[OSM-dev] New minutely replication test
One of the services the Operations Working Group (OWG) runs is the replication feed for OSM changes. This is used to broadcast changes which are picked up by data consumers. We are running a test of new software to power the replication feed, osmdbt (https://github.com/openstreetmap/osmdbt). The existing software, osmosis, relies on quirks of PostgreSQL that do not apply to more recent versions. We are looking to have users test this new feed for correctness to make sure it works with their update processes and software. To use it adjust your replication URL from https://planet.openstreetmap.org/replication/minute/ to https://planet.openstreetmap.org/replication/test/minute/. We intend to offer hourly and daily update streams like are provided right now. When using the new diffs, there are several things to be aware of - There is no simple relationship between the old sequence numbers and the new ones. Use the timestamps to work out what the new sequence number is. - These are experimental and we're trying to find bugs. It might be necessary to stop them or run other maintenance. - The current URL is not a long-term one. It will change. We are particularly interested in reports from people who have written custom OSM diff fetching code, as they may find different bugs in osmdbt or their own code that the standard software does not. If you have questions or issues, please comment on https://github.com/openstreetmap/operations/issues/475 or here. ___ dev mailing list dev@openstreetmap.org https://lists.openstreetmap.org/listinfo/dev
Re: [OSM-dev] Restarting the EWG
Replying to Jochen and Frederik together here for compacting things. Thanks for the replies, although i would still be very much interested in the board's thoughts on the matter these comments already provide some valuable views. Since neither you nor me are pure hobby volunteers in OSM, us discussing how convincing your replies to my first question are is pretty academic. Actual hobby volunteers will ultimately have to answer that. Regarding the second question both of you seem to see no problem with the preference for "people whose work we know and enjoy" paradigm regarding money spending and personnel selection choices in an organization like the OSMF. That is where our views seem to fundamentally differ. A deeper open discussion on the matter within the OSM community would be fundamentally important (and i tried to incite such a discussion in the past on several occasions) but this list is certainly not the right place for that. Since Jochen brought the discussion a bit on personal motivation for volunteer work a few notes on that from my side - to maybe make it easier for others to understand why i choose to engage in certain volunteer activities but not in others. For volunteer contributions in OSM i chose those fields where i can make difference through the quality of my work - either craftsmanship in practical work or solid analysis, arguments and reasoning in more abstract matters. I deliberately and fairly categorically stay out of matters where i would have to negotiate with the interests of others independent of arguments and reason w.r.t. the common good. If that disqualifies me for a leadership position or even fully in having a substantial influence on decisions in the prevailing organizational culture of the OSMF i am fine with that. -- Christoph Hormann http://www.imagico.de/ ___ dev mailing list dev@openstreetmap.org https://lists.openstreetmap.org/listinfo/dev
Re: [OSM-dev] Restarting the EWG
Simon, agree that’s a danger. I think the starting point is not a list of technical tasks, but strategic goals that have technical implications and to which technical needs can be linked. Connect up the tech to overall purpose. For instance, the data privacy for gdpr compliance is a osmf goal. To achieve that there’s a specific implementation need. The EWG needs to look at all means to achieve it — thus far putting out a cfp (repeatedly) has not turned up any rails devs. We need to think about why, and evaluate changes of tactics. At yesterday’s board meeting we talked about various MWG needs to connect civicrm and OSM.org. So another example. It’s not all “boring”. There’s always a swirl of ideas to refresh OSM.org landing page. Primarily that’s a communication and design question, but for ewg the question is how ready is the rails app for implementing new designs. Mikel On Friday, November 20, 2020, 6:04 AM, Simon Poole wrote: Without at least some guidance from the board on purpose and scope I see a real danger of this turning in to yet another iteration of "lets make a top ten list of stuff that might attract devs" with more money, aka not just GSOC, thrown in as the sole change. With the boring stuff that "actually needs to be done" (tm), being ignored. it isn't as if we don't have the experience of numerous failed EWG reboots. Examples: - the data privacy related work that needs on the API, the website and data distribution, this is probably the best defined and scoped work that has ever existed in the history of OSM, still it has made zero progress over the last three years, - putting a system in place to manage third party sources, permissions to use them and provide attribution in a scaleable fashion (realistically just providing the mechanics for this wont be enough, as the clean up itself has to be organized and that could easily require multiple man years of clerical work). I'm sure there are other similar items from operations and communications that are simply never going to make any kind of list without the EWG actually being made -responsible- for clearly defined outcomes instead of a lot of hand waving that will simply gyrate to projects that result in the largest amount of back patting (iD etc). Simon Am 19.11.2020 um 17:09 schrieb Paul Norman via dev: > The OSMF Board is looking at restarting the Engineering Working Group > with a revised scope to include handling paid software development. > This scope needs to be developed with existing and new volunteers, but > my ideas are that it would include > > - Google Summer of Code, > - managing development to be paid by the OSMF by contracts and grants, > and > - collaborating with other organizations for multi-organization grants. > > It would do this by by > - placing calls for proposals for paid work on topics like top ten tasks; > - accepting other proposals; > - defining an approximate distribution of OSMF funds for development > between primary/secondary/tertiary services, external services, and > new services; > - encouraging applications from skilled individuals who aren't > professional developers, professional contractors, companies, and others. > > Once the scope and funding distribution guidelines are defined we > would want to start with low-risk projects involving experienced > people who need less management. > > If you are interested in changing the EWG to handle these roles, > please let me know. > > > ___ > dev mailing list > dev@openstreetmap.org > https://lists.openstreetmap.org/listinfo/dev___ dev mailing list dev@openstreetmap.org https://lists.openstreetmap.org/listinfo/dev ___ dev mailing list dev@openstreetmap.org https://lists.openstreetmap.org/listinfo/dev
Re: [OSM-dev] Restarting the EWG
Without at least some guidance from the board on purpose and scope I see a real danger of this turning in to yet another iteration of "lets make a top ten list of stuff that might attract devs" with more money, aka not just GSOC, thrown in as the sole change. With the boring stuff that "actually needs to be done" (tm), being ignored. it isn't as if we don't have the experience of numerous failed EWG reboots. Examples: - the data privacy related work that needs on the API, the website and data distribution, this is probably the best defined and scoped work that has ever existed in the history of OSM, still it has made zero progress over the last three years, - putting a system in place to manage third party sources, permissions to use them and provide attribution in a scaleable fashion (realistically just providing the mechanics for this wont be enough, as the clean up itself has to be organized and that could easily require multiple man years of clerical work). I'm sure there are other similar items from operations and communications that are simply never going to make any kind of list without the EWG actually being made -responsible- for clearly defined outcomes instead of a lot of hand waving that will simply gyrate to projects that result in the largest amount of back patting (iD etc). Simon Am 19.11.2020 um 17:09 schrieb Paul Norman via dev: The OSMF Board is looking at restarting the Engineering Working Group with a revised scope to include handling paid software development. This scope needs to be developed with existing and new volunteers, but my ideas are that it would include - Google Summer of Code, - managing development to be paid by the OSMF by contracts and grants, and - collaborating with other organizations for multi-organization grants. It would do this by by - placing calls for proposals for paid work on topics like top ten tasks; - accepting other proposals; - defining an approximate distribution of OSMF funds for development between primary/secondary/tertiary services, external services, and new services; - encouraging applications from skilled individuals who aren't professional developers, professional contractors, companies, and others. Once the scope and funding distribution guidelines are defined we would want to start with low-risk projects involving experienced people who need less management. If you are interested in changing the EWG to handle these roles, please let me know. ___ dev mailing list dev@openstreetmap.org https://lists.openstreetmap.org/listinfo/dev OpenPGP_0x4721711092E282EA.asc Description: application/pgp-keys OpenPGP_signature Description: OpenPGP digital signature ___ dev mailing list dev@openstreetmap.org https://lists.openstreetmap.org/listinfo/dev
Re: [OSM-dev] Restarting the EWG
Christoph, On 11/19/20 20:41, Christoph Hormann wrote: > * why i as a pure hobby OSM contributor with experience in the field of > development should volunteer my time to manage the paid development work of > others on my own unpaid time. You make it sound like this was something new, but the OSMF which largely consists of unpaid hobbyists is already managing paid contributions in various forms. That doesn't make your point an invalid one - from the very first time the OSMF was paying for development this question was on everyone's mind, and many other Open Source projects who use bounties or Summer of Code or other means of compensating developers in addition to attracting volunteer contributions are faced with the same problem. Speaking as a member of an existing OSM working group, the DWG, I can say that I could imagine a couple of projects where I would like to put some of my DWG volunteer time into managing paid development work that would in the end make my life in the DWG easier, and it would not diminish my DWG engagement at all - on the contrary, probably. So yes, if not handled well then using money to pay for stuff can be a turn-off, but it certainly doesn't have to be! > * how i as someone with a business or professional career interest in the OSM > context would be able to contribute to this work without universally having a > massive conflict of interest with every decision of substance that is being > made. That remains to be seen. Obviously we wouldn't want someone to bring in their spouse and kids, or even their friends, as contractors. Then again, "I have worked with this guy in the past and I am confident he can do what we need here" could be a very valuable piece of information. As always, the thing about conflicts of interest is that they need to be properly declared and managed (instead of covering everything in sticky "we all want the best for OSM so where's the conflict" sauce), but if that is done well then they can be handled. Bye Frederik -- Frederik Ramm ## eMail frede...@remote.org ## N49°00'09" E008°23'33" ___ dev mailing list dev@openstreetmap.org https://lists.openstreetmap.org/listinfo/dev
Re: [OSM-dev] Restarting the EWG
On Thu, Nov 19, 2020 at 08:41:15PM +0100, Christoph Hormann wrote: > > Paul Norman via dev hat am 19.11.2020 17:09 > > geschrieben: > > > > - managing development to be paid by the OSMF by contracts and grants, and > > For better understanding of how the board envisions this to work, could you > explain: > > * why i as a pure hobby OSM contributor with experience in the field of > development should volunteer my time to manage the paid development work of > others on my own unpaid time. > * how i as someone with a business or professional career interest in the OSM > context would be able to contribute to this work without universally having a > massive conflict of interest with every decision of substance that is being > made. > > Please note although these might sound like rhetorical questions they are > not, i am honestly interested in how the board envisions this to work. Why does someone want to join the EWG? Maybe because they don't want to sit at the sidelines but contribute to OSM, take responsibility and shape the future of OSM with their ideas and their values? I think it is totally awesome that the barrier of entry to contribute to such a working group is basically that you have to have an email address and nothing else. No "You have to be this tall to ride the rollercoaster". And you get to steer the rollercoaster, too. At least a little bit. Of course sysadmin experience would be great, but it is absolutely not required. And of course there is a lot of work involved if you actually want to make a difference and a lot of talking to people and compromising. But how many places are there in the world where it is so easy to start making a difference? I am the first to tell you that this kind of work can be really frustrating at times, but it is also enormously rewarding to work with others and see that your ideas can make the world a tiny bit better. Full disclosure: I am currently paid by the OSMF for software development (the EWG was not involved in that at the moment, but might be in the future) and I have been paid by others, too. I am not going to join the EWG. That conflict of interest would be too large. But that doesn't mean I can't contribute my spare time somewhere else. I am on the board of directors of the FOSSGIS e.V., the local chapter of OSMF in Germany. FOSSGIS is employing somebody for organizational work, I am basically their boss. This gives me a bit more opportunities to see my ideas and values brought into action. Although far less than you might expect. We are not hiring minions, but people who think and act on their ideas, their values, and their interests. And that's a good thing. What brings OSM (and, really, anything else) forward is the shared work of people with diverse interests and backgrounds. Of course there are more (potential) conflicts of interest there. Theoretically I have through FOSSGIS and OSMF some influence on EWG policy who might be my boss in the future. And I have some influence because I know people in EWG. And because I write this email. Why should that be a problem? I basically started having conflicts of interest the moment I entertained the idea of making my OSM hobby into a profession. I have done that more than 10 years ago. And I think about that basically every day. With every project I start (or not start), with every software I write (or write differently), with every email I write. Like this one. That's just part of my life. And anybody's life really. We all have to find a way to make our hobby life and professional life work together and work together in this project. Yes, it can be very complicated sometimes, but very often it isn't. You do your best, disclose potential conflicts of interest, keep away from something if needed. And move on. And just to explode everybody's mind who wants to think about the layers and meta-layers of conflicts of interest I write about in this mail I write this next sentence: I am available for contract work, software development or general OSM consulting. Jochen -- Jochen Topf joc...@remote.org https://www.jochentopf.com/ +49-351-31778688 ___ dev mailing list dev@openstreetmap.org https://lists.openstreetmap.org/listinfo/dev