April 2018 - Pharo-users - lists.pharo.org

Re: [Pharo-users] SortedCollection>>reverse answers an inconsistent object in Pharo 6

by Denis Kudriashov

Hi Richard. I agree with your proposal. But it force me to think that we should completely move to SortFunction's. In that case SortCollection will have sortFunction variable instead of sortBlock. And for your scenario reverse operation will be simple expression: "sortFunction := sortFunction reversed". 2018-04-23 3:09 GMT+02:00 Richard O'Keefe <raoknz(a)gmail.com>: > Test case: > #(1 2 4) asSortedCollection reverse add: 3; yourself > <print it> > The answer *should* be aSortedCollection(4 3 2 1) > but *is* aSortedCollection(4 2 1 3). > This works in Squeak. > The problem is that SortedCollection does not define > #reverse[d] but simply inherits it(them), and the > inherited code pays no attention to the sortBlock. > > I propose adding the following two methods to > SortedCollection: > > reverseInPlace > |a z| > a := firstIndex. > z := lastIndex. > [a < z] whileTrue: [array swap: a with: z. a := a + 1. z := z - 1]. > sortBlock := sortBlock > ifNil: [[:x :y | y <= x]] > ifNotNil: [[:x :y | sortBlock value: y value: x]]. > ^self > > reversed > ^self copy reverseInPlace > > The ANSI method is #reverse, not #reversed, but Pharo > defines #reverse to call #reversed, and OrderedCollection overrides > #reversed, so SortedCollection *must* override #reversed. > > #reverseInPlace is the name Squeak uses for the other > method. It also has a definition in SequenceableCollection, which is not > but is > equivalent to > > reverseInPlace > |a z| > a := 1. > z := self size. > [a < z] whileTrue: [self swap: a with: z. a := a + 1. z := z - 1].l > ^self > > r > >

April 25, 2018

Re: [Pharo-users] SortedCollection>>reverse answers an inconsistent object in Pharo 6

by Erik Stel

Why does the sortBlock has to answer true for the same (identical) object? I have not been able to find where it is specified. And in different locations I see #> being used as a comparator for the sortBlock. So why is it bad (as you wrote)? (It seems you describe that SortedCollection should create a collection with a Total Order, whereas I describe a situation where a Strict Order is also valid.) -- Sent from: http://forum.world.st/Pharo-Smalltalk-Users-f1310670.html

April 25, 2018

Re: [Pharo-users] Launcher not launching

by Jan BlizniÄenko

Hello Thank you for quick reply. Yes, it seems like the same bug and the linked current alpha version works fine. Switched to it and did not encounter even any other problems so far. Thank you. Regards, Jan demarey wrote > Hello, > >> Le 25 avr. 2018 Ã 11:57, Jan BlizniÄenko < > bliznjan@.cvut > > a Ã©crit : >> >> Hello >> >> On Windows 10, I updated launcher from the one on >> https://ci.inria.fr/pharo/view/Launcher/job/Package-and-Publish-Launcher-st… >> (1.0.1) to the one now on official website http://pharo.org/web/download >> (1.1), but the new launcher does not open any images. >> I can create, rename, remove images or do whatever with them, but they do >> no >> launch ... their VM gets downloaded, launcher closes (if it is set to >> close >> on launch - it does not close if not set so), then nothing - the image >> does >> not open, no new Pharo.exe process in list of processes in Windows. >> Tried disabling AV, (re)installing to different directory, selecting >> different directories for images and vms, trying various Pharo versions >> (from 3 to 7) of images, still nothing. It does not work with both old >> and >> new images (downloaded by old working launcher or by this new one). >> There are no exceptions, no crash dumps, nothing in stderr. >> >> Those images can be opened by double clicking on them or by dragging and >> dropping them on C:\Users\User\Documents\Pharo\vms\61-x86\Pharo.exe (or >> similar - the same executable as Launcher should use). Also if I install >> back the old launcher (1.0.1) again, it works. >> >> The new launcher does not work on both of my machines (both Windows 10 >> that >> are quite similar in config and installed applications). >> >> For now, I will stick with 1.0.1, but I would like to find a reason for >> this >> problem. > > I think the problem is > https://github.com/pharo-project/pharo-launcher/issues/68 > Could you try the latest version for windows? > http://files.pharo.org/pharo-launcher/win-alpha/ > Iâm waiting to finish app signing on windows before publishing it. > > Thanks for the report, > Christophe -- Sent from: http://forum.world.st/Pharo-Smalltalk-Users-f1310670.html

April 25, 2018

Re: [Pharo-users] Launcher not launching

by Christophe Demarey

Hello, > Le 25 avr. 2018 Ã 11:57, Jan BlizniÄenko <bliznjan(a)fit.cvut.cz> a Ã©crit : > > Hello > > On Windows 10, I updated launcher from the one on > https://ci.inria.fr/pharo/view/Launcher/job/Package-and-Publish-Launcher-st… > (1.0.1) to the one now on official website http://pharo.org/web/download > (1.1), but the new launcher does not open any images. > I can create, rename, remove images or do whatever with them, but they do no > launch ... their VM gets downloaded, launcher closes (if it is set to close > on launch - it does not close if not set so), then nothing - the image does > not open, no new Pharo.exe process in list of processes in Windows. > Tried disabling AV, (re)installing to different directory, selecting > different directories for images and vms, trying various Pharo versions > (from 3 to 7) of images, still nothing. It does not work with both old and > new images (downloaded by old working launcher or by this new one). > There are no exceptions, no crash dumps, nothing in stderr. > > Those images can be opened by double clicking on them or by dragging and > dropping them on C:\Users\User\Documents\Pharo\vms\61-x86\Pharo.exe (or > similar - the same executable as Launcher should use). Also if I install > back the old launcher (1.0.1) again, it works. > > The new launcher does not work on both of my machines (both Windows 10 that > are quite similar in config and installed applications). > > For now, I will stick with 1.0.1, but I would like to find a reason for this > problem. I think the problem is https://github.com/pharo-project/pharo-launcher/issues/68 Could you try the latest version for windows? http://files.pharo.org/pharo-launcher/win-alpha/ Iâm waiting to finish app signing on windows before publishing it. Thanks for the report, Christophe

April 25, 2018

Launcher not launching

by Jan BlizniÄenko

Hello On Windows 10, I updated launcher from the one on https://ci.inria.fr/pharo/view/Launcher/job/Package-and-Publish-Launcher-st… (1.0.1) to the one now on official website http://pharo.org/web/download (1.1), but the new launcher does not open any images. I can create, rename, remove images or do whatever with them, but they do no launch ... their VM gets downloaded, launcher closes (if it is set to close on launch - it does not close if not set so), then nothing - the image does not open, no new Pharo.exe process in list of processes in Windows. Tried disabling AV, (re)installing to different directory, selecting different directories for images and vms, trying various Pharo versions (from 3 to 7) of images, still nothing. It does not work with both old and new images (downloaded by old working launcher or by this new one). There are no exceptions, no crash dumps, nothing in stderr. Those images can be opened by double clicking on them or by dragging and dropping them on C:\Users\User\Documents\Pharo\vms\61-x86\Pharo.exe (or similar - the same executable as Launcher should use). Also if I install back the old launcher (1.0.1) again, it works. The new launcher does not work on both of my machines (both Windows 10 that are quite similar in config and installed applications). For now, I will stick with 1.0.1, but I would like to find a reason for this problem. Thank you Jan -- Sent from: http://forum.world.st/Pharo-Smalltalk-Users-f1310670.html

April 25, 2018

[Ann] Iceberg 0.7.3

by Guillermo Polito

Hi all, We just introduced Iceberg 0.7.3 into Pharo. This new version solves several issues people had when contributing to Pharo in general and others. I've made a screencast on how to contribute to Pharo, and added some tutorial and glossary in Iceberg's wiki. Below the links and the changes log. Thanks for making your issue reports ^^. We will schedule a backporting to Pharo6.1 in a couple of weeks, once we have another iteration. Cheers, Guille # Contribute to Pharo with Iceberg 0.7.3 - wiki page: https://github.com/pharo-vcs/iceberg/wiki/Contribute-to-Pharo-with-Iceberg-… - Screencast: https://www.youtube.com/watch?v=PK2yCu2rWCc&feature=youtu.because This screencast shows how to use Iceberg 0.7.3 to contribute to latest Pharo 7, by cloning from scratch your own (potentially outdated) clone. This video will guide you through: - solve the "Missing local repository" by cloning - solve the "Fetch required" that appears because the fork is out of date - solve the "Detached working copy" by creating a new branch - use the pharo plugin to create an issue branch - make some changes, commit them and push them to your fork - make a pull request If you already have a clone, you can just import an existing repository from your disk. The rest of the instructions should be the same. If you don't have a fork, create one before doing anything else. New videos to come with more explanations. # Iceberg Glossary https://github.com/pharo-vcs/iceberg/wiki/Iceberg-glossary # 0.7.3 ChangesLog - #693 <https://github.com/pharo-vcs/iceberg/pull/693>: fixing problem when the announcement of changes in a package is nil. (bug, important) - #682 <https://github.com/pharo-vcs/iceberg/issues/682>: update update script (Pharo6-backport) - #681 <https://github.com/pharo-vcs/iceberg/issues/681>: Author name is not correctly decoded (UI, enhancement) - #680 <https://github.com/pharo-vcs/iceberg/issues/680>: Repair actions dialog should explain the situation (UI, enhancement, important) - #679 <https://github.com/pharo-vcs/iceberg/issues/679>: Create branch repair action should show create branch from issue if available (Pharo plugin, UI) - #678 <https://github.com/pharo-vcs/iceberg/issues/678>: Iceberg does not correctly recognize packages (bug, important) - #677 <https://github.com/pharo-vcs/iceberg/issues/677>: Unknown commit can only be solved with repair action (Model, bug, important) - #676 <https://github.com/pharo-vcs/iceberg/issues/676>: Unknown commit state is not properly shown (UI, bug, important) - #667 <https://github.com/pharo-vcs/iceberg/issues/667>: Merging branches with file deletion doesn't seem possible (bug, important, merge) - #666 <https://github.com/pharo-vcs/iceberg/issues/666>: Iceberg sometimes add a $ to commit message (Model, bug) - #662 <https://github.com/pharo-vcs/iceberg/issues/662>: The Pharo plugin has to do the fetch only if required. (Pharo plugin, enhancement) - #660 <https://github.com/pharo-vcs/iceberg/pull/660>: Add a star badge in the push button in case of a new branch () - #658 <https://github.com/pharo-vcs/iceberg/issues/658>: Progress Bar during checkout of Repository. (UI, enhancement) - #657 <https://github.com/pharo-vcs/iceberg/issues/657>: Confusing UI: Ghost Test maybe should be default? (UI, enhancement) - #656 <https://github.com/pharo-vcs/iceberg/issues/656>: Typo in Progress Bar (UI, bug) - #651 <https://github.com/pharo-vcs/iceberg/issues/651>: When pushing a new branch the list of commit is empty (UI, bug) - #650 <https://github.com/pharo-vcs/iceberg/issues/650>: "Create new branch from issue" is leaving image in detached working copy state () - #649 <https://github.com/pharo-vcs/iceberg/issues/649>: The VM is crashing when trying to use Iceberg from a saved image (bug) - #645 <https://github.com/pharo-vcs/iceberg/issues/645>: IceLibgitRepository should use "utilities" protocol instead of "utils" (clean up) - #644 <https://github.com/pharo-vcs/iceberg/issues/644>: Unnecessary temp in IceTipGitHubRepositoryPanel>>#getGitHubRepository" (clean up) - #643 <https://github.com/pharo-vcs/iceberg/issues/643>: LGit_GIT_ERROR: error authenticating: no auth sock variable loading a Tonel based baseline dependency in smalltalkCI (bug, important) - #639 <https://github.com/pharo-vcs/iceberg/issues/639>: Create pull request fails if branch has not been pushed (Github plugin, bug) - #636 <https://github.com/pharo-vcs/iceberg/issues/636>: The pull button should have a star badge when it has no upstream (UI, enhancement, important) - #634 <https://github.com/pharo-vcs/iceberg/issues/634>: Travis is not building Pull Requests (CI, bug, important) - #624 <https://github.com/pharo-vcs/iceberg/issues/624>: add "add remote" to remote picking on pull/push browsers (enhancement) - #610 <https://github.com/pharo-vcs/iceberg/pull/610>: Fix for issue #608 <https://github.com/pharo-vcs/iceberg/issues/608> and #607 <https://github.com/pharo-vcs/iceberg/issues/607>. (clean up) - #609 <https://github.com/pharo-vcs/iceberg/issues/609>: New branch repair action does not refresh dirty packages (UI, bug) - #583 <https://github.com/pharo-vcs/iceberg/issues/583>: Creating a tonel project does not commit the properties file (Model, bug) - #542 <https://github.com/pharo-vcs/iceberg/issues/542>: Bug during commit of a new package (Model, bug) -- Guille Polito Research Engineer Centre de Recherche en Informatique, Signal et Automatique de Lille CRIStAL - UMR 9189 French National Center for Scientific Research - *http://www.cnrs.fr <http://www.cnrs.fr>* *Web:* *http://guillep.github.io* <http://guillep.github.io> *Phone: *+33 06 52 70 66 13

April 25, 2018

Re: [Pharo-users] Right repo for TaskIt and features?

by Holger Freyther

> On 24. Apr 2018, at 23:31, Santiago Bragagnolo <santiagobragagnolo(a)gmail.com> wrote: > > > Yes. But with more work than the workers can handle the queue will grow. Which means the (median/max) latency of the system will monotonically increase.. to the point of the entire system failing (tasks handled after the external deadlines expired, effectively no work being done). > > > Normally the worker pool adjust to the minimal needed workers (there is a watch dog checking how much idle processes are there, or more workers are needed, and ensuring to spawn or stop process regarding to the state). > So, the number poolMaxSize is just a maximal limit. This limit should be set for ensuring that the tasks that are running concurrently are not incurring into too much resource consumption or into too much overhead leading to kind of trashing. > I am not really friend of setting only a number for such a complex problematic, but so far is the only approach I found that it does not lead to a complex design. If you have better ideas to discuss on this subject, i am completely open. (the same to deal with priorities by general system understanding rather than absolute numbers) I think we might not talk about the same thing. Any system might end up being driven close or above its limits. One question is if it can recover from it. Let me try to give you a basic example (and if one changes from 'dev' to a proper work pool one just needs to adjust timings to show the same problem). The code schedules a block that on each invocation takes about one second to execute. But the completion time is monotonically increasing. | completions | completions := OrderedCollection new. 1 to: 1000 do: [:each | | start | start := DateAndTime now. [ (Delay forSeconds: 1) wait. completions add: (DateAndTime now - start) ] schedule. (Delay forMilliseconds: 200) wait. ]. completions Now why is this a problem? It is a problem because once the system is in overload it will never recover (unless tasks are being stopped). The question is what can be done from a framework point to gracefully degrade? I am leaving this here right now. holger

April 25, 2018

Re: [Pharo-users] Right repo for TaskIt and features?

by Holger Freyther

> On 25. Apr 2018, at 08:42, Andrew Glynn <aglynn42(a)gmail.com> wrote: > > Generally to avoid this I've used the Synapse micro service bus. It also allows the creation of an unlimited number of queues, allowing higher priority tasks to "jump the queue". ' Backpressure' is precisely what message buses avoid in distributed computing. Can you elaborate and point to which Synapse you are meaning? If you use transport protocols like TCP (in contrast to QUIC or SCTP) there will be head-of-line blocking, how do you jump the queue on a single TCP connection?

April 25, 2018

Re: [Pharo-users] Right repo for TaskIt and features?

by Andrew Glynn

Btw I think you meant "thrashing", not "trashing'. Â Trashing is what my team leads do when they read my code. Â ð. AndrewOn Tue, 2018-04-24 at 15:31 +0000, Santiago Bragagnolo wrote: > > > On Tue, 24 Apr 2018 at 16:18 Holger Freyther <holger(a)freyther.de> > wrote: > > > > > > > On 24. Apr 2018, at 20:16, Santiago Bragagnolo > > o(a)gmail.com> wrote: > > >Â > > > Hi Holger!Â > > > I respond in bold > > > > hehe. And in the reply I am back to non rich text. Let me see if I > > quote it correctly. > hahahaha, non rich?, how come?Â I will still bolding hoping that if > you need to ensure the content you will check in a rich text client > :DÂ > > > > > > > >Â > > >Â > > >Â > > > On Tue, 24 Apr 2018 at 12:00 Holger Freyther <holger(a)freyther.de> > > wrote: > > > Hey! > > >Â > > > > > > > I wondered if somebody thought of remote task execution? > > >Â > > > *If you mean something else, I would need more information :). > > > > > > > > > > > When you do [ action ] schedule / [ action ] future, both created > > tasks are scheduled into the default runner. The default runner is > > a working pool with a default 'poolSizeMax' on 4, meaning, limit 4 > > processes working over the tasks. (this is a dynamic configuration, > > you can change it byÂ > > > TKTConfiguration runner poolMaxSize: 20. ) > > > > Yes. But with more work than the workers can handle the queue will > > grow. Which means the (median/max) latency of the system will > > monotonically increase.. to the point of the entire system failing > > (tasks handled after the external deadlines expired, effectively no > > work being done). > > Normally the worker pool adjust to the minimal needed workers (there > is a watch dog checking how much idle processes are there, or more > workers are needed, and ensuring to spawn or stop process regarding > to the state).Â > So, the number poolMaxSize is just a maximal limit. This limit should > be set for ensuring that the tasks that are running concurrently are > not incurring into too much resource consumption or into too much > overhead leading to kind of trashing.Â > I am not really friend of setting only a number for such a complex > problematic, but so far is the only approach I found that it does not > lead to a complex design. If you have better ideas to discuss on this > subject, i am completely open. (the same to deal with priorities by > general system understanding rather than absolute numbers)Â > > Â > > > > For network connected systems I like to think in terms of "back > > pressure" (not read more from the socket than the image can handle, > > eventually leading to the TCP window shrinking) and one way of > > doing it is to have bounded queues (and/or sleep when scheduling > > work). > > > > I can see multiple parts of a solution (and they have different > > benefits and issues): > > > > * Be able to attach a deadline to a task (e.g. see context.Context > > in go) > > * Be able to have a "blocking until queue is less than X elements" > > scheduleÂ (but that is difficult as one task might be scheduled > > during the >>#value of another task). > > > I wouldn't mind to have a second type of queue with this behaviour, > with a mean of configuration for setting one or other queue with it's > specific management encapsulated. > > Personally, in my domains of usage ( crawling, querying and in > sensor/actuator) i personally wouldn't use it. But I suppose you have > a better domain for this case. It would be good to discuss it to have > a better understanding of the need.Â > Â > > > > > > > Are there ideas how to add remote task scheduling? Maybe use > > Seamless for it? > > > Since you speak about seamless here, i suppose two different > > images, doesn't matter where.Â > > > It's not a bad idea to go seamless, but i did not go through the > > first restriction of remote executions (if the remote image can or > > not execute the task and if both images share the same semantic for > > the execution), then i did not yet checked on which communication > > platform to use for itÂ > > > > Right it would need to be homogeneous images (and care taken that > > the external interface remains similar enough). > I would like to understand a bit better what you are trying to do.Â > I have the hunch that you are looking for a multiple images solution, > for load balance in between images. TaskIT is mean to plan tasks into > process, regarding to the local image needs. You seems to need > something for planning tasks of a general system, beyond one image, > and maybe taking in care a process / network topology. > > If it's more that side, we should discuss in what extensions we can > do into taskit to be suitable of usage in this case, but surely I > would be inclined to do a higher level framework or even middleware > that uses taskit, than to add all those complexities in taskit. The > good news is that i may be needing something similar, so I will be > able to help there.Â > > > > > > > > > Have workers connect to the scheduler? Other ideas? > > > what do you mean by connection to the scheduler? The workers we > > use do not know their pools, if that is what you mean.Â > > > > Let's assume scheduling a task is simple (read something from a > > socket) but the computation is expensive (database look-up, math, > > etc). Hopefully one will reach the point where one image can > > schedule more tasks than a single worker image can handle. At that > > point it could be neat to scale by just starting another image. By > > inverting the launch order (workers connect to the scheduler) > > scaling can become more easy. > hahaha, now i re-read this paragraph and is not a hunch anymore. You > are looking for multiple images :).Â > > > Â > > holger > > > > > > > >

April 25, 2018

Re: [Pharo-users] Right repo for TaskIt and features?

by Andrew Glynn

Generally to avoid this I've used the Synapse micro service bus. Â It also allows the creation of an unlimited number of queues, allowing higher priority tasks to "jump the queue". Â ' Backpressure' is precisely what message buses avoid in distributed computing. One of my never-have-time-for projects is to port Synapse to Pharo. SST has a 'start a slave on another node and route to it' methodology but it's hella complex, especially in terms of distributed garbage collection etc. Â For real-time systems SST is great, not really necessary to get into that kind of complexity otherwise though. AndrewOn Tue, 2018-04-24 at 15:31 +0000, Santiago Bragagnolo wrote: > > > On Tue, 24 Apr 2018 at 16:18 Holger Freyther <holger(a)freyther.de> > wrote: > > > > > > > On 24. Apr 2018, at 20:16, Santiago Bragagnolo > > o(a)gmail.com> wrote: > > >Â > > > Hi Holger!Â > > > I respond in bold > > > > hehe. And in the reply I am back to non rich text. Let me see if I > > quote it correctly. > hahahaha, non rich?, how come?Â I will still bolding hoping that if > you need to ensure the content you will check in a rich text client > :DÂ > > > > > > > >Â > > >Â > > >Â > > > On Tue, 24 Apr 2018 at 12:00 Holger Freyther <holger(a)freyther.de> > > wrote: > > > Hey! > > >Â > > > > > > > I wondered if somebody thought of remote task execution? > > >Â > > > *If you mean something else, I would need more information :). > > > > > > > > > > > When you do [ action ] schedule / [ action ] future, both created > > tasks are scheduled into the default runner. The default runner is > > a working pool with a default 'poolSizeMax' on 4, meaning, limit 4 > > processes working over the tasks. (this is a dynamic configuration, > > you can change it byÂ > > > TKTConfiguration runner poolMaxSize: 20. ) > > > > Yes. But with more work than the workers can handle the queue will > > grow. Which means the (median/max) latency of the system will > > monotonically increase.. to the point of the entire system failing > > (tasks handled after the external deadlines expired, effectively no > > work being done). > > Normally the worker pool adjust to the minimal needed workers (there > is a watch dog checking how much idle processes are there, or more > workers are needed, and ensuring to spawn or stop process regarding > to the state).Â > So, the number poolMaxSize is just a maximal limit. This limit should > be set for ensuring that the tasks that are running concurrently are > not incurring into too much resource consumption or into too much > overhead leading to kind of trashing.Â > I am not really friend of setting only a number for such a complex > problematic, but so far is the only approach I found that it does not > lead to a complex design. If you have better ideas to discuss on this > subject, i am completely open. (the same to deal with priorities by > general system understanding rather than absolute numbers)Â > > Â > > > > For network connected systems I like to think in terms of "back > > pressure" (not read more from the socket than the image can handle, > > eventually leading to the TCP window shrinking) and one way of > > doing it is to have bounded queues (and/or sleep when scheduling > > work). > > > > I can see multiple parts of a solution (and they have different > > benefits and issues): > > > > * Be able to attach a deadline to a task (e.g. see context.Context > > in go) > > * Be able to have a "blocking until queue is less than X elements" > > scheduleÂ (but that is difficult as one task might be scheduled > > during the >>#value of another task). > > > I wouldn't mind to have a second type of queue with this behaviour, > with a mean of configuration for setting one or other queue with it's > specific management encapsulated. > > Personally, in my domains of usage ( crawling, querying and in > sensor/actuator) i personally wouldn't use it. But I suppose you have > a better domain for this case. It would be good to discuss it to have > a better understanding of the need.Â > Â > > > > > > > Are there ideas how to add remote task scheduling? Maybe use > > Seamless for it? > > > Since you speak about seamless here, i suppose two different > > images, doesn't matter where.Â > > > It's not a bad idea to go seamless, but i did not go through the > > first restriction of remote executions (if the remote image can or > > not execute the task and if both images share the same semantic for > > the execution), then i did not yet checked on which communication > > platform to use for itÂ > > > > Right it would need to be homogeneous images (and care taken that > > the external interface remains similar enough). > I would like to understand a bit better what you are trying to do.Â > I have the hunch that you are looking for a multiple images solution, > for load balance in between images. TaskIT is mean to plan tasks into > process, regarding to the local image needs. You seems to need > something for planning tasks of a general system, beyond one image, > and maybe taking in care a process / network topology. > > If it's more that side, we should discuss in what extensions we can > do into taskit to be suitable of usage in this case, but surely I > would be inclined to do a higher level framework or even middleware > that uses taskit, than to add all those complexities in taskit. The > good news is that i may be needing something similar, so I will be > able to help there.Â > > > > > > > > > Have workers connect to the scheduler? Other ideas? > > > what do you mean by connection to the scheduler? The workers we > > use do not know their pools, if that is what you mean.Â > > > > Let's assume scheduling a task is simple (read something from a > > socket) but the computation is expensive (database look-up, math, > > etc). Hopefully one will reach the point where one image can > > schedule more tasks than a single worker image can handle. At that > > point it could be neat to scale by just starting another image. By > > inverting the launch order (workers connect to the scheduler) > > scaling can become more easy. > hahaha, now i re-read this paragraph and is not a hunch anymore. You > are looking for multiple images :).Â > > > Â > > holger > > > > > > > >

April 25, 2018