2014 10_Meeting_IRC_Log

Michal Čihař edited this page Apr 14, 2016 · 2 revisions
Clone this wiki locally
(12:03:20) Marc9: Start of 2014-10 meeting
(12:03:33) Marc9: Item 1: Committee membership voting
(12:03:35) madhuracj: Thanks smita and dstorm
(12:03:56) Marc9: Voting by alphabetic order :)
(12:04:25) Marc9: Which current committee member is in favor of accepting Chirayu?
(12:04:31) madhuracj: I vote in favor
(12:04:32) Marc9: I am in favor, and Michal too
(12:05:06) ibennetch-mobile: I also vote in favor for both.
(12:05:17) smita_: me too in favor :-)
(12:05:39) Marc9: Sorry Smita, you cannot vote yet
(12:06:20) Marc9: By the way, the quorum is 50% of current members, which means 4 members. As Michal already voted, the quorum is there
(12:06:42) Marc9: Chirayu is accepted, congratulations
(12:06:44) madhuracj: Great!
(12:06:50) madhuracj: Congratulations!
(12:06:59) ***dstorm thanks everyone
(12:06:59) Marc9: Now, who is in favor of accepting Smita?
(12:07:05) Marc9: I am in favor and Michal too
(12:07:08) ibennetch-mobile: Excellent. Thank you Marc9 for that information.
(12:07:35) Marc9: dstorm can vote now
(12:07:37) ***dstorm thinks he can vote now being member
(12:07:54) ibennetch-mobile: To be clear, I also vote in favor.
(12:08:12) dstorm: I also vote in favor
(12:08:14) madhuracj: In favor
(12:08:26) Marc9: Smita is accepted, congratulations
(12:08:40) ibennetch-mobile: Congratulations to you both! And welcome.
(12:08:41) Marc9: Item 2: Roadmap to version 4.3
(12:08:53) smita_: thank you team :-)
(12:09:06) Marc9: I see two show-stoppers for 4.3.0-alpha
(12:09:20) Marc9: First, the regression about importing SQL
(12:09:36) Marc9: Second, the "insanely slow" problem with thousands of databases
(12:10:06) madhuracj: I agree, specially the import problem
(12:10:08) dstorm: I can notice significant slowness in QA_4_2 and master even with 3-4 databases
(12:10:09) Marc9: The first one has already been discussed but I believe we should remove the mbstring functions in this part of the code
(12:10:27) Marc9: dstorm, talking about import or in general?
(12:10:32) dstorm: in general
(12:11:05) Marc9: I did not notice that; which MySQL version?
(12:11:36) dstorm: 5.5.38-0ubuntu0.12.04.1
(12:11:47) Marc9: and is it since Hugues has merged the PMA_String stuff all over?
(12:12:16) dstorm: I also encountered execution timeout at home page (after login)
(12:12:24) Marc9: But first, are there other problems blocking 4.3.0-alpha?
(12:12:30) dstorm: most probably after that PMA_String stuff
(12:13:00) madhuracj: I do not see anything major other than those two
(12:13:13) Marc9: dstorm, if you could check to confirm that, and add a comment to the relevant commit in github
(12:13:25) smita_: I think we have got many reports as well for slow issue .. after PMA_string
(12:13:27) Warped [~Warped@] a rejoint le salon.
(12:13:31) dstorm: Sure. I'll check that
(12:13:48) Marc9: smita: for importing or other things?
(12:14:04) ibennetch-mobile: I also am not aware of any other blockers, but both of those are rather important issues to resolve.
(12:14:06) smita_: btw I get execution timeout on home page many a time due to git version update..
(12:14:22) dstorm: those slow issues we are getting since 4.2 maybe
(12:14:44) dstorm: but still I can see difference in 4.2 and master branch
(12:14:49) Marc9: smita, be careful when you switch your git repo, better close and reopen your browser
(12:14:50) ibennetch-mobile: I also have been noticing some strange behavior relating to the git version check.
(12:14:56) smita_: no, in general.. not specifically import
(12:15:32) dstorm: yeah, in git version check. just got one now. It is pointing to StringMB class
(12:15:39) smita_: yeah marc, right. mostly i get when  change the repo. thanks, 'll keep in mind .
(12:15:49) Marc9: so this should be reported as a bug please
(12:16:21) smita_: Sure, I'll report.
(12:16:24) Marc9: dstorm, can you see the differences also on demo.phpmyadmin.net ?
(12:16:46) dstorm: IIRC, I saw something similar already reported. I'll see that's same or something else
(12:17:57) dstorm: not on demo.phpmyadmin.net
(12:18:23) Marc9: dstorm, so it might be specific to your environment; please post more details in a bug ticket
(12:18:25) dstorm: I think its because of caching on demo.phpmyadmin.net
(12:18:45) dstorm: Performance tends to improve after a while of caching but at start it is slow
(12:19:27) madhuracj: Anyway it's best to have a bug report to keep track of the issue and we probably can discuss more about the problem there
(12:19:33) Marc9: To recap: PMA_String has introduced a regression (import) and a possible slowdown in general (to be confirmed)
(12:19:47) Marc9: madhuracj: right
(12:20:17) Marc9: About the other problem (thousands of db): this is not regression but I believe we should make an effort for 4.3
(12:20:41) madhuracj: I believe it's there since version 4.0
(12:21:08) Marc9: By the way, I started to test a version from labs.mysql.com that includes a new dictionary, which is supposed to help information_schema performance
(12:21:12) dstorm: Marc9: Is there any particular reason why it is slow with huge amount of db's in newer version
(12:21:39) Marc9: dstorm: because we stopped using "SHOW DATABASES" in favor of polling information_schema
(12:22:10) dstorm: madhuracj: 4.0 is pretty fast, as per bug reports.
(12:22:24) Marc9: So we could, in 4.3.0, offer an option that reintroduces using "SHOW DATABASES", but reduces the feature set
(12:22:45) dstorm: Marc9: how will that affect the feature set?
(12:22:59) madhuracj: Marc9: I see. Yes, that is an option
(12:23:11) Marc9: even if the new version (for MySQL 5.7 ?) is better, we should help users who will stay on MySQL 5.5 for years
(12:23:25) Marc9: dstorm: for example, the db filter in the navi panel
(12:24:06) ZweiSteinSoft [590fed21@gateway/web/freenode/ip.] a rejoint le salon.
(12:24:16) dstorm: Hi Ann+J.M
(12:24:16) Marc9: dstorm, you can try "git grep information_schema" to get an idea; not all of these queries are translatable to SHOW DATABASES
(12:24:23) ZweiSteinSoft: Hi everybody
(12:24:27) Marc9: Hi JM
(12:24:30) ibennetch-mobile: I like this suggestion. Users will be stuck on slow MySQL installations for years.
(12:24:31) madhuracj: Hi JM
(12:24:38) ibennetch-mobile: Hello JM
(12:24:42) smita_: Hello JM
(12:24:52) ZweiSteinSoft: Hi team, we're here short because we're on mobile.
(12:25:05) dstorm: Marc9: yeah i'll try that
(12:25:13) ZweiSteinSoft: We vote accepting both. We should go for GSoC 2015.
(12:25:31) Marc9: So is everyone in favor of putting back an option to exclusively use SHOW DATABASES (off by default) ?
(12:25:41) ZweiSteinSoft: 4.3.0-alpha.1: It depends on Marc's tests with the shared hosting server...
(12:26:13) Marc9: JM: there are still not 150000 grants on this server and I'm still waiting for your answer about which query takes minuts
(12:26:16) Marc9: minutes
(12:26:37) ibennetch-mobile a quitté le salon (quit: Remote host closed the connection)
(12:26:40) dstorm: I am in favor of performance over some features
(12:26:49) Marc9: By the way, I believe that Madhura will need an access to this server
(12:27:10) madhuracj: Indeed, if I am to work on these issues
(12:27:22) J_Steen [~J_Steen@ip51cdd230.speed.planet.nl] a rejoint le salon.
(12:27:39) Marc9: madhura, we'll see this during item 4 :)
(12:27:42) smita_: yeah, I m in favor. better than having slow performance ..
(12:27:42) ZweiSteinSoft: We'll be back tomorrow evening on broadband, answering all the mails, this one too.
(12:27:46) ibennetch-mobile [~ibennetch@mobile-166-171-059-184.mycingular.net] a rejoint le salon.
(12:27:47) madhuracj: :)
(12:28:00) Marc9: Anything else to say for item 2?
(12:28:41) Marc9: Note that the slow performance for thousands of db is only for root, not for their users
(12:28:42) J_Steen a quitté le salon
(12:29:21) Marc9: Item 3: GSoC 2015
(12:29:33) Marc9: It has been officially announced
(12:29:46) Marc9: If we participate, what would be the goals?
(12:29:49) ZweiSteinSoft: :)
(12:30:09) madhuracj: I think we should still take part with the aim of attracting new developers
(12:30:11) dstorm: We should. We have lot of bug reports and rfe
(12:30:16) Marc9: (taking into consideration that we have a full-time developer)
(12:30:36) dstorm: +scrutnizer issues and test coverage too
(12:31:11) ibennetch-mobile: This is a good question. I was just thinking about this because madhuracj is fixing many of the things I would want to mentor :-)
(12:31:15) Marc9: Do we have a way to know which code is not covered?
(12:31:37) Marc9: ibenneth: you mean in the feature requests, right?
(12:31:43) dstorm: yeah, coverage reports provide good detail on that
(12:31:49) ZweiSteinSoft: Code Coverage report? in Scrutinizer perhaps?
(12:32:03) dstorm: coverage-html being very user friendly showing every line of code i.e. covered or not
(12:32:10) Marc9: ok thanks
(12:32:26) dstorm: phpunit with coverage-html
(12:33:16) ZweiSteinSoft: We will read the rest of this chat tomorrow evening... have a nice time (around the world—from India over Europe + the States to Canada…)
(12:33:45) ZweiSteinSoft: Until soon! —Ann + J.M.
(12:33:47) ibennetch-mobile: Yes I mean feature requests
(12:33:56) madhuracj a quitté le salon (quit: Read error: Connection reset by peer)
(12:34:15) ibennetch-mobile: Doofus ZweiSteinSoft
(12:34:16) ibennetch-mobile: Gah! Stupid autocorrect.
(12:34:21) Marc9: The problem is, we almost never got a project idea from a student, it came from the team
(12:34:33) ibennetch-mobile: I mean goodbye!!! Sorry about that.
(12:34:55) ZweiSteinSoft: ibennetch-mobile: :)
(12:35:04) Marc9: Bye JM
(12:35:19) dstorm: Marc9: for a project idea from student, he must be well known with phpMyAdmin as user.
(12:35:23) madhuracj [~kvirc@] a rejoint le salon.
(12:35:35) Marc9: dstorm, right
(12:35:37) ZweiSteinSoft a quitté le salon (quit: )
(12:36:12) smita_: we may even have responsive themes for UI .. as current themes are not responsive
(12:36:15) Marc9: So, we could have one project about improving unit tests coverage, and ???
(12:36:33) dstorm: refactoring, maybe
(12:36:52) Marc9: smita, I believe we discussed producing themes for a GSoC project and it was not acceptable
(12:37:40) ***dstorm wonders how we will accomodate so many things on mobile display
(12:37:56) smita_: hmm .. is there any specific reason?
(12:38:02) Marc9: Ok, so we'll open new GSoC 2015 pages and everyone who wants to mentor, can add his/her idea
(12:38:39) Marc9: Just a quick question: who is available/willing to be a mentor?
(12:38:50) ***Marc9 is
(12:38:56) ibennetch: I am also
(12:39:08) ***dstorm is not sure
(12:39:25) smita_: Since I'll still be student so I might be participating ..
(12:39:31) Marc9: Michal is probably available
(12:39:31) madhuracj: I am also
(12:39:48) Marc9: Smita: good point
(12:39:49) ***dstorm is not more a student :(
(12:40:35) Marc9: Anything else about GSoC 2015?
(12:40:49) smita_: BTW how about having some app, android or iOS ? if possible
(12:41:46) Marc9: Smita, this was discussed but we really did not know if it was appropriate
(12:41:55) smita_: may be for tablets n iPad for managing few stuffs with limited features..
(12:41:56) dstorm: maybe we need RESTful APIs for it.
(12:42:16) smita_: yeah .. we'll certainly need API's
(12:42:25) Marc9: It's also a question of maintenance for the future code
(12:42:38) madhuracj: We might have problems maintaining android and iOS apps
(12:42:45) madhuracj: exactly :)
(12:42:45) ibennetch: Tablets and phones were discussed some time ago, I believe as part of an IRC meeting. We decided we didn't have the resources or user interest to make it worthwile at that time.
(12:43:12) Marc9: Isaac: I remember this, but I don't remember which meeting
(12:44:01) smita_: marc9: agree, we'll need active android/iOS developers for maintainance
(12:44:11) Marc9: Item 4: Suggestions for contract developer
(12:44:21) Marc9: Madhura, we are listening...
(12:44:41) madhuracj: Your ideas are most welcome in order to prioratize work
(12:45:22) Marc9: I would like to release 4.3.0-alpha ASAP so the two subjects above would be my priority, so
(12:45:28) madhuracj: During the first week I attended mostly to bugs, security issues and some refactoring
(12:45:39) ibennetch: About the tablet version, I think this was it: https://wiki.phpmyadmin.net/pma/2014-01_Meeting_IRC_Log
(12:45:52) dstorm: I would say (in order), Bug fixing should be given priority then test code
(12:45:54) Marc9: if you can work with Hugues about the PMA_String thing (evaluate the impact, fix the import bug)
(12:46:22) Marc9: and then add the option for deactivating information_schema, this would be fantastic
(12:46:46) dstorm: Marc9: that would be really great
(12:46:54) Marc9: After, there could be some bug fixing to prepare 4.2.10 in two weeks
(12:46:57) madhuracj: Marc9: Sure I will work on them
(12:47:21) Marc9: thanks Isaac
(12:48:56) ***Marc9 has a dream: an empty bug tracker
(12:49:06) madhuracj: :)
(12:49:25) ***dstorm too (zero defect = quality)
(12:49:39) Marc9: Of course, some old tickets can be closed because they are out of date or we got no answer
(12:50:02) Marc9: dstorm: we noticed that since the start of refactoring effort, quality has improved
(12:50:05) madhuracj: I started attending to those age old ones
(12:50:05) ibennetch-mobile a quitté le salon (quit: Quit: Colloquy for iPhone - http://colloquy.mobi)
(12:50:08) ibennetch: I've been very pleased with the progress in this first week and think the others have good suggestions.
(12:50:32) ibennetch: I suggest prioritizing 4.3.0-alpha, of course, then going through to eliminate some outdated bugs.
(12:50:36) ibennetch: *bug reports*
(12:50:54) ibennetch: Which I've noticed you doing anyway.
(12:51:02) Marc9: You see, once 4.3.0-alpha (or beta) are released, we will get more bug reports about it
(12:51:33) ibennetch: I also think there are some improvements to be made on the error reporting server, though I'm not clear on what the issues are with it.
(12:51:50) madhuracj: Marc9: Indeed, the bug rush will begin
(12:51:51) Marc9: also, I would like Madhura to add new features for 4.4, not for 4.3. This means a shorter release cycle for 4.3.0.
(12:51:55) smita_: what about giving some time on testing the release version before the release ofcourse ?
(12:52:04) smita_: that we can even do ..
(12:52:19) Marc9: Smita: we always have alpha, beta and release candidates
(12:52:56) Marc9: and I will put -alpha in production at my work ;) once the two issues are fixed
(12:53:20) dstorm: Marc9: brave :)
(12:53:22) madhuracj: So I will not take and rfes until we have QA_4_3
(12:53:46) Marc9: madhura, yes I think it would be better this way
(12:53:54) madhuracj: any*
(12:53:56) ibennetch: If I may add on here, I think we could overhaul the priority tag of the trackers.
(12:54:32) ibennetch: That might help madhuracj prioritize his work, also - but right now I feel like we're not clear on what the difference between the priorites are.
(12:55:00) Marc9: Well, we only use 5 and 1
(12:55:32) Marc9: What do you suggest?
(12:56:20) glebihan_ a quitté le salon (quit: Remote host closed the connection)
(12:56:20) nisargjhaveri [~nisargjha@] a rejoint le salon.
(12:56:26) Marc9: (meeting ends in 4 minutes)
(12:56:43) ibennetch: We can use the status tag to mark when something's been fixed or released and change the priority tag to actually show the priority (1 would be release-blocking, 5 is wishlist). Something like that.
(12:57:13) ibennetch: That's the gist of my suggestion; I can take it to the mailing list so the meeting can end on time.
(12:57:17) Marc9: Ok, we can discuss this on phpmyadmin-devel list
(12:57:33) Marc9: If there is nothing else, I smell the end of meeting...
(12:57:49) ibennetch: I have nothing. Thank you all, this has been quite productive
(12:57:59) Marc9: Thanks to all
(12:58:14) Marc9: END OF MEETING