Here is another presentation I hold during the Wikimedia CEE Meeting in Armenia this summer. Unfortunately the audience was not technical, so it was not as well received as the one from last year. Still, I hope it will be interesting for other people.
Yesterday I decided to give Mapillary a second try (I’ve tested it once before just after launch and it was pretty much unusable). And since I was planning to go over Translapina, I thought: what a great way to see how the app handles various issues one might encounter on mountain road, such as tight curves, changing light and poor GPS cover while discovering a new road (for both me and Mapillary). Southern Romania is pretty well covered in imagery, at least on the main roads, but once you go in the mountains, the coverage drops significantly.
The road in itself was nothing to write home about. I know some of my colleagues are crazy about it, so I’ll probably get some angry looks at work, but the scenery is just not comparable with the Transfăgărășan. There are bends and valleys and mountains in clouds, but the vegetation reminds me more of the desert than the mountain. Plus, as a driver, you always have to look out for rocks and trees falling from the side of the road, as the construction quality is not the best and the earth in the area is very friable as well.
The app was not impressive either. It has come a very long way from the previous version I’ve tested, but it’s still very slow even when it works, which is not always. The app crashed often, especially when the map was on, or simpli freezed for no apparent reason and no warning to the user. I had to pull over, close the app and reopen it. Also, some images are blurred, probably because the app did not wait for the camera to focus. When taking the pics in “distance mode”, no pictures should be taken when the car is stopped. However, the app sometimes failed and continued to take 1 or 2 pictures after the app was completely stopped.
The settings are not uniform (some need to be disabled, but there is no checkbox next to them, instead one simply clicks on them, other need to be enabled) and the different picture modes are insufficiently expained – I had to deduce what each of them does based on many different settings.
Also, a very important problem of the app is that it takes up the screen from nagivation apps. If it is to be used on unknown roads, it should have a “background mode” where it takes pictures while the user is in another app, or, at the very least, allow the map to be configurable in size and position.
Now, I know that some of the issues I mention can be caused by the phone I had, a OnePlus One. It’s not the newest phome you can buy, but if Mappilary developers expect their app to be used only on flagships, they might be in for a nasty surprise.
I will continue to test the Mapillary app for the next few days, but my feeling is it’s not yet ready for prime time.
Luna trecută am fost invitat de prietenii de la ApTI să țin o prezentare scurtă, de 10 minute, despre libertatea de panoramă, în cadul unei dezbateri despre reforma drepturilor de autor organizate pentru a face cunoscute inițiativele Comisiei Europene în domeniu. Participanții erau din medii diverse, de la ORDA, patronatul editorilor, geo-spatial.org etc.
Prezentarea este disponibilă la Wikimedia Commons în format pdf.
Last autumn I participated in the Wikimedia CEE Conference in Estonia, where I held a presentation about the list management bot used for maintaining Wikipedia’s List of Historic Monuments in Romania. The presentation was quite technical, so I didn’t get a lot of audience, but I did get some good feedback both during and after the conference.
Pentru amatorii de filme românești, Cinepub este locul unde pot urmări legal filme românești. Deși majoritatea nu sunt foarte cunoscute publicului larg, sunt și câteva din cele care au ajuns în sălile de cinema în ultimii ani.
Cum în ultima vreme am folosit numai Kodi pentru a urmări filme, am zis că n-ar strica să fac un plug-in care să-mi permită să urmăresc filmele pe TV. Cinepub folosește youtube pentru găzduirea filmelor, deci aș fi putut teoretic să fac un playlist ce conținea toate filmele, dar ar fi fost mai puțin interesant. Codul e pe github, feel free to fork and improve.
- Descărcați arhiva
- Deschieți Kodi și mergeti la System->Settings->Add-ons->Install from zip file și alegeți arhiva.
- Dacă instalarea a avut loc cu succes, găsiți filmele la Video->Add-ons->cinepub.ro
Asta-i tot, enjoy!
In the last few days there was quite of a storm in the free-content world raised by an amendment introduced by the French MEP Jean-Marie Cavada in the InfoSoc evaluation report made by Julia Reda. The amendment (one of more than 500 proposed in the commission that discussed the report) basically forbids Freedom of Panorama (FoP) in Europe, by allowing only non-commercial uses of reproductions of copyrighted works in public places.
The longer story: In 2001, the European Council and Parliament adopted directive 2001/29/EC “on the harmonisation of certain aspects of copyright and related rights in the information society”, short the InfoSoc directive. In late 2014, the Internal Affairs Committee appointed Julia Reda to create a report on the implementation of this directive and to offer advice on how should the European Commission tackle a new copyright directive. So far, the report was presented to the Committee and it received over 500 amendments; the Committee voted on them on June 16th and some (including Mr. Cavada’s amendment on FoP) were adopted. The next step is for the Parliament to vote on the report in early July. Then, in late 2015 or early 2016, the Commission will begin drafting the new directive which will pass through the European Council and Parliament.
In a long blog post, Mr. Cavada justifies his amendment. The gist of the post is this:
Le combat […] est […] mené avant tout pour permettre aux monopoles américains tels que Facebook ou encore Wikimédia, d’échapper au versement des droits aux créateurs.
Here is my translation in English:
This war is waged in order to allow American monopolies such as Facebook or Wikimedia to skip paying royalties to creators.
I thought a short FAQ from a Wikipedian specifically regarding this blog post would be useful. The questions are the ones I see asked around me, on Facebook or blogs and the answers are exclusively my own opinion. This post assumes you know a bit about copyright and Freedom of Panorama. You can find more generic information written by Wikimedians on this page.
Q1: What is the difference between Wikimedia and Wikipedia?
A1: See here.
Q2: Is Wikimedia a monopoly? How about Wikipedia?
A2: Wikimedia is most definitely not a monopoly. Beyond Wikipedia, the movement expanded in many different areas, such as Public Domain original works (Wikisource), tourism guides (Wikivoyage), community journalism (Wikinews) etc. In mst of these fields the Wikimedia websites are not even close to being leaders, let alone a monopoly.
Wikipedia on the other hand is a different story. Thanks in no small part to Google’s rating algorithms, Wikipedia has become the dominant player in the area of general information on a subject. While this is still far from a monopoly (other websites do exist and they do get traffic), one can understand how this dominant position might bother some players.
Q3: Is Wikimedia an American monopoly?
A3: The simple answer is that, with the 250+ language versions, with volunteers from almost every country and with chapters (local NGOs) in almost 100 countries, Wikimedia cannot be considered “American” and Mr. Cavada is simply playing with terms in order to align Wikimedia with Facebook and other established content publishers.
However, this answer is far for complete. There are many frictions withing the Wikimedia community (this NY Times article is relevant) and many of those are about the fact that the Wikimedia Foundation (WMF), that hosts Wikipedia and the other sites is based in the US (see this thread, for instance). My personal opinion is that this kind of argument will remain present no matter where the Foundation is hosted.
Also, the influence of chapters is not as high as some believe. In 2011, their ability to fundraise using the Wikimedia trademarks was severely limited (see the WMF’s executive directors recommendations at the time and the other pages linked from there), concentrating all the movement’s funds in the hands of a single entity and making the global chapters dependent on the decisions of a funding committee. Leaving aside the personal pride of the chapters and the fact that this limited the ways one could donate (no more phone donations, higher costs for wire transfers etc.), this is obviously bad news for the volunteers in countries that are under sanctions from the US (because money transfer to and from these countries are forbidden) for the WMF. Again, my opinion is that this is not a US-specific problem, but an internal issue of the WMF.
Q4: Is Wikimedia financed by monetizing content? Is it making a profit?
A4: Definitely not. Wikimedia is maintained by an NGO, curated by volunteers and financed from donations.
Q5: Mr. Cavada says Wikimedia requests high-quality images that can be used for commercial purposes, thus preventing the right-owners from collecting royalties. Is that right?
A5: I am not sure what he refers to, but it sounds like he talks about the GLAM partnerships (Galleries, Libraries, Archives and Museums). While there are many types of partnerships (wikipedian in residence, content donation, exhibit tagging etc.), they are all based on the free will of the partners and, more often that not, refer to works already in the Public Domain. The Cultural Entities that donate high-quality content to Wikipedia do so because they understand that spreading and reusing their content makes them known to the public and can attract future (paying) visitors. We also have to keep in mind that usually GLAMs do not own the copyright on the works they host; at most, they are entitled to the database creator’s rights.
For the “normal” users that make their own pictures and upload them to Wikipedia, there is no such request enforced. They can contribute as they see fit, as long as they respect other people’s copyrights.
Q6: Does Wikipedia hurt European copyright holders?
A6: This is not about Wikipedia vs. the copyright holders, but about free content and balances. As I said in A4, the Wikimedia sites are maintained by a non-commercial entity, so they could in theory use Mr. Cavada’s proposal to keep using images of buildings under copyright.
However, at the very core of Wikipedia is the respect of the user’s freedoms (now you know what “free” stands for in “The free encyclopedia”), so we want to make sure reusers of Wikipedia content can do anything with our content, as long as attribution is offered. This means that some people that were simply monetizing the original works in question instead of building on them to offer added value to consumers will have something to lose.
On the other hand, the fact the Wikipedia content is free also means that European entities that are prepared to embrace the change can win from it. For instance, architects could use free images to build a portfolio instead of employing a professional photographer. Cultural entities can complement and improve their exhibits by using free content (see the examples in A5), bringing in more visitors and thus more revenue.
Finally, for the end-users the main benefit is not quantifiable in money, but it undoubtedly exists: the access to higher-quality, legal content.
In conclusion, we can say that while the usual way of earning money from copyright work is somehow affected by FoP, the overall benefits far outweigh this loss (the references cited here should convince you of that). Even copyright holders can adapt and compensate the losses by using free content to their advantage.
Q7: Can’t the Europeans create their own Wikipedia to compete with the “Americans”?
A7: It’s not that simple. Starting from 0 would imply way too much financial and human effort. One could, of course, start by using Wikipedia’s content (since it is freely licensed). However, Wikipedia’s license (CC-BY-SA 3.0) is what’s called “strong copyleft“, meaning that any good changes in the new project could be integrated back into Wikipedia. In order for the new project to succeed, it would need to convince a critical mass of users to move from Wikipedia to the new project. Historically, this has proven tedious. Here are a few examples from Romania:
- The project documenting wooden churches in Romania begun at the Romanian Wikipedia with 4 or 5 very active members, one of which was an architect with a PhD in the area. Because of infighting the project now has a single active member; he’s always saying that he hates the way Wikipedia works (including the license), but there is no other project that would offer his images the same exposure to the public as Wikipedia.
- Another example is Enciclopedia României. It was started by wikimedians leaving the project in 2007 and was published under a non-commercial license, just like Mr. Cavada’s proposal would impose. The project now has only 5000 articles (compared to hundreds of thousands in the Romanian Wikipedia), mainly because the founders could not increase the contributor base – they simply did not have anything over Wikipedia.
The European Commission did in fact start a project meant to bring the European heritage in the spotlight: Europeana. While it was not designed to directly compete with Wikipedia, but with commercial initiatives like the Google Art Gallery, it is an interesting case study.
The project was aimed at reusers, not end-users and imposed a drastic license (CC-0, which is basically “no copyright”) for metadata and descriptions, but kept the original license for the images and texts published. This allowed content exchanges with Wikipedia: In 2012, Europeana integrated more than 12.000 free images of monuments in Romania from the Wiki Loves Monuments contest with help from the CIMEC. Since then, other countries have used Wikipedia content to enhance Europeana. The CC-0 Europeana descriptions were in turn used to generate articles in Wikipedia, just like it would happen to an European encyclopedia.
Q8: So what do you make of Jean Paul Cavada’s post overall?
A8: That’s a very difficult question. While it raises legitimate questions regarding the responsibility of publishers such as Facebook or Wikipedia, that have appeared time and again since the dawn of the Internet, the approach taken seems flawed. Not only Wikipedia is nothing like Facebook, forbidding all commercial use of copyrighted content in public places even in countries that currently allow it would most likely hit more the local, European, reusers than global Internet companies, which have the knowledge and the money to avoid these regulations.
It’s far more likely that his amendment will hit the souvenir shop next to you than Facebook or Wikimedia.
Q9: Aren’t you biased? Why should I trust you?
A9: You shouldn’t believe anyone blindly, but rather think for yourself. Check out (with a critical eye) the sources available on the Internet and try to answer the following questions for yourself:
- How much are the original creators (sculptors, architects etc.) earning in countries without FoP?
- Are the original creators the ones earning or losing that money or do they go to big businesses? Think of the image of the Eiffel Tower at night.
- How many court decisions have there been against big re-users in countries without FoP?
- How much are the re-users losing by not being able to use those monuments?
- How much are the final users losing both in terms of money (monopoly implies a premium) and non-financial value by not having access to the creative works that could appear if FoP existed?
- Considering all the answers above, is the FoP bringing value to society or not?
Questions about the situation in Romania
Q10: What’s the FoP status in Romania?
A10: Reproductions of copyrighted works in the public space can be used for non-commercial purposes. See this page for more details.
Q11: What’s the Romanian MEPs’ position on this?
A11: I have contacted all 32 MEPs and so far I have received
- a Socialist representative assigned an assistant to look into the matter; he assured me that the objections regarding the ambiguity in “non-commercial usages” will be taken into account
- an ALDE MEP assured me that she will look into the matter and will discuss it with her colleagues from the same European political family
- finally, I have received another email from an assistant confirming the receipt of the email
- Update: Another independent MEP has told me that “he values the freedom of speech, but also copyright and the right of each country to decide on the best way to protect this according to the local cultural landscape” and he will vote “according to all available data and the requests received from Romanian citizens”. So I guess he recieved more emails regarding the subject, which is a good thing.
What other questions do you have regarding Jean-Marie Cavada’s blogpost or the larger issue of FoP in the European legislation? Ask in a comment and I’ll do my best to provide an answer based on sources available on the Internet.
După cum știu unii dintre voi, nu sunt de loc prietenul taximetriștilor. De aceea, acum câțiva ani, când au început să apară parcările de lungă durată pe lângă Otopeni am fost foarte bucuros.
Sigur, pentru bucureștenii care plătesc maxim 100 de lei dus-întors taxiul (cu aparat de taxare corect), ele nu sunt încă o soluție pentru vacanțe, ci doar pentru plecările mai scurte, gen city-break. Eu încerc totuși să le folosesc de câte ori am ocazia și am zis să vi le prezint și vouă pe cele mai cunoscute, împreună cu câteva date și impresii personale. E posibil ca și alte hoteluri din zonă să ofere asemenea servicii dar să nu fi aflat încă de ele.
So, parcările pe care le-am folosit până acum sunt:
- Tarife: variabile, conform listei
- Locație: în fața terminalului plecări (P3/P4)
- Timp de transfer: ~3 min la plecare/până la 10 min la sosire (pe jos)
- Rezervare: nu
Parcarea oficială a aeroportului ar trebui să fie cea mai interesantă datorită apropierii de terminal. Din păcate, dezorganizarea și indicatoarele incomplete te fac să pierzi o grămadă de timp, cel puțin prima dată când o folosești.
În primul rând, la intrarea în aeroport dinspre București nu există indicator decât pentru P3 – de P4 trebuie să știi dinainte. Apoi, parcarea respectivă a fost transformată dintr-o parcare normală prin închiderea unor intrări și ieșiri, dar fără a le scoate și indicatoarele. Ai deci toate șansele să te trezești față în față cu niște blocuri de beton când încerci să intri sau să ieși de acolo.
Datorită aglomerației și a sistemului de plată inflexibil (dacă ai intrat, trebuie să plătești 40 de lei ca să mai ieși) mulți ajung să parcheze pe iarbă sau pe drumurile de acces. Un efect secundar este că nu te mai poți baza pe numărul de locuri libere afișat de sistem. Mi s-a întâmplat să văd locuri libere deși sistemul indica că parcarea e plină.
Alte probleme țin de numărul insuficient de jetoane (se termină jetoanele din aparatul de la intrare) și de cât de complicat este să ajungi din terminalul sosiri la parcare (trebuie să folosești pasajul de la etajul 1 al terminalui sosiri, care este semnalizat prin niște indicatoare fără vizibilitate și câteodată chiar neaduse la zi). De asemenea, nu există supraveghere video.
- Tarife: variabile, conform listei
- Locație: în fața terminalului plecări, dincolo de DN1
- Timp de transfer: 5-15 min, cu microbuzul (gratuit)
- Rezervare: online și telefonic
Park&Fly a fost prima parcare pe termen lung apărută la aeroport. Prețurile sunt un pic mai mari decât la aeroport, iar în preț este inclus și transferul la și de la aeroport cu microbuzul. În mod normal drumul nu durează mai mult de 5 minute, dar dacă ai ghinion și tocmai a plecat cu alți clienți, e posibil să ia ceva mai mult.
Impresia generală a fost de profesionalism, dar prețurile mari îl recomandă în special celor veniți de departe sau care își pot deconta cheltuielile.
- Tarife: 20 lei/zi
- Locație: pe DN1, sensul spre București
- Timp de transfer: 10-15 min, cu microbuzul (gratuit, merge la fiecare oră)
- Rezervare: telefonic, doar pentru grupuri
Chiar dacă are un preț scăzut, plasarea pe sensul spre București și transferul doar o dată pe oră fac această locație mai puțin interesantă. Dacă sunteți genul care ajunge cu mai mult de 2 ore înainte la aeroport și eventual vă permiteți să pierdeți un microbuz, ar putea fi o soluție.
- Tarife: variabil, conform listei
- Locație: pe DN1, sensul spre Ploiești, în Săftica
- Timp de transfer: 15-45 min, cu mașina (gratuit)
- Rezervare: online și telefonic
Motelul Săftica oferă cel ma redus preț dintre cele 4 parcări, dar este și cel mai îndepărtat de aeroport, fiind în Săftica, după lac. Datorită locației și a faptului că transferul se face cu 1 singură mașină de 4 locuri, s-ar putea ca transferul să dureze mai mult (dacă mai sunt și alți călători șoferul trebuie să facă mai multe curse). La întoarcere am așteptat aproape jumătate de oră după mașină. Cred că folosirea unui microbuz s-ar impune.
O altă problemă pentru șoferii corecți o reprezintă faptul că primul loc unde se poate întoarce regulamentar spre București este tocmai în Vlădiceasa, la câțiva kilometri buni spre Ploiești. Șoferii de transfer folosesc o trecere pentru pietoni pe post de loc de întoarcere.
Partea bună e că și această parcare este disponibilă 24/24 și la prețul unui transfer cu taxiul poți staționa până la 6 zile.
Sper că lista de mai sus vă va ajuta să vă faceți o idee asupra opțiunilor de parcare disponibile în apropierea aeroportului. Voi unde parcați când mergeți cu mașina voastră la Otopeni?
I took advantage of the 11/11 discounts from Aliexpress to get a mobile USB battery pack. I knew the quality will be realtive, so I went for the so-called 20.000mAh model. When it arrived, I decided to make some tests on it. For all tests I used the included USB A-microUSB B cable, which seems capable of providing the advertised 2.1A, but has the data wires disconnected, presumably in order to comply with the Chinese standard for mobile phone chargers.
What 20,000mAh means is that, in theory, the 3.7V Lithium battery could provide a 1A CC current for 20 hours. This means the battery can hold 3.7V*20Ah=74Wh. However, the USB standard requires 5V CC, so the user will only have 74Wh/5V=14,800mAh available.
Furthermore, the technical details on the back of the battery (Fig. 2) state the capacity is only 48Wh, which means 9,600mAh, or less than half of the advertised capacity for actually loading some devices.
For charging the battery I used a 1A charger, since this is the maximum current accepted by the battery. It took about 12h to load, with the current going down from 750mA to 400mA (Fig. 3). It is only a guesstimate, since I don’t have the exact current curve, but I suspect it can’t have loaded more than 5V*600mA*12h = 36Wh, or 3/4 of the advertised capacity.
The problems started to pile up when trying to use that energy to actually load some devices. I used a Nexus 4 (2100mAh battery) and a OnePlus One (3100mAh) battery. I started charging when the phones had about 20% from the battery left.
It quickly became obvious that I could not leave the phones unsupervised: the battery would get hot really quick (surface temparature above 50°C) and the loading would stop until I restarted the battery. I guess I should be happy that there was a protection circuit, otherwise the loading circuit could become a fire hazard. The problem would become more and more serious as the phones’ batteries charge, so I could not get a definitive loading time for the OPO. All I can say it got from 20% to 83% in 2h10′. This means ~2000mAh in 2.25 hours, or more than a quarter of the presumed capacity of the battery.
With the Nexus 4 things went more smoothly: after leaving the battery to chill for a while, I was capable to load the phone 3 times with one charge of the battery, with some energy left in the battery. This would mean a capacity of at least 80%*2100mAh*3=5040mAh.
Another problem was the serios variation of both voltage and current during loading. It was at least 4 times as large as the one observed with a quality charger (see the movie at the end of the article). I don’t think most modern electronic devices would have any problems with that, but some older devices might encounter issues.
In conclusion, I think we can say I have a battery with a capacity of about 5-6000mAh, which gets really hot pretty fast. It can be a life saver, but it’s not a product one can depend on. I will probably go for a Xiaomi or Anker external battery next – I’ve seen some really good reviews on the Internet about them. For the curious, I have added some additional pictures at the end of the article, with the associated explanations.
Large variations in voltage and current can be observed when loading from the protable battery
Postul ăsta a început ca o discuție pe Facebook; mulțumesc celor care s-au implicat în ea.
Când am cumpărat Raspberry Pi-ul am considerat că are nevoie de o conexiune cu fir, așa că am rămas cu un adaptor wireless “în plus”. Zilele astea am experimentat un pic cu el la laptop, pentru că e mai bun (cel puțin teoretic) decât cel încorporat. E vorba de un EDUP 802.11n “certifcat” (de Teguna) că funcționează cu RPi.
Ulterior, datorită încărcării în banda de 2,4GHz (Fig. 1 este făcută în perioada de Crăciun, în mod normal e și mai aglomerat), am investit și într-un adaptor Edimax 802.11ac, care funcționează și pe frecvența de 5GHz (unde routerul meu e singur deocamdată).
În primul rând, diferența dintre benzi este vizibilă imediat: lipsa perturbațiilor permite adaptorului să atingă viteza maximă teoretică pe un canal de 80MHz (433Mbps). Din păcate pare destul de greu să configurezi (cel puțin în Linux) sistemul să se conecteze mai întâi (dar nu exclusiv) în banda de 5GHz. Poți să-i spui să folosească doar această bandă, dar asta face imposibilă conectarea la aceeași rețea pe banda de 2,4GHz. Dacă aveți idei despre cum se poate rezolva această problemă, do let me know.
Pe partea de acces Internet, principala observație este că, spre deosebire de 90% din ISP-urile de pe lumea asta, RDS consideră că e mai ok să-mi lase uploadul (mult) mai mare decât downloadul (Fig. 2 și 3) În curând sper să trec la viteza superioară de acces (200Mbps), ceea ce ar trebui să mă apropie și mai mult de line rate, chiar în lipsa unei conexiuni fizice gigabit.
Diferență dintre noile adaptoare și celelalte adaptoare pe care le-am mai avut e doar la upload: până acum n-am reușit să ating 72Mbps în vreun sens. Cu Dell Wireless 1703 obțineam cam 50Mbps upload, la fel și cu Intelul vechi.
In other news, ruterul Alix făcut “pe genunchi” pare să comute line-rate fără probleme dacă îl scutesc de wireless. Păcat că nu vine și în versiunea gigabit, mi-ar fi plăcut să văd cum se comporta. Restul rețelei e prin switchuri gigabit, deci nu prezintă foarte mult interes din punct de vedere al testării.
La seminarul Geo-spațial 2014 de la Timișoara am avut două prezentări. Prima dintre ele a abordat problema organizării unor evenimente cu foarte mulți voluntari și problemele întâmpinate de noi la organizarea Wiki Loves Monuments.
A doua a ținut loc de introducere pentru expoziția WLM, care a fost prezentată timp de două săptămâni în holul mare al Universității de Vest.