3 ms·
TLDR: Facebook built the app responsible for their entire ad platform on a chrome-only tech (WebSQL), and when it was deprecated, instead of dropping in a widel
by supermatt 6y ago
TLDR: Facebook built the app responsible for their entire ad platform on a chrome-only tech (WebSQL), and when it was deprecated, instead of dropping in a widely-used polyfill, they decided to do a major 3yr rewrite risking 25% of their revenue in the process.
I guess this is the stupid shit that being "too big to fail" permits.
- asplake 6y agoI didn’t know that such a polyfill existed but even so, that was pretty much my reaction. Facebook would be quite capable of developing one, and potentially a route there to something more service-based once the panic is dealt with.
- kps 6y agoAnother option would have been to use a chromium build with WebSQL still in place.
- PeterisP 6y agoThey would have to get all their advertiser customers to use that custom chromium build, which would be possible but would not be good for their ad sales.
- protomyth 6y agoGiven how long some of us had to hang on to Internet Explorer 6, I'm pretty sure the advertisers would have required their IT staff's to comply and install the "Facebook Browser".
- ADefenestrator 6y agoWouldn't have necessarily had to be nearly as bad as IE6 for security, either. Just lock the special browser to facebook.com only, or even just the ads API, and there's not a lot of room for exploit.
- joe_the_user 6y agoThere are a lot of big, messy, "terrible" applications with dumb dependencies that seem to live far longer than they "should". It's mostly in large companies but not really related to these companies. These same applications usually have unique functionalities and user who have become expert in the use of apps, including knowing the "stupid parts". Writing these apps is hard and it's a completely rational decision to jury-rig deprecated libraries to make such an app work however irrational it might look from the outside.
- supermatt 6y agoOh, I'm totally aware of technical debt, etc. I have worked on both maintaining and rewriting a number of such enterprise apps myself. It is just in this case, if the revenue WAS ever at risk (don't forget we are talking billions here), then you would need to be a bungling idiot to not fix the issue at hand before embarking on a rewrite. Personally, I think the article is a dramatisation - I would certainly be too embarrassed to write such a thing if it were true.
- ska 6y ago> I guess this is the stupid shit that being "too big to fail" permits. Or just how technology actually mostly works out in the wider world.