We wants to revive the good, old stats.gentoo.org, and for this we've build a new gentoo-stats-project, the gentoo-buildtime-database (gbd). You'll find the unofficial projectpage here: http://www.gentoo-stats.org/index.php Reproducible: Always Steps to Reproduce:
No-one commit the uhinv ebuilds please... When uhinv is at a decently releasable state I'll add it to the tree. Right now it's rather, er, incomplete :)
Created attachment 37564 [details] The ebuild for the gbd-client itself
Created attachment 37565 [details] the ebuild for the uhinv-tool
I've had a look at the client code, and some things probably need improvement. If you want I can give a few tips. Is there a cvs repository or a tarball with the server sources, I'd like to look at that as well.
The server source is very clean. So only the client code needs some improvement.
Well, every help is welcome! Just because the client... it is/was my first public bash-script, so it cant be optimal. :)
Created attachment 37574 [details] New version of the gbd-client I've (hopefully) fixed the bug which causes negative GUs. Sorry, but I cant test this, because I have never stopped a emerge before it was finished. But it should be fixed now. :)
Created attachment 37622 [details] New version of the gbd-client This ebuild and the next one (uhinv-0.4.ebuild) are the newest versions of the gbd-client. Now with PCI-Support.
Created attachment 37623 [details] New version of the uhinv-tool
Created attachment 37652 [details] Latest version of gbd-client Now the bug which causes negative or very huge GUs should definitly be fixed!
*** Bug 74641 has been marked as a duplicate of this bug. ***
*** Bug 74639 has been marked as a duplicate of this bug. ***
Well, I see that Ciaran is still tracking this issue as duplicates get marked as such :) The question is: are you going to do anything about this? Will gbd-client be eventually added to the portage tree?
Maybe you could just give me one single cogent reason for not adding it, and another one for marking something new - maybe take the time and look at versions and my comment inside of the bug, i guess it wont kill you - as a dublicate. I did open another bug on purpose, you see? Another thing i noticed, is that not only this but several other bugs, never left the 'upstream' and died, because of some <whatever here> devs. I dont understand what the problem is when there are 3rd party tools related to Gentoo that dont even touch anything of the system and therefor can hardly damage it. You have no right to claim any monopoly.
Created attachment 46225 [details] 1.3.7 17.12.2004
Created attachment 46226 [details] uhinv 0.4 17.12.2004
Daniel, I'm sorry, but I don't think that blaming gentoo devs is a good idea. They can just get offended and start intentially ignoring your ebuilds. That's a road to nowhere. I understand gentoo developers, they have great responsibility and blindly adding everything to the portage tree can actually be harmful. For example, your gbd-client should be carefully reviewed to be sure that it is not a spyware and it does not collect more information than it is required (that's not that I don't trust you, gentoo developers just take all the responsibility when adding anything to portage as stated at http://www.gentoo.org/proj/en/devrel/handbook/handbook.xml?part=3&chap=1). I understand you too, you have made an excellent resource actually trying to help gentoo developers and everyone else. And ignoring your efforts, gentoo developers surely make you feel offended. I hope that current situation will change for better. Probably Ciaran is just too short on free time or he just does not think that having good statistics is that important and focuses on other tasks that seem to be more important. But I do think that it is very important and a lot of other people share this opinion. We just need to hear what Ciaran thinks and what is the status of this bug. Maybe he just needs more time or he needs more people to come here and "vote" for adding gbd-client to portage to be sure that it is really needed not only for you but for everyone else.
Please reopen when we get an upstream who doesn't go around making absurd accusations and creating conspiracy theories at every given opportunity. We have enough people like that working with Gentoo already, there's no room for more.
Serge: They know it is working and even tested it themselves iirc. I wouldnt say anything against it if Ciaran could give me just one single cogent reason. Also, there is not much to care about, since it - as said above - doesnt touch any configuration files in contrast to ufed and similar. But no dev shows real interest which lasts longer than half an hour. Ciaran: Even expected such a childish reaction, instead of just giving any reasons - which might help. It looks like "Gentoo" has no room for criticism also. It is a bad fact that there is one single person who's able to prevent all others from even noticing an ebuild - I'd like to know what the avg. dev would say about it. It is not, and has never been my goal to 'create a conspiracy theorie', I simply expressed what not only me thinks. It just looks like Gentoo doesnt like 3rd party tools that mention a word like 'Portage' anywhere.
In addition, a statement like: "The server source is very clean. So only the client code needs some improvement." (Pieter Van den Abeele) shouldnt prevent it from beeing added. "some improvement" is nothing critical, but just mentions parts of the code beeing inelegant - or maybe I totally misunderstand and it means "Throw it away and never look at it again. ugly." This has nothing to do with any conspiracy or anything.
We know the existing stats system sucks. We're going to replace it whenever someone comes up with a viable solution. I've already explained to you what viable solution means in this case. In particular, we need to be sure that we're going to have Gentoo developers actively supporting it. This means we either need to recruit one of you as developers or find an existing developer who has time. I'm not going to stick something in the tree which includes "Gentoo" or "Portage" in the name without that. There's already a huge mess with a load of app-portage tools that no longer work and no longer get fixes, I'm not adding to that.
I believe we can stick something in the tree as long as its clear that the project is currently not an official gentoo project and that we (gentoo) currently don't have resources to support the development of the project ourselves. (Besides somebody maintaining the ebuilds). Since none of the ebuilds have 'gentoo' or 'portage' in their name, I have no arguments against adding the ebuild in question to the tree.
The new name will be basc (Buildtime And Statistics Client), while the version basc-1.3.8, which is ready yet, will be released in sync to including it into the official portage tree.
Incvs
Please report bugs etc. to me.
