1024MAK wrote: ↑Wed Nov 15, 2017 11:06 pm
All the information on WoS that I have referred to is already available on WoS or it's forums. Or was in the now discontinued Wikipedia page on WoS. Or is available on other public sites (like the internet archive).
Lee has already said that he helped Einar with the WoS data. So Einar did start ZXDB from scratch, but he did have help with the WoS data and it's format. He has also had help from other people I believe (although to be honest, I have not followed his every step, so I don't know any details).
When Lee took over WoS, WoS was already in trouble (lack of updates, forum software faults and server problems). As Martjin was not spending much time on it any-more. If you can find them, within WoS forums there are gaps in time due to the forums or the whole site going off-line. This happened multiple times (and no, I did not record the details). Then WoS crashed off-line completely. Lee did what he could and he got something back-up.
I am aware of all of this, and my conclusions on technical architecture and approach used to build a replacement still stand.
1024MAK wrote: ↑Wed Nov 15, 2017 11:06 pm
If you were tasked with recovering and rebuilding an entire large web site, that was using a mishmash of old obsolete software glued together with custom code, how would you do it?
In very much the same way Einar has approached it. In terms of prioritisation:
- Preserve the existing data
- Render that preserved data as a website
- Provide the simplest collaborative process to update/correct data
- Iterate and improve
Einar saw the importance of first preserving the existing data, and shared the fruits of his work. That's goal 1 achieved. Peter and Daniel independently prototyped a website using that data, because Einar shared that data. That's goal 2 achieved. Hikoki -- or was it hikaru? -- collated data on new games released after 2012 that are not on WoS, in a spreadsheet. That was quickly imported into ZXDB. That's goal 3 working.
What remains to be seen is whether the current good-enough solution of Einar being the gatekeeper for updating ZXDB is sustainable. The data is in a much better position. It's in the form of MySQL import statements, which means anyone with general web development skills can take this data and be their own gatekeeper.
Also, note the approach. Started with just Einar. Then Peter joined by building a website on-top of Einar's data. Then other people offered their help to Peter, so now Peter has a small team helping him produce the Spectrum Computing website. Daniel started with his own website (ZXInfo), and is now also sending Einar updates and corrections, so there's at least a team of 2 working on the data updates. Before there was just Martjin, now there's a core team of 3 solidifying around the ZXDB/SpectrumComputing model.
If someone decides a CMS with fine-grained permissions is necessary, he can grab a copy of Einar's MySQL import files, spin up a database and go ahead with building out this CMS. The advantage of Einar's approach is that neither the website nor the data are dependent on having a CMS. In this model, it is not a critical must-have feature.
Einar and Peter's agility has created a viable alternative. One that's also been conducive to growing a network of collaborators. That's starkly different to Fogarty's approach, which supposedly started with a world-class development team, but is now looks to be just one guy doing the development alone.
1024MAK wrote: ↑Wed Nov 15, 2017 11:06 pm
And keep in mind that Lee wants members of WoS to help maintain the site. He does not want to do everything himself.
Lets say Fogarty's approach hasn't been conducive to that. Help has been offered by multiple people over the past few years, all rebuffed and declined. Einar bent over backwards to help Fogarty, even offering writing the import scripts to import ZXDB to whatever database schema Fogarty has.
1024MAK wrote: ↑Wed Nov 15, 2017 11:06 pm
Whoever took on the WoS site was going to have a lot of work to do.
Yes, Einar and Peter have spend a good deal of time to get to where they are, and within a year they are further ahead than Fogarty has been in 3. Einar has been open about his work, so the WOS data is no longer in danger of being lost. Peter has a small team developing this site.
The key point is that the most important goal was to preserve the existing data in a state it could be used. Migrating it into a MySQL database, and making the MySQL dump itself available to the community, damn that was smart!