Is Wikimedia an American monopoly?

Posted by Strainu on July 01, 2015
Wikipedia / No Comments

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.

General questions

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:

  1. How much are the original creators (sculptors, architects etc.) earning in countries without FoP?
  2. 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.
  3. How many court decisions have there been against big re-users in countries without FoP?
  4. How much are the re-users losing by not being able to use those monuments?
  5. 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?
  6. 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 3 responses:

  • 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.

Share and Enjoy:
  • Facebook
  • Twitter
  • Identi.ca
  • email
  • LinkedIn
  • Reddit
  • StumbleUpon
  • Add to favorites

Parcări de lungă durată pentru Aeroportul Otopeni

Posted by Strainu on June 02, 2015
Weekend Trips / No Comments

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:

Parcarea aeroportului

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.

Park&Fly Otopeni

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.

Rin Airport Hotel

  • 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.

Motel Săftica

  • 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?

Share and Enjoy:
  • Facebook
  • Twitter
  • Identi.ca
  • email
  • LinkedIn
  • Reddit
  • StumbleUpon
  • Add to favorites

No-name chinese battery pack

Posted by Strainu on February 07, 2015
Gadgets / No Comments

Fig. 1: Battery pack

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.

Fig. 2: Technical specs

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.

Fig. 3: Input current at the beginning and end

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

Fig. 4: With no load, the Lithium battery voltage of 3.7V is visible.

Fig. 5: Under load, the maximum voltage is about 5.15V.

Share and Enjoy:
  • Facebook
  • Twitter
  • Identi.ca
  • email
  • LinkedIn
  • Reddit
  • StumbleUpon
  • Add to favorites

Tags: , , , , ,

Teste pe RDS

Posted by Strainu on December 27, 2014
Gadgets / No Comments
Încărcarea spectrului

Fig1: Încărcarea spectrului de 2,4GHz

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).

Fig. 2: Speedtest cu EDUP

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.

Fig. 3: Speedtest 802.11ac

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.

Share and Enjoy:
  • Facebook
  • Twitter
  • Identi.ca
  • email
  • LinkedIn
  • Reddit
  • StumbleUpon
  • Add to favorites

Provocări și rezultate într-un proiect de voluntariat la scară mare în România

Posted by Strainu on December 02, 2014
Publications, Romana, Wikipedia / No Comments

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.

Share and Enjoy:
  • Facebook
  • Twitter
  • Identi.ca
  • email
  • LinkedIn
  • Reddit
  • StumbleUpon
  • Add to favorites

Tags: ,

Colectarea și verificarea coordonatelor pentru un proiect liber

Posted by Strainu on April 10, 2014
Publications, Wikipedia / No Comments

Weekendul trecut am fost la Cluj, la seminarul “Soluții libere open source pentru prelucrarea și reprezentarea datelor geospațiale” organizat de geo-spațial și am ținut o prezentare despre cum am strâns și verificat coordonate pentru Lista monumentelor istorice de pe Wikipedia. Prezentarea este disponibilă la sfârșitul articolului.

Pe lângă reîntâlnirea cu colegii de la geo-spațial, care au avut un rol important în promovarea Open Government Partnership la nivelul Guvernului României (și implicit la existența portalului date.gov.ro), am avut și ocazia să fac o aproximare a nivelului de cunoaștere a activităților Wikipediei în limba română și a OpenStreetMap România.

Deși mulți dintre studenți auziseră de cele două proiecte libere, destul de puțini erau conștienți că ele pot fi editate și mai mult, că există un interes pentru date geografice. Din păcate puțini erau familiarizați cu data-miningul sau cu licențele libere, așa încât prezentarea mea a fost percepută ca fiind prea tehnică. Am primit totuși câteva întrebări după prezentare în legătură cu interacțiunea între datele publice de pe date.gov.ro și articolele Wikipedia.

În ceea ce privește celelalte prezentări, mi-a atras atenția prezentarea “Rețeaua națională de localități: actualizarea listei de localități viabile din România“, în care Sorin Rusu de la TeamNet a făcut o analiză a localităților-fantomă din România. În prezent sunt în discuții cu el pentru a avea acces la setul de date, informațiile putând fi foarte utile pentru Proiectul Localităților din România.

Am mai aflat de asemenea că serverul eGISpat al INP ar conține zonele de protecție corecte și complete ale monumentelor istorice din 8 județe, iar proiectul de ridicare a coordonatelor continuă. Se impune probabil o nouă ofensivă la nivelul Ministerului Culturii pentru a obține publicarea datelor respective pe portalul date.gov.ro.

Am avut de asemenea ocazia să discut cu un coleg de la en.wp (care a dorit să rămână anonim) și care se plângea de tonul abordat la Cafeneaua ro.wp, considerat de el ca un pumn în gură pentru cei din afară. Mă bucur să văd că nu-s singurul care a constatat asta. El a mai ridicat și alte probleme, fără să mă convingă că situația de la en.wp ar fi mult mai bună.

În concluzie, deși prezentarea a fost considerată prea tehnică, discuțiile ulterioare au deschis câteva piste interesante de explorat pentru a obține mai multe date pentru Wikipedia.

Share and Enjoy:
  • Facebook
  • Twitter
  • Identi.ca
  • email
  • LinkedIn
  • Reddit
  • StumbleUpon
  • Add to favorites

Downloading images to TI Tiva C Launchpad from Keil in Wine

Posted by Strainu on February 07, 2014
Embedded, My Projects / 2 Comments

I needed to run some programs on my Tiva C Launchpad board from Linux. Normally this is very simple, as shown in a previous post. However, I needed to use the ARM Keil IDE, in order to get credit for my work.

Keil runs fine under Wine, but Wine is not that good with unknown USB devices. However, I managed to find a way around that and download the image using lm4tools instead of the default flasher. The instructions were tested using Tiva C, but work just as well on Stellaris Launchpad.

1. First, we gotta be sure we can run it:

git clone https://github.com/utzig/lm4tools.git
cd lm4tools/lm4flash
make
sudo make install
echo 'ATTRS{idVendor}=="1cbe", ATTRS{idProduct}=="00fd", GROUP="users", MODE="0660"' | \
  sudo tee /etc/udev/rules.d/99-tiva-launchpad.rules

The last command is optional and ensures the command can be used as a regular user. The Vendor and Product id are obtained by running:

lsusb | grep Luminary

and are the same on Stellaris Launchpad and Tiva C

2. Then, we have to make the script callable from wine. This can be done by creating a shortcut in the fake C:\ drive used by wine:

ln -s /usr/bin/lm4flash ~/.wine/drive_c/windows/system/lm4flash.exe

To check if it works, just start cmd from wine:

$ wine cmd
c:> c:\windows\system\lm4flash.exe

3. lm4flash needs a bin file and Keil ony generates an axf file. To get the bin file, go to project options, user tab and paste the following in the “Run programs after Build/Rebuild” section:

C:\Keil\ARM\ARMCC\bin\fromelf.exe  --bin HelloLaunchPad.axf --output HelloLaunchPad.bin

Ofcourse, you need to customize the paths.

4. The final step is to use lm4flash to download the image. Go to the project options, Utilities tab and select the “Use External tool for Flash programming” radio button from the “Configure Flash Menu Command” section.

Command: C:\windows\system\lm4flash.exe
Argument: HelloLaunchPad.bin
Run Independent checked (this is so the IDE won’t block if the script fails for some reason)

5. That’s it, make a small change to the code (so you know if the new code is active) and click on Flash->Download. Hope this helps someone :)

Share and Enjoy:
  • Facebook
  • Twitter
  • Identi.ca
  • email
  • LinkedIn
  • Reddit
  • StumbleUpon
  • Add to favorites

Tags: , , , ,

IDEs for developing for the Tiva C board.

Posted by Strainu on February 05, 2014
Embedded / No Comments

Now that we have a working environment for the Tiva C, we might as well install some IDEs and make it easy to develop complex programs. The main options on Linux are:

  1. TI’s own Code Composer (an Eclipse IDE with some changes); it can also be installed within an existing Eclipse instance, but it only works with versions 3.8 to 4.0.
  2. Eclipse itself can be configured so it works with the Linux tools
  3. If for some reason you want to use Keil, the IDE from ARM itself, you will need to use Wine, as there is no Linux version.

In this post we will talk about the first to IDEs and leave Keil for another day, as the configuration is a little bit more complex in its case. If you want a step-by-step guide for Keil (and Code Composer), I recommend this excellent article.

Code Composer

Code Composer conveniently offers a dependency script the one can use to determine if all necessary packets are installed. I strongly suggest to run it first, as debugging afterwards can be difficult. You can find the full instructions below

#precheck
wget http://software-dl.ti.com/dsps/dsps_public_sw/sdo_ccstudio/scratchpad/depends_0.2.tar.gz
mkdir ticheck
cd ticheck
tar -xzvf ../depends_0.2.tar.gz
./check_depends.sh
#install anything needed
#download the CCS
./ccs_setup_5.5.0.00077.bin

You will need to choose the following options: Custom-> Tiva C Series ARM MCUs -> Select All -> Select All

If you could not install the drivers from the installer, you need to run another script: sudo /opt/ti/ccsv5/install_scripts/install_drivers.sh

You are now ready to start Code Composer. If you are familiar with Eclipse development, you should have no problem compiling, downloading and debugging code for the Tiva board.

Eclipse

In order to use Eclipse itself, you need to configure it to use the arm compiler and OpenOcd debugger. You can find an excellent tutorial on this page, so I won’t go into further details here.

Similar approaches can be taken with other IDEs, like Netbeans. If you are using another IDEs, do leave a comment and let us know how it’s been working for you.

Share and Enjoy:
  • Facebook
  • Twitter
  • Identi.ca
  • email
  • LinkedIn
  • Reddit
  • StumbleUpon
  • Add to favorites

Tags: , , , ,

Programming for the Tiva C Launchpad on Linux

Posted by Strainu on February 04, 2014
Embedded / 2 Comments

I’ve recently acquired a Tiva C launchpad from Texas Instruments, that I plan to use to complete an online course much like PM was during my faculty years. However, unlike 7 years ago, my only OS is now openSUSE, which meant having an additional (although small) challenge in creating a development environment.

Below is a checklist of what you will need. Prerequisites include being able to download and compile programs from source code (but you probably know that if you’re trying to program an embedded board, right?).

1. First and foremost, you will need a toolchain for compiling the code you write. Tiva C has an ARM Cotex-M processor, making it relatively easy to find a great number of toolchains.

Many people on the internet recommend Summon-arm, but this toolchain is no longer maintained. Instead, the author recommends GNU Tools For ARM Embedded Processors (AKA GCC ARM Embedded), so I went with that one. You can download either a precompiled archive, or the sources and do the compiling yourself. As my computer is fairly old, I decided on the binary archive:

wget https://launchpad.net/gcc-arm-embedded/4.8/4.8-2013-q4-major/+download/gcc-arm-none-eabi-4_8-2013q4-20131204-linux.tar.bz2
tar -xvf gcc-arm-none-eabi-4_8-2013q4-20131204-linux.tar.bz2
sudo cp -r * /usr

2. Then, you need the Texas Instruments libraries and headers. You will not find any Linux kit on the TI website, but fear not: the Windows installer is just a zip file that contains a working Makefile. The kit can be downloaded from this page (or direct link). You will need a TI account to download – the same one used to order the Launchpad. Compiling it is also a breeze:

mkdir tilib
cd tilib
unzip ../SW-EK-TM4C123GXL-2.0.1.11577.exe
make

3. You will also need something to download the binaries on the device. lm4flash is the recommended way of doing this on Linux. This is also compiled easily and quickly:

git clone https://github.com/utzig/lm4tools.git
cd lm4tools/lm4flash
make
sudo make install

echo 'ATTRS{idVendor}=="1cbe", ATTRS{idProduct}=="00fd", GROUP="users", MODE="0660"' | \
sudo tee /etc/udev/rules.d/99-stellaris-launchpad.rules

The last command is optional and ensures the lm4flash command can be used as a regular user. The Vendor and Product id are obtained by running:

lsusb | grep Luminary

and are the same on Stellaris Launchpad and Tiva C

You can check that lm4flash works by downloding one of the examples included in the TI archive (see step 2). After a successful download, the link should blick while the color varies.

cd tilib/examples/boards/ek-tm4c123gxl/blinky/sourcerygxx
lm4flash blinky.bin

4. Debugger – a must if you write more than very simple programs for the Tiva C. The OpenOcd package is the way to go.

If you want to fetch the source on a Deb-based Linux (Debian, Ubuntu etc.) using git, here is what you must do:

sudo apt-get install git-core libtool autoconf texinfo libusb-dev libusb-1.0 screen
git clone http://git.code.sf.net/p/openocd/cod

On my openSUSE, I prefered to download the tarball:

sudo zypper in git-core libtool autoconf texinfo libusb-dev libusb-1_0-0 libusb-1_0-devel
wget http://downloads.sourceforge.net/project/openocd/openocd/0.7.0/openocd-0.7.0.tar.bz2

In both cases, compiling is the same:

./configure –enable-maintainer-mode –enable-ti-icdi
make
sudo make install

5. While TI offers its own startup libraries, some people prefer a totally free (as in FLOSS) linker. Instructions on installing such a linker on the Stellaris Launchpad (the forerunner of Tiva C) can be found on this page. I have not tried it, as for my requirements the TI libraries are enough, but due to la similarities between the platforms, this should also work just fine on the Tiva.

This is it, you can now compile and download your own programs on your new board. However, if you’re not a vim-or-emacs person, you might want to install an IDE.

Share and Enjoy:
  • Facebook
  • Twitter
  • Identi.ca
  • email
  • LinkedIn
  • Reddit
  • StumbleUpon
  • Add to favorites

Tags: , ,

New voyages map

Posted by Strainu on January 19, 2014
Geeks, Personal / No Comments

It’s been a while since I updated my visited countries map. As in 2013 I’ve been to 5 new countries, I thought it was time for an update.


visited 27 states (12%)
Create your own visited map of The World

Share and Enjoy:
  • Facebook
  • Twitter
  • Identi.ca
  • email
  • LinkedIn
  • Reddit
  • StumbleUpon
  • Add to favorites