digitalmars.D - Why I won't fix phobos bugs anymore
- Basile B. (8/8) Jun 02 2016 Yesterday I've took the decision not to propose anymore PR for
- Andrei Alexandrescu (9/10) Jun 02 2016 You mean @andralex? You are right. I am sorry, I'm coming off an
- Seb (9/21) Jun 02 2016 Can't we have someone that can dedicate a fixed amount of his
- Basile B. (4/27) Jun 02 2016 Eventually I'll come back to bugfix if they take Jake, but not
- Basile B. (3/15) Jun 02 2016 Yep, I mean "@andralex".I was clearly speaking about you.
- Andrei Alexandrescu (7/14) Jun 02 2016 Also meant to add: email should help. I am currently nursing 3-4 emails
- Patrick Schluter (4/25) Jun 02 2016 That would be Sisyphus, Midas was the king that starved because
- Pie? (9/35) Jun 02 2016 Kinda silly, right? He could have just easily paid someone to
- ixid (3/9) Jun 03 2016 I think a part of the point of the myth is that everything is
- w0rp (5/5) Jun 03 2016 I think a healthy option might be to find one or more
- Andrei Alexandrescu (4/6) Jun 03 2016 Yah I was thinking every email I touch generates another email. Better
- qznc (9/17) Jun 02 2016 Hm, I just got on PR [0] into Phobos in 8 days. The experience
- Andrei Alexandrescu (5/12) Jun 02 2016 There's always room for another PR. Let me make sure that gets crazily
Yesterday I've took the decision not to propose anymore PR for Phobos bugfixes, even if most of the time it's easy. 1) It can take up to 2 or 3 weeks until a "phobos bugfix" get merged. Even a straight one. 2) Once a pr gets the label " andrei". It basically means that "it's dead".
Jun 02 2016
On 06/02/2016 03:41 PM, Basile B. wrote:Once a pr gets the label " andrei". It basically means that "it's dead".You mean andralex? You are right. I am sorry, I'm coming off an unprecedently busy spring spent mostly evangelizing D at various conferences, or doing contract work that will pour money in the Foundation's coffers. This is not work I can delegate, but is poised to have great impact (more on that later). I thought leaving Facebook would free my time, but things have gotten really crazily busy. And look at me - I spend most of my time on the autodecoding thread. Andrei
Jun 02 2016
On Thursday, 2 June 2016 at 20:17:32 UTC, Andrei Alexandrescu wrote:On 06/02/2016 03:41 PM, Basile B. wrote:Can't we have someone that can dedicate a fixed amount of his professional time to maintain the D infrastructure? There is so much to do - reviewing and categorizing PRs is just the tip of the ice berg. Ideally it would be a full-time position, but if a company would dedicate 20% of an employees time to start be able to contribute to D, that would be an awesome step forward.Once a pr gets the label " andrei". It basically means that "it's dead".You mean andralex? You are right. I am sorry, I'm coming off an unprecedently busy spring spent mostly evangelizing D at various conferences, or doing contract work that will pour money in the Foundation's coffers. This is not work I can delegate, but is poised to have great impact (more on that later). I thought leaving Facebook would free my time, but things have gotten really crazily busy. And look at me - I spend most of my time on the autodecoding thread. Andrei
Jun 02 2016
On Thursday, 2 June 2016 at 20:23:37 UTC, Seb wrote:On Thursday, 2 June 2016 at 20:17:32 UTC, Andrei Alexandrescu wrote:Eventually I'll come back to bugfix if they take Jake, but not you Seb. For a reason or another I don't like you wilzbach.On 06/02/2016 03:41 PM, Basile B. wrote:Can't we have someone that can dedicate a fixed amount of his professional time to maintain the D infrastructure? There is so much to do - reviewing and categorizing PRs is just the tip of the ice berg. Ideally it would be a full-time position, but if a company would dedicate 20% of an employees time to start be able to contribute to D, that would be an awesome step forward.Once a pr gets the label " andrei". It basically means that "it's dead".You mean andralex? You are right. I am sorry, I'm coming off an unprecedently busy spring spent mostly evangelizing D at various conferences, or doing contract work that will pour money in the Foundation's coffers. This is not work I can delegate, but is poised to have great impact (more on that later). I thought leaving Facebook would free my time, but things have gotten really crazily busy. And look at me - I spend most of my time on the autodecoding thread. Andrei
Jun 02 2016
On Thursday, 2 June 2016 at 20:59:52 UTC, Basile B. wrote:Eventually I'll come back to bugfix if they take Jake, but not you Seb. For a reason or another I don't like you wilzbach.You are frustrated. I get that. Don't make this personal for others, please. Maybe you should ignore this thread for today?
Jun 02 2016
On Thursday, 2 June 2016 at 21:04:46 UTC, qznc wrote:On Thursday, 2 June 2016 at 20:59:52 UTC, Basile B. wrote:My POV is that it's easy to fix some phobos bug. But to get the easy fixes merged is a PITA. PRs that fixe a bug are hard to get merged. Why ? I don't know. Sometime we have to act like a bully to get a PR merged and that's not normal. Perso I give up.Eventually I'll come back to bugfix if they take Jake, but not you Seb. For a reason or another I don't like you wilzbach.You are frustrated. I get that. Don't make this personal for others, please. Maybe you should ignore this thread for today?
Jun 02 2016
On Friday, 3 June 2016 at 04:12:55 UTC, Basile B. wrote:On Thursday, 2 June 2016 at 21:04:46 UTC, qznc wrote:No need to get emotional over code. But I do agree that bug fixes should be fast tracked. How hard would it be to set up a priority queue? Say 1. bug fixes 2. performance enhancement 3. ...On Thursday, 2 June 2016 at 20:59:52 UTC, Basile B. wrote:My POV is that it's easy to fix some phobos bug. But to get the easy fixes merged is a PITA. PRs that fixe a bug are hard to get merged. Why ? I don't know. Sometime we have to act like a bully to get a PR merged and that's not normal. Perso I give up.Eventually I'll come back to bugfix if they take Jake, but not you Seb. For a reason or another I don't like you wilzbach.You are frustrated. I get that. Don't make this personal for others, please. Maybe you should ignore this thread for today?
Jun 03 2016
On Thursday, 2 June 2016 at 20:17:32 UTC, Andrei Alexandrescu wrote:On 06/02/2016 03:41 PM, Basile B. wrote:Yep, I mean " andralex".I was clearly speaking about you.Once a pr gets the label " andrei". It basically means that "it's dead".You mean andralex? You are right. I am sorry, I'm coming off an unprecedently busy spring spent mostly evangelizing D at various conferences, or doing contract work that will pour money in the Foundation's coffers. This is not work I can delegate, but is poised to have great impact (more on that later). I thought leaving Facebook would free my time, but things have gotten really crazily busy. And look at me - I spend most of my time on the autodecoding thread. Andrei
Jun 02 2016
On 06/02/2016 03:41 PM, Basile B. wrote:Yesterday I've took the decision not to propose anymore PR for Phobos bugfixes, even if most of the time it's easy. 1) It can take up to 2 or 3 weeks until a "phobos bugfix" get merged. Even a straight one. 2) Once a pr gets the label " andrei". It basically means that "it's dead".Also meant to add: email should help. I am currently nursing 3-4 emails that are phobos-related among some 50 other important and urgent emails. There's some Midas effect - every day I wake up thinking "I'll halve my inbox today" and by the evening I still have 50 emails. It's maddening. But please don't take offense. Do email and I'll get to your work. Thanks! -- Andrei
Jun 02 2016
On Thursday, 2 June 2016 at 20:20:58 UTC, Andrei Alexandrescu wrote:On 06/02/2016 03:41 PM, Basile B. wrote:That would be Sisyphus, Midas was the king that starved because everything he touches becomes gold.Yesterday I've took the decision not to propose anymore PR for Phobos bugfixes, even if most of the time it's easy. 1) It can take up to 2 or 3 weeks until a "phobos bugfix" get merged. Even a straight one. 2) Once a pr gets the label " andrei". It basically means that "it's dead".Also meant to add: email should help. I am currently nursing 3-4 emails that are phobos-related among some 50 other important and urgent emails. There's some Midas effect - every day I wake up thinking "I'll halve my inbox today" and by the evening I still have 50 emails. It's maddening. But please don't take offense. Do email and I'll get to your work. Thanks!
Jun 02 2016
On Friday, 3 June 2016 at 05:18:49 UTC, Patrick Schluter wrote:On Thursday, 2 June 2016 at 20:20:58 UTC, Andrei Alexandrescu wrote:Kinda silly, right? He could have just easily paid someone to feed him with all the gold he made, right? Also, Sisyphus must not have been too crafty! If he spend all that time digging out the hill then it would have been lower in gravity and he wouldn't have to carry it for eternity... just give it a nudge and it would roll down. Hen he could use all that dirt to build a barricade to keep the boulder from rolling away(unless it could magically go through the dirt).On 06/02/2016 03:41 PM, Basile B. wrote:That would be Sisyphus, Midas was the king that starved because everything he touches becomes gold.Yesterday I've took the decision not to propose anymore PR for Phobos bugfixes, even if most of the time it's easy. 1) It can take up to 2 or 3 weeks until a "phobos bugfix" get merged. Even a straight one. 2) Once a pr gets the label " andrei". It basically means that "it's dead".Also meant to add: email should help. I am currently nursing 3-4 emails that are phobos-related among some 50 other important and urgent emails. There's some Midas effect - every day I wake up thinking "I'll halve my inbox today" and by the evening I still have 50 emails. It's maddening. But please don't take offense. Do email and I'll get to your work. Thanks!
Jun 02 2016
On Friday, 3 June 2016 at 05:51:02 UTC, Pie? wrote:Also, Sisyphus must not have been too crafty! If he spend all that time digging out the hill then it would have been lower in gravity and he wouldn't have to carry it for eternity... just give it a nudge and it would roll down. Hen he could use all that dirt to build a barricade to keep the boulder from rolling away(unless it could magically go through the dirt).I think a part of the point of the myth is that everything is returned to its previous state over night, including earthworks.
Jun 03 2016
I think a healthy option might be to find one or more professionals Andrei can respect and trust to delegate the work of approving pull requests and such to. As always with open source contributions, it's largely up to someone stepping up to the game. Then maybe bug fixes can be approved at a greater speed.
Jun 03 2016
On 6/3/16 1:18 AM, Patrick Schluter wrote:That would be Sisyphus, Midas was the king that starved because everything he touches becomes gold.Yah I was thinking every email I touch generates another email. Better yet: cornucopia - no matter how much email I archive away from inbox, it stays at 50... -- Andrei
Jun 03 2016
On Thursday, 2 June 2016 at 19:41:30 UTC, Basile B. wrote:Yesterday I've took the decision not to propose anymore PR for Phobos bugfixes, even if most of the time it's easy. 1) It can take up to 2 or 3 weeks until a "phobos bugfix" get merged. Even a straight one. 2) Once a pr gets the label " andrei". It basically means that "it's dead".Hm, I just got on PR [0] into Phobos in 8 days. The experience was good. Andrei was heavily involved and responded timely. All delays were deserved and I clearly needed the reviews and comments. It even was a little too quickly, since it got merged before I could remove a somewhat outdated comment [1]. ;) [0] https://github.com/dlang/phobos/pull/4362 [1] https://github.com/dlang/phobos/pull/4362#discussion_r65607429
Jun 02 2016
On 06/02/2016 04:52 PM, qznc wrote:Hm, I just got on PR [0] into Phobos in 8 days. The experience was good. Andrei was heavily involved and responded timely. All delays were deserved and I clearly needed the reviews and comments.I just had some pent-up ideas from a while ago.It even was a little too quickly, since it got merged before I could remove a somewhat outdated comment [1]. ;) [0] https://github.com/dlang/phobos/pull/4362 [1] https://github.com/dlang/phobos/pull/4362#discussion_r65607429There's always room for another PR. Let me make sure that gets crazily delayed :o). Andrei
Jun 02 2016