votes has the true power. |
Weaver Projects
whats the difference between 9/11 and a cow?
u stop milking a cow after 10 years |
|
| Back to top |
|
 |
bonsaikitten Apprentice

Joined: 01 Jan 2003 Posts: 213 Location: Aachen, Germany
|
Posted: Mon Mar 31, 2008 9:33 am Post subject: |
|
|
| dleverton wrote: | | Naib wrote: | yr tests are skewed and irrelvant
chroot for each 3 from a stage3 tarball or you are skewing! |
Nice to see you don't have any answers to my points, and continue to repeat what I've already told you misses the point entirely.
| Naib wrote: | | So you're saying that because we all get consistent results they must be wrong? Sorry, can't follow. | Not really. Just saying that your results aren't the same as other people's. Computers are funny like that.
|
So, would you please try to reproduce the numbers that have been comfirmed by others? I wonder how your machine is more special and manages to do it differently.
Also, it is quite amusing to see you backpedalling, flinging poo and then complaining that others fling poo back. Very coherent  |
|
| Back to top |
|
 |
zixnub n00b


Joined: 27 Dec 2007 Posts: 63 Location: Brasschaat, Belgium
|
|
| Back to top |
|
 |
nox-Hand n00b


Joined: 14 Jan 2006 Posts: 5
|
Posted: Mon Mar 31, 2008 9:51 am Post subject: |
|
|
| dleverton wrote: |
| Quote: | | [*]"lack of documentation" - lie |
|
Thankyou for that quick assessment, but that is technically his opinion, for whilst you may like that cute and fluffy documentation there currently is, if he does not find it enough, it's lacking. Calling him a liar for that is about as stupid as installing yum on Gentoo..
| ShiRhoHaiSun wrote: |
[*]"And it spews so much output that you won't be able to see what just happened" - lie. |
Yeah, perhaps Naib doesn't appriciate getting his eyes splattered with output in a manner that makes him notice how low refreshrate his expensive LCD has? OPINION!
| ShiRhoHaiSun wrote: |
[*]"make it extremely hostile to new users" - lie[/list]
|
See above. Can't be arsed to comment further.
Excuse me for nitpicking, but...
| Ha1f wrote: | | tylerwylie wrote: | | Ha1f wrote: | | energyman76b wrote: | | poopludis |
I fuckin' lol'd. | Same, just trying to say it out loud. |
Oh God, I lol'd again  |
Most. Useless. Sideremarks. Ever.
*gets out some popcorn and a good couch*
This'll only get better! _________________ rm -rf / |
|
| Back to top |
|
 |
tylerwylie Guru


Joined: 19 Sep 2004 Posts: 455 Location: /US/Illinois/Champaign
|
Posted: Mon Mar 31, 2008 10:00 am Post subject: |
|
|
| dleverton wrote: | | energyman76b wrote: | | poopludis | Hey, look, I can do that too! "poopage" and "poopgcore"! See how clever and funny I am?!  | Fail. _________________ POLITICS, n. A strife of interests masquerading as a contest of principles. The conduct of public affairs for private advantage. |
|
| Back to top |
|
 |
steveL Veteran

Joined: 13 Sep 2006 Posts: 1441 Location: The Peanut Gallery
|
Posted: Mon Mar 31, 2008 10:02 am Post subject: |
|
|
| dleverton wrote: | | steveL wrote: | | That's the bit that worries me; so effectively once you've used paludis for a while, that's it: you can't go back to the official package-manager. | Not as far as I'm aware. "Incompatible" was probably an exaggeration.
EDIT: more specifically, perhaps he would have needed to update his portage configuration files, or remove any installed packages that aren't compatible with portage (BTW, paludis has an option to mask those). |
No offence, but I don't see you as disinterested, given what I've seen of your behaviour. I accept what you're saying on good faith, but this whole project and the dev behind it have been incredibly divisive, all with the excuse of "telling others they're wrong."
| Quote: | | steveL wrote: | | And sorry, but --long-opts-just-because-you-can-are-a-PITA-for-usrs | Nope. Use bash/zsh completion or $PALUDIS_OPTIONS. And it's not "because we can", it's because there are only so many short options available, especially if we want to keep them understandable. |
Whatever; I ain't using it so it doesn't bother me what crap it does.
<Sput> just tried zsh completion btw
<Sput> paludis --dl-reinst <tab>
<Sput> took 25 seconds of intensive hdd usage until that was completed
<Sput> zsh completion works fine otherwise
And no this isn't scientific benchmarking (btw your setup is much further away from that than Patrick's method.) So poke holes all you want.
| Quote: | | steveL wrote: | | although paludis devs have a long history of pouring scorn on users | Nope. |
You did the same sort of response to Patrick's blog, simply calling things "lies." If you don't know why ciaranm was kicked out, I suggest you research it; same attitude he gives to devs on ml now he loves handing out to users, and has done much worse in the past. IMO he's the reason it still seems to be considered "leet" to look down on users amongst some devs. It's just a shame seemingly-intelligent people can't see the value in teamwork.
| Quote: | | steveL wrote: | | usually by denigrating the QA his co-lead was supposed to be in charge of, though he never did anything | Perhaps you're missing the point of the QA team. |
Perhaps you're missing the almost continuous stream of snipes at QA by your leader; he used to alternate between that and attacks on portage, til it was pointed out to him that spb ran QA. Now we just get the snipes at portage along with the 10-email exchanges because he can't communicate without playing some weird guessing-game to prove his superior knowledge of the PMS he wrote. This is presented as "authoritative" for all Gentoo devs despite being no such thing, and written by someone none of them could stand to work with, who clearly likes playing political games instead of promoting an ethos of collaboration.
| Quote: | | steveL wrote: | | where I'm sitting it's clear you at least were not being malicious | He called me a ricer and accused me of using Sabayon, both without the slightest shred of evidence whatsoever. |
It was a joke ffs. Everyone else could see it, whereas you, who had already taken the conversation down to that level, assumed he was doing the same as you, when he was just trying to lighten the mood (admittedly poking fun, but so what? Get over it.) |
|
| Back to top |
|
 |
nox-Hand n00b


Joined: 14 Jan 2006 Posts: 5
|
Posted: Mon Mar 31, 2008 10:05 am Post subject: |
|
|
| tylerwylie wrote: | | dleverton wrote: | | energyman76b wrote: | | poopludis | Hey, look, I can do that too! "poopage" and "poopgcore"! See how clever and funny I am?!  | Fail. |
Not sure what is most fail. Writing the comment, reading the comment, or actually replying to the comment.
Wait, I know.. replying to the comment about the initial comment. FAIL >_<.
Now get back to slaughtering each other about package managers!
.... or you could spend all this time improving your beloved favourites instead - I am sure you could help them win the future debate by NOT tattling about here? _________________ rm -rf / |
|
| Back to top |
|
 |
tylerwylie Guru


Joined: 19 Sep 2004 Posts: 455 Location: /US/Illinois/Champaign
|
Posted: Mon Mar 31, 2008 10:05 am Post subject: |
|
|
YUM > * _________________ POLITICS, n. A strife of interests masquerading as a contest of principles. The conduct of public affairs for private advantage. |
|
| Back to top |
|
 |
bunder Bodhisattva


Joined: 10 Apr 2004 Posts: 5213
|
|
| Back to top |
|
 |
nixnut Administrator


Joined: 09 Apr 2004 Posts: 10951 Location: the dutch mountains
|
Posted: Mon Mar 31, 2008 5:01 pm Post subject: |
|
|
| ShiRhoHaiSun wrote: |
Haha, u need to test on gentoo system, not sabayon. U silly ricer. -funroll-loops -funroll-all-over-floor !! |
Banned for two weeks from OTW. Calling someone a silly ricer we may choose to overlook, but accusing someone of being a sabayon user is really going too far
I'm leaving this locked, since I see no new arguments in the last page and a half, only more pm bashing of which we've seen quite enough already. _________________ Please add [solved] to the initial post's subject line if you feel your problem is resolved. Help answer the unanswered
talk is cheap. supply exceeds demand |
|
| Back to top |
|
 |
|
|
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
|
|